how to convert current output of sound chip to voltage?

Go To Last Post
5 posts / 0 new
Author
Message
#1
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Hi,

I'm building an analog synth controlled by 2 atmega 1284s. Been making slow but good progress - have midi in and out, scanning for a 61 note velocity keyboard, extra eeprom memory to store settings and a touch lcd for all the controls.

Working on getting the connections right for an old CEM3396 wave shaper and vcf chip now. I can get several waveforms out of it, but can't get them to a audio line out without major distortion.

 

The chip spec sheet just shows connecting it a inverting op amp input with an unspecified resistor feeding back and the non inverting input grounded. Regular current to voltage converter from what I have on op amps.

I use op amps as voltage amps and a voltage differencer (to get my DAC 2V output amplified to 7.5v and then shifted down to -2.5 to +5, the range I needed for all the controls for this chip) but I can't get a current to voltage arrangement to work.

 

If I hook the current output of the chip (scope shows centered around ground 4v with no load) to an op amp with any feedback (with or w/o resister) the signal just goes to ground - completely dead.

If I try to treat it as a voltage and hook it to the non inverting input and feedback direct to inverting input to make a voltage follower, it over amplifies, goes to 8 volts and badly clips the signal.

Same thing happens if I feed it into non inverting adjustable gain amp. Even at gain one output is too high and clipped.

But if I feed it into the inverting side, the side with any feedback, it is just flat 0.

 

The closest I've come was to hook the signal to the center of two 100K resistors hooked to +5 and -5. and then ran this into a voltage follower as a buffer and then an inverting gain op amp where I can adjust the line out level.

Some signals worked, but as I changed the waveshape to a saw (non square) the signal just becomes a very narrow pulse wave. Badly distorts the waveforms.

 

Can anyone help me turn this current out (max +- 500 uA per the spec sheet) into a clean small voltage that I can buffer and amplify through an op amp?

 

My background is digital (from the 80's) so the lcd interface and sound generation up to making the pitched square wave triggers (via hardware counter chips) for the analog circuits has gone well, but I'm trying to learn more about analog to do the last part.

Having fun with op amps lately, but have been stuck on this problem for the last week.

I'm trying to test the control ranges that actually work with the chip and how to adjust them as frequency changes, but I need to be able to hear them at this point.

 

Thanks,

David

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

well the basic ohms law states Voltage = current * Resistance.

 

so in the basis you only need a resistor to get from current to voltage.

 

With opamps you need to watch out that when you get to the voltage rails the output will not linear follow the input as it clips.

Also the opamp speed is an important factor.

 

I think you need to get the signal going step by step.

first of all what is the impedance of your scope?

1Meg or 10Meg and what is the specified capasitance of the probes?

with 500uA and 1Meg you would make 500V.

with 100K you would do 50V.

 

What I also think is that your amplifier stages are coupled DC wise, you should make sure that you have separation caps between each stage, or be very confident that a Dc set point for one stage will not influence another stage.

 

 

 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I'd be looking at the circuits for a prophet 5 or ppg wave synth to see what they  did

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Looking at the schematics for the 3 synths I can find that use this chip - Oberheim Matrix 6/6R, Matrix 1000 and Cheetah MS6 - they all have the same arrangement:

The same current to voltage arrangement I tried, but with a 470pf cap added in parallel to with the 10K feedback resistor, plus a 100K resistor to ground from the inverting input (which my circuit did not have).

So I will try this tonight and see if the additional components was what I was missing.

 

But what is odd, is that none of them go straight to this op amp. They all go through a 4053 analog switch first, even if it is always on. I've seen a lot of these 4051-4053 switches in old synths for multiplexing input sliders and output voltages,

but I thought they were a simple block or pass through. It looks like they are also a type of buffer??

I ordered some of the 4053's (triple 2 to 1 switches) from Ebay to try them out and see what effect they have on these signals. I have some old 4051 8 ways, but I figure I'll try both.

 

Thanks for the suggestions.

 

David

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I am not aware of the 4051 et al doing any buffering. Maybe the intent was to switch the input to the opamp to minimise crosstalk further down the analog chain.