Digital potentiometer as volume control

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
There are several similar ready built Chinese modules with PGA and CS3310 digital pots. My kit was from Yuan Jing company. I just wanted to try it and have a taste of digital pots with idea that in the future I build something more ambitious with digital pots. For instance, my kit was made with PGA2320 chip that works with +/-15V analog part rails, but Chinese company made the kit with just two 5V regulators instead of two 15V regs and one 5V reg (for digital circuit). They simplified thing as much as it was possible. It's not big problem because I usually use attenuation between -30dB and -25dB. There is no way that I will ever use amplification so headroom will not be a problem. But still I'd like to use all the potential of PGA2320.
 
Very interresting thread. I'm currently finishing my Mezmerize B1-Buffer by Salas and i was wondering if i could replace the Alps potentiometer with a digital one. I also want to select the digital channels via switches. Looking forward to see what you guys will come up with.

What i would love to see is some kind of modular concept, since everyone has a slightly different setup and need for a digital volume control.

e.g.
- one central unit
- various display addons (in my case a 20x2 display)
- various input selector addons (in my case 6 switches)
- vraious volume control addons (in my case a rotary encoder)
- varoius potentiometer addons (in my case a 20 kOhm logarithmic stereo potentiometer)
- various input selector addons (in my case a relay or digital switch board to control the relays already present on my preamp)
 
Last edited:
if you are using a buffer (and need no gain), there are resistor-only parts (i.e. no internal opamps) that might be options for you.
Two examples (there are others) are:
Muses MUSES72320
Maxim DS1882

You may want to review the the Hafler 915 preamp schematic. It is basically a B1 with CMOS switches for input selection. It does use a conventional pot for volume and balance, however.

good luck,
mlloyd1

Very interesting thread. I'm currently finishing my Mezmerize B1-Buffer by Salas and i was wondering if i could replace the Alps potentiometer with a digital one. I also want to select the digital channels via switches ....
 
Very interresting thread.

Are those chips as good (as forgiving about sound) than the discrete ladder attenuator made with cms résistors ?

Is the most non programing natural chip to choose is the queitest dB one with the best channel separation ? Or are there some other things to look at (does it mean a dead quiet PS to power those chips or do they have already a complex inputt power noise filtering ? (All the Ebay's kits the same for instance if the chip is the same ?)
 
You will always find folks that say "no way IC device can be better than discrete" but we won't start that one today.
:hypno2:;)
Instead, I'll say you can search this board and see that:
1. Pass Labs (and other manufacturers) use the Muses device in their top line preamps.
2. Wayne of Pass Labs had good things to say about the Maxim device.
For what it is worth, those are reasonable "votes of confidence".
:cool:

mlloyd1
 
Thanks for your answer,

if it's good enough in the expensive pass pre it's enough good for me...

So it need at least a buffer at the output as the M. B1 for instance ? On the EBay kit seen above it seems there are just a direct outputt ? As far as there are no inputt or outputt serie cap, it should be good enough for me for a DAC perhaps !
 
In fact I built Chinese preamp with PGA 2310 because I wanted to hear how good it is without any buffers at the input or output (except for the one that is inside chip itself), to hear pure sound of digital pot. I was not disappointed. I think that digital pot is at least as good as the best mechanical pots, if not better.

It's nice that mlloyd1 informed us that there are examples from Maxim and Muse that do not have even internal opamp. The biggest obstacle for amateurs is programing of MCU and display interface. But the sound is excellent.

Another excellent digital pot is MAS6116.
 
Last edited:
Thanks for that , does a buffer after is really needed if the amp as high Z inputt (100K) ?

At least I like the idea of one in all (2 or 3 sources, double vol pot or one with a balance control) with a good quality price ratio.

For Good PS I assume the pcbs made by Mravica member are good enough...

