Well, the day finally came, I got all the IC's in, MN3001's from Small Bear,
MN3005's from DIYGUITARPARTS.AU, and an SA751N from Pedalhacker...
Plugged it all together and... nothing.
*side note, I'm not sure if it's possible to assemble the pedal as described in the build docs, I followed the drilling template exactly on a 1590BB and the DC Jack is blocked by a diode and capacitor. It's not like that in the included photo which appears to be an earlier version. I started a new enclosure in a bigger box, taking the LEDs and Pots off the pcb.
I've started going through the circuit with my audio probe, and found that the signal gets to Pin 6 of the SA751 (NE570), but there's nothing coming out of Pin 7 (and no signal downstream of that point).
I'm hoping someone can offer some help or suggestions in debugging before I order another IC (and hope that's the issue)
should I go to PedalHackers again?
does anyone have a source for another NE570 equivalent?
Here are my IC voltages (Madbean doc values in [...])
Clearly my Q's have crazy high voltage, and PIN7 of IC2 is way high...Does this indicate a specific issue? What is the next step?
- some of the voltages are not printed in the Doc the page seems to get cut off.
https://madbeanpedals.com/projects/_folders/Delay/pdf/ManOWar.pdf
Q1 (2N5088) IC1 (TL072)
C | 11.85 | [11.71] | 1 | 5.88 | [5.84] |
B | 3.68 | [5.32] | 2 | 5.88 | [5.86] |
E | 4.8 | [5] | 3 | 5.75 | [5.84] |
Q2 (2N5088)
C | 10.4 | [11.71] | 6 | 5.88 | [5.85] |
B | 10.73 | [2.96] | 7 | 5.88 | [5.85] |
E | 11.85 | [2.39] | 8 | 11.85 | [11.71] |
Q3 (2N5088) IC2 (SA571N - PedalHacker)
C | 10.97 | [11.71] | 1 | 0.74 | [0.88] |
B | 10.5 | [6.13] | 2 | 1.8 | [1.77] |
E | 11.84 | [5.55] | 3 | 1.8 | [1.77] |
Q4 (2N5088)
C | 10.29 | [11.71] | 4 | 0 | [0.0] |
B | 10.72 | [5.53] | 5 | 1.19 | [1.77] |
E | 11.85 | [4.94] | 6 | 1.17 | [1.77] |
Q5 (2N5088)
C | 0.5 | [0.472] | 7 | 11.11 | [2.97] |
B | 0.5 | [0.65] | 8 | 1.81 | [1.77] |
E | 0 | [0.0] | 9 | 1.81 | [1.77] |
Q6 (2N5088)
C | 0.5 | [0.472] | 10 | 4.03 | [3.98] |
B | 0.5 | [0.65] | 11 | 4.03 | [3.98] |
E | 0 | [0.0] | 12 | 1.82 | [1.77] |
Q7 (2N5088)
| C | 11.89 | [9.4] | 13 | 11.84 | [11.71] |
B | 11.58 | [N/A] | 14 | 1.8 | [1.77] |
E | 7.07 | [N/A] | 15 | 1.8 | [1.77] |
16 0.8 [0.832]
IC3 MN3005 IC4 MN3101 (Small Bear)
| 1 | 11.89 | [11.71] | 1 | 11.23 | [11.04] | | 2 | 3.4 | [5.54] | 2 | 3.38 | [5.54] | | 3 | 6.72 | [6.04] | 3 | 0 | [0.0] | | 4 | 6.33 | [N/A] | 4 | 7.93 | [5.51] | | 5 | 0 | [N/A] | 5 | 11.16 | [10.81] | | 6 | 7.91 | [N/A] | 6 | 0.5 | [1.7] | | 7 | 6.22 | [N/A] | 7 | 0 | [varies] | | 8 | 0.75 | [N/A] | 8 | 0.75 | [0.745] |
IC5 MN3005 IC6 MN3101 (Small Bear)
| 1 | 11.84 | [11.71] | 1 | 11.20 | [11.11] | | 2 | 5.62 | [5.58] | 2 | 5.62 | [5.54] | | 3 | 4.52 | [6.18] | 3 | 0 | [0.0] | | 4 | 4.51 | [6.16] | 4 | 5.63 | [5.57] | | 5 | 0 | [0.0] | 5 | 11.06 | [11.04] | | 6 | 5.63 | [5.58] | 6 | 0.49 | [1.69] | | 7 | 5.05 | [5.96] | 7 | 7.04 | [9.38] | | 8 | 0.74 | [0.745] | 8 | 0.74 | [0.755] |
Huge thanks to anyone who got this far! |
|
|
|
|
|
|
|
|
What are you using, voltage wise, to power it? Got any pics front and back and did you reflow everything?
