Coleco Telstar Combat! No Video Repair HELP!

[TheProgrammerIncarnate]

Here I added some more pictures to the ones I already took: https://drive.google.com/drive/folders/1wT2uEGUd0Weu8t7w5Pnv5nfFAgAe3Tmx?usp=sharing

 

Just tell me which components you can't see properly and I'll look at them closer

[ChildOfCv]

Thanks.  Well, I only see one new picture (fulltop).

 

One thing I'll probably have to consult on, though, is the values of the resistors.  For brevity I'll just start making a list.

 

R71 (upper left corner).  Is that brown/black/brown?  (100 ohm)

C(38?) (under C50)  What's the ID?  Also, they seem to be saying 33uF/16V, or maybe those numbers mean nothing and their values are on a different side of the component?

R7 - brown/black/yellow? (100K)

Just above R7:  What's that ID?  And is it Grey/Red/Orange?  (82K)

The transistor to the left of Q3, and most of the components in that area have their IDs hidden.  What is the part number of Q3 and its buddy on the other side?

I think the transistor below them is a 2N2021?

Also, the mylar capacitors (the green ones) will need their values read off.

 

[TheProgrammerIncarnate]

15 hours ago, ChildOfCv said:

R71 (upper left corner).  Is that brown/black/brown?  (100 ohm)

C(38?) (under C50)  What's the ID?  Also, they seem to be saying 33uF/16V, or maybe those numbers mean nothing and their values are on a different side of the component?

R7 - brown/black/yellow? (100K)

Just above R7:  What's that ID?  And is it Grey/Red/Orange?  (82K)

The transistor to the left of Q3, and most of the components in that area have their IDs hidden.  What is the part number of Q3 and its buddy on the other side?

I think the transistor below them is a 2N2021?

Also, the mylar capacitors (the green ones) will need their values read off. 

 

R71 is 100 ohms

The capacitor under C50 is C38, I can't see anything except 33/16 & "Tracon" on them. They look like pre-speaker filters, which should limit their possible values.

R7 is correct

The transistor near Q3 is Q4, across from it is Q5

Q5 says "1702M P731" on it, while Q3 & Q4 say "MPS A13" Google says the 1702M is an audio transistor which would make sense.

The mylar capacitors only say "1.50V" on them, so I'll have to test them along with the Tracon ones

[ChildOfCv]

9 minutes ago, TheProgrammerIncarnate said:

The capacitor under C50 is C38, I can't see anything except 33/16 & "Tracon" on them. They look like pre-speaker filters, which should limit their possible values.

It's looking likely, then, that 33uF is the value and 16V is the rating.  I tried a quick search for 16uF capacitors rated at 33V and found nothing.  They all seem to be used as starter capacitors for industrial motors, and about the size of a can of Coke, with ratings above 200V.

Edited August 6, 2019 by ChildOfCv

[TheProgrammerIncarnate]

Ok I messed around with the composite mod again and here's the results. I used a 470+100ohm voltage divider, plus a 100ohm resistor to match impedance (75ohm might be better, but I don't have any.) Schematic is below.

 

On the digital TV it looks great, except for some noticeable vertical lines which are based on the clock frequency.

 

On the CRT TV, the black components (right tank + mines) and grey background are hard to tell apart, the white areas also are not maximum brightness.

 

But when I set the CRT's brightness to 100 (instead of 50) it looks perfect. Which is making wonder if the composite mod is good but the TV isn't.

[eDoc]

1 hour ago, TheProgrammerIncarnate said:

Ok I messed around with the composite mod again and here's the results. I used a 470+100ohm voltage divider, plus a 100ohm resistor to match impedance (75ohm might be better, but I don't have any.) Schematic is below.

 

On the digital TV it looks great, except for some noticeable vertical lines which are based on the clock frequency.

 

On the CRT TV, the black components (right tank + mines) and grey background are hard to tell apart, the white areas also are not maximum brightness.

