I finally finished up my Aion Blueshift last night. The good news: It makes sound. The bad news: I think something wrong with the chorus path through IC9 (the lower MN3207 on the schematic). The trim pot didn't do a lot there when moving it like the trim pot for the IC3 path. There's also some warble in the effect. Through my time on youtube, the whole thing with the Dimension C was a lack of warble from the two seperate delay paths. Maybe the sound I'm hearing now is correct (it still sounds massive and lush), but the effect of the trimmer made me think there may be something else at play.
Here's a quick video of the effect I put together this morning. The order of the effect is switch #2, switch #1, no switch, switch #3. If you've built this effect, I'd be interested to hear whether or not mine sounds like yours or if something is amiss
I does sound a bit too cyclical (I think that's a word). I haven't been around mine in ages (I really gotta get me second boards built) but it sounds like there's too much of a standard chorus sound.
I agree with Gordo. I think you have identified the problem, there is only chorus path working. Maybe you can try swapping the position of each 3207 to see if the "no bias" thing happens at the other trim after switching the chips?
I've taken the entire pedal back apart. One of the things I was trying to get away with were some jack holes that I drilled not too well. I also removed all of the switches as they weren't attached the best; I noticed when putting the pedal through a fuzz that there was no noticeable change in sound with the middle and right switches. I'm going to re-flow all of the switches, and will also try swapping around the chips. I also think once I solder the switches in place, I'll wire the entire thing outside of the enclosure so I can test it better. I'm normally a rock before box guy, but this pedal makes that a lot harder :(.
Thanks for the replies! I'm glad I wasn't going crazy with thinking the sound wasn't quite right yet. I have another project in flight right now for a friend for Christmas. Once I get that knocked out, I'll get back to this.
Agreed, this one needs to be assembled in the box (at least the top PCB). Theres no good way to get the switches and the standoff all lined up without doing that.
Ok, I removed all the switches, removed all the wires, and re etched/painted/drilled/sanded an enclsore that fits the parts much better. I then took way more time wiring everything up and connecting the toggle switches (I didn't have those 3 nuts to match the length of the offset before; now it's so much easier), and tried it out.
Same deal. Some sort of pulsed modulation. No noticeable change in the sound with switches 3 and 4. Below are the things I have looked at.
* Audio probe pin 7/8 of IC3 and IC9 (the BBD). I am getting a strong modulated signal there.
* Swap IC3 and IC9. No change.
* Audio probe the two test points. Strong modulated signal, meaning it's making it's way back up to the signal board.
My next steps are to understand what the Compounder chip does, and follow that signal around until I get to the output jacks. I have all of the voltages from the Clock/BBD board but I'm not quite sure they're useful yet.
How are the jumpers set for the bypass mode?
They're set with the standard bypass mode, but I'll verify when I get home this evening.
I had the bypass jumpers on the components part of the board. I think it was wired correctly, but just to be sure, I re did the them to match the traces on the bottom of the board (so face up). The effect is definitely wired in the stock bypass mode. There's still warble that I think shouldn't be there.
I did a quick video again. This time I used a looper pedal. The order is bypassed, then mode 1, 2, 3, the bypassed again. The good news is that all of the switches seem to behave now! When I pass a test signal through it, I can definitely hear a difference in the effect.
Just spitballing here, I'm not an expert. Since you have one BBD that works and one that doesn't, maybe try measuring the voltages on both to compare. Might lead you in the right direction. Also, If your DMM had a frequency counter try measuring the clock pins on both BBDs.
I'll have to check my DMM so see if it has a frequency option. I measured voltages from every chip that's part of the clock and delay board. Since this circuit has 2 identical subunits, I wanted to see if there was a chip in particular acting up. Everything that should match did match.
Also, I think at this point, I need to hook it up to a good oscilloscope. I assume that TP1 and TP2, used for biasing, will have inverted copies of the same wave shape. If something appears amiss there, it could explain what's going on. Also, I assume that the problem is not in the creation of the chorus sound but the way they're combined back together. I'm not quite sure what role the 2nd compounder has in the circuit but I'll read through the data sheet to understand it tomorrow night.
