I have successfully de-CQM'd a CT3990 and CT3600 :

I also took a look at the CQM chip and it runs off 46.615120MHz, with 20x cycles per BCLK and 960 cycles per sample, giving 48557.416...Hz sample rate. Output format is 16bit i2s with 48x BCLK cycles per sample. Original OPL2/3 sample rate is 3.579545 / 72 = 49715.903...Hz.
YMF289 has 16bit Right-Justified format so no reuse of the TDA1387 DAC is possible, one will have to use the crappy Yamaha's DAC.

Tiido wrote:

YMF289 has 16bit Right-Justified format so no reuse of the TDA1387 DAC is possible, one will have to use the crappy Yamaha's DAC.

Good to know. Thanks!

Thanks Eep386 and everyone else for the information 😀 Going through the list of parts to change and the YAC516 datasheet, I've managed to switch my "Soundblaster 32" CT3620 and Soundblaster 16 CT2950 from CQM to the YMF289's OPL3.
Both of them are working well on my test PC, which is good because the CT2950 wasn't so useful otherwise. I'm really happy with the CT3620 - my AWE64 Gold broke and this fills the gap very nicely with 8MB ram fitted.

My main takeaways from this:

• The 10 ohm resistor needs to be a bit bigger apparently
• Cheap 33.8688 Mhz oscillators from China work fine and arrived quickly
• Removing the CQM chip and TDA chips requires a heatgun
• Fitting the YMF289 & YAC516 chips was easier to carefully hand solder
• YMF289 bought on ebay from China (notice it's missing the 'B' ?) mayyyy be a clone, but it's definitely a proper OPL3 clone if it is one
• YAC516 works well - note it's upside down on the CT2950

I ended up making some graphics to work from:

CT3990:

CT3620:

CT2950:

Fantastic! Maybe now we won't have to be afraid of so many CQM cards that are otherwise quite nice...

Hello,

Recently i pickup de-CQMed CT2950, but the previous owner removed the YMF289 chip from it because he couldn’t make the FM sound work. He attached original CQM and TDA to complete. So, i want restore CT2950 to original state. I soldered back CQM, TDA, resistors, but i need help with identify capacitors near CQM. Pls look at the pic below:

Orange rus text mean "broken" - C110 near TL064 have split pin. Now I temporarily put random capacitors:
C63,C64 - i put 12 pF,
C55, C65 and С66 - around 1 uF,
C76,C77 - pass through left and rihgt sound channels - 2,2 uF
broken C101- 100 nF.

Sound card is alive, FM music playing OK, but for greater authenticity, of course, I would like to solder the "correct" values of the capacitors in the harness ...
Thank you.

Hi there,

Do you know if the original owner posted about this last week? I saw a similar message but by the time I tried to reply I couldn't find the thread, maybe the post was deleted?

I checked the caps on mine by removing each capacitor and resoldering it afterwards. While measuring, C110 flew off my desk onto the floor and nearly got away 😁
c55 measures 100nF
c63 measures 3.3nF
c64 measures 3.3nF
c65 measures 125nF
c66: measures 125nF
c110: measures ~100nF (117nF)
C100 & C101 are 1uF

I'll update my picture so that it doesn't remove so many components. Looking at this picture of a CT2950 that came from the factory with the OPL3 fitted, it looks like removing C66 may have caused these problems.

Thermalwrong wrote on 2020-02-02, 18:59:

Do you know if the original owner posted about this last week? I saw a similar message but by the time I tried to reply I couldn't find the thread, maybe the post was deleted?

Hi,
I have not seen his posts on this forum. He created the topic in a Russian-language forum about the problem.

Thermalwrong wrote on 2020-02-02, 18:59:

I checked the caps on mine by removing each capacitor and resoldering it afterwards. While measuring, C110 flew off my desk onto […]
Show full quote

I checked the caps on mine by removing each capacitor and resoldering it afterwards. While measuring, C110 flew off my desk onto the floor and nearly got away 😁
c55 measures 100nF
c63 measures 3.3nF
c64 measures 3.3nF
c65 measures 125nF
c66: measures 125nF
c110: measures ~100nF (117nF)
C100 & C101 are 1uF

I'll update my picture so that it doesn't remove so many components. Looking at this picture of a CT2950 that came from the factory with the OPL3 fitted, it looks like removing C66 may have caused these problems.

Thank you! I will try to find suitable capasitors. But as before, two capacitors remain unknown - C76 and C77.

Another mistake on my part, the C76 & C77 capacitors are 1UF, they're the same as C100 & C101 😀

Thermalwrong wrote on 2020-02-03, 12:50:

Another mistake on my part, the C76 & C77 capacitors are 1UF, they're the same as C100 & C101 😀

Thank you!

You guys have done a fantastic job! Congratulations! I'm looking to do the same with a CT2940, although similar to the 2950, it is not exactly the same layout. Would any of you know the values of the capacitors and resistors?

OK, lemme see what I can recall of my fix-up of the CT2940... it's been some time and I lost my notes 🙁

X3 needs to be populated with a 33.868MHz oscillator.
R51 is 16 ohm, R52 and R53 are 47K, R57 is 3.9K, R65 is 1K.
C71, C73, C95, C96, C97 and C98 are 100nF caps.
I think C112 and C113 are 15pF caps.
After fitting the YMF289B and YAC516-E chips, you then remove the CT1978 chip at U3 and the TDA1387 DAC chip at U1.

That's all I can remember sadly; though I think some of the caps around the removed CQM and DAC chips measure around 100nF, you may be able to reuse them if you remove them carefully. I don't remember which though 🙁

I'm trying to find the correct resistors and capacitors to mod my CT2950, but I'm not sure about the ratings, such as wattage and ppm for the resistors, temp, voltage and flex for the capacitors. Do any of these specs matter?

