(1) Project: Man O' War DX powered by OneSpot 12V supply (actual reading: 11.73V)
(2) Issue: Unable to successfully calibrate. I'm stuck on biasing the first MN3005. I get a signal at pin 7 of IC3, however the signal at pins 3 and 4 of IC3 have a high-pitched whine mixed in and rather than getting a delay, the signal repeats and ramps up to a high-pitched oscillation.
I'm getting a very high-pitched, amplified signal mixed with the dry signal at pin 7 of IC2a.
Voltage at pin 1 of IC2 is around 1.2V (build doc has 0.88V) all other voltages before the first BBD are in spec.
(3) Steps to resolve issue: I probed the signal path from the input and confirmed part values. The dry signal gets to IC2 without anything remarkable. Reflowed joints. Removed Q2 and Q3 (NPN Si) and put in sockets. Originally had 2n5088 devices in each spot, same issue and voltages when replaced with 2n3904 devices.
(4) Substitutions: Feedback pot is 25kB instead of 20kB. MN3005 sourced from CabinTech; V571 from SBE.
Pictures attached. IC5 has been removed for calibration purposes.
There was a AD-900 with MN3008 build doc awhile ago on freestompboxes.org. It's, obviously, more complicated than the MN3005 version but still a good read. It assumes an oscilloscope. It still helps to visualize what the signal should look like, even when using an audio probe.
Here's the doc: https://drive.google.com/file/d/1MME5E1GFaOBb-6xN6fMEdc4ghryU8rEh/view (https://drive.google.com/file/d/1MME5E1GFaOBb-6xN6fMEdc4ghryU8rEh/view)
Quote from: tcpoint on December 31, 2019, 10:59:44 AM
There was a AD-900 with MN3008 build doc awhile ago on freestompboxes.org. It's, obviously, more complicated than the MN3005 version but still a good read. It assumes an oscilloscope. It still helps to visualize what the signal should look like, even when using an audio probe.
Here's the doc: https://drive.google.com/file/d/1MME5E1GFaOBb-6xN6fMEdc4ghryU8rEh/view (https://drive.google.com/file/d/1MME5E1GFaOBb-6xN6fMEdc4ghryU8rEh/view)
Thanks, I'll have to give that a look. I'm hoping to get an oscilloscope in the future, but I don't have one yet.
I haven't made a project with this compressor IC before, so I'm a little unsure about how to diagnose issues with it. It seems like that is the place to start, but I'm really at a loss.
Quote from: benny_profane on December 31, 2019, 11:14:42 AM
Quote from: tcpoint on December 31, 2019, 10:59:44 AM
There was a AD-900 with MN3008 build doc awhile ago on freestompboxes.org. It's, obviously, more complicated than the MN3005 version but still a good read. It assumes an oscilloscope. It still helps to visualize what the signal should look like, even when using an audio probe.
Here's the doc: https://drive.google.com/file/d/1MME5E1GFaOBb-6xN6fMEdc4ghryU8rEh/view (https://drive.google.com/file/d/1MME5E1GFaOBb-6xN6fMEdc4ghryU8rEh/view)
Thanks, I'll have to give that a look. I'm hoping to get an oscilloscope in the future, but I don't have one yet.
I haven't made a project with this compressor IC before, so I'm a little unsure about how to diagnose issues with it. It seems like that is the place to start, but I'm really at a loss.
I read through the document, and I don't think I'll be able to use much of that calibration procedure since it does rely so heavily on an oscilloscope.
Could this be caused by an issue with the V571? The output from that IC has the high-pitched signal (so it's not clock noise).
I've had more trouble with companders than BBDs. Here's a couple of articles of interest.
https://electricdruid.net/noise-reduction-with-companders/ (https://electricdruid.net/noise-reduction-with-companders/)
https://www.experimentalistsanonymous.com/diy/Datasheets/SA571%20AN.pdf (https://www.experimentalistsanonymous.com/diy/Datasheets/SA571%20AN.pdf)
Did you take out IC3 and test the signal at pin 7? Actually, testing the emitter on Q2 would be a better test.