Sent from my iPhone using Tapatalk
I'm using an "AMRENCE POWER SUPPLY" (cheap from Amazon), 12v 300ma
My build is kind of funny because I jammed as many expensive parts in there as I could - for fun?
I see a few joints in the pics that look sus, so I'm reflowing and cleaning up the board a little more (since taking the pics)
(https://uploads.tapatalk-cdn.com/20210414/ff9a4e0a40993a4cff42d4fb916e7faa.jpg)
Lots of solder joints that need solder. They should look like little Hershey kisses. Even after almost 10 years, I still have to go back over mine. I've started doing it along the way, but sometimes I still miss some.
Sent from my iPhone using Tapatalk
Thanks for looking at it. I'm a little sloppy sometimes
I have a proper Hershey's on each joint, and while this did have an effect on the noises I can get with the audio probe - louder, fuzzier, clicking... Still no good signal past the SA751 :(
There's a fuzzy helicopter kinda sound anywhere past PIN 6 of the SA751
Why is the signal reaching SA751 (PIN 6) and then not coming out PIN7?
(is it a clue that the voltage on this pin is 2x what it should be)?
It's interesting that the build documents say:
"Use an audio probe to probe pin7 of IC3 to verify that you have signal to the input of the first BBD.
If you do not, check pin7 of IC2a and the emitter of Q2 for output. Debug as necessary"
PIN 7 of IC2a (SA751) must be a critical junction, what does it mean that I'm getting right up to it and not out of it?
I think I just had an "aha!" moment.
I've used 2N5088's for Q1-Q4,. I think it's pinned for 2SC1815, which is a Common Collector E, C, B pinout. The 5088s are Common Base, E, B, C.
How can I use a 5088 in place of a 2SC1815? Is there a trick to cross the legs or something?
whelp, nevermind that idea. The Build Docs says the board is CBE so my 2N5088s are correct. (*read the footnotes)
Would my overzealous use of non-polar Caps be the problem?
I ordered a NE570 to replace the SA571
Have you checked all of your resistor values and cap polarities including the np cap?
Sent from my iPhone using Tapatalk
I have checked as many things as many times as I can.
One thing is the PCB has gotten damaged in a few places from desoldering and I've had to make a couple jumpers.
There's a chance there's a bridge somewhere on the top of the PCB I cannot see because it's under components.
What do you mean check the polarity of the np cap?
Check your polarities and that you used an np cap. Sorry about that.
Sent from my iPhone using Tapatalk
I used the green Bi-Polar caps where ever I had one of the correct value, just playing with the special looking parts in my parts box. Is that ok to do? Do Caps need to be polarized when the schematic calls for it?
Pin7, the output of IC2 (SA571N) is still like double what it should be. How can I diagnose this?
I have found that I do get a blip of good audio on Pin 7 right when I touch the probe to it, then it goes to *click, thrum, click* (and from there throughout the rest of the circuit).
What components would be responsible for high voltage on Pin 7 and a tiny blip of good sound followed by tick tick tick tick...
The sound is good through the output (C5), Q1 (good voltages and sound), and ok voltages and sound out of Pins 5, 2, and 3 of the SA571, someone on Facebook suggested it could be a "DC biasing" issue, what components are suspect?
Thanks!
Are you still getting the out-of-range voltage reading on pin 7? Based on what you've said about desoldering, broken traces, and the need for jumpers, I think that the board itself needs to be verified. Also, do you have any other circuits that use a compander so that you can verify the part?
I'm not sure if the issues I had with my MoW build are happening here--and I'm not convinced that the fixes there are going to help.
(https://www.madbeanpedals.com/forum/index.php?topic=30538.0)
I'm going to do a new build on the current batch of boards I have. I've looked them over this morning and do not see any trace errors. But, it's no big deal to re-verify in case something is amiss. I'll recheck against my recorded voltages and see if anything is up. I'll report back very soon so hang on just a bit.
Thanks Brian! I hate to put you through extra work, it will probably turn out to be something stupid on my part.
However, if you can update the build doc for the MOW (not DX) with the voltages, that page appears to be cut off on the bottom.
