Hi. I want my next preamp project to not have any moving parts, such as switches, pots, mechanical relays whatsoever. To eliminate the potentiometers, i am going to use digital pots. To achieve audio input switching, I have thought of using reed relays. To replace conventional switches, I want to use capacitive touch switches behind a black painted plaxiglass, which is going to be my front panel. Do I have any chance of making the aforementioned work at all? And how is the sound quality going to be affected by the reed relays and the digital pots?😕
Note: I'm not going to deal with DSP and stuff like that, I want my signal circuit to be as analog as possible.
Note: I'm not going to deal with DSP and stuff like that, I want my signal circuit to be as analog as possible.
Instead of Reed relays you could use a solid state input switch, although I'm not well versed on their performance.
A reed relay should provide a clean signal path but it is electromechanical in nature.
A reed relay should provide a clean signal path but it is electromechanical in nature.
If you just need to toggle between inputs (i.e. not directly select them), then you can control normal relays via transistors and a 4017 counter IC. Then you just need to provide a pulse for the counter which can come from a single switch or from a microcontroller.
Hi there,
Well, most of what you want to accomplish (and more) is possible with an Arduino. It will not be in the audio path directly nor will it affect the sound in an undesired manner, provided that you take care of the analog parts circuit.
One more thing, I am not sure why you want to use reed relays? They can be noisy when they reach the end of their lifetime and sometimes can behave unpredictably. Much better solution would be to use a rotary encoder (mechanical or optical) and one of the solutions that are already built and tested by other people.
But those are my opinions, feel free to explore these possibilities and be the judge yourself.
Good luck on this project. I'd love to keep track on the progress. You will post your progress here, right? 🙂
Cheers,
Jasenko
Well, most of what you want to accomplish (and more) is possible with an Arduino. It will not be in the audio path directly nor will it affect the sound in an undesired manner, provided that you take care of the analog parts circuit.
One more thing, I am not sure why you want to use reed relays? They can be noisy when they reach the end of their lifetime and sometimes can behave unpredictably. Much better solution would be to use a rotary encoder (mechanical or optical) and one of the solutions that are already built and tested by other people.
But those are my opinions, feel free to explore these possibilities and be the judge yourself.
Good luck on this project. I'd love to keep track on the progress. You will post your progress here, right? 🙂
Cheers,
Jasenko
Like ubergeeknz already mentioned, reed relays and normal relays are not allowed if you want to avoid all moving parts.
Electronic CMOS pass gate switches distort, but how much they distort depends entirely on how you use them. Basically the on resistance gets modulated by the voltage between the switch and its supplies, so to keep the distortion low, you can:
1. Try to keep this voltage as constant as possible, for example by connecting one side of the switch to a virtual ground node
2. Make sure that on resistance variations don't affect your circuit much by ensuring that one side of the switch sees a high impedance
Combine these two and you end up with the switching method used in this switched volume control:
A New approach of Volume Control: AAVA (Accuphase Analog Vari-gain Amplifier)
Electronic CMOS pass gate switches distort, but how much they distort depends entirely on how you use them. Basically the on resistance gets modulated by the voltage between the switch and its supplies, so to keep the distortion low, you can:
1. Try to keep this voltage as constant as possible, for example by connecting one side of the switch to a virtual ground node
2. Make sure that on resistance variations don't affect your circuit much by ensuring that one side of the switch sees a high impedance
Combine these two and you end up with the switching method used in this switched volume control:
A New approach of Volume Control: AAVA (Accuphase Analog Vari-gain Amplifier)
If you want your volume control as analog as possible then you'll want to avoid switches and stick with solutions like VCAs. My last volume control was built with balanced pairs of AD603, this did the job fairly well except that the control range is only just over 40dB. As the input impedance is rather low and the gain correspondingly high I used a step-down transformer right at the input. This also helps in rejecting common-mode noise from your source.
Of course I am going to post the progress but first I have to finish another project which is a 150w power amp. After that I will start building the preamp.
Also, I didn't know that reed relays where electromechanical. It seems that solid state audio switching is the most difficult part of the project. What if I used super clean jfets, like 2sk170 but cheaper, and formed a buffer for each input that turned on/off in order to toggle the input?
Also, I didn't know that reed relays where electromechanical. It seems that solid state audio switching is the most difficult part of the project. What if I used super clean jfets, like 2sk170 but cheaper, and formed a buffer for each input that turned on/off in order to toggle the input?