 

But when I set the CRT's brightness to 100 (instead of 50) it looks perfect. Which is making wonder if the composite mod is good but the TV isn't.

That looks great, especially on the digital TV.  Would you be able to take a picture on the actual board where you made the mod?

[TheProgrammerIncarnate]

5 hours ago, eDoc said:

That looks great, especially on the digital TV.  Would you be able to take a picture on the actual board where you made the mod?

The components are currently on a breadboard, so I used an alligator clip connected to the unmodulated video (blue circle), along with a pair connected to power/ground lines.

 

I forgot to mention it earlier, but a decoupling capacitor should be added to the mod. Without it, the brightness fluctuates whenever the sounds play.

 

[ChildOfCv]

7 hours ago, TheProgrammerIncarnate said:

But when I set the CRT's brightness to 100 (instead of 50) it looks perfect. Which is making wonder if the composite mod is good but the TV isn't.

You can always try different TVs, but that looks like a contrast issue.  Without scoping the signal I can only speculate, but it may be due to the signal not being perfectly 1V.  It may actually help, though, to modify the circuit slightly:

 

 

Oops, forgot to assign a value to the capacitor.  Try around 10uF.

 

The amplifier used for the composite mod is known as an emitter follower.  That's because whatever voltage is on the base, the transistor demands that the voltage be matched on the emitter (minus the small forward voltage drop that all silicon junctions cause).  So if the base is 3V and we go with the standard .7V drop, the emitter will be about 2.3V.  If the base is 3.7V, the emitter will be 3V.  And so on.

 

The problem with this circuit is that the load on the base will be affected by the emitter resistor and the gain of the transistor.  A small resistor (such as the 75 ohms that we want) will have only about a 7.5K input impedance, which you'll notice can throw the summing circuit way out of balance.  So I chose the darlington NPN transistor for output here because that increases the base impedance closer to 750K, which should affect the circuitry far less.

 

The capacitor removes the DC component from output.  However, TVs still expect a near-zero bottom and not a negative voltage.  So if it looks weird with the capacitor, the diode should act as a clamper to keep the overall signal on the positive side.

 

PS:  The capacitor may also aid in ridding the jail bars.

Edited August 7, 2019 by ChildOfCv

[ChildOfCv]

Here's some work done on the annotation of the board.  Red text is what I'm sure of, cyan is either unknown or in doubt. Could you fill those values in?

Edited August 7, 2019 by ChildOfCv

[SoundGammon]

I noticed that there is no bias voltage on the base, this would make the transistor behave erratically wouldn't it?  

 

Here's one of the many 2600 mods that have a resistor from the 5v to the base and then a resistor to ground to keep the voltage at one level.

[ChildOfCv]

Just now, SoundGammon said:

I noticed that there is no bias voltage on the base, this would make the transistor behave erratically wouldn't it?  

 

Here's one of the many 2600 mods that have a resistor from the 5v to the base and then a resistor to ground to keep the voltage at one level.

It's not needed if direct coupling, and if the input already has a high enough bias voltage.  Even if necessary, this forces you to adjust other resistors since the bias network will reduce input impedance and increase the load on the output.  This will in turn throw off the balance of the summing network that feeds it.

 

The more I think about it, I think it would actually be better, even for Atari, to use an op-amp with enough frequency response and sufficient output current to power a 75-ohm load.  This would have near infinite input impedance, meaning you don't have to modify the existing circuit much, if at all, and it would work even better than this basic circuit.

[SoundGammon]

Here's a link to a company that sells video op-amps!

 

https://www.analog.com/en/products/ada4891-4.html#product-overview

[ChildOfCv]

47 minutes ago, SoundGammon said:

Here's a link to a company that sells video op-amps!

 

https://www.analog.com/en/products/ada4891-4.html#product-overview

Yeah, something like that looks like it would work.  You'd use a circuit similar to this:

(Ignore the pin numbers--they don't have this exact chip in the database)

RV1 is only needed if the input signal is too strong.  RV2 is only needed if it's too weak.