*update*
The signal is clean going into pin7 of 1st 3005, the bias1 trimmer directly adjusts the voltage there.
The sound on pins3,4 (1st 3005) is helicopter noise only and changes speed with bias1 trim
pin2 (1st 3005) is very weak on both 3005 and 3101 - this is where I'm focusing my attention rn.
I think I might have killed the other 3101, I swapped them places to see if that did anything and I had 0v on pin2, swapped back and got 0.3v, at least something...
Q2, Q3 voltages look good
-that's as far as I've gone, signal hitting pin7 of 2nd 3005 is same helicopter as output of 1st 3005
Thanks for any further help!
I am determined to get this pedal working!
I'll get the voltages thing fixed up in the doc. Somehow that escaped me. Anyway, I'll be making a recording of voltages on the second build just for cross reference. I think I should be done sometime tomorrow morning.
When I observe those pcb snaps, I wonder, why you don't use the rosin while welding? I noticed it at the most of snaps of many builders here. Without sufficient rosin amount is no reliable welding. Try to add it while re-welding all points, and if you had to weld some points several times, because of lack of proper contact (no wonder), try if that pcb track is not already burned and cracked off. That's a basic thing. Then you can solve anything else.
Most of tin-tubes for welding of course contain inside some kind of rosin. But it mostly isn't enough.
Okay got my second build going and it's working just right. So, that's a relief. I'm going to get the final calibration done, make a record of the new voltages then I'll come back and we'll see about solving your problem.
One thing I remembered during this process is you do need to have IC6 (MN3101) installed to calibrate the bias of the first BBD. This isn't explicitly stated in the build doc, where I instead said "remove IC5 first" when you being calibration of IC3. So, that's something I will add to the doc: it should have been more clear.
Could only find one XVIVE chip so I put two MN3005 in this one.
Thank you Brian! Your build looks perfect, I'm jealous.
Here's the new list of voltages taken from the second build. They seem to be very consistent. The ones in yellow will be different for every build since those come from calibration (in fact I need to go back and take another pass at the bias for IC3 since it looks low).
So, based on this I don't see any issue with the boards. I decided to build the second one since my verification build was done on my original prototype from 2013. The production board is basically the same PCB just made prettier with minor layout adjustments. But, never hurts to be sure. Doesn't preclude the possibility of a single PCB having an issue but in my experience that's very rare. Generally you see several from the same panel have the same fault (like a microtrace connected between pads that shouldn't be there). Anyway, let's go with the assumption that the board was okay for now.
Looking at your reported voltages I'm seeing a lot of problems and some don't make much sense. Some of your transistor readings are nearly at DC (by that I mean close to the power supply level) across all pins. That indicates (1) some very defective parts (2) some suspect soldering or (3) the voltages were taken incorrectly. I think it might be the last one. Your soldering looks pretty good so I'm not concerned with it.
Your emitter readings are consistently DC where your collector and base readings are lower. Your base and emitter readings on Q1 appear transposed. Q2-Q4 are all completely off. The pin7 of the 571 is concerning. I don't know how that could be possible. Even with a faulty solder joint that pin is in an area not even physically close to the power supply traces so such a high reading is another reason I suspect some voltages may have been taken incorrectly. But, I think we can figure that out.
Also, the use of inappropriately sized parts is something that doesn't necessarily make for a bad build but it can make it a frustrating one especially when things go wrong. I'm not saying people shouldn't use what they have or can get because sometimes that's the only option. It's very cool to use some mojo. But, a bit of a headache inducer when you do that for something as complicated as an analog delay instead of fuzz or overdrive.
Anyway, that's a lot of preamble. And, it's not intended to be criticism. Just observation and hopefully a little bit of helpful guidance.
-----------------------------------
With that out of the way let's see what we can figure out. Hopefully we are still at a point this is salvageable. Before we even deal with the actual delay portion of the circuit we need to take another look at some of the transistors. Here are the first steps I suggest:
- Unload IC2 - IC6. Let's go through process of elimination so we are not even worried about the delay part now.
- Measure the voltage readings of pins 2,3 and 7 of where the 571 goes. Just read it right from the socket. You will likely have some readings in the mV region. For example, on mine I had about 68mV for pins 2 and 3 and about 230mV for pin7. What we are looking for here is if you get a very high reading in the V range instead of mV. That might indicate a problem area.
- If that goes okay, disconnect power, and re-insert the 571. Reconnect power. Now check pins 2, 3 and 7 again and compare to my lists. We'll also need the 571 in place to take proper readings of Q2.
