Hi all,
I’ve built Sam’s MS-20 filter (layout 1.0, without the lo/hi pass switch) on stripboard, and I get nothing on the output - I get some very very quiet unfiltered wave if I turn up the gain on my mixer to maximum, but the cutoff pot does nothing.
I have tried:
Checking power on all 3 rails - ok
Checking for shorts between power rails - ok
Swapping out the TL074 and LM13700
Checking the cutoff pot - swings between ± 12v at the 100k resistor.
Measuring the current between pins 1 and 16 of the LM13700 - Problem!
There should be some current flowing, but I have none
I think there must be an issue with the CV converter.
I tried reflowing all the joints around the transistors but no luck.
When I move the cutoff pot, the voltage on the other side of the 100k resistors swings between +0.2 and -0.2v. Is this an issue? Should this voltage be larger/smaller?
Are there any other points that I should check?
Bear in mind I have next to no knowledge of how an OTA works(they convert voltage to current?) or how the CV converter works, and only a vague understanding of filters themselves(thank Moritz Klein!).
Power is coming from 2 (admittedly noisy) DC wallwarts but I doubt they are the problem as all my other modules work fine.
I unplugged everything apart from the VCO and the filter to test it.
Here are some pics.
The bias input current pins (1, 16) are usually around two diode drops above your negative rail. So something like -10V - is that the case with our stripboard?
Max current into those pins should be something like 1mA - so you should be able to measure at most 10V across the 10k resistors that feed into them. Do you see any voltages across those?
In my experience it’s frustratingly easy to make the LM13700 go pop by feeding too much into the Iabc bias current pins.
Thanks for replying
There isn’t any voltage across the 10k resistors. Is there a simple test circuit I can use to make sure the 13700 isn’t fried? My instinct is ‘no go build a VCA’ but I could be wrong.
Edit: There is about 0.02v across the resistor when I squeeze the probe cables with my hand but I think that’s just noise.
If you’re not getting any Iabc current, the problem is the CV circuit, not the LM13700. It won’t let any signal through if there’s no control current. I’d start by checking that the transistors are mounted the right way.
(that, or the current is too low for your multimeter.)
You could try building the simple VCA from the lm13700 data sheet:@analogoutput beat me to it
But… if there’s nothing across the resistors, the next thing to check is the transistors that do the CV to current conversion. The more I look at the stripboard in the original post the more confused I get about that part of the circuit though so I’m afraid I haven’t much to suggest. Looks like a kind of current mirror but it’s confusing.
I just tried replacing the transistors - no luck
Next i’ll replace some of the resistors - i have a pile from the 90s which have been reliable so far but you never know…
Makes a lot more sense now, I was missing the opamp driving the common emitters…
So, it’s an exponential converter, right?
The negative feedback from the op-amp should be forcing the collector of the transistor connected to 470k to zero. So, with a -12V rail there should be something like 25µA through that 470k when all the CV inputs are 0. Then, if I’m thinking about it correctly, we should get 25µA out of the other transistor into the two 10k resistors - so each one should get something like 12.5µA - and hence 125mV of drop across the 10k resistors.
I replaced the resistor and now on the output all I get is switching PSU noise, and no sign of the original output. Time to build a decent power supply I guess.
