G90 front fan replacement step by step artwork [Text View]

JohnHWman

12-20-06, 01:05 PM

Hello to all G90 owners !

Since I've been asked for information on silencing the Sony G90 projector (-7dBA noise gain), I've decided to post this step by step procedure about the most difficult part of the job: The front fan replacement.

The four back fans are quite easy to replace with SilenX IXP-76-14...

http://www.silenx.com/ixtremaprofans.asp?sku=ixp-76-14
SKU: IXP-76-14
Rated dBA: 14 dbA
Dimensions:120x120x38mm
RPM: 1200 rpm
airflow: 72 CFM
nominal Voltage: 12V
Current Draw: 0,20A
Power Consumption: 2.40W
Bearings: Hybrid Immersion

The complete step by step artwork for these is described in this post : http://www.avsforum.com/avs-vb/showthread.php?t=781982

However the front fan replacement is not as easy as the four rear ones and needs some step-by-step instructions as follows:

The idea was to use a larger fan than the tiny 60x60mm (18CFM) located behind the PA board. Since larger fans do not have to spin as fast as smaller ones to move the same amount of air, they are quieter. I found SilenX IXP-52-14 (24CFM) fan that is considerably better and very silent:

http://www.silenx.com/ixtremaprofans.asp?sku=ixp-52-14
SKU: IXP-52-14
Rated dBA: 14 dbA
Dimensions:80x80x15mm
RPM: 1800 rpm
airflow: 24 CFM
nominal Voltage: 12V
Current Draw: 0,09A
Power Consumption: 1.08W
Bearings: Hybrid Immersion

However because this fan is larger than the one it replaces, it cannot been place in the Sony original 60x60mm fan position behind the PA board:
http://johnhwman.free.fr/La%20Construction/ventilateurs%20projecteur%20G90/Vue%20ventilateur%20de%20FAV.jpg
Instead, we will install it in the front like this:
http://johnhwman.free.fr/G90/Front_fan/Step10.jpg

The first step is to completely remove the older 60x60mm Sony fan and its aluminum holder by unscrewing the two screws.

WARNING : People MUST BE VERY CAREFULL that PA/PD board carry High Voltage on their PCB pads and it is very important to wait at least 2 hours after unit is switched OFF and before removing those PA/PD boards from the unit by hands :eek:

What follows are step by step instructions on how to install this new fan in front of the projector:

1°]Draw the aluminum grid panel circle cut-out using the fan as a ruler:
http://johnhwman.free.fr/G90/Front_fan/Step1.jpg
Note : This is the best fan position to ensure the best cooling of the PA and PD power supplies heatings elements.

2°]Update the IR sensor black plastic enclosure to decrease the depth of the lowest part:
http://johnhwman.free.fr/G90/Front_fan/Step2.jpg

3°]Cut out the circle area on the aluminum grid using a jig saw with metal blade:
http://johnhwman.free.fr/G90/Front_fan/Step3.jpg

4°]Clean the grid's circular border to remove all metal particles:
http://johnhwman.free.fr/G90/Front_fan/Step4.jpg

Next post to follow-on...

JohnHWman

12-20-06, 01:06 PM

5°]Drill the upper left rubber hole with a 4mm diameter drill bit and the fan's wire pass-through hole with a 5mm diameter bit. Use a rubber grommet to protect the cable and place the four rubber spacers within the grid:
http://johnhwman.free.fr/G90/Front_fan/Step5.jpg

6°]Place the fan on the rubber spacers and place the fan wires through the grid plate:
http://johnhwman.free.fr/G90/Front_fan/Step6.jpg

7°]At this stage, update the SilenX white connector housing wiring and shape it to fit the PA board connector by swapping the +12V and SENSE (Rotation in french ;) ) wires inside the housing to correspond to the Sony wiring:
http://johnhwman.free.fr/La%20Construction/ventilateurs%20projecteur%20G90/Croisement%20des%20fils.jpg

Turn the fan connector upside down and cut the two tables on the SilenX white connector housing with a cutter. With a small metal saw, make two slots in the white plastic housing :
http://johnhwman.free.fr/G90/Front_fan/Step7.jpg

Connect the SilenX connector to the PA board connector as shown:
http://johnhwman.free.fr/G90/Front_fan/Step8.jpg

Then reconnect all the others connectors back in their place:
http://johnhwman.free.fr/G90/Front_fan/Step9.jpg

8°]Slide-in the PA board and place the grid back in place and screw in the four black screws:
http://johnhwman.free.fr/G90/Front_fan/Step10.jpg

9°]We now need to modify the grey plastic housing to make the fan fit:
http://johnhwman.free.fr/G90/Front_fan/Step11.jpg

10°]We also need to modify the red IR plastic to make the fan fit:
http://johnhwman.free.fr/G90/Front_fan/Step12.jpg

11°] WARNING ! Put back the front plastic part on the G90 AND check is the fan is spinning fine (free of mechanical stress) when the unit is powered ON. This is a very important point. Since that the G90 spin detection hardware is sometime lacking of the SilenX mis-rotation signal detection, this could let the unit running while the blade is stopped by the 'misscuted' front plastic part (this will destroy your PA board because of overheat).

That's it! The fan should run fine when the unit is powered up. The big difference being how G90 is now even quieter than before (7dBA gain!) Enjoy!

John

kal

12-20-06, 02:10 PM

Nice work John!

Would you mind if I posted the instructions and pictures in Curt's "Advanced Procedures" section of his website? (It's here: http://www.curtpalme.com/Advanced_Procedures.shtm ). Full credit would go to you of course.

I'm slowing trying to accumulate all useful procedures into one location to make them easier to find.

Kal

JohnHWman

12-20-06, 02:40 PM

CIR-Engineering

12-20-06, 03:19 PM

Cool John! This is a great idea. I will definitely do it if I ever get to hang my G90 in the new house... it will be some time.

Also, I still plan to train a Spider2 for you with respect to my PR-650 Photoresearch. I just haven't set up any TV or other display in my new house yet. I am literally watching a 17" CRT direct view built 30 years ago. All it has for inputs is 75 ohm antenna leads ;)

craigr

JohnHWman

12-20-06, 04:16 PM

Thanks Craigr !

I'm waiting for the Spyder2 training file to full re-calibrate my first G90 unit and my second one as soon as I'll succeed in repair it.

John

kal

12-20-06, 05:20 PM

Yes, Kal no problem,

In fact, one of my G90s came from Curt :D

Don't forget to host the relevant pictures on Curt's server too.

This fan replacement makes the G90 front noise close to unhearable :) Since the noisiest fan on this G90 is the front one, the sound gain is very sensible and major...Excellent! And yes, the pictures will go too... I always want to make it as standalone as possible such that we're not dependant on any external sources.

I'm pretty busy this time of year, but I'll try and get it up there in the next couple of days...

Kal

kal

12-20-06, 10:14 PM

Ok, so I lied... :) I found some time to do it sooner than I thought:

http://www.curtpalme.com/G90_fan_mod1.shtm

Thanks John!

Kal

Frank D

12-20-06, 11:33 PM

Excellent posting and pictures. Very interesting.

Did you do any temperature testing before and after the mod to see if there was any difference?

Any testing before/after fan mod regarding noise levels?

dochlywd

12-21-06, 12:06 AM

John,

I will have to go back and redo mine with this updated version of the mod. I replaced the front fan with the correspoding small SilenX fan that fits inside the cage. I also seem to be having a slight drift problem and was thinking the fan mod might have something to do with it since Ken Whitcomb did the set up. We all know his reputation here and my G90 should be rock solid. Didn't you have some kind of issue with yours as well? Did you ever track down what was causing it?

As far the sound reduction. I didn't do any measuring, but my ears can hear a huge difference and this is the fan that made the biggest difference. The factory fan was much louder with a high pitch ringing to it. The SilenX is a quiet swoosh. Even my wife, who thinks we are all video dorks :D commented on how much more quiet it is.

Thanks!

Mike

JohnHWman

12-21-06, 04:14 AM

Hi Kal,

Thanks for the english explanation corrections ;) I'm only French :D The procedure looks pretty good now. I will update this post as well ;)

I can do the same kind of step by step procedure for the four rear fans, yes ? That can makes the G90 silencing procedure complete :rolleyes:

Hello Frank D,

I didn't make sound measurment before the update, that's a mistake :o I can easely revert the fan update and redo the missing noise measurment. I will post the results before/after (measured at 1 meter distance) soon.

I could also make temperature measurment before/after fan update as well but is is slightly more difficult to do for me since I would need 'K' probes to use with the thermometer and I need to order them... Also, I can already tell you that because of the new fan position (front) and because it is 24CFM instead of 18 originaly, the PA and PD heatsinks temperatures will be lower for sure ...

Mike,

My original red convergence drift on my first G90 unit was gone after I swapped the Red/Blue lens. After discussing through MP with lewis (Haydn who had the same issue with his own unmodified G90), it appears that this shift is associated with the way the lens attached to the LC housing. Maybe he can talk about his own experience. So I can confirm that your "slight drift problem" isn't relevant to the fan replacement.

