As I mentioned in a recent post, I’m building my first modules using Eddy Bergman’s excellent website as a guide.
Having 1. validated with the LMNC hive mind that indeed the modules I am trying to build should work together, and 2. come back to the project after a few months with fresh eyes, I realised that the pin-out assumed for the BC547 transistor was backwards on the schematic. After rotating them 180 degrees and triple-checking everything else, the good news is that my VCA module finally produces output.
A couple of (hopefully “final”) questions before I can screw this one into the case and move onto the next one:
Eddy’s website includes the following line for calibrating the module:
For trimmer A3 I advise to use a multi turn one. With this trimmer you trim away any DC Offset voltage on the audio output. Again you absolutely need a scope to do this but a cheap 20 dollar one will do nicely here.
What does “trimming DC offset voltage” actually mean in practice? After attempting to research “DC offset” online I can’t see any asymmetrical peaks on my scope output. What should I be looking for?
On the scope, the output of the module is transposed ~6v:
The same signal taken at the AUDIO_OUT jack (with the gain knob adjusted so that the amplitude of the input and output signals are roughly equal). Note the Vmax and Vmin figures:
Is this normal? If not, are problems 1 and 2 above related? (Adjusting the trim pot I have on the board doesn’t seem to make a difference to the offset).
Any assistance on these points would be appreciated
Note, the signal is 6v and not 10v because my VCO only outputs 6v for triangle and saw waves, and 8v for square waves. I think this is because I’m running it at 12v but I need to trouble-shoot it propertly (question for another thread).
With the 3340 chip, running on 12 volts, the waveforms are all unipolar: 0 to 4 V for triangle, 0 to 8 V for ramp, and 0 to about 10.5 V for square/pulse.
In some 3340 oscillator circuits there is scaling and offsetting of these waveforms before they go to the outputs. In the Digisound schematic, the ramp and pulse waves are just buffered and sent to the outputs while the triangle wave is amplified with a gain of 2. Either way no offset is added. So I would expect 0 to 8 V for both triangle and ramp. Your 6 V and 8 V seemingly are anomalously low.
Anyway, there is a DC offset since the waveform isn’t centered on 0 V. The VCA offset trimmer is probably intended to compensate for offsets added in the VCA circuit, not for offsets present in the original input, but it should alter the offset in any case. The trimmer is ±15 V (or I presume ±12 V in yours) through a 470k resistor in parallel with ground and a 100k resistor; the resulting voltage range at the op amp input pin should be about ±2 V ((12/470k)/(1/100k+1/470k)). That’s then amplified with a gain of 11 (1+100k/10k) for an output range of about ±22 V but of course the op amp output is limited to about ±10.5 V. So it should make a big difference in the output. If yours doesn’t seem to do anything there’s probably a mistake or defect in the build. ← Wrong, see below
If the trimmer doesn’t do anything check whether the trimmer isn’t faulty first! I recently got a faulty multiturn trimmer from a reputable shop that didn’t work at all. Fortunately, I realized this at the breadboarding stage and replaced it swiftly.
I just look and i think it’s not
in the schem and also in the @EddyBergman 's stripboard the Emitter (pin3) of Q1 & Q2 go to R 15 (33K) all seems good, and the same for Q3 & 4
Ah, interesting, thanks. To keep things clean, I’ll create a new thread for the VCO when I get around to looking at it and will paste this in.
So (“naive question”): the “DC Offset” is what I’m seeing on the scope, and the trimmer should simply move the wave in the second image up and down (i.e. you adjust the trim pot until the wave in the same place as the first image)?
Hmm, good point. I may need to order some more in case the trimmer doesn’t survive the desoldering process.
@EddyBergman 's schematic is correct, and the component symbol on my schematic is orientated correctly. What is incorrect on my schematic is the numbering of the pins - leg 1 should be the collector but the emitter is labeled as 1. This meant that the footprint on the PCB silkscreen was backwards which of course, I was blindly following…
Do some diagnostics first. Does a voltage appear on the wiper terminal and does it change when you change the trimmer? Same for TL074 pin 5? What happens at pin 7 with nothing plugged in to input or CV?
