Hi, I would like to ask if anyone has used digital pots in their amps/pre-amps and if they are worthwhile, or what the results were. 
[DS1669, MAX5160 and that ilk]
Or is a good quality rotary pot better for volume controll/attenuation.
Cheers
CB
[DS1669, MAX5160 and that ilk]
Or is a good quality rotary pot better for volume controll/attenuation.
Cheers
CB
I have used these (actually the older National Semiconductor pots) for guitar amplifier use. I used a seperate pot for input volume, master volume, bass, mid, treble, feedback level, and others. The advantage here is that you can store all of the settings in your microprocessors memory and call them up by stomping on a footswitch. I don't think that this feature is needed for stereo amp uses. If you go this route there are two rules that must be followed:
1) The audio signal input must always remain above 0 volts and less than the supply voltage to the chip. If the chip is running on a 5 volt supply, the audio signal must be considerably less than 5 volts peak to peak. This implies that the digital pot must be used in the input stage of the amp.
2) Care must be taken in the design and layout of the amplifier to avoid any digital noise getting into the audio signal circuitry, since this is near the amp input.
In my guitar amp I used a PIC microprocessor which reads the inputs from the rotary encoders (knobs), push buttons and footswitches, then programs the e-Pots and goes into sleep mode. In sleep mode the main clock stops, and the processor generates no noise. There is a wake up on change feature to wake up the processor when you turn a knob or hit a button. It programs the e-Pots, and goes back to sleep. No digital noise.
Since the tone stack circuitry has gain (2 12ax7 stages) I had to attenuate its output before going through the last e-Pot. Then there is an extra 12AX7 stage after the pot to get back to where I was (signal level) before the e-Pot. The programability is worth the effort in a guitar amp, i don't think it would be in a stereo. I would use a good quality pot or a stepped attenuator.
1) The audio signal input must always remain above 0 volts and less than the supply voltage to the chip. If the chip is running on a 5 volt supply, the audio signal must be considerably less than 5 volts peak to peak. This implies that the digital pot must be used in the input stage of the amp.
2) Care must be taken in the design and layout of the amplifier to avoid any digital noise getting into the audio signal circuitry, since this is near the amp input.
In my guitar amp I used a PIC microprocessor which reads the inputs from the rotary encoders (knobs), push buttons and footswitches, then programs the e-Pots and goes into sleep mode. In sleep mode the main clock stops, and the processor generates no noise. There is a wake up on change feature to wake up the processor when you turn a knob or hit a button. It programs the e-Pots, and goes back to sleep. No digital noise.
Since the tone stack circuitry has gain (2 12ax7 stages) I had to attenuate its output before going through the last e-Pot. Then there is an extra 12AX7 stage after the pot to get back to where I was (signal level) before the e-Pot. The programability is worth the effort in a guitar amp, i don't think it would be in a stereo. I would use a good quality pot or a stepped attenuator.
I was thinking about this kind of idea recently. PIC drives digital pot and so on.
As you say the digital pot does not like high voltages so you have to attenuate and then add another tube to get your signal back. One way around this may be to replace an existing gain stage with a beam deflection tube, 6AR8, and then drive the gain control grid with the digital pot. The gain control grid only needs low voltages.
But then I realized we no longer need the digital pot. Just drive an R-2R ladder network directly from the pins of the PIC micro to generate the control voltages for the 6AR8. BINGO a hybrid digital pot !
This would be nice if you have many channels to control simultaneously or if you want to make a remote for your tube amp.
Tube purists should be happy as the signal never enters solid state.
See here for 6AR8 info: http://basenjes.de/tubes/images/6ar8/6ar8.htm
As you say the digital pot does not like high voltages so you have to attenuate and then add another tube to get your signal back. One way around this may be to replace an existing gain stage with a beam deflection tube, 6AR8, and then drive the gain control grid with the digital pot. The gain control grid only needs low voltages.
But then I realized we no longer need the digital pot. Just drive an R-2R ladder network directly from the pins of the PIC micro to generate the control voltages for the 6AR8. BINGO a hybrid digital pot !
This would be nice if you have many channels to control simultaneously or if you want to make a remote for your tube amp.
Tube purists should be happy as the signal never enters solid state.
See here for 6AR8 info: http://basenjes.de/tubes/images/6ar8/6ar8.htm
Member
Joined 2002
Hi thanks for the replies,
The audio signal input must always remain above 0 volts and less than the supply voltage to the chip. If the chip is running on a 5 volt supply, the audio signal must be considerably less than 5 volts peak to peak
The audio signal input must always remain above 0 volts and less than the supply voltage to the chip. If the chip is running on a 5 volt supply, the audio signal must be considerably less than 5 volts peak to peak
Yes I agree, I've also seen this condition in cmos bi-lateral switches, am I right to assume that these devices use the same principle to switch resistor banks?
The switched attenuator idea is a possibility, I read an article (Many moons ago) on how to calculate a series of resistors for a "true log" attenuation curve but what rotary switch is ideal for the job if a good resolution (large numder of steps) is required. The largest I've heard of is a 12 step though I haven't looked very deeply into the subject thus far.
Some of the digital pots claim to have good log curves and have steps into the 100's, a good reason to employ them.
The puristsitting on my shoulder however tells me that the signal will pass through a solid state device.
I'm thinking that if I'm going to the trouble of building a valve pre-amp I should keep it valve, and therefore stepped or good 'ol rotary as tubelab.com suggests.
"replace an existing gain stage with a beam deflection tube, 6AR8" Interesting idea, I will explore this when I get my workshop sorted and if I can get a 6ar8.
Many thanks 🙂
CB
Digital Pot / 6AR8
Dynsdale,
Checkout John Swenson's 6AR8 preamp he posted on Audio Asylum last year. Might be able to combine a digital pot with the 6AR8 and make a sort of tube stepped attenuator hybrid.
Cheers -ALBQ
Also, my PGA2310 project
http://mywebpages.comcast.net/gillespie147/Junk/PGA2310-REV2.jpg
Link to John Swenson's post
http://db.audioasylum.com/cgi/m.mpl?forum=tubediy&n=58900&highlight=6AR8&r=&session=
Dynsdale,
Checkout John Swenson's 6AR8 preamp he posted on Audio Asylum last year. Might be able to combine a digital pot with the 6AR8 and make a sort of tube stepped attenuator hybrid.
Cheers -ALBQ
Also, my PGA2310 project
http://mywebpages.comcast.net/gillespie147/Junk/PGA2310-REV2.jpg
Link to John Swenson's post
http://db.audioasylum.com/cgi/m.mpl?forum=tubediy&n=58900&highlight=6AR8&r=&session=
- Status
- Not open for further replies.