The factory fan was much louder with a high pitch ringing to it. The SilenX is a quiet swoosh. Even my wife, who thinks we are all video dorks :D commented on how much more quiet it is. The new larger front fan is even less louder than 60mm SilenX previous solution :)

John

Don_Kellogg

12-21-06, 07:47 AM

Dude..

If you didn't do Temp checks how do you know this is not going to fry the thing? I realize the CFM is close to the same levels but. Just wondering.

Glad the issues with the Red drift are fixed, that is just strange I thought all 3 lenses were the same. Guess I better make them.

JohnHWman

12-21-06, 08:06 AM

If you didn't do Temp checks how do you know this is not going to fry the thing? Appart from its higher CFM, because the new fan place is more efficient than original Sony placement ;)

Frank D

12-21-06, 09:08 AM

Hi Kal,

My original red convergence drift on my first G90 unit was gone after I swapped the Red/Blue lens. After discussing through MP with lewis (Haydn who had the same issue with his own unmodified G90), it appears that this shift is associated with the way the lens attached to the LC housing. Maybe he can talk about his own experience. So I can confirm that your "slight drift problem" isn't relevant to the fan replacement.

John

Wow so because of the way the lens was attached it created drift. Interesting.

Was your lens attached different from all the rest of G90 owners out there?

Perhaps you and Haydn should do another post on this?

Frank

kal

12-21-06, 09:47 AM

Hi Kal,

Thanks for the english explanation corrections ;) I'm only French :D The procedure looks pretty good now. I will update this post as well ;) No problem! :)

I can do the same kind of step by step procedure for the four rear fans, yes ? That can makes the G90 silencing procedure complete :rolleyes: Sure! Any pics or procedures you add to this thread, I'll add to the procedures on Curt's site.

Kal

JohnHWman

12-21-06, 12:20 PM

Wow so because of the way the lens was attached it created drift. Interesting. Yes Frank and this is due to lens/tube housing/heatsink surface issue(s) as lewis discovered...
Was your lens attached different from all the rest of G90 owners out there? No, maybe just issue with pairing lens with tube housing/heatsink
Perhaps you and Haydn should do another post on this?yes, I would let lewis explain his own findings using my following thread
http://www.avsforum.com/avs-vb/showthread.php?t=665736

John

CIR-Engineering

12-21-06, 04:28 PM

John,

I would very much like to hear any and all info regarding your red shift and lens fix. This is very weird to me.

Thanks

craigr

JohnHWman

12-21-06, 04:50 PM

Sure Craigr,

I've asked lewis to share his findings on the relevant thread.

John

Chuchuf

12-22-06, 02:16 PM

John,

Very nice work indeed.

I just finished installing the 4 Silenx 90 CFM fans (the next size up from what you used) in the rear of my G90 as well as the front fan as you indicated above.

The noise reduction is pretty dramatic.

I had a hush box in my install to try and quiet it down but there was always the noise from that front fan. And even with the hush box it was pretty loud.

Now with the hush box off (mind you I have none of the covers on my PG at all, it is considerably quieter. With the hush box on, and everything off in my theater, I'd say it is quieter than my HTPC, which is VERY quiet. You can hardy tell the G90 is running.

What a pleasure!!!!

Now for some temp tests when I get a chance to insure the fans are doing their job (they seem to move a lot of volume).

Nicely done John.

Terry

Also in your drawing above of the stock Silenx connector they should be labeled starting from the top Gnd, +12V, Sense. I thing you may have these reversed.
The leads on the Sony stock fars are Red=12V, Black or Blue=Gnd and Yellow= Sense.

overclkr

12-22-06, 02:23 PM

John,

Very nice work indeed.

I just finished installing the 4 Silenx 90 CFM fans (the next size up from what you used) in the rear of my G90 as well as the front fan as you indicated above.

The noise reduction is pretty dramatic.

I had a hush box in my install to try and quite it down but there was always the moise from that front fan. And evem with the hush box it was pretty loud.

Now with the hush box off (mind you I have none of the covers on my PG at all, it is considerably quieter. With the hush box on, and everything off in my theater, I'd say it is quieter than my HTPC, which is VERY quiet. You can hardy tell the G90 is running.

What a pleasure!!!!

Now for some temp tests when I get a chance to insure the fans are doing their job (they seem to move a lot of volume).

Nicely done John.

Terry

Also in your drawing above of the stock Silenx connector they should be labeled starting from the top Gnd, +12V, Sense. I thing you may have these reversed.
The leads on the Sony stock fars are Red=12V, Black or Blue=Gnd and Yellow= Sense.

Uh oh, looks like I need to order some fans. :D ;)

Cliff

JohnHWman

12-22-06, 07:00 PM

Thanks Terry !

I see you go with 90CFM version for the four rear fans, as I said, I think that 90CFM is useless for the G90 since the original Sony stock fans are 77 CFM wich is very close to the 72CFM of the IXP-76-14 ... And using these will save you 4 dBA noise !

I'll do some sound measurments before and after all fan exchanges (I'm still waiting for IXP-76-14 fans).

About fan connector wiring between SilenX and Sony, I checked my HP and I'm sure about the following infos :

For the fours rear fans :

- from G90 side : the blue wire is "-B" power and this must be linked to the fan's ground wire.
- from G90 side : the Yellow wire is "SENS" signal and this must be linked to the fan's rotation info wire
- from G90 side : the black wire is "GND" ref. and this must be linked to the fan's +12V wire

For the front small fan :

- from G90 side : the black wire is "-B" power and this must be linked to the fan's ground wire.
- from G90 side : the Yellow wire is "SENS" signal and this must be linked to the fan's rotation info wire
- from G90 side : the red wire is "GND" ref. and this must be linked to the fan's +12V wire

In fact, fan's power supplies are negative vs ground, that is why they need to be linked to fan's ground and PJ ground to +12V fan's wires ;)

I hope it's clearer now for everybody.

John

dochlywd

12-22-06, 08:39 PM

Once the wire swap is done, you can use the adapter given with the SilenX fans as a jumper to connect the G90 lead to the SilenX fan connector. Just take an exacto knife and remove the three wires from the connector on the extra adapter. Then take a pair of needle-nose and push the pins so they are equal on both sides. Then insert the jumper into the G90 lead followed by inserting the other end into the fan lead. Wrap a piece of electrical tape around it and you're done. See attached pics.

Sorry for the poor quality. I can try to get better pics if needed.

Mike

Chuchuf

12-22-06, 10:05 PM

The reason I chose the 90CFM John is I didn't want to take any chances. Sometimes from what I have read, silent fan manufacturers are a bit overzelous with their specs. Plus I have a hush box to help dampen any additional noise.
Just watched a movie and the difference is fantastic!!

Terry

Don_Kellogg

12-23-06, 01:34 AM

Cliff let me know where you order them from I think I'll go that way as well. Not that it's all that loud but it would be nice since I don't know that I'll build a hush box.

Chuchuf

12-23-06, 07:28 AM

Cliff let me know where you order them from I think I'll go that way as well. Not that it's all that loud but it would be nice since I don't know that I'll build a hush box.

Don,

Compared to these new fans, it IS all that loud.

FYI I got mine at EastLuna dot com and FrozenPC dot com.

Terry

James McClellan

12-23-06, 08:31 PM

...I chose the 90CFM...
Terry, are these the fans you used? Model # IXP-76-18?

http://www.frozencpu.com/products/4274/fan-235/SilenX_120x120x38mm_-_18dBA_-_90CFM_iXtrema_Pro_Fan_IXP-76-18.html

http://www.eastluna.com/hardware/details.php?id=1965

James

Chuchuf

12-24-06, 08:54 AM

James,

Here is a "cut and paste" from what John originally sent me:
Here the best solution so far (quieter one) - The one I use on my G90 and the one I will use on my second G90 unit :

Four back fans change :

http://www.silenx.com/ixtremaprofans.asp?sku=ixp-76-14
SKU: IXP-76-14
Rated dBA: 14 dbA
Dimensions:120x120x38mm
RPM: 1200 rpm
airflow: 72 CFM
nominal Voltage: 12V
Current Draw: 0,20A
Power Consumption: 2.40W
Bearings: Hybrid Immersion

OR

http://www.silenx.com/ixtremaprofans.asp?sku=ixp-76-18
SKU: IXP-76-18
Rated dBA: 18 dbA
Dimensions:120x120x38mm
RPM: 1400 rpm
airflow: 90 CFM
nominal Voltage: 12V
Current Draw: 0,24A
Power Consumption: 2.88W
Bearings: Hybrid Immersion

Personnaly I would go for 72CFM version since this is the quieter solution with good airflow.

Front fan change by :

http://www.silenx.com/ixtremaprofans.asp?sku=ixp-52-14
SKU: IXP-52-14
Rated dBA: 14 dbA
Dimensions:80x80x15mm
RPM: 1800 rpm
airflow: 24 CFM
nominal Voltage: 12V
Current Draw: 0,09A
Power Consumption: 1.08W
Bearings: Hybrid Immersion

I chose to go on the side of caution and use the IXP-76-18 90 cfm fans because I also run a hush box. According to John that isn't necessary and the IXP-76-14 72 cfm fan is all that is needed.