Thanks for the response. My findings based on your suggestion:
It does. I see +12v and -12v per the schematic when the pot is centered (slightly circular as I’m inferring “center” from where +/-12v is), and when I turn the pot the voltage on one side decreases and the other increases (e.g. the -12v side measures -4v and the +12v side measures +20v)
Yes. The voltage here ranges from ~8.5v to ~9.4v when the trim pot is turned to either extreme (assuming you mean “TL072” here? )
I see a voltage of approx -3v (+/- ~1v, depending on how the trim pot is set)
I’ve also attempted to measure the input signal at various points in the signal chain in case this helps reveal the culprit (AUDIO INPUT LEVEL and GAIN both set at 12 o’clock):
This doesn’t make sense. There should not be 20 V from any point to ground — ±12 V should be the maximum. You should be measuring from the wiper to ground. I’m guessing you were instead measuring from wiper to the end terminals of the pot.
As mentioned above, with no inputs plugged in the voltage at pin 5 (from pin 5 to ground, that is) ought to vary about ±2.1 V. Then the output (voltage from pin 7 to ground) should go about ±10.5 V.
This is very different from what I expect if it’s the voltage measured from pin 5 to ground and nothing is plugged in on the input.
This too.
Added: I was a little confused before about what’s supposed to happen. But I’ve done a little simulation and it’s saying with no CV or signal plugged in, pin 5 ought to go about 7.9 V to 8.9 V over the trimmer range, and pin 7 should go about -4 V to +1 V. The former’s close to what you said (measuring from pin 5 to ground, right?) but your output swing seems to be too small. That suggests the trimmer is OK but there’s a problem with the op amp or connections to it.
At pin 1 the wave is centered around 0 V which is what is expected assuming you included the input capacitors (C3, C4). But at pin 7 it’s offset below 0 V by about 2.5 V. The trimmer’s supposed to be able to take care of that. Earlier I understood you to say the trimmer had no effect on the output but now it sounds like it does have a small effect. It should have a larger effect.
What does the Balance trimmer do? I’m a little confused about what it’s supposed to do — from what Bergman says it sounds like it changes the DC offset, but then it would be duplicating the effect of the offset trimmer, so maybe that’s not what it’s for.
Indeed I was Apologies, misread your suggestion. When I measure the wiper to ground, the volage at the wiper varies from -12v to +12v, depending on where in the range the multi-turn is.
Re-reading the write-up on @EddyBergman 's page, it contains the following line which may indicate the root of the problem:
You could have an issue where one of the trimmers is turned all the way to its limit because of the lower voltage but I have no data on this so I can’t be sure.
When I turn the BALANCE trimmer all the way to one side (so the 1K trim pot measures 0R between the wiper connected to +12v and R8, and 1K between the wiper and R9), and the OFFSET NULL trimmer all the way to one side (+12v at the wiper), I can almost shift the curve back above the 0v line (images below)
Is solving this problem as simple as changing the values of R8, R9 and/or R23 so that the BALANCE and/or OFFSET NULL trimmers operate over a different range? If so, can anyone suggest new values?
(I’d experiment but having gotten this close to finally having a working module, I’m reluctant to start ripping out resisters and guessing what changes to make in case my doing so inadvertently breaks something else…)
What the module is designed to do is to AC couple the input, centering it on 0 V, and then any new offset introduced by the module can be removed with the trimmer.
What you apparently are trying to do is to take a unipolar signal as input, let the module center it, and then restore it to unipolar on the output. That’s asking more of the trimmer than it’s intended for; also, I expect a trimmer setting that moves a square wave back to unipolar would not be a suitable setting for other waveforms (or, more accurately, waveforms with other amplitudes).
Reducing R23 (the 470k) would increase the trimmer range, but for the above reasons it’s hard to recommend a particular value for it.
If you really must have unipolar output an easier way to accomplish that would be to just get rid of C3 and C4, install jumpers in their place. Then I’d guess the existing trimmer range would be good enough to take out any introduced offset, and might be fine for different amplitude waveforms too.
But unless you have some particular reason for wanting unipolar output you can just leave it as it is and trim for bipolar output centered on 0 V. It’s probably what you want to send to your sound system at the end anyway, and by staying further from the rails is less likely to be clipped or distorted when you don’t want it to be.
Ah, in which case, the module is probably finished then I was working on the (apparently false) assumption that “input waveform == output waveform” and the only expected transformation was the amplitude. I can indeed center the waveform around 0v so no changes to any component values are required.
It still doesn’t answer my original question of what “trimming the DC offset voltage” (i.e. the purpose of trimmer A3) means in practice. Centering the wave involves adjusting both the OFFSET NULL and BALANCE trimmers, which both seem to have a similar (“identical”?) effect on the output. I have the result that I wanted however (“a VCA module that finally works”) so happy to leave this question open for the moment.
Thanks to everyone for your help. In the next few weeks I’ll amend the schematic, add references and attributions, find a suitable open license and then publish everything on GitHub
