New here, if this is the wrong place, please do move it to a better candiate location, Sorry.

The intro (and my situation) can be found in the 'I need an AV switch; Any recommendations?' thread near the end of it.



Symptom: Joytech 240c a/v switcher introduces problems on the 'Y' signal of component video. (Some problems are horizontal lines, other include flicker and whatnot)

Problem: The IC used to switch the signals was not made for this purpose. They are using a HEF4051 8x1 mux. This 30 cent IC is not capable of properly handling the sync signal in the 'Y' of component video. Theres 6 of them in the 240c (one for each component signal, and one for the composite video, and 2 for the analog audio).

Solution: Replace the IC used for switching 'Y' with a better, proper, ic.

The candidate IC is MAX4315 8x1 video mux/amp.


The fix...(until i have time to create a proper one with full pictures and whatnot)

Replace the HEF4051 with the MAX4315. But the 2 chips are not pin compatible, so this means you will need to solder several wires to very small pins on the MAX chip (not for the faint of heart).

Grab the datasheet for the MAX4315 chip, and for the HEF4051

You will want to remove the HEF4051 chip from the pcb, and a few other components (see the red circle in the pic to see the components removed)
Dont forget to also remove the 4.7uf capacitor on the underside of the board.

You will now wire up the MAX chip. Use the datasheets to match the pins.

Take note of this:
- 'VDD' on the HEF is NOT connected to VCC on the MAX, you need to take 5v from
the proper place (see the DSC03734 image - 1st one) and connect that to VCC on the MAX chip.
- The "Y" inputs on the HEF chip map to the "IN" inputs on the MAX (ie. Y0 -> IN0)
- VSS on the HEF is GND. The pad on the pcb corresponding to this connects to GND on the MAC chip.
- SHDN on the MAX is connected to 5V (VCC) via a 10k resistor (resistor is not super important, but it prevents the MAX from getting damaged by accidental mis connection)
- You will also want to connect a 0.1uf capacitor (ceramic is fine, non polarized) between VCC and GND on the max chip (as close to the max as possible). Again this is not required, but if the voltage regulator is not stable, this will make sure the MAX is running properly.
- You will need a 75ohm resistor connected between 'OUT' pin on the MAX and to the location indicated by the blue arrow on the picture.

On the second picture you will see a brown board and parts below the MAX chip, IGNORE it. Its there for some tests that I did at first, and are not required for the fix.

One more thing, try and make the wires as short as possible. This prevents interference.


Hopefully this is clear enough for the technically inclined to try out.
Im busy this week, and probably wont be able to draw up more formal information, next week Ill get on it (if people want it)
(Maybe ill make a small pcb that will make life easier)


Im running it for about 5 days now. Working beautifully.

This fix also works for any other switcher based on the 4051 ic, such as some of the low end 'XBOX 360' switchers.

Enjoy.

PS. If any more experienced AV designers wish to add to this mod please do so. Im by no means an expert in this field, I simply saw a need and found out how to fix it.

Useful links:
This is the circuit they based the 240c's video/audio switching on:
http://jap.hu/electronic/avswitch.html

Grab the datasheet for the MAX4315 chip,
http://datasheets.maxim-ic.com/en/ds/MAX4310-MAX4315.pdf

And for the HEF4051,
http://www.standardics.nxp.com/products/hef/pdf/hef4051b.pdf

A quick way to determine if the problem is inded the Y signal do this:

Connect your source to any of the inputs, connect the output to the TV. (As you normally would).

A problem should be visible.

Now, disconnect ONLY the 'Y' signal from the input and disconnect the output 'Y' cable as well, then using a RCA joiner connect the two directly.
(That is, Y is connected directly from source to TV, and Pb/Pr are connected thru the switch)

If the problem has now dissapeared, then replacing the IC used to switch 'Y' will fix your problem.

How has your Joytech been? Still working OK?
I'm having a problem locating the MAX4315 IC in Australia (any one help?)
I have emailed a few electronic shops with no luck so far.

Still running great.

The MAX4315 is a 'big' chip. Its retail is $4 in 1k lots.
Digikey has it for $8.
Its not something you can find at the local shop, but can usually be ordered.