But to answer your question the links you profided are the replacements to the 4 rear fans I used.

G70 owners should also be having a look at this thread as well.

Terry

JohnHWman

12-24-06, 10:56 AM

G70 owners should also be having a look at this thread as well.
True Terry ;)

The SilenX IXP-76-14 or IXP-76-18 can be used as direct replacement of the three rear G70's fans. I'll do it for my brother's G70 projector ;)

John

Chuchuf

12-24-06, 11:49 AM

What I find funny John is that the original fans are labeled in big letters "Silent Fan"......lol

Terry

secstate

12-24-06, 12:31 PM

James McClellan

12-24-06, 02:14 PM

Four back fans change :

http://www.silenx.com/ixtremaprofans.asp?sku=ixp-76-14
SKU: IXP-76-14
Rated dBA: 14 dbA
Dimensions:120x120x38mm
RPM: 1200 rpm
airflow: 72 CFM
nominal Voltage: 12V
Current Draw: 0,20A
Power Consumption: 2.40W
Bearings: Hybrid Immersion

OR

http://www.silenx.com/ixtremaprofans.asp?sku=ixp-76-18
SKU: IXP-76-18
Rated dBA: 18 dbA
Dimensions:120x120x38mm
RPM: 1400 rpm
airflow: 90 CFM
nominal Voltage: 12V
Current Draw: 0,24A
Power Consumption: 2.88W
Bearings: Hybrid Immersion

Front fan change by :

http://www.silenx.com/ixtremaprofans.asp?sku=ixp-52-14
SKU: IXP-52-14
Rated dBA: 14 dbA
Dimensions:80x80x15mm
RPM: 1800 rpm
airflow: 24 CFM
nominal Voltage: 12V
Current Draw: 0,09A
Power Consumption: 1.08W
Bearings: Hybrid Immersion

Terry

Thanks Terry. I'm not sure if I'll run a hush box or not, so I just went crazy and ordered
• 4 ixp-76-14's
• 4 ixp-76-18's
• 1 ixp-52-14.

I figure, I'll just sell the ones I don't use here on AVS. Yeah, I'll take a loss on my purchase, but I accept that.

James

wrighthenry

12-26-06, 12:31 PM

So, any before and after temp readings?

Frank D

12-26-06, 02:25 PM

Thanks Terry. I'm not sure if I'll run a hush box or not, so I just went crazy and ordered
• 4 ixp-76-14's
• 4 ixp-76-18's
• 1 ixp-52-14.

I figure, I'll just sell the ones I don't use here on AVS. Yeah, I'll take a loss on my purchase, but I accept that.

James

Great James! Let us know how your mod works out and related tips/issues/pictures.

Thanks
Frank

Don_Kellogg

12-26-06, 03:24 PM

Good deal then Thank you Terry, I will order some from frozenCpu I deal with them all the time. That much of a difference :) how many DB is it do you have a meter?

Chuchuf

12-26-06, 04:37 PM

Don,

I have a meter but didn't do any before.
The sound from these is totally different than the stock fans which had a "loud" woosh sound to them. The difference is not subtle, it is VERY obvious. And once again, in my hush box (which didn't quiet down the old front fan AT ALL) I have to make sure the PJ has turned off when I shut it down it is so quiet.

Wright, I haven't done any before and after temp readings but I can tell you that I notice no adverse effects on the PJ from 3 days of constant use. I will have to install a temp gague in my hush box and see if these has been an increase in temperature.

Terry

Chuchuf

12-26-06, 04:38 PM

Don_Kellogg

12-26-06, 05:03 PM

Opps sorry thank you John ")

Tedd

12-28-06, 08:39 AM

Nice fan swap.

I wonder if modding the rear of the fan casing to circular (by cutting off the square tabs) would allow it to be fitted into the new cutout recessed, and avoid the grill and ir modifications?

JohnHWman

01-02-07, 03:31 PM

No Tedd,

I don't think so since there is almost no room free at the back of the front grid (the PA and PD PCB edges are very close to the aluminium grid !).

All : I still waiting for the IXP-76-14 four fans to make my before/after noise measurment. They should be there soon now. I will report in here asap.

John

Stevoscopy

01-03-07, 05:39 AM

Just a note:
Been running SilenX on my 12xx for over a year now. They have not missed a beat.
I will use them again in the future.
Originally I was after Papst fans but model I wanted was not available in Aussie.

They may/might be noisier than the Papst but the dramatic difference from the stock standard fans made me forget that as soon as I hit the little green button.

Frank D

01-03-07, 09:41 AM

JohnHWman

01-06-07, 09:55 AM

All,

I've updated my step by step procedure in my second post of this thread to add three pictures of the SilenX original PC white housing update. This operation should now be more easy to do with pictures. kal : you may update Curt's HP as well !

As promise, I've received my IXP-76-14 four fans and installed them after making noise measurment before/after.
All noise measurements were made before and after the fan replacement at a 1m distance in front and back of the G90 with it's outer plastic shell removed (ie: worst case scenario).
http://johnhwman.free.fr/G90/Front_fan/Noise-measure.jpg

Here is the results :

- Noise floor : 34dBA
- With original Sony fans : Front = 44dBA and Back = 46dBA
- With new SilenX fans : Front = 37dBA and Back = 39dBA

Overall noise Gain : 7dBA not that bad ;)

John

mp20748

01-06-07, 10:01 AM

Overall noise Gain : 7dBA not that bad ;)

John

Very good. So without the mod, the noise is more than twice what you're getting now?

JohnHWman

01-06-07, 10:08 AM

Almost Mike,

It's mandatory that human hears noise perception is divided by a 2 factor for 10dBA less noise level.

John

mp20748

01-06-07, 10:16 AM

Almost Mike,

It's mandatory that human hears noise perception is divided by a 2 factor for 10dBA less noise level.

John

with that and looking at the distance (between the meter and G90), and by you showing the environment that the measurement was taking in. The difference is huge after your mod.

I would think based on this, that with the G90 mounted in a more acoustical environment (HT), then if the G90 was heard during subtle sound from a movie before this. It should not be heard at all after the mod.

Kudos!

kal

01-09-07, 09:26 PM

Ok, the Sony G90 Fan Mod procedure on Curt's site has been expanded to show you how to replace the back 4 fans in addition to the front fan. Make that G90 nearly twice as quiet!

See: http://www.curtpalme.com/G90_fan_mod1.shtm

Kal

JohnHWman

01-12-07, 03:37 PM

Don't forget to add the SilenX connector housing pictures that were added lately (point #7 : 3 more pictures now).

Picture always tells thousands words ;)

John

kal

01-12-07, 04:56 PM

Done.

That's enough modifying and expanding of your instructions (please). :)

Kal

JohnHWman

01-13-07, 09:07 AM

Thanks kal, that's perfect now !

KrisRoberts

01-28-07, 03:39 AM

JohnHWman

01-28-07, 04:55 AM

Hello Kris,

No, I didn't checked temperatures (before/after) fans mods on my second G90 unit since this unit is still under servicing (YA board failure, see :http://www.avsforum.com/avs-vb/showthread.php?t=735051&page=2)

On my first unit I mades some temperatures measurments after the fans mod.
http://www.avsforum.com/avs-vb/showthread.php?p=7483882#post7483882
My first unit is using both a circulation system from blowing out of a box (exhaust) and a hushbox.

John

kal

01-31-07, 12:47 PM

Though I am not arguing these fans are an improvement, SilenX has a bit of a reputation in the QuietPC community for exaggeration, particularly their noise ratings. The fans are high quality but don't believe for a minute that the sound rating is at the rated airflow. See this real world test for an example:

http://www.silentpcreview.com/article689-page4.html#silenx

Basically that 80MM fan (based on their testing results) is really no better than the one in my Barco 808 that sounds like an airplane (the later based on NMBs specs for the fan). It has the slighty less air flow and slightly less noise.

For those investigating fan replacements there are two really good recent indepedent tests of 80 MM and 120 MM fans at the same site:

80MM Fans:
http://www.silentpcreview.com/article689-page1.html
Hi! I have to agree that SilenX seems to be overzealous in their claims of low noise. In some cases the actual amount of air moved (CFM) doesn't match their rated spec either.

But the review you mention above is for an older version of the SilenX fans so I wouldn't use look at the conclusions too closely.

For example, here's a picture of the 80mm fan SilenX fan they used:

http://www.silenxusa.com/productcart/pc/catalog/sx_fan_8025.jpg

Here's the SilenX style that we're using now in our CRT projectors:

http://www.silenx.com/ixp-54-front-large.jpg

So you can see, they've redesigned it. With the new bullet nose and a few other changes, it "looks" like it may be quieter, but I still find that their actual dBA ratings are way to optimistic. Be weary of the CFM output as well. Make sure to take before/after temp readings. I replaced my 80CFM Barco808 fans with 90CFM SilenX models (the new style) and they're definitely quieter and the internal temperature is more or less the same (within a degree or so).

