or Connect
AVS › AVS Forum › Display Devices › CRT Projectors › VDC 8" HighRes Projector
New Posts  All Forums:Forum Nav:

VDC 8" HighRes Projector - Page 2

post #31 of 101
First time I've seen the schematic on the neck board. I can see a lot of work went into this. It's a real work of finely tuned electronics--great job!


Simply because C44 and R33 are in the circuit, the VIM being used is critical to the proper performance of these neck boards. I'm sure that's what tse was referring to with "compatibility"

That neck board is very well tuned, and its tuning is relevant to what's being used on the input of the card.


I can also see where the H1100 is also very critical to this design. Whatever it's original performance, it has been enhanced in this circuit. And that would mean, you can't put another IC in it's place. So the H1100 is perfect there, unlike most other circuits where you can swap in a different IC, this is not the case here. Anything other than a H110 could cause any kind of performance problems.


That is a very well tuned High Frequency Video Power Amplifier circuit...
post #32 of 101
There wasn't a real good reason for using the HFA1100 other than it works good enough and there was already a VDC part number and spec sheet for it.

The 81771 cards won't work with the originals because when spotkill is enabled the transistors on the original cards will not allow the voltage to go higher than about +0.7 or 0.8V. That will not go high enough to kick in spotkill on the 81771 cards.

Spotkill has been a problem with different configurations because of the small difference between enabled and not. The 81771 takes more than +1V to kick in spotkill which allows more noise and crap without false enabling.

The 02-270340-01 card has the same thresholds as original Electrohome cards so spotkill works.

Scott
post #33 of 101
I thought it might be one of the Protection circuits.
So wold it be like the issue i had on my one Vim we talked about before, where one of the circuits can act as a radio receiver and the high frequencies can cause the spotkill circuit to come on and off depending on the frequency content of the Video.

Thats why I wanted the service manuals theory of operation for that VNb.

Athanasios
post #34 of 101
Oh I'm sure you could sub in a different opamp for the 1100 and make it work without any problems. There's nothing special about the 1100 compared to other opamps in the same class. Also I'm sure (without looking) that there other opamps with the same or better DC performance specs. As for AC performance we already know there are better opamps. Like tse already stated these are not designed or made with the HT room in mind. As far as C44/R33 goes that's just to do a little BW peaking to probably compensate for alittle droop down the chain.
post #35 of 101
Quote:
Originally Posted by WTS View Post

Oh I'm sure you could sub in a different opamp for the 1100 and make it work without any problems. There's nothing special about the 1100 compared to other opamps in the same class. Also I'm sure (without looking) that there other opamps with the same or better DC performance specs. As for AC performance we already know there are better opamps. Like tse already stated these are not designed or made with the HT room in mind. As far as C44/R33 goes that's just to do a little BW peaking to probably compensate for alittle droop down the chain.


I agree, and we (you and I both) know there's much better chips. But as indicated, this is not a design with HT in mind. So for commercial applications, the H1100 is fine.

C44/R33 in that circuit is a lot different than general peaking in other circuits. That final stage is much like a big opamp that was tuned to do two things at once (best bandwidth/lowest noise). And doing this in the final stage is a real headache. that's why there's so many indcuctor/resistor networks along the way. Putting peaking there also means (as you also indicated) that a HF boost is needed to push the HF component along the chain.

The only thing I'm not seeing in that first stage is DC offset bias, which is a simple resistor on the stock boards. It not only sets the pedestal, but also makes sure the signal is equal on both top/bottom outs.

Good job though, considering what's still left out there to make this happen.
post #36 of 101
Now if they (VDC) were to build an HT version I wonder what their choice for opamps would be. Maybe Scott would like to answer that as I'm sure he has palyed around with a few of the newer ones available.
post #37 of 101
Hint Hint
Scott, Mike, plus anyone else who wants to play

OPA695 is the way to go.
or maybe something else he he



Cheers
Steve
post #38 of 101
Quote:
Originally Posted by HK-Steve View Post

Hint Hint
Scott, Mike, plus anyone else who wants to play

OPA695 is the way to go.
or maybe something else he he



Cheers
Steve

thats old new Stevie boy !!