Contact your local Maxim distributer, see if they can arrange for a free sample. Or try the website and see if you can order a sample there.


The alternative, use a MAX4314. This version is 4x1 (4 inputs).
If you only have 4 things that will be connected by component video, then this will suffice.
(It is also possible to connect 2 Max4314's, but it requires some extra logic to handle the selection bit A3)

I guess I am just looking for clarification on this before I go ahead and do this, I have the following pin-to-pin configuration from the datasheets, can you verify this is indeed correct?


4315 HEF
1 9 (A2)
2 10 (A1)
3 11 (A0)
4 +5V (Vcc) [Sourced from board]
5 13 (In0)
6 14 (In1)
7 15 (In2)
8 12 (In3)
9 1 (In4)
10 5 (In5)
11 2 (In6)
12 4 (In7)
13 7 (Vee)
14 to +5V via 10k resistor [See pin 4]
15 8 (VSS)
16 3 (Output)

I guess I am just looking for clarification on this before I go ahead and do this, I have the following pin-to-pin configuration from the datasheets, can you verify this is indeed correct?


4315 HEF
1 9 (A2)
2 10 (A1)
3 11 (A0)
4 +5V (Vcc) [Sourced from board]
5 13 (In0)
6 14 (In1)
7 15 (In2)
8 12 (In3)
9 1 (In4)
10 5 (In5)
11 2 (In6)
12 4 (In7)
13 7 (Vee)
14 to +5V via 10k resistor [See pin 4]
15 8 (Vdd)
16 3 (Output)



Almost, the following pin is correct, but the designation is wrong:
15 8 (Vdd)

Correct is :
15 8 (VSS)


Everything else is correct.

hah, You are correct.. I am pretty tired tonight, and I tend to mess up my Vcc/Vee/Vdd/Vss designations even when I am not tired! :) Thank you for double checking me.

WOW! I just pulled apart my generic component switcher this past weekend out of frustration, to see what was inside. If my memory is correct it also had three of the HEF4051 chips inside. :D

Thanks muchly for sharing the info - sounds like a project i'll be looking into!

Thanks to a fellow member I saw the insides of a 245c.
The 245C is similar to the 540 (360 switcher) - which has the same problems as the 240c.

The 245 is not fixable (as easily) using the method I outlined above.
The 245 would require removal of several ic's and much more complex wiring.

Yea, it turns out that my "240c" is not a 240 after all, it is marked 965 on the front, but seems to be a 245, from the PCB descriptions.

It uses cascaded Quad 2x1 video switches, and I am currently working on a solution for it.

The 245/540 can in theory be fixed. The problem is you will need to run wires from all the Green component lines back to the MAX chip, and then the output runnedto the output signal. Also will need to remove all the 2x1 switch ic's and corresponding circuit (essentially isolate the Y signals)

Looking into fixing this problem on my 245 and came to this thread from the one that spawned it.

NicolasG, you said there that:
"They did use better ic's, the TE330, but they did not use a proper amp on the output, thus limiting the bandwith."

Would that mean that the 245/540 problems could be fixed by the addition of an amp to the Y output? I'm far from an expert, and I may well have misinterpreted what you meant (in which case sorry). It would be relatively easy to add an amplifier circuit outside the box between the Y-output socket and the Y-plug on the cable, would that be sufficient?

That could even be what the fabled "adapter" that Joytech have told some they'll offer is, and not the plain-old resister that everyone seems to think it will be.

If the dongle was an amp, it would require external power.

the problem with the TS5V330 is that it is a passive switch with 10 ohm resistance. When you cascade enough of them, that 10ohms adds up. Also, the effective bandwidth is reduced every time you cascade a switch, so with that many, it ends up being no better then the HEF.

I did think of that - a dongle could easily get power from the DC supply with a pass-thru type plug into the external power jack. But that's all conjecture anyway.

The cascading is the bit I don't quite understand. From the photo that was submitted and the TS5V330 datasheet: the 330 is a 4PDT switch and the photo shows 4 of them, cascading them to provide a 6 way switch would require 5.