Kal

Earlyadopter

01-31-07, 03:57 PM

Has anyone done temp before/after checks? I've always wondered about this sort of thing from the perspective of the air pressure and flow. When you switch a circulation system from blowing out of a box (exhaust) to blowing into a box (intake) it seems like it changes the box from being a low pressure zone that draws air from a number of vents to a high pressure zone that blows air out many of those same vents instead of primarily through the one vent where the exhaust fan used to be.

The old exhaust fan pulled cool air from the vents on the other side of the box and forced it over the components. The new intake fan may blow more air in than the old exhaust fan, but as it raises the pressure in the box some of that air blows out the holes in the grill its mounted to and the actual amount of air going over the components and out the other side of the box might not be as much as it was.

If the overall airflow system actually draws from the original exhaust manifold with other fans down the line I guess that low pressure might help keep the circulation up.

I dont have my projector yet, but I'm certainly interested in getting it as quiet as possible without jeopardizing the equipment. If nobody has taken measurements yet, I'll do my best to collect some data when I do get it.

I had the same concerns about mounting the front fan on a perforated panel. I agree that pushing or pulling the air through a closed chamber should result in a similar cooling effect, but leaving the perforations in the front panel provides an additional outlet. When I performed the fan replacement on my g90 i didn't care as much for the asthetics of hiding the front fan as I did for providing the best cooling, so I positioned the fan directly in front of the heat sink on the PA board. This meant cutting more of the plastic front grill away, but I felt it provided better air flow over the heat sink and to the board below. As a side benefit, the IR receiver does not have to be modified. In addition, I used aluminum duct tape to seal the holes in the front panel. With this change, no air escapes from the front. The pressurized chamber now exhausts air through the old fan opening and all other openings that previously let air in.

Unfortunately, I am unable to make any internal temperature measurements, but the mod has been working for a couple months with no problems. A couple other methods I changed were: instead of cutting the fan hole with a saw, a 3" metal cutting hole saw works great, producing a perfect round hole. Instead of adding a hole and grommet for the fan wires, the fan wire can be rerouted out of the back of the fan housing through the 3" opening. Just be careful that it doesn't get pushed into the blades upon installation.

I also agree with you about not jeopardizing the equipment for the sake of less noise. I originally replaced the rear fans with the recommended IXP-76-14 (72 CFM) fans. I didn't take measurements, but the outlet air at the rear of the projector felt warmer than with the OEM fans. My projector is ceiling mounted with no way to bring cool air to the front intake area so I have to rely on the fans to do an adequate job of cooling. I found that after extended periods of operation (> 3hrs) the picture started losing vertical stability. Blowing cooler air from the floor up to the front of the projector solved the problem, but this was not an acceptable fix. I replaced the IXP-76-14 fans with IXP-76-18 (90 CFM) fans and this solved the cooling problem. The noise produced by the IXP-76-18 fans may be close to that of the original Sony fans, but to me the increased cooling is worth it.

Best of luck with your projector & mod!

wirespeed

02-07-07, 12:18 AM

Does the G90 have temp sensors? If I somehow screwup the fan replacement, will it gracefully shutdown, or will the magic smoke escape?

Earlyadopter

02-07-07, 11:06 AM

Does the G90 have temp sensors? If I somehow screwup the fan replacement, will it gracefully shutdown, or will the magic smoke escape?
There may be some thermal protection devices or circuits on some of the boards, but none are tied to the error code display. In other words, if a thermal shutdown or failure occurs, it won't be reported as such, only as a board failure. The only cooling monitoring that is done is the monitoring of fan speed through the sensor or tach, fan output. If the front fan stops, the projector shuts down and error code 6E is reported on the LED display. If any of the four rear fans stop, the system shuts down with error code 6D displayed.

Since Sony relies on the fans operating to sense proper cooling, replacing the fans with lower capacity fans can cause overheating. This is why I replaced the IXP-76-14 SilenX fans with IXP-76-18 90CFM units. I also don't trust the IXP-76-14 fan rating of 72 CFM, since my fans were identified as 60 CFM units. The SinenX website explains that the 60CFM rating was established on a pre-production model. I'm sure the flow isn't even near 72 CFM when installed in the Sony. The outlets are perforated metal panels that generate some backpressure, reducing the flow.

I doubt the magic smoke would escape, but reliability and stability could be affected, so I think the tradeoff of a few more db of noise for lower operating temperatures is a good one.

JohnHWman

02-20-07, 04:31 AM

Thanks for sharring your own experiments Earlyadopter,

Positioned the front fan directly in front of the heat sink on the PA board can be a good option, you're right. However please note that PD board SMPS inverter may need air cooling for its two heatsink (MOS transistor and dual power diode). Thus lowering the fan's position will decrease the airflow on these two PD board elements.

But yes, as a side benefit, the IR receiver does not have to be modified and this save artwork.

In addition, I used aluminum duct tape to seal the holes in the front panel. With this change, no air escapes from the front. The pressurized chamber now exhausts air through the old fan opening and all other openings that previously let air in. That is a good thing I should I've think of ;) I will do it and update this thread procedure / pictures.

About G90 sensor temp : Please note that according to G90 SM and theory of operation manual, there is one temp sensor located on the PE board (located behind the rear fans at the right). According to TOO, This sensor is used to monitor the internal air ambient temperature and adjust the rear fans rotation speed when temperature rize. Refering TOO page 12-12 (section 12-11-3) : the fans applyed voltage variation is : 10.5V@25°C until 13.5V@40°C.

I personnally measured 14VDC on rear fans when I tested the SilenX IXP-76-14.

But I should agree with you that hanging the G90 on the ceilling may require the use of IXP-76-18 fans rather than 76-14 to be secure on the top room temperature effect (especially in summer).

My G90 configuration is special since I use a special warm air exhaust duck with separated motor to extract the generated warm air from the G90's back : thus, I haven't any static pressure issue with the rear fans back grids within the unit.

John

KrisRoberts

04-01-07, 01:28 PM

How is this mod working out for people?

I got one of each, a 60x25mm 18cfm and a 80x15mm 24cfm from Silenx.

Last night I pulled out the stock 60x25mm fan and played with running each of them from a 12v power supply on my desk.

I put the stock fan back in the projector for the time being and put a thermal probe inside near the rear of the PA board. I'll run it in the stock config for a while and monitor the temperature.

Running the three fans on my desk, it was clear that the stock fan blows _MUCH_ harder than either of the Silenx fans. And honestly I'm pretty skeptical that either one would move enough air to be a satisfactory replacement.

I guess I'm thinking about replacing the stock fan with the 60mm and installing the 80mm fan up front and seeing how the noise and temperature are with both going as opposed to the single stock fan.

I'm sure the 80mm fan up front is much quieter than the stock 60mm fan, but honestly I'm not convinced that it moves as much air as the specs say sitting in the open on my desk and am even less comfortable with its static pressure capacity to force air through an enclosed space.

JohnHWman

04-02-07, 03:01 AM

Hello Kris,

Based on their respectives specs, original is 18CFM and suggested SilenX 24CFM. Even SilenX may little overate their products specs, the SilenX solution move more air than the orginal solution : that's sure.

You feel that stock fan blows much harder than SilenX but keep in mind that they are not the same size (60mm vs 80mm) then it's normal that you feel so, even the CFM are telling the opposit.

Also, the new fan position (in front of the heatsinks elements of the PA and PD boards) ensure that the cooling of such heating elements are better than before (the air static pressure on these elements are much much stronger than the original Sony Solution).

FYI, the best elements to probe temp. are the heatsinks geometrical centers.

Hope this makes things clearer ?

John

spatz

04-03-07, 03:27 AM

Dear John,

I have been performing noise mods on all G90s I sold and these are many. If the projector is operated in an environement where a human being still feels ok let us say maximum 35 degrees then you can just reduce the fan speed of the original fans getting even better results. The fans used in SONY G70 and G90 projectors are very high quality fans and the noise they make comes mainly from the amount of air they are moving. So if you install "same spec fans" and they are 7db less I doubt that they move the same amount of air. The noise from the bearings from a good quality fan is very low compared to the airflow noise.
These modified projectors have been all running for several years and thousands of hours including my own G90.
The reason why it is possible to make these reductions is simply that these machines have been designed to work also in rear screen cabinets where temperatures can easily reach 45-50 degrees centigrades.
Only the 1292 has developed sometimes tube failure being used in these cabinets with not enough cooling of the CRT surface resulting in partial delamination of the aluminium coating.
The small front fan however should be replaced by the big one as you did as smaller fans usually are more noisy because of much higher revolutions to achieve a minimum of air flow.
I have been speaking with a guy that had MARQUEES operating without any fans in a climate controlled environment to 20 degrees so I would not be too surprised if a SONY G90 has the same cooling headroom.
Probably it will be difficult to find someone that would give it a try :)
NECs however run very hot and I would not dare to try this with a XG version projector.

kawal

04-04-08, 06:14 PM

Late last summer I picked up a nice G90 with 1634 hours and new looking tubes. But since my G70 has been serving me so well, I haven't bothered to put up the G90 yet. I just got it into my shop the other day and turned it on, and I was quite disappointed in the amount of fan noise it had. That's when I came back to the old reliable avsforum, and found this thread that old reliable John started a while back. Good work, again, John. :)

I didn't even realize that the G90 had a front fan until I read this thread, but it was obvious most of the noise was in the front of the projector. The four rear fans weren't too bad.