That's Strange we saw some preamp shematic for few times (Linear Audio, Bruno Putzey...) : Do these ones avoid the issues of the usual carbon/cermet pots about floor noise level and channels separation) ?? Or do theywork just on the gain structure or to make the best buffer but with current pots (alps blue & co) ?

I don't understand where I should go to climb the next step to improve my 90s Yamaha pre... for a better swap ?!
 
Last edited:
These ebay offerings look just the ticket - my worry is their reliability - what happens in 5yrs from now, do I have to replace it with something that has a different footprint - what should we use if we want to build something that will be reliable long-term ? seems to me that the venerable pot or switched attenuator has the advantage there.
 
Thanks a lot for your response. Will definitelly look into this.

Yesterday i was playing around with my Arduino. I want to solve all my requirements first seperatelly (1. Arduino/Teensy controlled Potentiometer, 2. Input switching, 3. Rotary encoder and IR, 4. LCD Display and UI design).

The Potentiometer has for now highest priority. I'll use one of those prototype boards first, before moving on. Final result should be a PCB where I can plug-in a teensy micro-controller and attach all external components. If I ever get there I will be more than happy to share my results and look for improvements. My goal is just to create more out of the box comfort for our DIY-Audtio projects.

if you are using a buffer (and need no gain), there are resistor-only parts (i.e. no internal opamps) that might be options for you.
Two examples (there are others) are:
Muses MUSES72320
Maxim DS1882

You may want to review the the Hafler 915 preamp schematic. It is basically a B1 with CMOS switches for input selection. It does use a conventional pot for volume and balance, however.

good luck,
mlloyd1
 
Hi guys,

I'm working on an amplifier for my home.
On the preamp side I'm using a 2-way crossover (315hz) based on ESP Project 9, and for the power amp I'll use TDA7293...

It will have a significant level of digital stuff, like SPDIF, Toslink, display, usb, and bluetooth...

I was wondering why didn't anyone here mentioned the MCP41HVxx, are they really bad for sound? In fact, I didn't find any information about the THD of this folk...

But the good point of them is that they are +-18V supply capable, 8-bit resolution and are pretty cheap (1.51 on digikey)

I'm going to use 3 of them after the crossover: L R and subwoofer

What you guys think?
 
I'm currently re-building an amplifier with old ILP pre amp and main amp modules, bit would like to replace the mechanical pots with digipots.

Rather unusually, the volume control needs to be 100 kOhm linear, I also need 3 x 10 kOhm for tone and balance.

I've looked at various digipots, most have very low rails, +/- 2.5V, the pre-amps are at 25V, so dropping to 15 isn'T an issue. The LM1972/3 and PGA2310 look interesting, but I can't find their end to end resistance.

I don't need any buffering or pre-amp functionality.

Any suggestions would be great fully received.


Tom
 
The LM1972 and PGA2310 are not digital pots, they are effectively programmable gain amplifiers.

Effectively, these chips use a resistor ladder network which is switched by multiplexers.

I was beginning to suspect that, there's no end-to-end resistance values quoted.

I have since found the AD5204/6 devices that come in 10, 50 and 100 kOhm options and are 4 and 6 channel respectively. These look promising, not least because they come in PDIP format so can easily be soldered. The biggest downside I can see is there's no zero crossing built in.

Tom
 

PRR

Member
Joined 2003
Paid Member
What is the actual physical principle used on these “pots”? The block diagram shows a pot with a wiper but is it really an R2R network with a bunch of addressable FET switches?..

I am pretty sure these things are usually a "pot with 99 taps"(*) and 99 FET switches. (Some older ones had a few taps and an interpolator to get in-between taps, but I think FETs got too cheap to be so clever.)

Note the "wiper resistance" of 250 Ohms is about what you would expect for small FET switches.

R/2R does not lead to simple log-pot implementation. A string of 1dB 2dB 3dB 6dB pads makes good log, but needs a heap of buffers or a lot of low-R switches.

(*) 63 taps, 127 taps, whatever.......
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.