Athanasios
post #39 of 101
Quote:
Originally Posted by Mike Parker View Post

The only thing I'm not seeing in that first stage is DC offset bias, which is a simple resistor on the stock boards. It not only sets the pedestal, but also makes sure the signal is equal on both top/bottom outs.

Mike is this the V-clamp? I dont see that circuit on the new Schematic .It comes in from the pin 9 on the connector but now it goes to a resistor then to ground. Looks like Scott hid that on us !

Athanasios
post #40 of 101
Quote:
Originally Posted by nashou66 View Post

Mike is this the V-clamp? I dont see that circuit on the new Schematic .It comes in from the pin 9 on the connector but now it goes to a resistor then to ground. Looks like Scott hid that on us !

Athanasios

No, he got the V clamp section very well done (in my opinion).

What I'm referring to is 'offset"

That circuit is a mirrored cascode amplifier (if I remember it correctly). So think of it like a push pull audio amplifier where there's a top part of the sine going to the top amp, and the bottom part of the sine going to the bottom amp. What happens here is the signals are out of phase to each other from the G1 and cathod. And that's the mirror part, but the entire signal is out of phase unlike a push pull amp, that only the top part of the signal goes to one final while the bottom half of the signal to the other. It's two complete signals out of phase with each other.

With that, the section that splits or does the most to makes sure the signals are even in both amplitude and linearity would be the preamp (or in audio - differential). And since with this design, the chip that's being used in the preamp should have exceptional and special characteristics to make this happen properly. I know he has this happening, it's just that I'm not seeing where it's happening in the chain. mainly because I'm not an engineer. But I know it's being done.

On the stock boards, they're using two resistors to do this. And if you change with a different chip, you'll also have to change those resistor values to make sure you have the best gain and linearity. You have to use a scope to look at both signals. And the best way to determine the resistor values is the use a varible resistor while watching the scope for best gain, balance and linearity.

I'm a mod tweaker, so I'm sure I'm a bit over the board here, but I've learned a lot scoping video signals through the video chain. And have learned with HDTV signals how important it is to keep the pedestals and amplitude as linear as possible through the chain.
post #41 of 101
Oh, it was mentioned somewhere about the large caps. He's still using the 22uf caps. And for the longest I used to think they needed to be a larger value. But have learned that it's the perfect value. You just nee to make sure the voltage is larger, and that they're 105 temps.

If you notice in the image, I'm using two 10 uf caps on each voltage rail. That makes for a total of 20 uf, which is lower. I found the lower values work best. I even have a board that I'm using a single 10 uf on each rail.


post #42 of 101
Oh I think i understand, so in the new design the V-clam from the Vim is not needed and its taken care of on the VNB itself Right in the circuit.

Ic so one your board in the pic where the normal 22uf 160v cap is you put a 10uf cap and then down the chain closer to the 400ohm 15w resistor you put another 10uf cap. Probaly parallel to the 100nf smd cap, if my memory of the board is correct and looking at your pic.

I see where this could be advantageous to split up the 22uf as further down the line it can pic up some unwanted noise and the 10uf catches that along the way... or is it something else?

Athanasios
post #43 of 101
Quote:
Originally Posted by nashou66 View Post

Oh I think i understand, so in the new design the V-clam from the Vim is not needed and its taken care of on the VNB itself Right in the circuit.

I'm not 100% sure what tse is doing here, but I do follow where he's using the V clamp into the VIM. I'm not saying that he's not doing anything, it's being done for sure, he's just so good at things that he's presented another learning curve for me.


