3086 Tremolo (needs some help)

[soggybag]

This is a tremolo from the old stompboxology news letter. I have the whole thing built. I'm getting audio when the effect is engaged. I hear a bit of a boost. So it seems to be working but it isn't modulating.

I checked IC1 (TL064) with the multimeter and I don't see any fluctuations around pins 8, 9, and 10.

I copied the schematic into Eagle from the original and had a PCB made. The original schematic was a little confusing. I may have made a mistake here. Below are the original and my schematic.

There seemed to be a mistake in the original schematic where the depth pot was missing. I added this into my schematic after SW1. 





Since I'm getting audio my guess is that the control voltage is not reaching the control pin of IC3 (pin 3 LM3086).

Poking around with the multimeter it seems like the oscillator is not oscillating (IC1C, TL064 pins 8, 9, and 10).

Here are the voltages that I read from IC1.

1.   0.77
2.   2.14
3.   0.81
4.   9.50
5.   4.71
6.   4.73
7.   4.73
8.   0.00
9.   0.80
10. 0.68
11. 0.00
12. 0.81
13. 8.58
14. 0.70

 

[anotherjim]

IC1a pin 3 should be fed from the preamp IC1b pin7. Node B on the original. Brave to have gone straight for a PCB. You may have to bend out IC1a pin3 so you can run a wire to the right spot.
Could be other problems but this one stood out.

[soggybag]

Thanks! You're right!

Looks like IC1d (pin 12) needs to go to 1/2V (B) also.

[soggybag]

Thanks Jim! You were spot on! I added a jumper from pin 3 to 7 of IC 1 and it's working!

[soggybag]

Here is the finished build. Note the jumper added on the back not hard to do. It's the story of my life one connection shy of a complete circuit.

I have a few of these boards. If anyone wants one DM me and I'll mail you one. I got these done inexpensively at a service: PCBWay.com

[duck_arse]

the opamp designers manual tells us to delete C14 as bad practice. but carry on!

[soggybag]

the opamp designers manual tells us to delete C14 as bad practice. but carry on!

Thanks for the tip. This is pretty easy to test.

Is this because the op-amp is a DC buffer connected to the bases of the three transistors which are high impedance?

[PRR]

> Is this because the op-amp is a...

No. Don't hang capacitors to ground on opamp outputs.

We've all seen sign language (seems to have got faddish with the pandemic). Nimble fingers. Now tie a 5 pound weight to each finger. The signer will "mumble", "slur her words", or go "spastic" from the strain.

Me, I don't quite see why that plan needs a dual opamp Vref buffer. And 10uFd is enough to upset the opamp, not enough to force Vref to be STABLE. Two Rs and a fat C ought to be a fine Vref.

[soggybag]

> Is this because the op-amp is a...

No. Don't hang capacitors to ground on opamp outputs.

We've all seen sign language (seems to have got faddish with the pandemic). Nimble fingers. Now tie a 5 pound weight to each finger. The signer will "mumble", "slur her words", or go "spastic" from the strain.

Me, I don't quite see why that plan needs a dual opamp Vref buffer. And 10uFd is enough to upset the opamp, not enough to force Vref to be STABLE. Two Rs and a fat C ought to be a fine Vref.

This came from the old Stompboxology new letter. Most of the projects from these seem a little over engineered.

Sounds like you're suggesting I could drop the single op-amp and use IC1-B in its place along with a resistor and cap voltage divider for the 1/2v ref.

[duck_arse]

you could delete R26, C13, C14 and IC2, circle C becomes circle B, and carry on.

[soggybag]

Thanks for the help! This is a great improvement it gets rid of a whole chip and related parts which simplifies things.

Here is an updated Schematic:

[soggybag]

I built two of these. Using the schematics below. Has the extra bias buffer the other doesn't Removing the extra bias buffer and C14 leaves me with a ticking/thumping noise. The circuit with the extra buffer doesn't have this sound.

In the circuit without the extra buffer I can see a distinct triangle at the collector of Q6. Which seems where the sound is coming from. The trim pot R33 reduces this triangle as I turn the resistance to 0. But even at 0 there is still significant ticking.

Another question that cam up was the substrate. I had read that this should grounded. But this circuit leaves pins 12, 13, and 14 of the LM3086 un connected. I saw this extra transistor and wondered if it could used in place of Q1 or Q6 to save a part?

Here are the two schematics. The original:



The modified schematic. Here I have removed IC2, R26, C13, and C14.

[soggybag]

If anyone is interested in building one of these. I have some extra PCBs. I ordered 10 of each and they sent me 11. The cost from PCBWay.com was about $1 each. I have 22 boards and I've built 3 and gave a couple away.

I'm going to do an updated board when resolve these issues and order an updated batch. Message me with your address and I'll mail you one or both.

The red one is the original circuit as described in Stompboxology. This one has a mistake that requires a jumper. The white one is modified to remove the single op-amp but suffers from the ticking/thumping noise.

[soggybag]

Anyone want to weigh in on why removing the extra op-amp causing LFO noise here?

Problem: Take a look at the two schematics I posted above. The first uses an extra op-amp to buffer the bias voltage. This seems redundant but removing introduces LFO ticking/thumping noise.

[anotherjim]

I have a suspicion the LFO noise is getting into the signal path via LM3086 pins 3 & 4.
The remaining buffer opamp shares the same IC package as the LFO. The extra TL071 buffered version benefits from an input filter R26/C13 and a separate IC package from the LFO.
It's a bit of a mess, the LFO and the LM3086 Vref supplies are better served separately.
Looks like the designer wanted a very clean and low AC source impedance for the Vref "C" feeding the LM3086 Q1 & Q2 bases via the 1k resistors, hence the extra TL071 with a cap on its output.

[soggybag]

Thanks, that makes sense.

It looks like you should be able to use the same source for all of the bias voltages. If I measure the probe the bias voltage "C" at the op-amp with the scope it, it doesn't show any fluctuations.

Would it help to put R26 and C13 between "C" and R21 and R23?

[duck_arse]

I suppose it's too late now, but, I wonder if you dropped the second buffering oppie, and connected the oscillator section "B" bias(s) to the R26 100k, without the electro, whether that would allow the osc to waggle its bias at a distance from the rest of the audio section.

[soggybag]

That's an idea. I have a few extra boards I'll see if I can implement it.

Another question. What about the substrate? Looks like original design leaves pins 12, 13, and 14 unconnected. I seem to remember reading one of those pins is substrate and should be connected to ground?

[PRR]

...pins 12, 13, and 14 unconnected. I seem to remember reading one of those pins is substrate and should be connected to ground?

Let's go to the datasheet.
http://www.mit.edu/~6.301/LM3086.pdf

Substrate should tie to the most negative point of any element in the array. (In single +9V this may be "ground".)

[anotherjim]

I think bias "C" needs to be a low AC impedance. I don't see how it can easily avoid an opamp buffered 4.5v supply which is probably why the original went to the trouble of using the TL071. The R of an RC filter from the TL062 buffer would need to be of low resistance to not drop any voltage by the needs of the LM3086 circuit. You might get away with 100R and 100uF but the ground return current is in the same chip as the LFO and that can bypass your filter.

The 9v power input capacitor might be larger than 10uF and I'd have 100uF at least. Further, because LFO "ticks" contain high frequencies, a 100nF ceramic cap should be across the TL064 power pins 4 & 11. Fit another 100n across the 470uF C2 to clean supply "E" for good measure.