New dual VCA module troubleshooting

KOSMO MODULES
1

Hey folks,

I’ve been adapting one of the YuSynth/Eddy Bergman designs for a simple hardy VCA and I seem to have hit a wall and I cannot figure out why.

Basically I have built one of the designs from Eddy’s site in the past and I use it quite often, but I wanted to use a TL074 instead of a TL072 and have a dual all in one with indicator lamps so I can see when it’s being triggered (more flashing lights = better right?).

I’ve used basically the same schematic design but adapted it to my own stripboard setup, and after I’ve tested this I’ll whip it into a PCB for anyone who’d like it.

My problem however is that it doesn’t appear to behave the same way as the original, and although I’ve spent all day carefully running over it with far too many coffees I cannot figure it out. I know that Eddy’s design inverts the output by running through another op-amp, but I don’t think that should have anything to do with what I’m doing here?

What happens is that at the point of fading in the input signal to the output, it “pops”. If I use the initial level trimmers to bring it just to the edge, it’s like it’s jumping up a cliff, kind of popping as it passes the threshold, almost like a dry joint might sound.

On my oscilloscope I can see that the waveform is being distorted, and instead of a square wave but at a lower level, I can see it’s jumping in at a certain level, and kind of angling the intial upward edge of the square wave. It’s quite a distortion and sounds bad.

If it were just one side I’d suspect a dud connection or faulty TL074, but I’ve checked all of this over and switched the IC also, both sides act the same way.

Here’s the schematic:

Here’s my stripboard layout:

I’ve also removed the LEDs in case they were somehow interfering with things, but no luck

Any pointers for my tired brain would be great
Jesse

2 Likes
2

Welcome in the forum :slight_smile:

i don’t already look all your stripboard, but to add some led it’s maybe better to add a transistor

something like this on the CV input jack

led indicatrice schema

for your base, if you take the @EddyBergman version, he said they had mistake on the old version

have you used the last one ?

2 Likes
3

Of course :blush: I was being lazy and was going to add something along those lines later on, but in this case I’ve removed the LEDs from the circuit and am still having the problem.

On a side note: the very popular CGS circuits, where does the name cat girl synth come from? Was it a performance name of the author or something?

Also, I have to ask you about your “leech” fuzz circuit in another thread, I was getting almost nothing from it

1 Like
4

CGS it’s the name of a synth projest created by Ken Stone

personally I would not advise building this passive distortion, it’s really not very effective
but you can also ask your question in his thread :wink:

5

I would try setting the CV to some constant voltage.

First - set the CV to zero volts and apply your signal, then adjust all the appropriate pots so that you just hear nothing i.e the VCA is closed.
Then, set your CV to whatever the max should be and check the behaviour of the circuit - is the signal going through cleanly, or is it getting distorted somewhere. There might be some gain/offset adjustments that need to be made. Then, if everything behaves as you expect, you can start applying dynamic CV.
I’ve found that I get some pops on my own VCAs when using a gate signal directly, but a smoother envelope helps alleviate that.

Cheers

3 Likes
6

Heya - apologies for the slow reply, I’m in Australia.

Thanks @jaredical Yes this was my first go-to - I removed CV input and adjusted the initial level trimpot so that the output only was just off with the gain pot down, and then applied CV both initially straight in, and then from my envelope generator.

The problem is at the transition point between off and on, rather than fading in with a slow envelope, it kind of pops and distorts the lower level until fully on. It functions as a VCA, but the transition is poppy and low level is distorted initially. This popping even happens if I just use the gain pot to bring the level in.

So is there a difference with using a TL074 than a TL072 that I don’t know about that would require me to change components?

Like I said I’ve built the single channel version of this with a TL072 and I haven’t faced these problems before, (in fact I went back to it to check and I don’t get these transition pops/distortions) so something about this dual setup is either incorrect, or I’m missing something.

7

Your replies should get here pretty quick seeings as Australia is in the future…

Did you use the other units on the TL074? If not, you should ground the non-inverting input and connect the output to the inverting input on the unused op-amps in the package - maybe something funky there?

3 Likes
8

@jaradical hahaha not really, we’re about 30 years in the past on most issues here…

No the other was with a TL072, the schematic and layout is above for the current build, it’s all 4 channels used of the TL074.