The 330 has an enable input (pin 15) so you could switch 8 ways with only the four switches in a parallel layout. Since an individual switch can select left/right/neither you can select between 8 signals as long as you only enable one switch at a time.

That being said the circuit design skills here (or testing for that matter) really don't seem to be up to much. I really need to open mine up and have a poke around since the photo isn't quite clear enough to see exactly where even the topmost the traces go.

The output related problem with the TE330 is based on a post by Videostorm (who oviously knows much more about this than I do)
"TS5V330 is a 10 ohm passive switch. They would need to follow it with a high bandwidth amp in the box if they use that device. "
http://www.avsforum.com/avs-vb/showthread.php?p=9638698&&#post9638698

I dont have the 245/540 in hand, so I have not traced it out. But from the pictures I received of them, there is 6 of them. On the 540, 3 are on the top board and 3 are on the bottom board. I would venture and guess the 245 is the same, 6 chips.

Do note I have not delved into how the optical is being switched, given their design choices, they very well could be using the 330's for that too.

They are also using a hc04 logic inverter, which I assume is to aid in the control logic for the switching, and they are using two serial-parallel ic's (hc595).

Find attached 2 pics of the 540, displaying the ic's and locations. (let me know if you want hi-res versions)

Im very happy that this is gaining momentum. Sure, were fixing a product that should work in the first place, but given the price (versus other switchers) the fix is very reasonable.

I'll try to get some decent photos of it this weekend, the 245 is very stackable and since it produces no heat so it's at the bottom of the stack so extricating it will take some time. :)

One other thing to consider - by my reckoning it has to switch 6 video channels (Y/Pb/Pr, Y/C and composite) so I don't know if the TS5V330s are only used for component. I think I may need to do a bit more maths on the number of chips vs channels though.

It's annoying that it isn't totally up to spec, but as you said it is cheap especially for the functionality you get.

BTW on over at the Trusted Reviews site in their gaming section (can't link to it because I'm under the post limit) the review of the 540C states:
Update 31/01/2007: Joytech has been in contact informing us that they’re working on a discrete adapter to solve the Nintendo Wii output issues, as mentioned in the review. These will be given out free to anyone who needs one, and the changes will be incorporated into future manufacturing runs of the Control Center 540C.

Hi,

I also have a 245c with the same issues...

I recently opened it up and found that the audio L/R still use the HEF4051.

Would it be crazy to rewire one of these and use for the Y signal?

The HEF4051 is bad for Y switching. So feeding the Y signal thru these instead of the TE330 wont be of much use.

Give it a try though, connect Y to one of the audio and do the same on the ouput.
It most likely wont work, specially since the audio connections dont have the proper circuit to handle video, but its worth a shot.

sorry for my lack of explanation...

i meant, replace one the L/R HEF4051 chips with a MAX4315 then run the Y signal through.

Actually having read up on the datasheets I think I have a solution that will not even require opening the case.
The TS5V330 can pass a signal in either direction, so you send a signal (possibly just a simple DC voltage) into the Y "output" of the switch, it'll show up on the currently selected Y "input", do a bit of logic processing and you have the binary number representing the currently selected input. This can then be used to switch a MAX4315.

Granted that would require a lot of wiring but no desoldering or even opening the case - might not even invalidate your warranty ;)

The logic should be fairly simple, the circuit I've got drawn up uses a few 4-input ORs but I think there's probably a simpler way.

You could even add opto-isolators to the circuit and use it to control things based on which input is selected.

Wow... that is pretty sad that a user of the product (that was out for a long time) came up with a solution and the people that make the things couldn't! Good work anyway. :)

Sorrt for the bump, I just thought I'd give some kudo's on the fix! I'm not good enough with the solder iron to try it myself though. :(

boon_army_cadet are you saying to use the systems input/outputs to determine which input is selected so that you can change the input on a secondary switch ?

Yes thats very possible, but in that case, i think its easier to have the secondary device receive IR signals, same ones as the main device, and switch based on that.

Which one should we get, MAX4315ESE or MAX4315EEE
and in QSOP or SOIC or uMAX(?)

thanks.

soic is the biggest pin pitch possible. so go with that.

as for ESE or EEE, doesnt matter.

thanks. So with this fix, does this mean the 240c will be one of the best HD switchers? I have heard it had issues, however I am not sure if this fix mean that it will have no issues.