Quote from: tcpoint on January 02, 2020, 03:44:45 PM
I've had more trouble with companders than BBDs. Here's a couple of articles of interest.
https://electricdruid.net/noise-reduction-with-companders/ (https://electricdruid.net/noise-reduction-with-companders/)
https://www.experimentalistsanonymous.com/diy/Datasheets/SA571%20AN.pdf (https://www.experimentalistsanonymous.com/diy/Datasheets/SA571%20AN.pdf)
Did you take out IC3 and test the signal at pin 7? Actually, testing the emitter on Q2 would be a better test.
I just took IC3 out and tested the signal at pin7 of the compander as well as the emitter of Q2. The high-pitch signal isn't there, but there is a low hum now. It's introduced at Q1 (2n5088), and is there with the signal up to the output of the compander.
I replaced Q1 (previously a 2n5088) with a 2n3904. The signal now passes fine up until the compander. At pin 7, there's a high-pitched signal with the dry signal very faintly present. The emitter of Q2 is the same.
I think I'm going to try swapping the compander. This seems to be the root of the issue.
What's your feedback set at? Your description sound like run away oscillation.
Quote from: HamSandwich on January 06, 2020, 01:01:45 PM
What's your feedback set at? Your description sound like run away oscillation.
Per the calibration instructions, the feedback, mix, rate, and depth knobs are at min. Delay is at 12 o'clock.
This is all happening at the compander it seems. The signal is fine going in, then turns to a high-pitch whine at the output. All voltages are okay for the compander except for pin 1, which is close to twice as much as what's specified in the doc (i.e., ~1.5v rather than 0.88v).
The first half of the compander chip is being fed with return of the loop of the feedback as well, so I'm not wholly convinced it's impossible for it to be something further down the line.
Does twiddling any of the knobs cause changes in the noise?
It would be great to check another commander chip, but you could try someone radical. Take the chip out, stuff a makeshift jumper from a resistor lead and jumper pins 6 and 7 together and also 14 and 11 together. I think that should remove the compander from the equation and allow you to get signal through. Feedback might be wonky, but if that isn't really the case, turning it all the way down should negate it.
Hope that doesn't blow up your MN's :o
Quote from: HamSandwich on January 06, 2020, 09:17:24 PM
The first half of the compander chip is being fed with return of the loop of the feedback as well, so I'm not wholly convinced it's impossible for it to be something further down the line.
Does twiddling any of the knobs cause changes in the noise?
It would be great to check another commander chip, but you could try someone radical. Take the chip out, stuff a makeshift jumper from a resistor lead and jumper pins 6 and 7 together and also 14 and 11 together. I think that should remove the compander from the equation and allow you to get signal through. Feedback might be wonky, but if that isn't really the case, turning it all the way down should negate it.
Hope that doesn't blow up your MN's :o
I have some replacement compander devices coming in to try out. I'm going to do a full verification of the circuit again to see if I've missed anything in the meantime. I'll give your suggestion a try as well.
If any eagle-eye folks out there want to play from home, I'd really appreciate the extra eyes on this.
I too am having problems calibrating my MOW Dlx.
In saying that I have never successfully built and calibrated an analog delay.
Have tried Aquaboy, total recall and MAW dlx and have not been able to calibrate and get any of them to work.
I have built every other complicated build with success (polychorus, infinitephase, Ada flanger etc.)
Why I can't build an analog delay successfully has got me beat (I don't have access to an oscilloscope either)
Anyway I will keep watching this thread and start troubleshooting my Man o war dlx build as well. Good luck with yours!
Update: Swapped the compander with a new device. Although I was hoping it'd be that easy, the behavior is the same as before.
I'm thinking that some of the slight voltage drifts are more important than I originally thought. I'm going to reverify my components and provide a comprehensive list of voltages.