2 Likes
9

i dont know where your problem is but i just checked your stripboard version with the schem and everything looks good.

EDIT :

1 Like
10

@Dud thanks for looking through it - it’s nice to know I’m not going mad!

I’ve definitely run over the board with a fine tooth comb to ensure all the joints are nicely flowed, and also I can see the behaviour is exactly the same on both channels so they’re both identically flawed which points us away from it being a bad joint etc.

I’ve just spent some time with it with the oscilloscope and I can see where the pop comes from - it’s jumping to a negative DC offset with the output when opened, and when closed, it jumps back up again.

So this gets more interesting. Looking at the schematic, to me that would mean that the balancing act between the -12V and +12V over the long tailed pair of BC547 that feed the output op-amp is somehow out.

Am I reading this right?

11

Sorry i couldn’t answer that.
on the other hand I am thinking of something else, and i don’t know if that could do this sort of problem but, have you matched your transistors ?

12

No problem of course I understand!

Yes that’s something I definitely always do (as close as I can).

1 Like
13

It’s not a calibrate adjust to do ?

" Calibrating the VCA:

Before we start, do all the measurements on the ‘Normal Audio Output’, not the AC one. I added the AC output as an afterthought to block any DC offset voltages from coming through but normally you can use the the Normal Output. So connect your probe there for calibrating the circuit.

With trimmer A2 you set the initial bias voltage on the base of transistor Q3. This influences the working of the Gain potmeter and you should set it in such a way that with the Gain potmeter all the way closed the signal is just muted. If you turn Gain up the last played note will then become audible without pressing any keys on the keyboard. So Gain is normally closed.
I found that the audio signal initially starts at 10V and then drops to about 6V. To counteract this you need to open up the Gain potmeter a little and then trim again so the signal is muted with Gain slightly open, This will stop the voltage drop of the signal and keep it at full power all the time. You will see this soon enough if you start testing it and connect this circuit to a scope. It’s easy to counteract and it is in itself not a real problem because you hardly hear the voltage drop but you know, I strive for perfection ^___^
Trimmer A1 is the Balance trimmer. You set it so that the part of the signal that is above the zero Volt line has the same level as the part below the zero Volt line. In other words, you set it so that the signal has the same amplitude in both the positive and the negative part of the wave. This is best done using a triangle or sinewave on the input, together with an oscilloscope.
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."

(on @EddyBergman infos)

2 Likes
14

It shouldn’t be a TL072 vs TL074 thing, they should behave the same way. (In the datasheet they’re all lumped together with the same electrical characteristics.)

image
image713×166 22.4 KB

Have you tried more than one TL074? Preferably from different orders from different sources? Just in case you have a defective one.

How are you doing your transistor matching?

1 Like
15

I also note that there is no buffer at the (audio-) input, so maybe it could be some weird impedance problem? Can you try with a different input module, one that has a buffer at the output? Just guessing :slight_smile:

2 Likes
16

Signal input impedance is 221k, a little higher than the typical 100k but I’d think it’d be fine. CV is 50k, a bit low but also probably fine. That’s without the dodgy LED.

2 Likes
17

How do you calculate that? I only read a mention of a Hybrid-pi Model on electro smash for that purpose, but I did not understand it…

18

For the signal it’s just 220k+1k resistance to ground. The transistor base impedance can probably be assumed to be large compared to 1k but even if it were 0k that would only bring the total down to 220k.

For the CV, with the LED removed, there’s 50k to ground, with the bottom of the attenuator in parallel with 100k–150k to virtual ground which doesn’t matter much.

With the LED in place that’s parallel to a 1k to ~+1 V and that’ll mess up the CV, though I don’t see how it could do anything other than weaken the CV effect.

3 Likes
19

Speculating…
Doesn’t that LED have an almost infinite impedance, up to the point where it start conducting and then it becomes suddenly very low ?
So the CV can “get by” the LED up to the voltage where it start conducting, then it drops to whatever is left once the LED has taken it’s toll.

3 Likes
20

Yeah, a CV less than about a volt would be basically unaffected, while a large CV would lose nearly half.

4 Likes