Quote:


Ic so one your board in the pic where the normal 22uf 160v cap is you put a 10uf cap and then down the chain closer to the 400ohm 15w resistor you put another 10uf cap. Probaly parallel to the 100nf smd cap, if my memory of the board is correct and looking at your pic.

I see where this could be advantageous to split up the 22uf as further down the line it can pic up some unwanted noise and the 10uf catches that along the way... or is it something else?

You got it!

using two caps can be better than one, especially if you put the second one nearer the chip in the right place.
post #44 of 101
This is great news
post #45 of 101
Quote:
Originally Posted by mp20748 View Post


The only thing I'm not seeing in that first stage is DC offset bias, which is a simple resistor on the stock boards. It not only sets the pedestal, but also makes sure the signal is equal on both top/bottom outs.

Mike did you noticed that there is unsymmetricall supply for that frist IC (and also some another parts also) Ic is powered by +6v / -5V. i havent tested how op-amp behaves with supplyvoltages like that, but maybe that is what does that DC offset in this new desing?
post #46 of 101
JArmo, also look at the prorection ciciut U1, it has a +15 and a -5 power supply to it.

Athanasios
post #47 of 101
Quote:
Originally Posted by 1031 View Post

Mike did you noticed that there is unsymmetricall supply for that frist IC (and also some another parts also) Ic is powered by +6v / -5V. i havent tested how op-amp behaves with supplyvoltages like that, but maybe that is what does that DC offset in this new desing?

Not sure why that circuit uses the odd voltage rails to the chip. I wonder if that could cause offset.

I know it's being done somehow/somewhere, but I'm not seeing it. I know because he made great progress with the low end, therefore things have to be near perfect to make that happen.

Controlling HD signal pedestals is a piece of work. Almost any offset screws up the pedestals along the chain, especially when the circuit is direct coupled.


Proper pedestals is the reference to proper black levels and true color range..
post #48 of 101
The attachment shows the basic amp. It no longer has the v-clamp circuit like the old card. This amp is just a direct coupled amplifier. The video is clamped (DC restored) on the VIM so doesn't need to be clamped again here.

The video offset is controlled by the BRITE circuit placing a DC voltage on the base of Q12. Here it is -1.3V. With the output of U5 at 0V (video level black) the -1.3V is enough to slightly turn on Q12 and Q19 which causes the outputs to move away from the rails. As the output of U5 increases in voltage Q12 and Q19 conduct more causing the outputs to move closer to 0V turning the CRT on harder.

This is the low frequency scenario. The high frequency portion is the reason for the non symmetrical rails on U5. Black output is 0V which gives 5V of headroom from the -5V rail. Full white is about +1.7V which leaves 4.3V of headroom from the +6V. That's plenty for low freqs but the peaking network made up of C44 and R33 put some pretty good overshoots on the output of U5 so a little extra headroom was called for. The overshoots help extend the bandwidth by overdriving the output amplifier when its bandwidth starts to drop off.

Scott
LL
post #49 of 101
Quote:
Originally Posted by tse View Post

The attachment shows the basic amp. It no longer has the v-clamp circuit like the old card. This amp is just a direct coupled amplifier. The video is clamped (DC restored) on the VIM so doesn't need to be clamped again here.

The video offset is controlled by the BRITE circuit placing a DC voltage on the base of Q12. Here it is -1.3V. With the output of U5 at 0V (video level black) the -1.3V is enough to slightly turn on Q12 and Q19 which causes the outputs to move away from the rails. As the output of U5 increases in voltage Q12 and Q19 conduct more causing the outputs to move closer to 0V turning the CRT on harder.

This is the low frequency scenario. The high frequency portion is the reason for the non symmetrical rails on U5. Black output is 0V which gives 5V of headroom from the -5V rail. Full white is about +1.7V which leaves 4.3V of headroom from the +6V. That's plenty for low freqs but the peaking network made up of C44 and R33 put some pretty good overshoots on the output of U5 so a little extra headroom was called for. The overshoots help extend the bandwidth by overdriving the output amplifier when its bandwidth starts to drop off.