After some delay, here are the voltages off of my ICs. The voltages on the control board (so the OPA2134s and especially IC2 NE570) look kind of odd to me.
IC1
1. 4.68
2. 4.68
3. 4.29
4. 0
5. 4.67
6. 4.68
7. 4.68
8. 9.37
IC2
1. 1.25
2. 1.86
3. 1.86
4. 0
5. 1.86
6. 1.86
7. 2.97
8. 1.86
9. 1.86
10. 1.86
11. 1.86
12. 1.86
13. 9.37
14. 1.86
15. 1.86
16. 0.76
IC3
1. 0
2. 3.14 - 3.18
3. 3.38
4. 6.12 - 6.14
5. 6.55
6. 3.11 - 3.16
7. 4.69 - 4.70
8. 4.69 - 4.70
IC4
1. 6.56
2. 3.11 - 3.16
3. 0
4. 3.13 - 3.18
5. 4.7 - 5.8
6. 0.7 - 1.7
7. 3.8 - 4.9
8. 6.14
IC5
1. 3.6 - 4.9
2. 1.0 - 1.5
3. 1.2 - 1.9
4. 0
5. 3.3
6. 3.3
7. 2.6 - 3.9
8. 6.34
IC6
1. 2.6 - 3.9
2. 3.36
3. 2.1 - 4.5
4. 0
5. 3.36
6. 1.0 - 1.5
7. 3.8 - 4.9
8. 6.34
IC7
1. 0.5 - 6.0
2. 3.32
3. 2.5 - 4.2
4. 0
5. 3.35
6. 3.36
7. 2.7 - 4.1
8. 6.71
IC8
1. 6.56
2. 3.13
3. 0
4. 3.13
5. 4.3 - 5.8
6. 0.7 - 1.8
7. 3.7 - 5.0
8. 6.1
IC9
1. 0
2. 3.15
3. 3.38
4. 6.12
5. 6.54
6. 3.13
7. 4.80
8. 4.80
IC10
1.
2.
3.
4.
5.
6.
7.
8.
IC11
1. 1.27
2. 1.87
3. 1.87
4. 0
5. 1.87
6. 3.11
7. 3.11
8. 1.87
9. 1.87
10. 3.09
11. 3.09
12. 1.87
13. 9.37
14. 1.87
15. 1.87
16. 1.27
IC12
1. 4.68
2. 4.68
3. 4.68
4. 0
5. 4.68
6. 4.68
7. 4.68
8. 9.37
I'm on vacation until Monday. If you still need some voltages measured, I'll be happy to do that on Tuesday. A week away from my soldering iron and I'm jonesin' a bit.
I don't own a proper multimeter. I own one of those diy $20 kits. However, I decided to give it a shot with the signal generator from my testing rig from JMK. After some tweaking around with trimpots while watching the scope, I found the oscillation had a notable drop. I then re measured pin 3 of both BBD and the voltage was 4.1V. I plan to hook this up to a proper scope and input the signal from the build doc, but I'm starting to feel that the 3.4V suggested starting voltage must be too low for the parts I sourced. I know the build doc said it may go as high as 3.8V. I feel my parts may be in the higher voltage.
I spent last evening watching some Waza Dimension C videos. Mine definitely isn't that motionless.
You can download a signal generator on your phone, I use Function Generator PRO. You need to make a 1/4 inch adaptor. If you have kids, there's likely a broken pair of headphones around that you can use ;D
I'm not sure from your post if you have a scope or not. If you do, you don't need to measure the voltage to bias the BBD, the scope is better. For multimeter I'd suggest VC97, it's ~$30 on Amazon, it has a frequency counter.
I have quite a few multimeters and a cheap diy scope. I need to run the 8kHz signal through it and tune it from there. The signal I mentioned using earlier was in the 440Hz range. I've used that function generator to build my Lab series preamp. I don't know how to make the signal generator output a signal that's 2V peak to peak.