Edited August 10, 2019 by ChildOfCv

[TheProgrammerIncarnate]

On 8/7/2019 at 3:01 AM, ChildOfCv said:

Here's some work done on the annotation of the board.  Red text is what I'm sure of, cyan is either unknown or in doubt. Could you fill those values in?

Here's your picture with the values filled in. I just noticed that all of the green capacitors (except the small one in the upper right) are 0.1uF. Sorry for the late reply, I've been busy.

 

[ChildOfCv]

Cool.  I updated the schematic and my image with those corrections.

 

Are you working on the second picture with the remaining callouts?  Also, it would be nice to know exactly how that trim pot on the right side is connected into the circuit.

[TheProgrammerIncarnate]

22 hours ago, ChildOfCv said:

Are you working on the second picture with the remaining callouts?  Also, it would be nice to know exactly how that trim pot on the right side is connected into the circuit. 

Second picture? There's only one up there. I'll get back to you on the trim pot.

[ChildOfCv]

7 hours ago, TheProgrammerIncarnate said:

Second picture? There's only one up there. I'll get back to you on the trim pot.

Weird.  I could swear I uploaded a second picture.  Maybe I forgot to use the submit button.  Let's try that again.

Of course you don't have to provide the audio side values again

[TheProgrammerIncarnate]

Ok I confirmed the values of everything except stuff in the far upper-right. There's a close-up of that region, so I'll mark that picture instead as this one is blurry and has things in the way. A few things I wasn't able to confirm/find out values on: the two inducters/coils inside the modulator, the two inductors to the left of the power switch, and "L1" which appears to be a ferrite choke. None of these had values on or near them. Two of the dark brown film capacitors & a transistor inside the modulator were too tightly packed for me to read their values. The transistor is probably a standard or video-specific NPN, while the capacitors are probably "39" (nanofarads?) like all of the others of that type. Lastly the two blue pots don't have values on them and appear to have been set at the factory, as they've been glued in place.

 

I also found the manual in slydc's archive, cleaned the controllers, and took some notes on the gameplay. That'll be useful for whenever I start looking at the AY-3-8700's die.

[ChildOfCv]

Cool.  Here's the newest schematic and the newest version of the parts pic.

 

You can get an approximation of CR20 by turning it on and measuring the voltage across it.  Of course, there is also a part number printed on it, but you'd have to pull it and roll it around to find the whole thing.

 

To get the value of the trim pots, you'd also have to pull them and measure the resistance across the outer pins underneath the gunk.  It's pretty much impossible to measure while in circuit.

 

Combat.zip

[TheProgrammerIncarnate]

Here's the picture of the upper-right corner. I think I labeled all hidden/obscured parts. Also, the speakers are 0.2W, 8 Ohm for the record.

 

Unfortunately I'm going to be away from home again so I won't be able to properly inspect those pots or inductors. Just mark them with "untested" for the time being. Also, I'm going to store the schematics in my 1st gen game system REing git repo, you won't mind as long as I credit you?

[ChildOfCv]

That's fine.

 

Hmm, I see there's a W5 wire too.  That wasn't apparent in the other pictures.  Where does it go?

 

Could you get the part number of Q6 though?  I'm especially fixated on that one because its apparent use in the circuit makes no sense if it's NPN.  eDoc's board has it as a NPN transistor, but his board has been (badly) repaired too, so it may be an incorrect replacement.

 

For that matter, Q5 also seems suspicious.  As is, it's incorrectly biased.  There ought to be a voltage divider resistor on the emitter, or something, to give it some voltage difference there.  It might also make sense if R57 were tied to ground instead of power, though repeated scrutiny of the tracks tells me that it's connected to R64 on that side.

Edited August 15, 2019 by ChildOfCv

[ChildOfCv]

Updated schematic.  PCB is also populated, so you can see where components are.  This also helps to confirm that the schematic matches the traces, since I followed the same paths as pictured for each connection.

 

Combat.zip