Hi all,
I was wondering if anyone has any experience using 'digipots' as part of an audio circuit.
The specific application I have in mind is a guitar buffer/clean boost circuit. Apologies for the terrible sketch I have attached.
Ignoring the line receiver part of the drawing for the purposes of this post, I have built the guitar buffer part of the circuit and it sounds great!
What I wanted to do next was make the gain controllable via midi so I bought a 100k digipot and placed this in the place of the Rf resistor in the sketch (I also changed Rg to a fixed 20 something k resistor to keep the gain roughly the same)
The digipot is controlled from an arduino based 'midi receiver'.
In theory all good. but it sounds awful. Replacing the digipot with a regular 100k resistor works fine so its not the resistor values.
On closer inspection, the digipots 'potentiometer' are only rated to 5v (which I thought might be enough for this cct) so I am thinking this might be the issue.
Vss for the circuit is +12v, so was I crazy thinking that I wouldn't see anything like 5v in the feedback loop across the digipot?
My next step was to try find a higher rated digipot (there are not many 🙂 ) anybody had any similar problems?
I was wondering if anyone has any experience using 'digipots' as part of an audio circuit.
The specific application I have in mind is a guitar buffer/clean boost circuit. Apologies for the terrible sketch I have attached.
Ignoring the line receiver part of the drawing for the purposes of this post, I have built the guitar buffer part of the circuit and it sounds great!
What I wanted to do next was make the gain controllable via midi so I bought a 100k digipot and placed this in the place of the Rf resistor in the sketch (I also changed Rg to a fixed 20 something k resistor to keep the gain roughly the same)
The digipot is controlled from an arduino based 'midi receiver'.
In theory all good. but it sounds awful. Replacing the digipot with a regular 100k resistor works fine so its not the resistor values.
On closer inspection, the digipots 'potentiometer' are only rated to 5v (which I thought might be enough for this cct) so I am thinking this might be the issue.
Vss for the circuit is +12v, so was I crazy thinking that I wouldn't see anything like 5v in the feedback loop across the digipot?
My next step was to try find a higher rated digipot (there are not many 🙂 ) anybody had any similar problems?
Bigtim-I don't think you would be reaching the 5V threshold for the digipot, especially if you are coming straight from guitar into the buffer circuit. The feedback voltage depends on the signal voltage going into the opamp. You may have over looked something in the wiring. Which digipot are you using and do you have a diagram or photo of how it is connected?
A few things to check would be: How close the clock line is to the signal line. The high frequency clock line could introduce noise into the signal. Some digipots have reset and shutdown pins. Make sure you don't leave these floating, tie them to 5V. Make sure everything, digipot, arduino, and circuit are correctly grounded.
The digipot needs to be biased around 2V5.
It wont go below zero volts or else it will distort heavily.
It wont go below zero volts or else it will distort heavily.
http://ww1.microchip.com/downloads/en/DeviceDoc/22059b.pdf
Hi guys, thanks for the replies, above is a link to the pot I used (I ordered a higher voltage ad5290 last night see if that makes any difference, but as SD says I was expecting the voltage to be fine. I have tried the cct coming directly from the guitar and coming via a balanced receiver (unity gain) and get the same fault.
The digipot is wired in place of the Rf resistor in my sketch with the wiper and one of the terminals (A I think) as shown in the sketch.
Interesting what nigel is saying about biasing the resistor, so are we saying these pots are directional (won't go below 0v?) Heavy distortion is exactly what I am getting. I did wonder about this but isn't the 100k resistor network on the input of my buffer doing this? (Sorry for the dumb questions guys 🙁 ) I had thought that these resistors were biasing the op amp to the mid point of +Vss and 0v is this a wrong assumption?
Hi guys, thanks for the replies, above is a link to the pot I used (I ordered a higher voltage ad5290 last night see if that makes any difference, but as SD says I was expecting the voltage to be fine. I have tried the cct coming directly from the guitar and coming via a balanced receiver (unity gain) and get the same fault.