- Now re-take the readings on Q1 - Q4 and compare to the two lists I have. It's not drawn on the board but for almost all my transistor parts the square pad is collector (or drain for JFETS). Certainly they are here. All the collectors for Q1 - Q4 should read at the supply which in your case is 11.85v. Base will be lower and emitter lower still. Make sure you have a solid ground connection on your DMM's black lead, too (maybe clip it to the dc jack ground if you can, whatever is convenient). It's really easy to get a wrong reading.
Let's start from there and build out. If we can eliminate these issues in order I think we'll have something. Also, sorry for the missing voltage readings in the doc. I will correct those later on. I just completely missed it because I have little attention to detail sometimes!
Also, try taking the transistor voltage readings from the bottom of the pcb. This usually ends up being easier.
Thank you again, and I really appreciate your help. I was struggling for a week or so.
I appreciate your fixes on the Documents, I did have both 3101's in while attempting to bias.
One thing I might still have wrong, I have my LED's and Pots all wired to the board, I see that your voltages are taken with "no bypass LEDs active" Could there be something happening with the way I have my test build (rock it before you box it) setup?
For clarification on where things stand currently:
The voltages I first posted were truly weird - then I discovered the disconnect between R18 and pin7 of my compander. Once I jumpered that connection the voltages (on the early part of the circuit) mostly came in line with the build docs.
The signal is loud and clear through the NE570 and Q2 and into 1st 3005 (pin7), but output on pin3,4 of 3005 is only helicopter (does react to bias1 trim), everything downstream from there is same helicopter.
Low voltage on pin2 of the 1st 3005/3101, and Q5, Q6 look way low (see next post for current voltages).
Not seeing any reaction on pin2 from any trim pots.
Clearly there's a disconnect or bad chip somewhere in that area. Any idea where to look?
I will continue to scour the board for bridges or bad solders, checking continuity between every component as I go.
Thanks gang, Brian, I am actually enjoying learning through this nightmare, but I do hope it ends!
The voltages as they are now:
Ahh, great. I didn't realize that some of this had already been sorted. So, the good news is I think we know where to go from here. The problem appears to be in the clock itself. There might be another issue with the MN3005 output but let's figure the clock out first. Your voltages on Q5-Q7 are the start of the problem. Below are your voltages as compared to my newest set. Some of the voltages do vary slightly of the turn of the Delay pot (for reference I took these with Delay at max) but your readings are way way off.
----the boring stuff----
The clocks (MN3101) are controlled by an oscillator which is set externally via resistors and capacitors. It reads that oscillator, does some transformations then outputs a 2-phase clock that is read by the BBD which tells it how quickly to pass the audio signal through all its internal transistors (so, the delay time). But, we don't want a fixed delay time b/c that's not useful. We want to vary the clock in a precise way so that the delay time can be controlled. And that's what all that stuff between R45 up to pins 5 and 7 of IC6 (MN3101) is doing. The MOW is more complicated in its clock manipulation than say, a DM-2 or even the DMM, and it also uses a master/slave configuration (excuse the terminology for anyone who objects to that type of description). IC6 (master) sends out the read oscillator frequency to IC4 (slave) so it duplicates the same 2-phase output to drive its BBD (IC3). All this is to point out that a failure along any point in that sequence has a downstream effect which is why we are seeing a completely wrong voltage on IC4 pin2. It's most likely originating either at the delay pot or soon after.
Here's a much more precise and learned version of what I am describing: https://www.diystompboxes.com/smfforum/index.php?topic=99316.0
---what to do---
The places we need to check are
(1) junction of R45/D3 (you should be reading around 8v)
(2) delay wiper (middle lug). At CCW you should also read about 8v. At CW you should have a couple hundred mV.
(3) R47 - this is connected to Q5 CB which you are reading way low. But, check the other side of R47. You should be getting 11.83v there.
(4) Q7 E is connected directly to VC, but you are reading 5.54v. So that's definitely a problem.
I've attached another image that might be helpful (traces diagram). Everything connected to the fat red trace should read VC which is 11.83v for you. The green highlighted pads should all read about 8v. Use this as a guide as well. If you are not getting the correct readings in either of those then I suggest using the continuity checker on your multimeter to make sure they are physically connected. For instance check that R29 has continuity with Q7 emitter, or R53 or even pin1 of IC5. Do continuity checks with the circuit un-powered.
