Boss DD-3 Problem

[Unluckyfett]

Started on another forum, posting here because the other forum doesnt seem very active.

Schematic: https://www.badcaps.net/filedata/fetch?id=2255465

Pink label Square chip made in Taiwan. Bypass works, direct out works. Delay signal is not right. Another person described it as " constant a noise that can be clearly shaped by Delay Time, Volume and Mode, but very subtle by Feedback" Clock seems to be working. I replaced all 10uf electrolytics just because. No change. Based on sound and experience, should I be looking at the AD part or the DA part? I found a breakdown of how the microcontroller samples 12bit audio, I have a scope, I plan on doing more probing.


Update 1:

I pulled C049 to help isolate the problem, and that killed the ticking garbage noise but no delay (obviously). So I used the tone probe back to IC8. I am not sure how the whole DA thing is accomplished but I know it's done by RA1 and IC8b. Looking before IC7, pin 3 of IC5 has clean signal. If I understand correctly, pin 7 of IC5 should be serial pulses, which it seems to be. So now I'm wondering if its a problem with RA1 and IC8b or if theres a problem with the micro controller. For shiggles I checked the adjustment of the clock as shown in the documentation and it says the pulse should be 5v and 12.5us rising edge to rising edge. I only had the cheap chineasieum scope today, the clock signal is present but it was only reading .5v P-P. I double checked all the voltage rails and everything is close enough 8.8v 4.2v and 4.8v. I'm sure the battery is weak after all the troubleshooting. I checked the memory chip and there is clean serial bitstream on the scope. another thought I had was the compander. Is all the noise "normal" and the compander removes it in normal conditions? I checked the tants attached to IC2 and both check good, usually they let the smoke out so Im not surprised that they test fine.

Lots of testing and no real progress.

I apologize for the scattered thoughts. My brain is a series of Robot Chicken Sketches.

Tomorrow I am going to break out the Tektronix if I have time and double check the clock. Besides throwing components at it, I am running out of ideas

Please someone let me know if my thought process is entirely wrong

Update 2:Again, not 💯 on the whole AD-DA process but it looks like things start to breakdown at IC8. I ordered one and we'll see what happens. With IC8 removed there is no noise (but no effect obviously).

Anyone have any thoughts on the wave forms?

[Slowpoke101]

Replace the 1uF capacitors C045, 49, 50 & 51. Then check the quality of the audio on pin 14 or 15 of IC2. This is the delayed audio after it has been put back together and filtered but before it is sent into IC2's expander. You can then see what its expanded delayed audio output looks like on the pins of C045.

The 10uF caps are usually the main cause of troubles with the DD-2 and 3 delay pedals but you have changed these but do make sure that you got all of them and that you have them install with the correct polarity.

Sometimes damaged memory can cause glitch or ticking noise to be heard in the delay signal. Generally if you increase or slow down the delay rate length, the glitch repetition rate or duration changes. The only solution to this problem is to replace IC6. It is a 4 bit 64 Kb dynamic RAM and they are a bit hard to find but they can be found.

Also, check the 5 volt power rail. Make sure that it is at 5 volts and that it is clean. Check it with your oscilloscope. C019 and C027 may need to be replaced (both 47uF) if the rail is noisy.

[Unluckyfett]

Replace the 1uF capacitors C045, 49, 50 & 51. Then check the quality of the audio on pin 14 or 15 of IC2. This is the delayed audio after it has been put back together and filtered but before it is sent into IC2's expander. You can then see what its expanded delayed audio output looks like on the pins of C045.

The 10uF caps are usually the main cause of troubles with the DD-2 and 3 delay pedals but you have changed these but do make sure that you got all of them and that you have them install with the correct polarity.

Sometimes damaged memory can cause glitch or ticking noise to be heard in the delay signal. Generally if you increase or slow down the delay rate length, the glitch repetition rate or duration changes. The only solution to this problem is to replace IC6. It is a 4 bit 64 Kb dynamic RAM and they are a bit hard to find but they can be found.

Also, check the 5 volt power rail. Make sure that it is at 5 volts and that it is clean. Check it with your oscilloscope. C019 and C027 may need to be replaced (both 47uF) if the rail is noisy.