The digipot is wired in place of the Rf resistor in my sketch with the wiper and one of the terminals (A I think) as shown in the sketch.
Interesting what nigel is saying about biasing the resistor, so are we saying these pots are directional (won't go below 0v?) Heavy distortion is exactly what I am getting. I did wonder about this but isn't the 100k resistor network on the input of my buffer doing this? (Sorry for the dumb questions guys 🙁 ) I had thought that these resistors were biasing the op amp to the mid point of +Vss and 0v is this a wrong assumption?
the data sheet does tend to suggest that the a,b and w can be +ve/-ve with respect to each other?
Tried SD's suggestions tonight. All pins are tied to something ( +5v or 0V accordingly) and when playing, the clock signal was 'off' still no joy 🙁 going to order and Analogue devices 'high power' chip I think, these (unlike the one I have) do specifically say in the application notes that they are good for 'audio volume control'
Digipot
Hi, In my experience with digipots they can not have the signal drop below the ground level and most work best on a split supply of 2.5V . The input level is normally not over 1 V. If you go outside these limits you will get wonderful distortion. Depending on the actual IC most have rather complete instructions on the web. I would check out the limits of the one you are using.🙂
Hi, In my experience with digipots they can not have the signal drop below the ground level and most work best on a split supply of 2.5V . The input level is normally not over 1 V. If you go outside these limits you will get wonderful distortion. Depending on the actual IC most have rather complete instructions on the web. I would check out the limits of the one you are using.🙂
Bigtim1966;3979692 said:
I designed a mixer using Microchip digital pots and found you have to bias the inputs around 2v5. The pots wont work outside of the range 0 to 5 volts.
If you do you are probably fighting protection diodes on the inputs.
To be fair the datasheet does specify that the analog terminals range is Vdd to Vss, no real surprise there (And you will likely find the same thing with most of the other parts out there).
Regards, Dan (Who was designing one into a limiter sidechain earlier today).
Regards, Dan (Who was designing one into a limiter sidechain earlier today).
Thanks Nigel and Bruce. Your ideas do give me hope for the Analog devices chip (if my understanding of the buffer cct is correct)
So I think my buffer is biased at half its supply voltage which is 9v, so the microchip chip is seeing 4.5V bias with no input signal (hope this is correct) Which might explain why it is distorting when I give it a signal.
The AD5390 will handle +/-15v or +20v single supply so if the above is correct, hopefully it will work. 🙂
So I think my buffer is biased at half its supply voltage which is 9v, so the microchip chip is seeing 4.5V bias with no input signal (hope this is correct) Which might explain why it is distorting when I give it a signal.
The AD5390 will handle +/-15v or +20v single supply so if the above is correct, hopefully it will work. 🙂
I firstly assumed the voltage range was only applicable to the digital side of the pot. This is wrong of course.
So I made a 1/2 vcc supply with 2 10K resistors and a 100uf capacitor.
From this I took a 47k to the digipot input to bias it.
I also wanted to use a couple of digipots as a tone control.
I just biased the opamp at 2v5 so the signal was always in bounds.
I also found if I switched off the digipot I still had to bias the input to 2v5 or I got terrible noise breakthrough even though the pot was off.
I think I made the mistake Nigel is referring to above. I had assumed that the only limit on the analogue side of the digipot was the Vdd maximum. The advice of you guys who have done a better job than me of reading the datasheet is much appreciated.
Here's hoping the AD chip will do the job.
Would very much appreciate a view on my assumption that my buffer (and at unity gain) therefore my digipot is biased at +4.5v if Vdd in the sketch in my first post is 9V.
Here's hoping the AD chip will do the job.
Would very much appreciate a view on my assumption that my buffer (and at unity gain) therefore my digipot is biased at +4.5v if Vdd in the sketch in my first post is 9V.
The digipot you are using also can have a negative supply.
That would simplify matters for you.
If you just want to run it off 9volts then a 4.5 volt bias supply will work fine.
That would simplify matters for you.
If you just want to run it off 9volts then a 4.5 volt bias supply will work fine.
- Status
- Not open for further replies.