TV Jungle chips and B/W picture. Vintage Arcade Repair

[Karnov]

I trying to get an vintage Atari COPS arcade game up and going. It's an late laserdisc based game and working ones are very rare.
The issue(one of many, but the bigger issue) is that the picture is black and white. There is a board that takes the NTSC composite signal from the laserdisc and converts it to RGB+sync for the arcade monitor. It is using a Sony CXA1013AS jungle chip to do this.  The datasheet does not exist on the Internet, but the Sony KV-19TS20 uses the same chip and I have that service manual.
The waveforms I scope out looks to be about the same as indicated in the TV service manual. When I connect my PicoScope to pin 8 (XTAL) on the CXA1013AS (probe set to 10x), the scope is reporting it running at 9.40MHz. Not 3.58Mhz. Voltage is spot on at 100mv p-2-p as indicated by the TV manual and it has a nice sine wave. I have replaced the CXA1013AS chip and tested a bunch of stuff on this board, but everything appears to be working fine.
Am I measuring this wrong? Any other tips?

Any help is appreciated. This is one of those rare games we really want to get operational for the arcade museum for the public to play.

[fzabkar]

The only time I have seen a crystal oscillate at a frequency which was not a harmonic was in an IBM PS/2 Model 25. In my case I suspect that there may have been a dag on the crystal when it was trimmed, and this dag subsequently fell off during normal usage. Can you hear any rattling when you shake the crystal?

You might also want to check the LC passives.

[vk6zgo]

Are you just relying upon the frequency readout of the 'scope, or have you measured the duration of a cycle in the time domain?
It should be about 279ns for 3.58MHz, & 106ns for 9.4MHz.

[floobydust]

I think the colourburst osc might be ok, it's just getting loaded down by the scope input capacitance?

Do you see chroma signal (colour burst) at the Jungle IC input pin 46 from the composite video? Just to make sure you have a colour signal coming in. It gets filtered out for luminance signal split to pin 43. You can compare the signals at those two pins to see what's going into the jungle IC. The IC is also in KV-2137RS, KV-2127RS
Look at the Saturation control voltage at pin 39 3.5VDC that it's not 0V.

[james_s]

I assume you did verify that you are getting color video out of the laserdisc player? Many players have Svideo outputs too which will be separate chroma/luma signals, ie B&W video and a separate color signal. With old arcade games it's pretty typical to find all kinds of random hacking from past techs.

[Karnov]

link=topic=256782.msg3273254#msg3273254 date=1602457659]
The only time I have seen a crystal oscillate at a frequency which was not a harmonic was in an IBM PS/2 Model 25. In my case I suspect that there may have been a dag on the crystal when it was trimmed, and this dag subsequently fell off during normal usage. Can you hear any rattling when you shake the crystal?

You might also want to check the LC passives.

[/quote]

No rattle when shaking the crystal.  I don't have a cap tester that goes as low as 15p, but I ordered some replacement caps anyway. I am not sure how to match up that candy style inductor, but I ordered a wires wound one along with all other passives based on best guess.

Are you just relying upon the frequency readout of the 'scope, or have you measured the duration of a cycle in the time domain?
It should be about 279ns for 3.58MHz, & 106ns for 9.4MHz.

I first came up with approx 100ns on an old analog scope. I then put the PicoScope on it and it shows 9.4MHz. I believe it is accurate, but I plan to check on my Tek465 (highest Mhz sope I have) when I tackle the project again next weekend. I could be off on my counting.

I think the colourburst osc might be ok, it's just getting loaded down by the scope input capacitance?

Do you see chroma signal (colour burst) at the Jungle IC input pin 46 from the composite video? Just to make sure you have a colour signal coming in. It gets filtered out for luminance signal split to pin 43. You can compare the signals at those two pins to see what's going into the jungle IC. The IC is also in KV-2137RS, KV-2127RS
Look at the Saturation control voltage at pin 39 3.5VDC that it's not 0V.

I suspect the scope input capacitance is loading it down, but I have no way to validate that.  There is a comb filter that is buffered with a half dozen transistors before the luma and chroma. I pulled out the transistors and tested them all and they are good. I am getting a chroma and luma signal to the jungle chip. Thanks for the information on the other TVs that use this chip. I have a watch on ebay for the PCBs. I checked all the pots for color adjustment and they appear to be working fine.

I assume you did verify that you are getting color video out of the laserdisc player? Many players have Svideo outputs too which will be separate chroma/luma signals, ie B&W video and a separate color signal. With old arcade games it's pretty typical to find all kinds of random hacking from past techs.

Yep. Player works great. The monitor has color when I feed it an RGB test signal.

[james_s]

I have had a couple of crystals oscillate at the wrong frequency, it's rare but not impossible. Given how cheap crystals are it might be worth changing it out to see if it fixes it. I would also check the loading capacitors.

[Karnov]

No luck on changing the crystals, caps or inductor. Still not 3.57Mz

[Karnov]

So the TV diagrams say the C IN should be 2.4v or 2.8v p2p, depending on what model I look at. I am seeing 1.6v at best.

So I diagrams the circuit starting after the HCF0053 comb filter to the C IN (Croma In) on the CXA1013. The signal does not change much until you get to the emitter of N11. Then it really drops to almost nothing. The signal is boosted back up by the time it gets to the base of N12.

But only to 1.6v. But I don't know if that is enough.

What is N4 and N11 doing anyway?