Scott

So this is with the VNB that is compatible with the Vims we mostly use now?
So on the VNB's we have now is there still a need for the Clamp(DC restore)?
can that be eliminated on our VNB's and we can have them directly coupled or will it cause problems? it sounds like one more circuit out of the way is a good thing. And if the DC restore was only needed once, why did the original design call for it on the VNB?

Athanasios
post #50 of 101
Quote:
Originally Posted by nashou66 View Post

And if the DC restore was only needed once, why did the original design call for it on the VNB?

Athanasios

It's not mainly on the neck boards. The video chain is basically DIRECT coupled all the way to the CRT's. The only place on the VIM where it's not direct coupled is before the coupling caps (0.1uf) on the first stage. At that same stage, the signals are clamped using the V Clamp and clamp IC (right after the first stage IC on the VIM)

The direct coupling of the Marquee makes it a better design, mainly because there's no need to use several clamp pulses in the chain to control reference. And when you use that design, they can fight with each other.

However, when dealing with HD signals, there's something with it's transcoded digital tri-level sync (all HDTV signals are transcoded digital component), that can create another problem along the signal chain...
post #51 of 101
Quote:
Originally Posted by tse View Post

The attachment shows the basic amp. It no longer has the v-clamp circuit like the old card. This amp is just a direct coupled amplifier. The video is clamped (DC restored) on the VIM so doesn't need to be clamped again here.

Ok, this is where you've improved on the low end. Good move!

The more you try to control the reference black level by using more than one clamp or correction circuit/pulse, the more difficult it is a get/maintain good low end control.



Quote:


The video offset is controlled by the BRITE circuit placing a DC voltage on the base of Q12.
Scott

I saw that. What I was referring to was the output of H1100 when scoping for linearity (using a ramp pattern).

On the old stock boards, they are using R90 to even out the ramp. So they are using that resistor to bias that chip for best video linearity. Without that resistor the ramp would most likely distort. It would look somewhat like a non linear gamma curve.
post #52 of 101
Quote:
Originally Posted by mp20748 View Post

It's not on the neck boards. The video chain is basically DIRECT coupled all the way to the CRT's. The only place on the VIM where it's not direct coupled is before the coupling caps (0.1uf) on the first stage. At that same stage, the signals are clamped using the V Clamp and clamp IC (right after the first stage IC on the VIM)

The direct coupling of the Marquee makes it a better design, mainly because there's no need to use several clamp pulses in the chain to control reference. And when you use that design, they can fight with each other.

However, when dealing with HD signals, there's something with it's transcoded digital tri-level sync (all HDTV signals are transcoded digital component), that can create another problem along the signal chain...

Are you sure mike? I know the V-clamp is on the Vim but from there is comes to the VNB via pin 9 . From there it looks like it activates mosfet q27 that allows sample signal from transistor Q1 to be combined with "bright"
and SpotKill via U4 and U1 as signal outs_ample(all on page 1 of the VNB schematics). this then goes the the one side of the cascading amp for G1(page 2 of VNB schematics) I assume this is the second DC restore Scott mentioned, Help me to understand this a bit more... I love learning how all this works

It is that Circuit that is missing from the new Schematics, now it is not called out_sample, it is Bright and SpotKill that come in separately and are combined Via U2 on the new Schematics.

I have a headache now

Athanasios
post #53 of 101
OK, I'm seeing and trying to understand something new here. Where in the video chain (on the stock neck board) is clamping also being used other than first stage VIM?

I was looking at Q26, and it shows "sample" in the drawings. That makes me think clamping, but it's also the BRIGHTNESS control on the neck boards.