As long as the wattage of the resistors is at least 1/8W I don't think the other specs will be too important. You want 0805 (2012 metric) size parts for this application.
Though generally you'd want the lowest ppm and tolerance range for the resistors; in this circuit those specs are not too critical.

Eep386 wrote on 2020-12-12, 16:25:

As long as the wattage of the resistors is at least 1/8W I don't think the other specs will be too important. You want 0805 (2012 metric) size parts for this application.
Though generally you'd want the lowest ppm and tolerance range for the resistors; in this circuit those specs are not too critical.

Thank you for the quick response. Here are the parts I am ordering. Thoughts?

Good luck with it 😀 Although I didn't say in my graphic, the SMD capacitors are all 0805 kinda size, 16/50 volts for caps with 0.125 or 1/8 watts for the resistors. Only a few of the resistors that I noted as 1/4 watt or 0.25 watt need to be bigger.

That price for the 1uf capacitors is kind of extremely high. 2 of the 1uf capacitors I just moved from C76 & C77 to C100 & C101. The last 1uf capacitor, I used from a cheapy ebay SMD 0805 capacitor selection book, which has proved handy for a bunch of projects. You probably don't need 5% spec capacitors for this application.

Man, looking at this card 18 months on, I'm glad I put together the graphic, I barely remember what I swapped around and my work looks clean enough (with the flux cleaned) that I almost can't tell what was factory and what was my work. Not sure where I sourced the 0.25w 10ohm resistor, probably another soundcard.

Eep386 wrote on 2020-12-11, 02:12:

OK, lemme see what I can recall of my fix-up of the CT2940... it's been some time and I lost my notes :( […]
Show full quote

OK, lemme see what I can recall of my fix-up of the CT2940... it's been some time and I lost my notes 🙁

X3 needs to be populated with a 33.868MHz oscillator.
R51 is 16 ohm, R52 and R53 are 47K, R57 is 3.9K, R65 is 1K.
C71, C73, C95, C96, C97 and C98 are 100nF caps.
I think C112 and C113 are 15pF caps.
After fitting the YMF289B and YAC516-E chips, you then remove the CT1978 chip at U3 and the TDA1387 DAC chip at U1.

That's all I can remember sadly; though I think some of the caps around the removed CQM and DAC chips measure around 100nF, you may be able to reuse them if you remove them carefully. I don't remember which though 🙁

Awesome, thank you so much! I'm waiting for the oscillator to complete it and I'll let you now how it went. Thanks!

Superpuissant wrote on 2020-12-28, 21:35:

Eep386 wrote on 2020-12-11, 02:12:

OK, lemme see what I can recall of my fix-up of the CT2940... it's been some time and I lost my notes :( […]
Show full quote

OK, lemme see what I can recall of my fix-up of the CT2940... it's been some time and I lost my notes 🙁

X3 needs to be populated with a 33.868MHz oscillator.
R51 is 16 ohm, R52 and R53 are 47K, R57 is 3.9K, R65 is 1K.
C71, C73, C95, C96, C97 and C98 are 100nF caps.
I think C112 and C113 are 15pF caps.
After fitting the YMF289B and YAC516-E chips, you then remove the CT1978 chip at U3 and the TDA1387 DAC chip at U1.

That's all I can remember sadly; though I think some of the caps around the removed CQM and DAC chips measure around 100nF, you may be able to reuse them if you remove them carefully. I don't remember which though 🙁

Awesome, thank you so much! I'm waiting for the oscillator to complete it and I'll let you now how it went. Thanks!

Just an update, the surgery was a sucess, the yamaha chip has been sucessfully installed and works perferctly (although I have butchered a bit the card while removing the CT1978 chip... I really need to buy a hot air gun...)! This would never have been possible without your help, thank you so much!

Glad to be of help! 😀
That's another ViBRA16 saved from CQM cancer.

Tiido wrote on 2019-03-14, 02:14:

I have successfully de-CQM'd a CT3990 and CT3600 : http://www.tmeeco.eu/BitShit/PCschit/CT3600-deCQM.jpg http://www.tmeeco.eu/Bi […]
Show full quote

I have successfully de-CQM'd a CT3990 and CT3600 :
http://www.tmeeco.eu/BitShit/PCschit/CT3600-deCQM.jpg
http://www.tmeeco.eu/BitShit/PCschit/CT3990-deCQM.jpg

I also took a look at the CQM chip and it runs off 46.615120MHz, with 20x cycles per BCLK and 960 cycles per sample, giving 48557.416...Hz sample rate. Output format is 16bit i2s with 48x BCLK cycles per sample. Original OPL2/3 sample rate is 3.579545 / 72 = 49715.903...Hz.
YMF289 has 16bit Right-Justified format so no reuse of the TDA1387 DAC is possible, one will have to use the crappy Yamaha's DAC.

Apologies for being a bother about a relatively old thread, but I've got a CT3600 that I'd like to do this to, and while the resistor values required are clearly visible, I'm having a bit of trouble figuring out which capacitors go where, exactly. If I had to guess looking at sizes, I'd say that C46, C50, and C52 are 1uF, and C45, C46, and C47 are 100nF? Other cards seem to require ~12pF capacitors in some places, but I'm guessing that's to do with the clock crystal, which is handled by the clock generator in the case of the CT3600?

Perhaps either Tiido or Eep386 can shed some light on this for me, if they're feeling merciful.

C46, 50 and 52 should be as high as you can find, I used 10µF. Bigger = better, less noise etc.
C45, 47 and 48 are 0.1µF but can also be 10µF or something else big. I took the parts from around CQM and its DAC stuff.
C41 I moved from C37, R13 from R9. This moves the CQM clock parts to OPL3L side, the MicroClock PLL generates the right stuff.