I read every post in this thread, and as usual, I learned a lot. I pulled the front fan and looked up the specs. It is a Japan Servo TUDC12Z4S and that little bugger turns at 4300 rpm - so no wonder it's so noisy! Like the rear fans and those in the G70, these are all very high quality fans and mounts, but also like the G70, it seems that the cooling design from Sony may be somewhat overkill for our home theater applications.

I noticed in this thread that there are a lots of questions, concerns, and comments about replacing the OEM fans with the SilenX fans. Last year, when I saw the similar thread regarding the fan replacements for the G70 projector:

http://www.avsforum.com/avs-vb/showthread.php?t=788782

I did a bunch of research into fan design, specifications, and fan performance. All of my findings and references are in that thread. But to summarize, I found that the SilenX specifications are way overrated, as well as incomplete. Providing the maximum airflow, even if it was accurate, is only part of the information needed, because maximum airflow is only achieved when the fan is sitting by itself on a test bench. Once the fan is installed in its environment, the restrictions in the projector case reduce the airflow to an amount depending on the fan's static pressure rating. SilenX doesn't even provide static pressure ratings, but their slower speed generally results is lower pressure ratings than the faster OEM fans. Think of the static pressure as the pressure you would feel if you put your hand over the entire fan, thereby stopping all airflow.

The bottom line is that when SilenX quotes an airflow rating like 72 cfm for its IXP-76-14 fan, it won't be anywhere close to the airflow of the OEM Japan Servo fans. Even the "90 cfm" IXP-76-18 fan, which runs at 1400 rpm instead of the 1200 rpm 76-14 fan, won't compare to the OEM fan. I even had SilenX provide me with a special set of three 1600 rpm fans, and they do not put out the same about of air as the original fans in the G70 which run at around 2000 rpm (spec is 2650 rpm).

With all of that said, I am running a set of SilenX fans in my G70, and I haven't had any problems with my projector overheating. They are MUCH quieter than the OEM fans, so even though they don't perform as well, they do perform adequately.

Now that I have a G90 to put up in place of my working G70, I plan to install the three SilenX fans from my G70 in the G90. These run at 1600 rpm, but I just ordered an IXP-76-18 (1400 rpm) to replace the fourth rear fan in my G90. And I ordered the IXP-52-14 80mm fan that John recommended for the front fan. Although I doubt the 24 cfm rating of the SilenX fan (1800 rpm) will perform as well as the 19 cfm rated OEM fan (4300 rpm), I expect it too will perform adequately for the purpose. And from all of the previous comments on this thread, I'm sure all of these fans will be a heck of a lot quieter than the original fans.

When you look at the OEM fans and their quality, then look at the cheap plastic SilenX fans that don't even have ball bearings, you really wonder if you want to put these into your top quality Sony projector. The SilenX fans have tiny motors and huge blades and take half of the power of the OEM fans to run them, but they do the job and you don't see them, so why continually put up with all of that fan noise?

So again, John, you've made a good practical recommendation. :)

Brian Feldman

04-05-08, 05:30 PM

There is some great information in this thread regarding reducing the fan noise in the G90. Rather than swapping fans with another type wouldn't it be a lot easier to put in some dropping resistors so that we can reduce the RPM's of the fans? This was a popular mod to the NEC XG projectors and it worked great.

If someone can figure out what value resistor we would need to reduce the stock fans by 1/3, this in theory should reduce the amount of noise without having to go through the expense and effort of sourcing out and replacing all of the fans in the projector.

Any thoughts on this???

Fellenz

04-05-08, 06:13 PM

Reducing the RPM's would reduce the airflow and may lead to overheating. you will be better off replacing the fans with quieter ones that move the same volume of air.

KrisRoberts

04-05-08, 08:15 PM

I took a shoebox sized clear plastic storage box that sealed tightly and made two holes in it. A 60mm hole on one end with the stock fan and am 80mm hole on the other side. It was shocking to me how much less air the SilenX fan moved than the stock fan. I dont have equipment to measure the air flow, but a tissue over the exhaust hole gives a pretty good relative comparison.

The arguments about the 80mm fan moving more air but it being less apparent because its larger kind of go out the window when you are comparing both fans in an enclosure with the same exhaust route.

Another consideration is that the stock fan blows the air out of the power unit enclosure, creating a low pressure zone that pulls air from a number of places and across the components. The SilenX fan on the front of the power unit blows air in, creating a positive air pressure zone and some of the air goes out the same exhaust where the stock fan was but some of it may go out the holes in the metal right next to the fan.

Finally, the mod instructions show how to remove material from the plastic grill for the SilenX fan to fit on the front. When installed, the rest of the plastic grill is immediately adjacent to the fan and seems like it must be obstructing the airflow into the fan. When I was experimenting with mine, it sounded clearly different with and without the front plastic grill. If it sounds different, that suggests to me that the fan speed is different and so is the air flow.

Many people seem to be happy with this fan mod and not many have reported failures. I would encourage anyone considering a fan mod of any kind to put in one or more thermal probes and measure the temperature before and after the mod.

In the end, I stuck with the stock fan but did as Brian suggests above. I wired in a variable fan control module and turned down the RPMs slightly. My thermometer tells me the main heat sink is about 10 degrees warmer than when the fan runs at full speed and I'm comfortable with that. When I experimented with the front mounted 80mm SilenX fan, the temperature got much warmer than that very quickly.

Brian Feldman

04-05-08, 08:48 PM

Kris, since there are a number of fans in the G90, how did you use an individual potentiometer to control the speed of all of them? If this is possible, it would be great to be able to manually control the speed of the fans so that we can greatly eliminate the noise the fans make.

In my particular situation, the room has a 12' celeing and the g90 hangs down aprox 4' from the ceiling and has a a/c register that blows cold air right on top of it. In this scenario I think that it would be pretty safe to be able to slow down the stock fans and still provide the proj with plenty of cooling.

KrisRoberts

04-05-08, 10:44 PM

I only slowed down the small fan in the front. I have not changed or slowed down the four large fans in the rear. In my theater, the higher pitch noise from the small fan in the front was predominant. I used a fan control module that is intended to be used in a PC and wired it with the same style connectors from the sony fans. The wires on the control unit are long enough that it stays outside the projector as does the thermometer. That way I could easily monitor the temperature and adjust the fan so the heat stayed in the range I was comfortable with.

Who knows what the actual tolerances are with these projectors? Its been suggested that they were designed to be used in commercial installations with several units sharing cramped rear projection enclosures in which the temperatures would have gotten much higher than anyone would tolerate in their home theater. Its quite possible that operating at 40 degrees above what I observe with my stock fan is perfectly acceptable. But until I have a donor machine for spare parts on hand I'm personally opting to tolerate more noise.

JohnHWman

09-29-08, 08:07 AM

FYI, I've updated the procedure to add step#11 (fan spin check) since there is a high risk leaving the fan mechanically stressed by a 'misscutted' plastic front panel (in step #9 and "10)...

See full story here : http://www.avsforum.com/avs-vb/showthread.php?p=14752225#post14752225

This is an important point.

Thank you

John

kawal

09-29-08, 12:29 PM

Note: I posted this same information in a seemingly related thread:

http://www.avsforum.com/avs-vb/showthread.php?t=1070026

I hope I'm wrong, but...

This is all very interesting to me, because recently my G90 image started shaking once in a while, but since it is intermittent, it is hard to tell what is happening. At first I just thought it was the source video (Dishnet HD-DVR), but internal test patterns would also shake.

A couple of weeks ago, my front SilenX fan started howling. I had replaced all five fans per John's thread - on my previous G70 I had also replaced the three fans with no problems. But I'm wondering about this front fan on the G90 as it seems to have very little airflow compared to the OEM front fan. In my case, I had decided to tape up all of the front holes in the G90 case to insure that the front SilenX fan would pull all of the air through it and force it back through the chassis.

When the fan started howling, I hit the front of the cabinet and it stopped howling. I assumed it was running since I got no error codes. This howling happened again - only once more - and I "solved" the problem by hitting the case again.

But these recent threads worry me and I am going to pull the front off to insure the fan is running at the very least!

Most of the time my picture is very stable, but still once in a while I get this shaking image - not awful, but noticeable.

I'm thinking perhaps I should put the OEM fan back in, but that noise is awful. How can I slow the fan down and keep the front SilenX fan in place? Just using the resistor for the OEM fan and tap into the wires for the Silenx fan?

These problems do seems related, don't they?

Ray

kawal

10-04-08, 02:21 AM

Since I am concerned about the ability of the SilenX fan by itself to provide adequate cooling, I decided to wire in both the SilenX fan AND the OEM fan. But I slowed down the OEM fan to about 3000 rpm. It normally runs about 4500 rpm and at that speed, it is really noisy.