I may check out the VC97, I don't have a frequency counter.
The app I mentioned above will output a a waveform at any frequency, you can adjust the volume of the waveform in the app. You will measure the input jack of the pedal with your scope and adjust the volume in the app until you see the 2V p-P you are looking for. Then you can look at the output pins of the BBD and adjust the bias.
I passed an 8kHz waveform into the pedal at -20.0dB (2V) and used my super cheap DYI kit to make the signal equal. However, it's still pulsing way too much. I'm going over to a friend's house who can let me use their considerably better oscilloscope, and we'll see if there's any difference there once I calibrate it with something more professional.
I was able to run some tests through a Rigol 1054Z at work today. I discovered that the trim pot settings I dialed in with my cheap DYI scope were nearly spot on. The wave shape looks the same on the top and the bottom. I did notice that Aion's wave picture looks more rounded at the bottom; mine looks steeper.
It also may be a function of me not knowing how to dial in the scope correctly, but when I probed both testing points, one wave was still and the other was constantly moving. I had expected to see the waves as inverses of one and other.
The bad news is that I still have way too much motion in the effect. I've moved chips around on the smaller board to rule out any issue there. I'm going to begin looking at the compounder ICs I think next.
Can someone tell me if the output of the 2nd compounder is supposed to be motionless? I think you could audio probe it. I'm trying to find the place where they WO waves are combined back together. I assume it's there.
I am finally getting around to troubleshooting a few troublesome builds. I ordered 2 new CoolAudio 3207s and 3102s from Smallbear a few weeks back and decided to change both of the delay parts of the circuit to see if I could find the cause of the issue. The results were...not helpful.
Now, when I put my oscilliscope on the testing points, both waves look like this:
(https://i.imgur.com/zNCRG8U.png)
I am still noticing IC3 has considerably less impact on the sound, so I may try changing out the Opamp that drives it.
I also noticed what I can best describe as a motorboating sound when I passed the 8000kHz signal through the circuit for testing. I'm not sure if that's expected.
If anyone has a complete set of voltages for the various ICs and transistors I would be very interested at this point. This circuit is so complex and difficult to troubleshoot due to the dual board nature that coming up with places to begin investigation has proven to be hard.
IC2 (NE570 lower right in Blueshift)
1: 1.03V
2: 1.84V
3: 1.84V
4: 0
5: 1.84V
6: 1.84V
7: 3.11V
8: 1.84V
9: 1.84V
10: 3.03V
11: 3.03V
12: 1.84V
13: 9.86V
14: 1.84V
15: 1.84V
16: 0.52V
IC11 (NE570 top center in Blueshift)
1: 0.95V
2: 1.84V
3: 1.84V
4: 0V
5: 1.84V
6: 3.02V
7: 3.02V
8: 1.84V
9: 1.84V
10: 3.03V
11: 3.03V
12: 1.84V
13: 9.86V
14: 1.84V
15: 1.84V
16: 0.91V
Note: Pin voltages of clock and BBD will vary by bias voltage (MN3207 pin 3) so these aren't necessarily good targets, it just matters that the BBD is properly biased.
"~" denotes oscillating voltage.