Power Supply: OneSpot – 11.82v
Transistors
Q1 Si NPN (2n3904)
C 11.53
B 4.49
E 4.04
Q2 Si NPN (2n3904)
C 11.54
B 2.81
E 2.19
Q3 Si NPN (2n3904)
C 11.54
B 4.4
E 3.78
Q4 Si NPN (2n5088)
C 11.54
B 3.75
E 3.16
Q5 2n5088
C 0.511
B 0.511
E 0
Q6 2n5088
C 7.35 (Drains when probed)
B 0.521
E 0
Q7 2n5087
C 8.37
B 11.39
E 11.55
Integrated Circuits
IC1 4558
1 5.71
2 5.71
3 5.7
4 0
5 5.7
6 5.72
7 5.72
8 11.56
IC2 NE570
1 1.07
2 1.85
3 1.85
4 0
5 1.85
6 1.85
7 2.84
8 1.85
9 1.85
10 4.4
11 4.4
12 1.85
13 11.56
14 1.85
15 1.85
16 0.98
IC3 MN3005
1 11.56
2 5.47
3 4.72
4 4.71
5 0
6 5.47
7 5.65
8 0.73
IC4 MN3101
1 10.9
2 5.47
3 0
4 5.47
5 10.85
6 0.945
7 0
8 0.727
IC5 MN3005
1 11.56
2 5.49
3 4.48
4 4.48
5 0
6 5.49
7 5.57
8 0.734
IC6 MN3101
1 10.93
2 5.48
3 0
4 5.49
5 10.88
6 0.935
7 8
8 0.734
These voltages look pretty good to me. There some variation but that's not unexpected.
What about your MN3101? Where did those come from?
Do you have a frequency measurement option on your multimeter? It might help if we can verify the range of you clock frequencies across the min and max delay times.
It's not that unusual to hear a bit of clock frequency in the BBD depending on where the clock trimmers are set, but given that there is no delay output at all the problem must reside in either the BBDs or the MN3101.
Quote from: madbean on January 16, 2020, 07:15:08 PM
These voltages look pretty good to me. There some variation but that's not unexpected.
What about your MN3101? Where did those come from?
Do you have a frequency measurement option on your multimeter? It might help if we can verify the range of you clock frequencies across the min and max delay times.
It's not that unusual to hear a bit of clock frequency in the BBD depending on where the clock trimmers are set, but given that there is no delay output at all the problem must reside in either the BBDs or the MN3101.
Thanks for taking a look at this, Brian. I appreciate it.
Good to know that the voltages look alright. None of the variances seem to indicate anything incorrect with the passives? I've gone over everything multiple times, but I'm not seeing anything wrong there.
I got the MN3101s from Paul at diyguitarpedals.com.au and the BBDs from cabintech.
I don't have a frequency option on my DMM, unfortunately.
There is delay when I probe the BBDs, however, it swells to runaway oscillation and is in no way under control—biasing to a 'clean' delay while probing is not possible. I bought four 3101s, so I can try the others. I don't have backup BBDs..
Should I start swapping the clocks and go from there?
I'm getting the same response with the other MN3101 clocks from the previous order lot. I now have replacements coming in from SBE. I've never had an issue with MN3101 clocks from them before. If that's the issue, I'm hoping the replacements will be the solution.
I bought 10 MN3101 from diyguitarpedals.au around 6 months ago and at least 3 were bad. I feel I still came out ahead because he was selling them so cheap so I didn't complain to him. But a percentage of the MN3101s he's selling are bad.
All the BBDs from Cabintech I have bought have been good ICs.
As a side note, all these ICs are very static sensitive. I hope you're using safe handling techniques with these ICs because they are very easily zapped.
Good luck with your troubleshooting.
Quote from: danfrank on January 23, 2020, 06:13:16 PM
I bought 10 MN3101 from diyguitarpedals.au around 6 months ago and at least 3 were bad. I feel I still came out ahead because he was selling them so cheap so I didn't complain to him. But a percentage of the MN3101s he's selling are bad.
All the BBDs from Cabintech I have bought have been good ICs.
As a side note, all these ICs are very static sensitive. I hope you're using safe handling techniques with these ICs because they are very easily zapped.
Good luck with your troubleshooting.
Cheers. I've got some new devices on the way. I've never had an issue with cabintech either and Paul has been very courteous and more than accommodating when I contacted him—even though I'm not sure it's the clocks at this point.