I felt a good compromise would be to use both fans, and the results are pretty decent. Even at 3000 rpm, the OEM fan is evident, but far quieter than at full speed. Not as quiet is with only the SilenX fan, but I have more confidence in the cooling capacity with both running.

It took some experimenting with various resistors to slow down the OEM fan, because the power supply won't maintain 12 volts to the fans with both connected. The details of what I ended up with and the results of this modification are provided on this site:

http://www.kawal.net/G90%20front%20fans.htm

Also included are some graphs to show what happens to the fan speeds as voltage to them is varied.

Overall, I think I have found a good compromise between reasonable projector noise and safe cooling capacity.

Ray

kawal

10-11-08, 03:18 PM

As I noted in another thread, I decided after a week of trial that even the slowed down OEM fan was still too noisy.

So I purchased another SilenX fan, a 60mm IXP-34-16, and replaced the OEM fan with the SilenX fan in the OEM fan housing. Updated details are described at this site:

http://www.kawal.net/G90%20front%20fans.htm

With both a 60mm and an 80mm SilenX fan combination, the projector is much quieter and I believe I have adequate cooling for the circuit boards.

Ray

Richardc

10-13-08, 01:10 AM

Ray
Did you take into account that the Japan Servo model fan on the PA board is run from a -15 volt supply through a 33 ohm resistor.
And the Fan prot 2 detection is in front of the resistor.
Richard

kawal

10-13-08, 12:01 PM

Ray
Did you take into account that the Japan Servo model fan on the PA board is run from a -15 volt supply through a 33 ohm resistor.
And the Fan prot 2 detection is in front of the resistor.
Richard

Richard,

After reading your post, I just did look into your comments and I looked at the schematic in the Service Manual. Here is the schematic:

http://www.kawal.net/front%20fan%20schematic.jpg

All I did was replace the OEM front fan with two SilenX fans connected in parallel, with the speed sensor circuit connected to the 60mm fan as it was on the OEM fan.

I see that the two fans are in series with the 33 ohm resistor R14 and that the sensing circuit is downstream from the resistor. Do you see this as a problem? The working voltage with the two SilenX fans is only 10.2 volts instead of around 12 or so. Would this have an effect on the sensing circuit? And perhaps make it inoperative? Should R14 be replaced with something else? I have seen some comments that perhaps the sensing circuit may not be working well (i.e. shutting down the projector with a code) with the SilenX fans.

I'm a mechanical engineer, not an electrical engineer, so I don't have the knowledge to be able to recommend a modification.

Perhaps JohnHWman could weigh in here as well and suggest his recommendation?

Ray

Richardc

10-14-08, 04:13 AM

Ray
I Have done some research into both types of fans.
These Silenx type are not compatible with the Japan Servo type S, the fault conditions upon stall (stop condition) are different.
The Silenx fans are basically computer case fans and send out square waves on rotation to the cpu for speed (rotation etc)
The Japan Servo Type S TUDC12Z4S are Lock detection type, and the output is open collector.
When fan is running the output is low all the time, when stalled the ouput goes high telling the YA pcb to close down the G90.
I have attached data sheet for the Japan Servo type.
I spun up the IXP-34-16 on the bench to see what was comming out of the sense line, and it was square wave pulses.
The Japan Servo make professional fans for all types of applications, Silenx have a bad name in the computer world as cheap -rap
I also did the mod to my G90, and the front Silenx was noisy (bearing noise) right from the start, as I said in another post there just was not enough air out of these fans and now that I have done some research on the front fan am glad that I pulled all the fans out and went back to the OEM ones.
So maybe someone can find a Lock detection type fan to replace the front one.
Ray you will need all the cooling you can get when you go to 1080p 72Hz.
Richard
http://rapidshare.com/files/153841028/TUDC_3.pdf

Richardc

10-14-08, 05:13 AM

CIR-Engineering

10-14-08, 10:13 AM

kawal

10-14-08, 01:04 PM

Richard,

Your comments were all very informative, and your testing seems to confirm why the SilenX fans will not always shut down the projector as some have noted.

When I had my G70, I did quite a bit of practical airflow testing as I compared the OEM fans to the SilenX fans. I was even working with SilenX Tech dept to see if they could build a quiet fan with more airflow. They were successful in a very limited way, increasing the speed (and airflow) from 1400 rpm to 1600 rpm for their 120mm fans. Once they sent some higher speed fans (2000+ rpm) for more airflow, the design was different from their Extrema design and basically similar in both design and noise to "normal" fans. So I gave up trying to find any more "quiet" fans.

Yes, the SilenX fans are "el cheapos", and they don't compare in either quality or airflow to the Japan Servo fans. And they outrageously claim the same airflow, which is ridiculous. But what this all comes down to is what sort of cooling capacity do these high quality projectors really need? The Sony projectors, as you know, are extremely high quality overall and it's no wonder they originally cost many thousands of dollars. The high quality OEM fans were consistent with the high quality overall design.

But still, do we really need this "overdesign" for our home theaters? I find the OEM fan noise to be pretty objectionable - first in the G70, and even more so in the G90 with its five fans. The only real way to know how much cooling we really need is a practical test in the real world. If the projector survives, the cooling is OK. I had the G70 running with the three SilenX fans for a couple of years with no problems noted, and it was much more enjoyable to watch. I've only had the G90 up for a few months, and I don't THINK I've had any problems with the SilenX fans. But I realize I've increased the risk.

So right now, I'm enjoying the quietness of four SilenX fans in the rear and two up front. JohnHWman also seems to have had good luck with the modification. To my knowledge, only Craig has had a problem and it seems to have been caused by a non-working fan. So everyone must consider the quietness vs. cooling issue and decide for himself what to do.

From what I have learned so far, I sort of doubt whether there is a better substitute for the OEM fan than the "el cheapo" SilenX, when also considering the noise reduction. I think other "high quality" fans will likely be just as noisy.

For now, I'll keep the SilenX fans in place, but I still found your comments and test results very interesting and informative. Thank you.

Ray

kawal

10-14-08, 04:36 PM

And I am pretty sure my PA board melt down was because the SinlenX fan was stopped and the G90 did not know it.

What I am planning to do for a fan mod is similar to what Spatz suggested. I think he makes a valid point that these projectors were designed to also be used as rear projectors and would be subject to 50 degree C environments that they will never see when on the ceiling in a home theater. On the ceiling, the G90 should need less cooling than when it is enclosed in a rear projection box.

I plan to go back to all stock fans, but I plan to put a resistor in line with every one of them to slow them down. This way I will be using the OEM fans, just at a slower speed. That way the quality and sense capability of the Japan Servos will be intact and I will also have a quieter projector.

I will not be using any SilenX or extra fans, just the Japan servos running at maybe 3,000 RPM instead of the 4,5000 (or about) in the stock config. I will do temp probing, but think it is probably ok even if the temp goes up a little.

craigr

Craig,

One of the questions you will need to answer for yourself is what speed you want to lower the fans to. It's a similar question to whether or not to replace the fans with a SilenX fan - i.e., what is a "safe" amount of cooling capacity?

What I found when testing the OEM front fan is that the original speed was around 4500 rpm. I decided (rather arbitrarily) to choose 3000 rpm as my "safe", but quieter speed. I also found that when you slow down the fan, there is a point at which the OEM fan won't start at all - so you need to choose a speed (and corresponding voltage) to be sure you are above that minimum point. Even at 3000 rpm, the OEM front fan was still noisier than I would like. The OEM fan, even running at 3000 rpm, has sort of a high pitched whine. The SilenX fan, of course running considerably slower, is much more of a "whoosh" sound, making it much more tolerable.

The rear fans on the G70 ran at about 2600 rpm, as I recall. I didn't bench test the G90 fans, but I could do that easily (since I have them out) if it would help you. I could provide a voltage vs speed curve so you could decide at what speed you want to lower them to. I would suspect that they would be similar to the G70 fans in performance. Since adding the resisitor will change the voltage to the fans, you may have to experiment with the resistor to end up with the speed you want. For my bench testing I simply vary the voltage to the fan with a power supply and use a very accurate optical tachometer to measure the speed.

It would be nice to know what is a really safe degree of cooling capacity, but unless we are willing to put up with the noise of the OEM fans running at full speed, it seems we have to compromise and hope our decision is the right one.

Ray

kawal

10-14-08, 04:48 PM

Hi again Ray
I have done some more testing tonight, if you stall the Silenx fans with your finger the output on the sense line can either be Hi or low depending where the blades stop.
So the Capacitor C11 220uF could be held discharged allowing the G90 to carry on (No fault) also if one fan stops and the sense line is low same deal.
The reason things appear to work is the speed of the fan pulses on the C11 Cap holding it low.
I would not be happy with this situation on my G90
Richard

Richard

Richard,

I understand your conclusion is that the speed sensing circuit is likely ineffective with a SilenX fan, of course adding some risk that the projector will not shut down in the event of a fan failure.