IC3 (MN3207)
1: 0
2: 3.00V ~ 3.10V
3: 3.62V
4: 5.90V ~ 5.95V
5: 6.32V ~ 6.37V
6: 2.94V ~ 3.08V
7: 4.30V ~ 4.34V
8: 4.30V ~ 4.34V
IC4 (MN3102)
1: 6.32V ~ 6.37V
2: 2.94V ~ 3.08V
3: 0V
4: 3V ~ 3.1V
5: 4.62V ~ 5.65V
6: 0.74V ~ 1.74V (no connection to PCB)
7: 3.83V ~ 4.34V
8: 5.90V ~ 5.95V
IC8 (MN3102)
1: 6.31V ~ 6.36V
2: 2.94V ~ 3.10V
3: 0V
4: 2.95V ~ 3.10V
5: 4.19V ~ 5.55V
6: 0.85V ~ 2.09V
7: 3.66V ~ 4.14V
8: 5.88V ~ 5.93V
IC9 (MN3207)
1: 0V
2: 2.95V ~ 3.10V
3: 3.39V
4: 5.88V ~ 5.93V
5: 6.30V ~ 6.36V
6: 2.94V ~ 3.10V
7: 3.97V ~ 4.02V
8: 3.97V ~ 4.02V
The other ICs are just normal op-amps, so they should typically be around 4.8V (1/2 supply voltage) +/- 0.5V for all pins except 4 (0V) and 8 (9.6V).
Thanks Kevin! The good news is that my IC2 and IC11 look like they're in the ballpark of what you have here. I think I was using a cheaper power supply through my testing rig and had slightly higher values but they are all relative to yours. However, since I changed out IC2, 3, 8, and 9, I'll need to re-measure those.
I do have a question for you about the optimal wave pattern of the two delays. When comparing the wave shapes from TP1 and TP2, if I put them on a scope at the same time, should they have some relatable shape to one and other? When I did my initial testing I saw that they were totally off from one and other but I'm starting to think that I had a bad clock/BBD chip in there. Now that I am relatively confident they are both good I'll test again.
They should look very similar to each other, though not at the same time if you were to take a snapshot (the phase is inverted from each other).
Ok, the inverted phase was what I was hoping to hear. I'll do some more advanced scope testing once I go through all of my new ICs.
I'll dig out my notes and compare voltages. I noted EVERYTHING before I boxed it. I used my crappy little o'scope kit and found that I'd pretty much nailed it by ear. I seemed to have a decent sized sweet spot right at about the middle of the trimmers.
My PC is also back together so I'll see if I can't record a few snippets. Haven't tried it stereo yet.
More voltage measurements over the weekend: The ICs Kevin listed above pretty much match what I just measured. I think if Gordo can post some sound samples I can confirm whether or not what I'm hearing is what I should expect.
I also just orders a Rigol 1054Z Oscilloscope. I spent quite a bit of the weekend reading through the topology of BBD and Clock design in circuits and plan to use those 4 channels to ensure all of that is working as designed. I needed one anyway to do the Lola Phase setup (DeadEndFX's Mutron Biphase project) and this seemed like a good excuse to pull the trigger.
Excuse the lame playing but I wanted to keep it simple so you could hear the effect. This is a Music Man Silhouette Special on the middle and neck pickups into the the Blueshift into a Scarlett 2i2 (1st gen) into Amplitube (the default American Tube Clean 1) dry.
https://www.gordomusic.com/Madbean/Blueshift.mp3 (https://www.gordomusic.com/Madbean/Blueshift.mp3)
I paused between switch settings and they are as follows:
Dry
All switches down
Switch 1 up
Switch 2 up
Switch 3 up
Switch 1 and 2 up
Switch 2 and 3 up
Switch 1, 2, and 3 up
I realized now that I didn't get just 1 and 3 up but you get the point. The differences are subtle, more so in a clean setting like this.
Here's 1&3 a little bit gainier. I pulled down the top end a bit as the MM tends to be a bunch on the bright side.
https://www.gordomusic.com/Madbean/Blueshift2.mp3 (https://www.gordomusic.com/Madbean/Blueshift2.mp3)
Ok, that sounds like it has much motion as mine does now!
By chance could you just pass some sort of sine wave into it if you have a function generator? I notice the movement a lot more when I'm not using an instrument.
BTW, your effect sounds really good. It's got move more movement than the new Waza Dimension C which I think is a good thing.
Here's a mono version of the old original.\:
https://youtu.be/gK8rZHKwizQ (https://youtu.be/gK8rZHKwizQ)
I agree though, at least in the Mike Hermans video the Waza version is more flanger sounding with a lot less movement. Maybe a shorter delay time?