However, I'm hoping that's the issue because I've gone over this to the point that I've either hypnotized myself into missing something very obvious, or it's a bad IC. On my end, I've been particularly careful—but it's been cold and static electricity certainly becomes more of an issue when the RH drops. I'll report back when I've auditioned the other devices, but I'm taking a break from it for now.
(Seriously though: if anyone has run out of
Where's Waldo books, please tell me I've done something boneheaded with population!
I'm going to complicate matters and say I don't think it's the MN3101, clocks either work or don't and your voltages seem fine and you are getting some delay so they must work, same goes for BBD's they either work, are extremely noisy when damaged or are dead. I think it's your compander output bias, I don't know why but for some people the low bias is fine and for others it causes issues, this isn't the first time that issues come up in a delay debugging thread. So I would adjust the two biasing resistors (can't bring up the schematic right now to get the designations, the two with an electro to ground in the middle) equally. IIRC @ 12v 36k should bias to half supply (33k would be fine if you don't have that value).
Quote from: Scruffie on January 24, 2020, 03:25:04 AM
I'm going to complicate matters and say I don't think it's the MN3101, clocks either work or don't and your voltages seem fine and you are getting some delay so they must work, same goes for BBD's they either work, are extremely noisy when damaged or are dead. I think it's your compander output bias, I don't know why but for some people the low bias is fine and for others it causes issues, this isn't the first time that issues come up in a delay debugging thread. So I would adjust the two biasing resistors (can't bring up the schematic right now to get the designations, the two with an electro to ground in the middle) equally. IIRC @ 12v 36k should bias to half supply (33k would be fine if you don't have that value).
Cheers. Thanks for the suggestion here. I'm more than happy to try this out; though, I'm a little unsure of where exactly you're referring. Would you mind updating this when you're able to reference the schematic? (I'd attach the compander section of the schematic, but attachments are still disabled.)
Between pins 5 & 7 if that helps
Quote from: Scruffie on January 24, 2020, 07:14:27 AM
Between pins 5 & 7 if that helps
Okay, right now pins 5 & 7 both connect to 10k resistors (R17 and R18) to a 10uF cap (C14) to ground. I'll try increasing both of those to 36k and see what happens. Cheers!
Them's the ones! It should raise the voltage on pin 7 so check it's close to 6v after the adjustment.
Quote from: Scruffie on January 24, 2020, 07:23:33 AM
Them's the ones! It should raise the voltage on pin 7 so check it's close to 6v after the adjustment.
Great, good deal. I'm pretty sure I only have 33k and 39k in both 1/8w and 1/4w—so I'll give the 33k a shot and see what happens. Cheers! I appreciate the suggestion.
Quote from: Scruffie on January 24, 2020, 07:23:33 AM
Them's the ones! It should raise the voltage on pin 7 so check it's close to 6v after the adjustment.
By the way, this is a significant increase in voltage on that pin—will BBD biasing also need to be compensated? It seems that a big aspect of this circuit is the low voltage leaving the compander.
BBD bias should be totally separate, although the input may be hotter. See if it works first
before looking at any adjustments but a BBD bias is specific to the chip, not what's surrounding it.
Quote from: Scruffie on January 25, 2020, 05:16:23 AM
BBD bias should be totally separate, although the input may be hotter. See if it works first
before looking at any adjustments but a BBD bias is specific to the chip, not what's surrounding it.
Okay, the build doc mentions that BBD biasing might need to change if those resistor values are altered so I wanted to see if there might be an issue there. Regardless, I'm going to socket those parts and use that as a start.
Three updates:
1) I socketed the replacement clock devices: no change.
2) I removed R17 and R18, installed sockets, and tried 33k resistors instead of 10k. The voltage at pin 7 of the compander increased to ~5.5v, but the same noise occurred when probing pin 7 of the compander as well as pins 3 and 4 of the first BBD.
3) I replaced the MN3005 with a new device: no change.
I think it's safe to rule out the compander, clocks, and BBDs at this point since they've all been replaced with devices from reputable sellers with the same behavior. Any other thoughts on what could be happening here?
Damn... well my final idea without a scope is that I see some 'nichion' caps by the compander, are they from Tayda? Not unheard of to get bad caps from them.
