THE ENGINEER'S THUMB... At last, a better compressor!

[merlinb]

there is no signal after D1 and I can't see anything going into pin 3 of U2.

Are you using DC coupling? You should see the DC go up and down in sympathy with the audio amplitude.

All your voltages look right except:

U2
Pin 1 3.89 V
Pin 2 4.45 V
Pin 3 3.9 V

Is this with a signal applied or not? With no signal I would expect pin 1 (Q1 base) to be closer to 5V (i.e. above the emitter). If the transistor is being held 'on' this could explain your volume drop... Is D3 the correct way around? Is R15 definitey 1Meg?

[merlinb]

Just wondering what it is about the design that means the threshold control isn't really worthwhile.

Basically because it is soft-knee compression. When you're operating in and around a soft knee, turning down the threhold has a very similar effect on the transfer function as as turning down the ratio. You can see that to the left of the red line (i.e. around the knee) both graphs look very nearly the same, so if your signal spends most of its time in this region, you won't notice the difference between using the threshold control or the ratio control.

[patricks]

Excellent, thanks!

[Yazoo]

Thanks for the suggestions. I checked and the resistor is 1M and the diode is the right way around. I've done the usual and scratched around the pcb to check for shorts but I couldn't find anything. I've swapped out the ICs just in case.

I am still getting a slight voltage/volume drop when I apply a test signal, a sine wave, with the ratio pot turned up and the level pot full on. With the ratio pot turned down and the level pot full on, the signal is at about half strength. I am just seeing a standard sine wave on the oscilloscope.

I should admit at this point that I really enjoy building but I don't really understand how the circuits work. Should I be seeing an AC wave on the oscilloscope after D1?

[merlinb]

With the ratio pot turned down and the level pot full on, the signal is at about half strength.

When you say 'turned down' do you mean higher resistance, or lower? How big is your input signal? It may be that the effect is working normally and what you're seeing is the compression. Try turning down the input signal level to 50mVpp -does the output signal amplitude then start to match the input signal?

Should I be seeing an AC wave on the oscilloscope after D1?

No, you should see DC, although it may be wobbly DC.

[Yazoo]

Many thanks! I had one last go and swapped out the LM13700 with yet another chip and it now works. So that's two dud LM13700s. 

There is no problem with the volume now.

[MC-Sammer]

I'm having some trouble getting this working; I'm attempting a 4 knob version (Attack 100k, Release 1M, Ratio 1M and Level 10K) with a layout of my own design.

The issue I'm noticing right now is that I can't turn the Ratio knob below 967K without getting clipping on the output, and if I go much higher than that there is no output at all.

Here are my pin outs for an equivalent TL074 based circuit http://jmkpcbs.com/wp-content/uploads/2014/08/Engineers-Thumb.pdf  (All controls maxed /w 100mv 1KHz input):
1: 4.96
2: 4.85
3: 4.46
5: 4.46
6: 4.53
7: 8.25
8: 6.30
9: 4.66
10: 4.44
12: 4.31
13: 4.53
14: 4.53

And the Output that is buffered by the LM13700 when ratio is set to min:
5: 4.47
6: 4.85
7: 4.85

I'd post my layout, but the attachment option isn't showing up for me; perhaps I don't have enough posts yet

[bluebunny]

Looks like you have no ground connection.  Check your wiring and/or soldering.

There is no "attach".  You need to have your pictures hosted somewhere in web-land (e.g. imgur), then embed using the Mona button and the URL of your hosted picture.  (Hit "Quote" to see how I embedded the Mona icon.)

[MC-Sammer]

Looks like you have no ground connection.  Check your wiring and/or soldering.

There is no "attach".  You need to have your pictures hosted somewhere in web-land (e.g. imgur), then embed using the Mona button and the URL of your hosted picture.  (Hit "Quote" to see how I embedded the Mona icon.)

I checked my wiring and did find that I missed one cut , putting the 4.5 bias on the OTA side of the 1K going towards the inverting input of the output opamp.

I've fixed that, but it still appears that the gain of IC1 is driving the output into clipping. It also does that when I have the LM13700 removed... Not sure what to make of that.



Ignore the board on the right, I was using it to spot check the cuts on my board.