Thank you for this treasure map. I needed the explanation of the MN3101s to follow the interactions, that helps a lot.
I found that Q6 was OL between B, C legs, I replaced it and immediately got about half way home.
I have about 2v (up from .2) on pin2 on IC3 IC4 and there is audible signal (with repeats) coming out of the 3005s! That is a big relief,
but should be 5.5v
However, the signal is muffled and there's a pretty big 120hz hum.
A few voltages are still lower than they should be (and a few higher), but I'm starting to believe.
I've tested continuity everywhere, and measured the resistors in place...
Maybe next I need to swap some of the capacitors?
Here's an update on my voltages, if that might help narrow down where else to look.
*maybe unrelated to the bigger issue here, but I noticed that Q1 has a lower Voltage on B than E pin (see attachment), that should never be the case right?
Hey guys, thanks for sticking with me!
I have literally replaced most of the components and tried socketing various resistors to help bias, not really helped.
I got another MN3101 today, (finally arrived from DIYGUITARPARTS.AU), popped it into IC4 position and now my Voltages are much better all around, not perfect but nothing really weird.
So far my issues have been a bad Transistor and a bad clock IC... ugh, how many hours of looking at solders and values to no avail...
The problem now, the signal out of IC3 (pins 3, 4) has a high pitched squeel and no repeats?
I can almost tune the squeel out with the bias trim and clock trim, but the repeats are nowhere to be found.
The pedal line out has a relatively clean sound, no repeats :/
Any clue what to look at next?
Thanks again! Future generations will remember your service!
You're a better man than me, I totally would've started over by now. Nicely done.
Sent from my iPhone using Tapatalk
Haha thanks Jimilee.
I don't know how, but I think I have it working! I just swapped out the ICs again, different NE570, different 3101, different places, and started fiddling the trims by instinct... the probe inside the circuit is whiny, noisey, or no delay, but the OUTPUT SOUNDS GOOD.
I'm going to mess with it a little more, but I think it might be ready to go in the box. holy cow what a journey.
I'll report back, or share my Build Report in the next couple days.
Thanks for the support and guidance :)
Quote from: half_smith on April 27, 2021, 09:05:56 PM
Haha thanks Jimilee.
I don't know how, but I think I have it working! I just swapped out the ICs again, different NE570, different 3101, different places, and started fiddling the trims by instinct... the probe inside the circuit is whiny, noisey, or no delay, but the OUTPUT SOUNDS GOOD.
I'm going to mess with it a little more, but I think it might be ready to go in the box. holy cow what a journey.
I'll report back, or share my Build Report in the next couple days.
Thanks for the support and guidance :)
Good on you. I went through some of this with the DMM pedal before I started over. While building the second I realized where I went wrong on the first one.
Sent from my iPhone using Tapatalk
Yeah, it's not quite right. I'm getting 1 nice sounding repeat, but the T1 trim or FDBK pot don't give me more repeats.
It's a huge relief to have anything coming out, but only a single repeat is kind of lame.
Any help with this?
My advice now: if you made proper welding with sufficient amount of tin and rosin, clear now ALL PCB points around with sharp needle from dirt from welding. I found there problematic points as I showed below (red pointers). Do it at both sides of PCB for all points. Maybe your problems will dissappear. Good welding and clarity is important. And PCB controll from crackles.
I managed to think clearly on this, and since I had 1 single repeat and no feedback, I traced the feedback loop and found that the same joint I had jumpered earlier (R17-IC2p5), probably my damage to the pad, needed a jumper to C37 to connect the FDBK loop.
Did that and it totally works and kicks major ass. After tweaking on the trims for a while (I found that bal1 can be set for full self-oscillation btw)
It's good. It's going in the box.
Brian, thanks for your help, I didn't mean to assume the board was faulty, I was lost in hell.
In the end I had to make 4 jumpers due to somehow losing traces, I will assume due to my roughness, and had to swap out 3 questionable components. All in all made this a real study, and I actually think I understand this circuit now and that was my ultimate goal anyway.
I'll post a build report eventually eventually eventually eventually
Wow man, nice job. I'm super impressed.
Sent from my iPhone using Tapatalk
So, I have been rocking the Man O' War for a few weeks, it's pretty cool, but there's a bit of "fuzz" or pfff on the first transient, the repeats are very clean. Any idea how I can clean up the initial note?
Try to play around Bias and Balance pots. Maybe too high or unbalanced values are going to output compressor NE570.