Quote from: gordo on March 02, 2020, 11:29:32 PM
I agree though, at least in the Mike Hermans video the Waza version is more flanger sounding with a lot less movement. Maybe a shorter delay time?
Maybe this is actually a disadvantage for me, but I have a friend who's got an original DM-2 who works at a store that sells Waza stuff. So I was always comparing the motion of my build against those two pedals. That said, that demo you shared of the original has motion more similar to mine (more pulsing sounds as you select higher modes) while my friend's old DM-2 sounds more like the Waza version.
NOD (new oscilloscope day) is tomorrow for me though so I plan to take a LOT of screenshots on both of the delay sections. I hope to use it to determine if there is something awry with mine or if it works correctly.
Thanks for all the help!!
I think the pulsing and "motion" are separate phenomena. In the Blueshift, the delays are modulated by opposite LFO signals, but there is a point in the modulation cycle where they line up. The result is a chorus that pulses. This is different from motion where you perceive a distinct pitch shifting up and down from a single modulation direction (a woozy, uneasy, seasick chorus effect). That is my understanding anyway.
I am starting to feel the same way about pulsing and "motion." This weekend I plan to put this thing through the paces with an oscilloscope, measuring the LFO, clock and delay signals, and comparing the two delay paths to see how much they are perfectly out of phase so I'll be able to back it up with some hard data.
A quick aside because I spent quite a long time staring at this schematic. Are the opposite LFO signals created when IC5 and IC6 ? It appears for the 'top' delay IC6A goes into IC6B. In the bottom delay IC5B goes into IC5A (so reversed). That was the only difference I could find in the between the two. Is there a connection difference I'm missing or is that what causes the phase difference?
IC6A is inverting, so the LFO gets flipped 180 for that channel, that's where it happens.
Here's shots of my scope:
This is the input signal. 8K at 2v p-p
(https://www.gordomusic.com/Madbean/BSIN.jpg)
This is TP1 biased
(https://www.gordomusic.com/Madbean/TP1.jpg)
This is TP2 biased
(https://www.gordomusic.com/Madbean/TP2.jpg)
And this is the output with all 3 switches up, variable waveform so just a snap of what I could get.
(https://www.gordomusic.com/Madbean/BSOUT.jpg)
Thanks to everyone who's commented so far. I bought a proper oscilloscope (A Rigol1054Z that I unlocked with all the options + 100 mHz bandwidth). On a side note, if you are becoming serious about building (and at this point, I would define myself as a serious hobbyist), make the investment in a Scope. It answers so many questions you may have.
Now onto my particular investigation. Using the scope (which generates a function signal itself), I was able to shape the waveforms coming out of TP1 and TP2 per the instructions. So good so far. I wanted to then put both TPs on the scope at the same time. However, I was unable to do this using standard edge triggering. One of the signals would remain stable while the other would never trigger. I spoke with an electrical engineering friend who thought the frequency of the signal was changing. And what do you know, I think he was right.
* The frequency on TP1 and TP2 both oscillate (from 7.95mHz to 8.06mHz)
* The voltage on TP1 and TP2 both oscillate (from 2.65V to 2.85V)
So looking through the scope triggering options, I found a method called delay triggering, which triggers if the 1st and 2nd signal are within some proximity of one and other in terms of shape. I chose to trigger for anything less than 1us. I then alternated between triggering the directions in opposite directions (out of phase) and the same direction (in phase). Both would trigger, then jump around, then trigger, then jump around. To me, this picture describes what I was seeing:
(https://i.imgur.com/Vua3Qd1.png)
There's something Gordo posted in his initial build where he mentioned that the pedal has a "vague modulation that doesn't really seem to start or stop." This sort of describes that. I still think my effect has a little more movement than the sound samples Gordo posted, but at this point I think the overall way my effect is working seems correct and I'm going to box it and call it good. However, if that voltage/frequency oscillation seems wrong, let me know because I'll have to rethink my hypothesis.