Since I plan to live with this risk and keep the SilenX fans - at least for now - I wouldn't mind speeding up the SilenX front fans a bit. When running two front fans, the voltage is about 10.2 volts instead of around 12 with the single OEM fan. The SilenX fans are so quiet that I could still live with a bit higher speed which would be achieved by increasing the voltage to around 12 volts.

As you noted, resistor R14 is a 33 ohm 3 watt resistor. Here's the schematic:

http://www.kawal.net/front%20fan%20schematic.jpg

I'm thinking if I replaced R14 with another 3 watt resistor around 18 ohms, I might achieve around 12 volts to the fans. But I don't know enough about electronics to know whether that might potentially cause other problems. I wouldn't think it would, but I would appreciate your thoughts - and perhaps opinions from others who might have a good electrical background.

Thanks,

Ray

Richardc

10-15-08, 04:53 AM

Hi Ray
2 X SilenX running in parallel, use 2 x 22ohm @ 1Watt in parallel will give you approx 12Volts at the fans. (or maybe 1 X 10 ohm @ 2 watts or 5 watts)
I bench tested this it gives about 12 Volts to the fans.
Also this is with the original SilenX fans, thats all I had on hand.
As for the Protect circuitry it won't hurt it going down to 10 ohms, the voltages will change on the semiconductors some what.
The worst that can happen it will shut down the G90.
But I don't believe this will happen.
Richard

Richardc

10-15-08, 05:44 AM

kawal

10-15-08, 11:39 AM

Ray
Also the other 4 rear fans are controlled by the PE pcb it is exactly the same setup as the front fan except these fans are speed controled also depending on the internal temp of the G90 if the G90 starts to raise in temp probably due to ambient increase then the four rear fans ramp up to cool the G90 internals down, they achieve this with a themo sense chip (LM35DZ) which controls a series regulator to the -B volts on one side of the four fans.
Richard

For the rear fans, I think I'm understanding from your description that the speed sensing circuit (protection circuit that may not work with the SilenX fans) would have no effect on this thermal sensing circuit. In other words, the fan speed is controlled by simply increasing the voltage to the fans - so it should continue to work OK?

Thanks,

Ray

Richardc

10-16-08, 04:18 AM

Yes
Regards
Richard

JohnHWman

10-16-08, 07:32 AM

Hi again Ray
I have done some more testing tonight, if you stall the Silenx fans with your finger the output on the sense line can either be Hi or low depending where the blades stop.
So the Capacitor C11 220uF could be held discharged allowing the G90 to carry on (No fault) also if one fan stops and the sense line is low same deal.
The reason things appear to work is the speed of the fan pulses on the C11 Cap holding it low.
I would not be happy with this situation on my G90
RichardGood schematic in-deep analysis Richard. I never had time to do this investigation myself and I apologize. This is clearly explaining how my own G90 unit stops when I tried to stall the front panel SilenX fan by hands : I've been lucky as C11 has the correct state at this time :cool: However, some had no luck (Craig) and the sense circuit failed to detect the fan stall with the known consequences...
I do think the same : this situation is not suitable as the G90 sense detection circuit is not designed to detect low (or no) speed of these 'PC-like' sense fans circuit...

I think I would recommand to keep the stock (noisy) 60mm Japan Servo front fan in place in the unit but slowed down by a resistor (with a Ecap in parallel to let it spin-up at full DC rail value) and add the SilenX front fan as explained in this DIY guide to compensate the loss of the Sony stock 60mm fan running at lower speed. Then, if the stock fan fail, it will be detected by the unit and if the SilenX fail, having the stock still running is safe.

I'm sorry: I did this mod. a little bit too fast, without enough design validation and I apologize for any inconvenience :o

John

CIR-Engineering

10-16-08, 10:13 AM

However, some had no luck (Craig) and the sense circuit failed to detect the fan stall with the known consequences...

I'm sorry: I did this mod. a little bit too fast, without enough design validation and I apologize for any inconvenience :o

John
LOL :D

I don't think it's your fault John. No one has to do the mod. If one does decide to apply your mod, one should be prepared for the consequences of known less cooling. Really, I should have been more careful and should have checked my fan. I should have also checked temperatures with a probe before and after the fan mod. I was too lazy and I paid with a melted PA board.

At least I can do component repair and my projector was back up and running in 7 days... so it could have been more unfortunate if it had happened to someone else. I wonder what Sony would have charged to fix my PA if they even still have the one guy who knows the G90's?

craigr

CIR-Engineering

10-17-08, 03:53 PM

So I am going to try and slow down the stock fan without adding a front fan at all. I placed a temp probe at the back of the heat sink on the PA board yesterday while watching a movie. The steady state temperature was 50.5 Celsius or about 120 Fahrenheit. I will slow down the front fan next and see what happens... I'll keep you posted.

craigr

CIR-Engineering

10-18-08, 11:22 AM

Well, with a 33 ohm resistor the steady state temp is 52 C / 126 F. That seems fine so I might slow it down a little more. The fan is quieter, but still kind of annoying, so I might have to rethink this approach all together.

I think I am going to take a temp reading with John's original SilenX fan mod... the one with the small SilenX fan in the stock location. I am curious to see how hot that set up got. I might take a temp reading with the front mounted SilenX mod as well.

I think this will help give me a better idea of what is actually safe.

craigr

cinema mad

10-18-08, 12:22 PM

Your A braver man than me Craigr :D ......

Cheers

CIR-Engineering

10-18-08, 01:04 PM

Your A braver man than me Craigr :D ......

Cheers

It's only 1.5 degrees C increase :p

Seriously, I think Spatz has a great point about the rear projection cabinet installations where the air on the intake is at around 50 degrees C... and that's before it passes over the heat sink. I bet these can run at 60 degrees C or more without problems from my measurement location.

But I must admit, it does make me nervous :eek:

craigr

kawal

10-18-08, 01:39 PM

Craig,

The info I posted on this site might help a bit:

http://www.kawal.net/G90%20front%20fans.htm

Take a look at the speed vs voltage curves and you will see what voltage you'll need for slowing it down. Without a tachometer, you really won't know how much you are slowing the OEM fan down. You'll notice that around 3000 rpm requires about 9 volts to the fan. But be careful that you don't slow it down much more than that or it will not start reliably.

I still found that even at 3000 rpm, the fan noise was annoying. That's why I ended up with two SilenX fans.

I may even increase the speed (voltage) on them a bit more just to be safe, because even at full speed the SilenX fans are pretty darn quiet.

I am interested to hear what you find with your temperature measurements, since I did not do any.

Ray

CIR-Engineering

10-19-08, 10:11 AM

CIR-Engineering

10-19-08, 10:45 AM

Good schematic in-deep analysis Richard. I never had time to do this investigation myself and I apologize. This is clearly explaining how my own G90 unit stops when I tried to stall the front panel SilenX fan by hands : I've been lucky as C11 has the correct state at this time :cool: However, some had no luck (Craig) and the sense circuit failed to detect the fan stall with the known consequences...
I do think the same : this situation is not suitable as the G90 sense detection circuit is not designed to detect low (or no) speed of these 'PC-like' sense fans circuit...

I think I would recommand to keep the stock (noisy) 60mm Japan Servo front fan in place in the unit but slowed down by a resistor (with a Ecap in parallel to let it spin-up at full DC rail value) and add the SilenX front fan as explained in this DIY guide to compensate the loss of the Sony stock 60mm fan running at lower speed. Then, if the stock fan fail, it will be detected by the unit and if the SilenX fail, having the stock still running is safe.

I'm sorry: I did this mod. a little bit too fast, without enough design validation and I apologize for any inconvenience :o

John

Hey John,

I haven't had all my coffee yet and am having trouble visualizing where a starting cap would go on a DC motor.... Would an electrolytic cap in parallel with the resistor for voltage control work??? That just doesn't seem correct at all.

craigr

kawal

10-19-08, 12:48 PM

Thanks for posting that info Ray. I actually went there last week and downloaded the page to my PC for further reference :)

I am going to try the two front SilenX mods that John made and see what the temps are. I think this will give a good idea of what is a safe temperature for these to operate at.

I will then slow down the stock fan more and see where I land. Also, Johns idea of a cap to start the fan at full voltage when using a lower voltage is a good one and will probably allow the fan to operate reliably at a lower voltage as well.

I'll update soon.

craigr

Craig,

I am currently running both the 60mm and the 80mm SilenX fans on the front at full voltage - i.e. no resistors. But the operating voltage is only about 10.2 volts, so the 60 mm fan is only running at about 1500 rpm instead of its "nominal" 2000 rpm. Actually, if you look at the speed vs voltage graph, even at 12 volts, that fan only spins at around 1800 volts.

http://www.kawal.net/front%20fan%20specs%203.jpg

My thinking is that John's original mod used only the 80mm fan and he is having good luck with that single fan. But for added assurance, the 60mm fan in addition should provide additional cooling (although they both run slightly slower than a single fan).

Speeding up both SilenX fans will help the cooling, and this would require replacing R14 (33 ohms) with a lower resistance. It would be very interesting to take the heat measurements that you are doing with two SilenX fans, running at 10.2 volts and something higher.