Thank you for the response. I will definitely look into all those points when I get back to my bench. This is my first time working with digital audio conversion, I'm not sure what the expanded delayed audio output should look like on the pins of C045.

If you look at the scope shots I posted,pin 5 of IC8 (output of RA1) looks like a reasonable D-A waveform to me but I'm not sure. The harsh clipping started at pin 7 of IC8.

 I'm definitely not ruling out a bad memory chip because of the glitchy sound. Mine doesn't have the IC6 that's listed on the scheme, it has a boss branded daughter board soldered to header pins where the original IC would go, and a DIP IC. I didn't take a picture but I definitely will Monday.

Thanks again for the response.

[Unluckyfett]

On the topic of the DRAM, would a M5M4464L-10 work for the M5M4464L-12? It looks like to me the 10 is faster memory with 30mw more consumption. My limited knowledge says it should work but I'm not sure if faster memory will work with the circuit designed for slower memory.

Thanks

 

[Slowpoke101]

Regarding the DRAM and the suffixes -12 and -10. Those refer to access speed, which means how fast the memory will access stored data and present it to the data pins in a stable form. The 12 means the data will be available within 120 nano seconds and the 10 means 100 nano seconds. You can use faster memory with no problems at all. Problems occur when you use slower memory but in a device such as this delay pedal which runs rather slowly, you could get away with using far slower memory with no ill effects whatsoever.

Now back to the glitchy noises. The output on pin 7 of IC8 doesn't look very nice at all but when you look at pin 4 of IC3, it's looking fairly good and starting to resemble an audio waveform. But still a bit noisy with a big glitch. If you then have a look at pin 14 or 15 of IC2 it should look far better. But make certain that the +5 volt rail is clean.

When you look at the expanded audio delay waveforms on the pins of capacitor C045 they should look like normal audio waveforms but not as fine as a pure audio signal. They will be slightly coarse due to the bit sampling depth. So it will be a little bit distorted but not badly.

Good luck.

[Unluckyfett]

I changed every electrolytic (less tantalums). While double checking I found a misprint on the silk screen.The polarity of C37 is reversed on the screen so I had it installed backwards. Screen with the scheme is correct. That didn't seem to make a difference. 5V rail looks noisy on the good scope, vs the cheap meter scope, Both posted. For shiggles I changed IC8 (bigger package) but the rail doesn't look any different. Could the noise be the MCU/DRAM drawing on the 5v rail? Clock noise? Maybe it's really minimal and I just don't know what's acceptable?

The DRAM is pictured, I'm curious if this is normal for these?

The last picture is how it currently sits. I think the more I test the more lost I get. So far nothing I've replaced has seemed to change the noise any. I'm really starting to look at the DRAM.

[Slowpoke101]

Very quick question. That 5 volt regulator IC6. Is it original, as installed by Boss or is it there because of a repair? Regardless, what voltage do you read on the two outer pins - meter negative to 0 volts of course. One pin should be about 9 volts or so and the other should be 5 volts. The 5 volt pin should be the one closest to the edge of the board.

[Unluckyfett]

Very quick question. That 5 volt regulator IC6. Is it original, as installed by Boss or is it there because of a repair? Regardless, what voltage do you read on the two outer pins - meter negative to 0 volts of course. One pin should be about 9 volts or so and the other should be 5 volts. The 5 volt pin should be the one closest to the edge of the board.

The 78L05 I replaced with a 7805 to rule out the regulator. 4.6v on outer pin.

[Slowpoke101]

Please put the 78L05 back in, for two reasons. First, a 7805 outer pins are reversed from what a 78L05 is and secondly, a 7805 will be noisy when supplying the small amount of current that this circuit requires. But do check the voltages again when you change back to the 78L05 please.

Edit: Forgot to say, The DIL (DIP) DRAM chip on an adaptor board is a modification option that the manufacturer may use when the offset SIL DRAM package type memory is unavailable. The standard chip works just the same.

[Unluckyfett]

Swapped the 78L05 back in. 4.7v stable on the scope. The new DRAM came in so I swapped that first and it's working perfectly now.

Thanks for all the help. Hopefully this thread will help people in the future.