BTW - Nicolas if you have the time, please draw up the formal instructions and parts list (possible where to obtain it, I know the US stores may be different in Canada, but it helps if you can link to a store in either country). Thank much.

Ive decided to leave this as is. All the technical details are there, just a matter of following them. For parts, radio shcak will have the simple stuff, and Maxim can send you free samples.

It does fix the problems with the video. Which is the majority of problems Ive seen on it.
I love it.

How many MAXIMs are required? From the instruction it appears to be at least 6. Am I reading this correctly?

Sorry for making the request, I am not very good at soldering stuff unless the instructions are detailed. The last time I did, I place a capacitor in the opposite polarity.


Thanks.

Ok I hate to ask but I've been trying to figure this out for months now. I Followed the directions outlined in this thread and for some reason after the modification only some equipment works. Now I've checked everything, and even redone the mod from scratch twice with new max chips. my xboxs, xbox1(multiple) and xbox 360 work through the modified switch perfectly as well as 2 separate DVD Players. My cablebox, Tivo(series3) and WII on the other hand do not, the pictures rolls and distorts, and basically looks scrambled. Does anyone have any ideas on what my be the cause?

How many MAXIMs are required? From the instruction it appears to be at least 6. Am I reading this correctly?

Sorry for making the request, I am not very good at soldering stuff unless the instructions are detailed. The last time I did, I place a capacitor in the opposite polarity.


Thanks.

No only 1 is needed. You are only replacing the 'Y' mux.

Ok I hate to ask but I've been trying to figure this out for months now. I Followed the directions outlined in this thread and for some reason after the modification only some equipment works. Now I've checked everything, and even redone the mod from scratch twice with new max chips. my xboxs, xbox1(multiple) and xbox 360 work through the modified switch perfectly as well as 2 separate DVD Players. My cablebox, Tivo(series3) and WII on the other hand do not, the pictures rolls and distorts, and basically looks scrambled. Does anyone have any ideas on what my be the cause?

Did you use a 4315 or a 4314 ?

Remmeber that the unit has 7 inputs, if you dont use the 8x1 maxim then you wont be able to use all of them. Try connecting the cablbox/tivo/wii to the xbox/360 connection and see if it works ?
Im guessing it will.

I'm using the 4315, I've tested the devices on all of the inputs, its not input specific. I redid some testing tonight since its been a while since I messed with it last, and it seems it has a problem with progressive signals. I discovered I'd left the xbox in 480i so I switched it over to 720P and it stopped working.

Very strange that it would cause said problems. Ive only tested it with a xbox, 360, dvd players, sattelite, from 480 to 1080.

Could be due to the sync not going properly, the circuit is very rough and is not adjusted for the horrid circuitry of the 240c, but in my case worked out well.
Playing with the VEE signal and proper grounding, and removing of the extra resistors on the board may help but its beyond me at this time.

Before you did the change, did you test those devices? Did you have the problem described.

Once thing you can do is use the max chip to control 4 inputs, and the original 4053 for the other 4 then you can have the best of both worlds.

Unfortunitly the Devices that aren't working with the mod are the ones that flipped out the most using the 4051, and yes all the devices worked before the mod, and are now working (with Y issues) on a 245C. What might you suggest in terms of playing with the VEE signal?

I have to make a decision now whether to get my money back for my Joytech AV Control Center 245C or try to get it fixed with the info given here (though I'm not sure if the info is complete for the 245C).
I live in the Netherlands and I have no alternative for this switcher.
What would you guys advise me to do? I don't think I have the knowledge and skills to open it up and apply some fixes just like that.

I also made some videos of the problems I have:
http://ntsc-uk.domino.org/showthread.php?t=68941

I'm kinda looking for someone around who could fix it.
Some info collected from here:
Thanks to a fellow member I saw the insides of a 245c.
The 245C is similar to the 540 (360 switcher) - which has the same problems as the 240c.