Hmm... If there's a second clamp in the video chain, i need to also look at removing it...
post #54 of 101
Quote:
Originally Posted by nashou66 View Post

Are you sure mike? I know the V-clamp is on the Vim but from there is comes to the VNB via pin 9 . From there it looks like it activates mosfet q27 that allows sample signal from transistor Q1 to be combined with "bright"
and SpotKill via U4 and U1 as signal outs_ample(all on page 1 of the VNB schematics). this then goes the the one side of the cascading amp for G1(page 2 of VNB schematics) I assume this is the second DC restore Scott mentioned, Help me to understand this a bit more... I love learning how all this works

It is that Circuit that is missing from the new Schematics, now it is not called out_sample, it is Bright and SpotKill that come in separately and are combined Via U2 on the new Schematics.

I have a headache now

Athanasios


OK, I'm seeing it now. I've seen that "sample" on Q1 but never paid much attention to it.

Interesting... Now I'll probably be up all night trying to figure out how to remove it..
post #55 of 101
Quote:
Originally Posted by mp20748 View Post

OK, I'm seeing it now. I've seen that "sample" on Q1 but never paid much attention to it.

Interesting... Now I'll probably be up all night trying to figure out how to remove it..


LOL , yeah that was bothering me for these past 2 days now! Scott says its not needed but it looks like the Bright and the SpotKill still have to get back to the G1 side of the amp with out the V-Clamp. We'd have to compare it to his new design. I am sure the needed signal voltage for the spotkill not to be activated has to be still there or not altered. that will be the tricky part.

Athanasios
post #56 of 101
Quote:
Originally Posted by nashou66 View Post

LOL , yeah that was bothering me for these past 2 days now! Scott says its not needed but it looks like the Bright and the SpotKill still have to get back to the G1 side of the amp with out the V-Clamp. We'd have to compare it to his new design. I am sure the needed signal voltage for the spotkill not to be activated has to be still there or not altered. that will be the tricky part.

Athanasios


OK, I got it now. they have a complete sample/hold circuit there. Much like what they're doing with the clamp chip on the first stage of the VIM. I agree with Scott in that it should not be needed, especially since the chain is direct coupled to the CRT's..

This should be easy..
post #57 of 101
Quote:
Originally Posted by mp20748 View Post

OK, I got it now. they have a complete sample/hold circuit there. Much like what they're doing with the clamp chip on the first stage of the VIM. I agree with Scott in that it should not be needed, especially since the chain is direct coupled to the CRT's..

This should be easy..

I expect a full report on my desk in the morning ! I have no idea where to begin on this!

Athanasios
post #58 of 101
Quote:
Originally Posted by nashou66 View Post

I expect a full report on my desk in the morning ! I have no idea where to begin on this!

Athanasios

and by the morning, I'll have a full report on this, but give me a few minutes and I'll have a partial report...
post #59 of 101
It's not going to be as easy as I had thought. So it looks like I'll be up until I make it happen.

Anyway, I'm all for making some changes in that area of the circuit. One of my problem areas on the neck board was the small ferrite bead near Q26. the bead is a noise stripper, but it fails big time In what I'm doing in my latest work. So I've changed the value and made other changes going back to those circuits. Still, there was more noise measured than I wanted to see. So these changes would put me where I would want to be by eliminating the use of that sample/hold circuit..

I'll be up with the sun...
post #60 of 101
Quote:
Originally Posted by mp20748 View Post

It's not going to be as easy as I had thought. So it looks like I'll be up until I make it happen.

Anyway, I'm all for making some changes in that area of the circuit. One of my problem areas on the neck board was the small ferrite bead near Q26. the bead is a noise stripper, but it fails big time In what I'm doing in my latest work. So I've changed the value and made other changes going back to those circuits. Still, there was more noise measured than I wanted to see. So these changes would put me where I would want to be by eliminating the use of that sample/hold circuit..

I'll be up with the sun...

I knew id peak your interest with that one !
I cant wait to see what you come up with and how difficult it will be.

Athanasios
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: CRT Projectors
AVS › AVS Forum › Display Devices › CRT Projectors › VDC 8" HighRes Projector