Also, the OTA bias line from the transistor looks confusing, but I'm just stretching the collector over the bias line and soldered it right into the bias.

[MC-Sammer]

For anyone who is curious, I think I narrowed it down to an issue with the 4.5V bias on the input opamps; now to track down the cause

[slashandburn]

I've found a board for this amongst my bits, it seems I must've etched Jon Pattons compact layout some while ago and then and then chickened out from building it. I'm going to give it a go soon.

I can't seem to find it in this thread, I'm sure it's buried here somewhere. Here it is for reference.


Looks like more of a challenge given that the components aren't numbered on the valve wizard schematic and I'd need to work a bit to work out whats going where.  Thought I was up for it. Fell at the first hurdle.

There are two pads underneath the IC labelled U3 (on Jons pcb layout). I'm guessing I leave them blank and probably shouldn't have bothered to drill them, but then I can't work out why they'd be there. 

A little help, please?

[merlinb]

Looks like more of a challenge given that the components aren't numbered on the valve wizard schematic

There is a numbered schematic here. It looks like Jon designed his PCB to match, kinda. Looks like U2 and U3 numbers have been swapped though.
http://valvewizard.co.uk/engineersthumb.pdf

Those pads look like surface-mount pads, corresponding to C4 and C5. I'm guessing Jon used surface-mount capacitors on the underside of the board to save space.

[slashandburn]

Looks like more of a challenge given that the components aren't numbered on the valve wizard schematic

I'm guessing Jon used surface-mount capacitors on the underside of the board to save space.

Oh. Bugger. Thats probably why I abandonned it the first time. I think you're right.  I'm staring at the board and thinking it was intended for a 1590A.  Damn thats a tight looking fit.  I'm neither skilled nor ambitious enough to pull that one off. No idea why I didn't etch the other board.

Thanks man, the numbered schematic really helps!  Doh!    Another coffee or two and I should know whether I'm feeling up to it.    Looks like just those two SMD components at a glance though.

Hmm, so for future reference: pads without drilling holes in the middle, don't drill them?

[midwayfair]

Those pads look like surface-mount pads, corresponding to C4 and C5. I'm guessing Jon used surface-mount capacitors on the underside of the board to save space.

It's redrawn using one of my layouts ... I have no idea what those pads are, but I definitely didn't use surface-mounted! slashandburn, my perf layout is somewhere in the depths of this thread, or in the library file linked on my Things I Make page if you need a parts reference.

[slashandburn]

Nice one thanks! Ill work it out!

Yup, found the original post after a lot of searching, then forgot to bookmark it. I'll have a look again another night, I think I'm short a few parts too.

[merlinb]

Hmm, so for future reference: pads without drilling holes in the middle, don't drill them?

Indeed. However, now that you have drilled them you could use a couple of through-hole disc capacitors. If you fold them over flat against the board then I doubt they'll get in the way (?).

[lars-musik]

Hi there,

looks like I've been sleepy lately. I completely missed first the post then the familiarity of the layout. I redraw Jon's layout for a 1590a ET.

Here's the build document:

https://www.dropbox.com/s/g7l603fsgfvsicg/Engineer_final.pdf?dl=0

Yours, Lars

[slashandburn]

Hi there,

looks like I've been sleepy lately. I completely missed first the post then the familiarity of the layout. I redraw Jon's layout for a 1590a ET.

Here's the build document:

https://www.dropbox.com/s/g7l603fsgfvsicg/Engineer_final.pdf?dl=0

Yours, Lars

Awesome!  Thanks man, that makes life much easier.    Did you really manage to squeeze it all in a 1590a?   Mad skills.  Way beyond my capabilities.

I have an idea for a prepainted 1590B that should suit this though. If I can get it finished, that is.  Prepare yourselves for more tedious questions from me on this thread over the coming days/weeks/months/years.

Those pads were indeed intended for (SMD?)  C4 and C5 then it seems?  I think I should be alright from there.

Cheers

[lars-musik]

Yes, those two are C4 and C5, both 1µF. Good luck!

[DK1]

Finally got it all together tonight. A little trial, and a lot of error, but it sounds great. Thank you!