Quote from: Scruffie on January 30, 2020, 05:46:34 AM
Damn... well my final idea without a scope is that I see some 'nichion' caps by the compander, are they from Tayda? Not unheard of to get bad caps from them.
All the passives are from mouser. I've rechecked all the cap values visually and they are correct, should I start swapping the caps around the compander with replacements?
No, if they're from mouser don't worry about it.
Go through the problem one more time and tell me exactly what you hear and where, please.
It should be along the lines of... Q1 Emitter - Dry Signal, IC1 pin-1 - high passed signal, Compander Pin 7 - compressed signal, Q2 emitter - compressed and low passed signal, IC3 balance output - delayed signal (noisy, whiny and crunchy), IC5 balance output - further delayed, Q3 & 4 emitters - low passed delay signal (less noisy, whiny & crunchy), Compander pin 10 - bright clean delay output.
I went back to the board to note the signal at each point you suggested. I'm now getting different behavior. Q1, IC1.1, compander pin 7, and Q2 emitter are all as described. I'm getting signal at IC3.7 and no oscillation as IC3.3/4—but I'm also getting no repeats. Also, the signal at IC3.3/4 sometimes is absent. I have six clocks and four BBDs—each permutation is the same now. I don't know what is going on (I doubt I've fried every part) each replacement has been done without power in the circuit, safe handling, etc. I'm thinking I might build another board up to if even just to have something for comparison. Other than that, I'm truly stumped.
UPDATE: I built a new board and got redundant ICs.
What I learned: This was my first BBD delay and these are a little strange to bias if you're unfamiliar with them. The new build was exhibiting the same behavior as the initial while biasing. I kept moving forward with the process even though I was getting odd results (i.e., clock oscillation/ramping of repeats). The trimmers are all highly interactive, and the default setup was giving me issues. So, I now have two viable builds.
New developments: I'm getting an almost 'ring-mod' sound on the repeats with the delay control close to max. I know that is is a biasing issue, so I'm still trying to dial it in. Is there any advice beyond 'try to get the cleanest delay at each BBD output' without an oscilloscope?
Ring mod sound suggests aliasing noise, a normal part of analog delay in some respects but it suggests either your minimum clock frequency is too low (too long a delay) or your filters aren't cutting the highs as much as they should be.
The cancel trims can help a little with removing this, probe the outputs with no input signal connected when adjusting them.
Quote from: Scruffie on March 13, 2020, 07:40:33 AM
Ring mod sound suggests aliasing noise, a normal part of analog delay in some respects but it suggests either your minimum clock frequency is too low (too long a delay) or your filters aren't cutting the highs as much as they should be.
The cancel trims can help a little with removing this, probe the outputs with no input signal connected when adjusting them.
Pins 3 and 4 of the MN3005 ICs? Thanks for the quick response! I appreciate it.
Not on the BBD output pins, you have to actually probe the output of the trimmer (balance in this case).
Quote from: Scruffie on March 13, 2020, 07:48:05 AM
Not on the BBD output pins, you have to actually probe the output of the trimmer (balance in this case).
Ah, got it. Thanks! I'll report back afterward.
Hey Benny, any luck?
I'm in a similar stuck point on my MOW, signal into the first XVIVE3005 pin7 but only clock or silence out pins3,4
I'm getting so used to probing this circuit, I'm kinda gonna miss it when it's fixed!
Quote from: half_smith on April 19, 2021, 09:31:10 PM
Hey Benny, any luck?
I'm in a similar stuck point on my MOW, signal into the first XVIVE3005 pin7 but only clock or silence out pins3,4
I'm getting so used to probing this circuit, I'm kinda gonna miss it when it's fixed!
I should've posted a final note to this thread and a link to the build report. Scruffie's advice about the clock and the filters is key. Noise is inherent to any BBD circuit. Also, the calibration is very interactive. The MoW seemed to be a little more sensitive to biasing and trim calibrations than some other BBD projects I've built.
Here's a link to the build report. (https://www.madbeanpedals.com/forum/index.php?topic=31052)
I'm not sure if your issues are the same as this one, but I'll respond to your tech help thread.