You slowed down your OEM fan with an additional 33 ohm resistor, but have you been able to check your operating voltage yet? From the graph, that should tell you about how much you've slowed it down. When I was experimenting with the OEM fan and the 80mm SilenX fan, here is what I found:

With a 33 ohm resistor in line with the OEM fan, the voltage with BOTH fans running was only 9.6 volts at the SilenX fan and only 7.2 volts at the OEM fan - the OEM fan would not start under these conditions. I disconnected the SilenX fan and the voltage increased to 12.2 volts (ahead of the resistor) and 9.1 volts at the OEM fan. 9.1 volts should give you around 3500 rpm on the OEM fan. Since I think this is what you also did (adding a 33 ohm resistor to the OEM fan), you should have the approximate same voltage and speed. But you also feel the noise is still annoying - the same as I did.

I eventually slowed down the OEM fan to around 3000 rpm (with the SilenX fan), but it was still too noisy for me. But I couldn't slow down the fan much more, or I would risk not starting it.

At least if you try to slow down the OEM fan a bit more, hopefully the OEM fan will still have a working detection circuit that will shut down the projector. But interestingly, when I first tried both the OEM fan with a 33 ohm resistor and the SilenX fan, the OEM fan did not start and yet the projector did NOT shut down! The sense wire was connected to only the OEM fan, so it SHOULD have shut the projector down, but it did not. So my only conclusion would be (and I'm not an electrical engineer) that lowering the voltage to 9.6 volts (ahead of the resistor) may have prevented the detection circuitry from working properly.

http://www.kawal.net/front%20fan%20schematic.jpg

In any case, my concern is not only the lack of a working detection circuit, but also the amount of cooling that I now have. I THINK the cooling is OK, but if you run any cooling tests with the SilenX fans, I will be very interested in your results.

At least I'm enjoying a relatively quiet projector. :)

Ray

CIR-Engineering

10-19-08, 04:55 PM

kawal

10-19-08, 07:32 PM

Well I have a little more temp data.

With the SilenX 60x60 fan in the oem location, just running that fan resulted in 138 F / 59 C. This is 18 and 9 C higher than the stock fan. John said he ran his projector in this config for quite some time so it is probably safe. However, I find it disconcerting.

I added a 47 ohm resistor in place of my original 33 ohm trial and put the Japan servo back in. After about one hour the temp was 132 F / 55.5 C. At this fan speed, the Japan servo is quieter, but still annoying.

I think I will try taking a temperature reading with Johns front fan mod only and see what that yields.

My new idea is to actually cut a hole through the bottom (top on the ceiling) of the projector for an 80x80 mm fan. I might take one of the eom Japan servos and use it on top of the projector and NOT use a fan in the stock position, or in front of the PA. I think I will cover the opening between the PA board and the chassis at the front of the PA with gaffers tape. This I think will force tons of air over the heat sink on the PA board as the air gets sucked over the PA towards the back of the PA. I find it likely that this will provide better cooling than stock. I will then probably slow down the large fan on top until I like the sound, or until I start seeing temperature increases.

I haven't been too worried about the fan supply voltage as I have been more concerned about the actual temperature I observe at the heat sink.

craigr

Craig

Interesting idea about the fan in the bottom of the projector.

Regarding the voltage to the OEM fan, I found that when it was "too low", the fan might run but it wouldn't consistently start. This would cause me to worry, especially if the shut down circuit isn't reliable. I hope your 47 ohm resistor isn't lowering the voltage to around this critical area.

I found that by using the original adapter cable off of the PA board, it was pretty easy to get to it to take voltage measurements.

When I did John's mod (80mm fan on the outside), I decided to tape up all of the chassis holes on the front. My thinking was that allows all of the air through the front fan to be forced through the heat sink area back to the area where the OEM fan is installed. Without doing this, some of the air through the fan could come right back out the chassis holes and never get to back of the heat sink.

If you were only getting a 20 degree increase with the 60mm SilenX fan by itself, I would think that adding the second 80mm Silenx fan should provide adequate cooling.

Ray

JohnHWman

10-20-08, 07:50 AM

I haven't had all my coffee yet and am having trouble visualizing where a starting cap would go on a DC motor.... Would an electrolytic cap in parallel with the resistor for voltage control work??? That just doesn't seem correct at all. It is ;) The idea is to let the Ecap shorting the resistor at start so the fan will see higher DC voltage value to get a safe start. Then, once the fan started, the Ecap will have its DC voltage increasing (to the Delta_V value) because of the current drop accross the resistor. Then, the DC voltage applied to the fan will be almost the (PS - Delta_V).
The fan will start at full speed (almost no serial resistor since shorted by the cap.), then it will be slowed down by the serial resistor once the capacitor will be charged.

Just an idea...

john

CIR-Engineering

10-20-08, 11:05 AM

It is ;) The idea is to let the Ecap shorting the resistor at start so the fan will see higher DC voltage value to get a safe start. Then, once the fan started, the Ecap will have its DC voltage increasing (to the Delta_V value) because of the current drop accross the resistor. Then, the DC voltage applied to the fan will be almost the (PS - Delta_V).
The fan will start at full speed (almost no serial resistor since shorted by the cap.), then it will be slowed down by the serial resistor once the capacitor will be charged.

Just an idea...

john

hmmm...

But in any event, I don't think I would want the fan to be spinning that low RPM anyway. Based on my temperature readings, I think my idea will not provide good enough cooling with the stock fan slowed down to a tolerable sound level.

craigr

CIR-Engineering

10-20-08, 11:12 AM

BTW, yesterday morning I took all the C and E boards out of the back of the G90, all four back fans, and some of the frame components... all to get to that frame component at the back with the square mesh that blocks the rear fans blowing on the E boards.

Once i got the frame part out, I cut and filed all the mesh out of the frame part so that the air from the four back fans can now flow without restriction over all the boards at the back of the G90. I can feel much higher airflow out the back of the G90 now. Also, the sound at the back seating positions behind the G90 is much quieter.

I have to drop the G90 from the ceiling soon to move the projector over 1.5" anyway. I will probably try my front fan mod at that time. I probably won't get to it until after my western states calibration tour ends on Nov 7th though.

craigr

CIR-Engineering

10-20-08, 11:14 AM

Craig

Interesting idea about the fan in the bottom of the projector.

Regarding the voltage to the OEM fan, I found that when it was "too low", the fan might run but it wouldn't consistently start. This would cause me to worry, especially if the shut down circuit isn't reliable. I hope your 47 ohm resistor isn't lowering the voltage to around this critical area.

I found that by using the original adapter cable off of the PA board, it was pretty easy to get to it to take voltage measurements.

When I did John's mod (80mm fan on the outside), I decided to tape up all of the chassis holes on the front. My thinking was that allows all of the air through the front fan to be forced through the heat sink area back to the area where the OEM fan is installed. Without doing this, some of the air through the fan could come right back out the chassis holes and never get to back of the heat sink.

If you were only getting a 20 degree increase with the 60mm SilenX fan by itself, I would think that adding the second 80mm Silenx fan should provide adequate cooling.

Ray

I figured that I would test voltages after I had determined that I could live with a given fan speed / temperature increase. This is just not going to work because the fan is still too noisy even with too much increase in temperature.

So I will try moving the fan to the top of the projector next.

Also, I will probably tape up the back of the PA chassis case so that air can not be drawn in through the stock exhaust hole used by the 60x60 mm oem fan. I don't want to suck hot chassis air into the PA compartment.

craigr

kawal

11-01-08, 04:19 PM

Craig,

Did you ever get a chance to try the two SilenX fan approach - one 60mm fan in the OEM location and one 80mm fan in front? That's what I'm using and the noise is quite acceptable. But I have not taken temperature measurements.

Ray

CIR-Engineering

02-19-09, 11:16 AM

I finally got around to doing something with the fans yesterday. I wrote it up on my web page.

http://www.cir-engineering.com/g90fanmod.php

craigr

Chuchuf

02-19-09, 12:42 PM

Good idea Craig.

Terry

WTS

02-19-09, 02:17 PM

About a year ago I did the same mod to my Zenith 1200x, I removed the small front fan and installed a larger one and slowed its speed down alittle. Result, more air and alot less noise.

kawal

02-19-09, 03:39 PM

That's nice write-up, Craig. And it sounds like pretty effective cooling as well.

As I've posted before, I changed out all four of my rear fans to SilenX fans and added a 60mm and an 80mm SilenX to replace the OEM 60mm front fan. The fan noise is quite acceptable, and I'm hoping the cooling is adequate - so far no problems. But I also share your concern about lack of automatic shutdown capability of the SilenX fans, due to their design. So I realize I'm trading additional risk for my relatively silent projector.

It sounds like you are using the OEM rear fans as well as one OEM (Japan Servo) rear fan (90mm) for the top fan. Since the four rear fans were also pretty noisy, are you running all five fans at full speed? I know that little 60mm OEM front fan was a screamer, but I assume you feel the new noise factor with five large fans is acceptable?

Your mod should make your G90 pretty safe from a heat related failure.

Thanks for sharing,

Ray