The 245 is not fixable (as easily) using the method I outlined above.
The 245 would require removal of several ic's and much more complex wiring.

The 245/540 can in theory be fixed. The problem is you will need to run wires from all the Green component lines back to the MAX chip, and then the output runnedto the output signal. Also will need to remove all the 2x1 switch ic's and corresponding circuit (essentially isolate the Y signals)

Looking into fixing this problem on my 245 and came to this thread from the one that spawned it.

NicolasG, you said there that:
"They did use better ic's, the TE330, but they did not use a proper amp on the output, thus limiting the bandwith."

Would that mean that the 245/540 problems could be fixed by the addition of an amp to the Y output? I'm far from an expert, and I may well have misinterpreted what you meant (in which case sorry). It would be relatively easy to add an amplifier circuit outside the box between the Y-output socket and the Y-plug on the cable, would that be sufficient?

That could even be what the fabled "adapter" that Joytech have told some they'll offer is, and not the plain-old resister that everyone seems to think it will be.


Hey Guys,

OK this is the response from Joytech in regards to the 540C

"The issue sounds like he is having the same problem with his DVD playback as some users are having with the Wii, that is that at times the "Y" channel signal is quite strong. We are currently working on an adapter that will add a resistor to this channel's input and will be available to give out to any customers that are having this problem. I know this does not solve the immediate issue, but as soon as we are near completion on the adapter I will let you know and you can then inform any customer's that require this adapter."

They have offered to swap the 540C for a 245C what do you guys think I should do?

Regards
Joe

Thanks for the picture. The 245C is NOT the same as the 240c.
I can quickly see they knew the problems of the 240 and thus resolved to fix them, but they failed.
The circuit used in the 245 is similar to that of the 540 (360 switcher) which again has the same problems. My guess would be they rushed the 245/540 at about the same time.

The problem with the 245 is the same as stated with the 540. They did use better ic's, the TE330, but they did not use a proper amp on the output, thus limitting the bandwith. (You can even see the location between the HEF4051's for the use of TE330's instead of the HEF4051's. They where expecting problems and had placed the pads for the different chips just in case - but the sound doesnt have problems so the chips didnt need switching)

The 245 wont be fixable in the way the 240 is, as the TE330's are daisy chained. (It may be fixable, but would be a much more complex fix)

A quick snaps of the reverse of the board would be handy.
And of the board on the bottom (the optical one).



Alternatively, I'm looking at the upcoming Joytech AV Control Center 640HD PS3. Maybe this one won't have the component problems anymore...

I know it's something of a thread-jacking, but it is a relevant question given the original parent thread -

IMHO, the Psyclone PSC01 *almost* has it all: it looks beautiful in a cabinet or in a stack, programmable LED display, discrete input selection via remote, and reasonably functional with 4 inputs. Affordable too, at around $100 (Best Buy).

Problem is, the damn thing cuts out when the picture/signal gets "too" bright/white.

I have read the threads; with little variation many people have experienced this problem. I think I speak for Psyclone owners everywhere when I say I would happily pay double the retail price for this same unit if it would just work reliably (reliably display a bright white picture).

The fact that Psyclone continues to produce generation after generation of PSC01 (I have bought, tested and returned them all) with the same guts and the same deficiency has made clear their intentions: reduce their costs and ignore the device's shortcomings. I guess from a business perspective I cannot blame them.

Nicolas (and others), you have mentioned a few times (in this thread and in the parent) that you're attracted to the prospect of building a switcher that hits the "sweet spot" that we AVS consumers feel has been missed by all manufacturers.

My proposition to you (and to whomever): how about we analyze and "fix" the PSC01? Am I wrong in thinking that the materials necessary to resolve its problem can't be more than sub $10? I am here to tell you, I'd gladly fork over $50 for a user friendly "kit" that made my PSC01 work.

I just thought I'd float this idea at you. If you care to pursue it, I am eager to assist. I ask for nothing more than a PSC01 that works :) You get 100% of the margin :) If it is insane, then say so. At least then we'll know that a year's worth of suffering a beautiful looking yet broken component switcher will drive a man to insane flights of fancy.

Got word from Joytech that the AVCC 640HD might be cancelled :(