For digitally controlled volume with analog signal path and low distortion, a vca ic such as THAT2181 is pretty much perfect.
Interesting, serious repurposing of parts not made for audio. The transformer is a great idea. How was the distortion performance re: audio specs?
Where would you get the control signal for a VCA from when you don't use any moving parts?
I don't yet have a working distortion meter - I have a National VP-7721 lined up for refurbishment but that'll take a while. Certainly I couldn't hear any distortion, the end customer for this is a vinyl user so perhaps his source eclipsed the volume control's distortion.
I don't have the aversion to moving parts the OP does. But I'd guess some kind of touch panel might do the trick. At a bare minimum, up/down volume functions from skin resistance sensing with interleaved PCB fingers?
Provided you isolate the inputs through large enough resistors and sum the outputs, again through large enough resistors, this will work. The resistors will add a bit of noise, but this is probably unimportant. The PS will still need to be commutated in some way, so you are only moving the unavoidable switching at the expense of complication.
A similar concept has been implemented commercially with valves, by having multiple parallel buffers and switching the heaters on or off depending on the desired input. Much easier with valves, as apart from pure capacitance the powered down inputs and outputs do not contribute any non linear distortion.
Perhaps the only clean way to achieve what you want without messing up the amplification chain is to use photocells for switching, as it is done in the Dartzeel pre. Some preliminary attenuation of the input signal prior to the switches may be desirable to minimise distortion.
I stated previously that all the switches will be capacitive switches located behind a plexiglass panel. These will be connected with an mcu, and the mcu will control the vca for attenuation, balance, mute etc
Haha, it's not that I hate moving parts. I just want to make something different and interesting, and not just take some preamp PCBs, solder them, connect them together and call it a day. That's too easy. But building something like that without ANY moving parts is somewhat arduous to achieve.
Photocells is a good idea, but they don't really isolate the channels; they just attenuate them (if by 'photocell' you mean LDR). After doing some research in the forum, I have found that many people have questioned on this topic, and two people have provided their designs:
Design #1 (Mooly's): https://www.diyaudio.com/forums/solid-state/119151-mosfet-amplifier-designed-music.html#post1452488
Design #2 (bcarso's): Source selector using Analog switches or multiplexers - opinions? posts 6&7
They are very similar, they use jfets for switching on/off and grounding the unused channel, for even better isolation. It is claimed that Mooly's design works perfectly and bcarso's looks very promising too, I think I'm going to try them out.
I actually looked at AD603 pretty hard as a studio compressor/limiter. The main annoyance, for that chore, is the low input Z and low output V, needing buffer and booster. (It's just a 100r 'pot' and a +/-5V amp.) The THD seems "maybe OK" from the little info given:
"For a 500 Ω load in shunt with 5 pF, the total harmonic distortion for a ±1 V sinusoidal output at 10 MHz is typically −60 dBc. The peak specified output is ±2.5 V minimum into a 500 Ω load." also Fig 12 Fig 13
https://www.analog.com/media/en/technical-documentation/data-sheets/AD603.pdf
-60dB is "only" 0.1%THD, and you can't write that on audio gear specs. However Fig 27 28 hint that THD may decline at lower frequency. If it only drops 10:1 before other nonlinearities dominate, 0.01%THD is fine for most uses (and beats burning tapes or smacking ADCs).
The 42dB range is ample for limiting speech/music (marginal for domestic playback volume knob). It's smooth, and simple to control with old-school analog thinking. But altogether the project was more complicated than other fine approaches. (I liked LDRs, but the recent abandonment of that industry does not look good going forward.)
Last edited:
Obviously. Don't have ANY controls ON the box. BluTooth receiver, control it with your smartphone.
_I_ would hate it. (I have a power-tool I have to set settings from my cellphone, and it's annoying.) But today it is often the easiest solution.
Or control it with your Alexa! Say "Alexa, 3 db less bass". Or play one of those headbanger tracks with the lyric "pump up the volume! pump up the volume!" over and over......
I misinterpreted the "as analog as possible" part; I thought you wanted everything as analogue as possible, but it only applies to the signal path, so a DAC in the control path is perfectly acceptable. Great, no problem there then!
Mind you, voltage-controlled amplifier ICs based on translinear circuits usually have far worse noise and distortion than stepped attenuators ("digitally controlled potmeters").
Then why did someone told me that I have to use a vca... Anyways, it seems that a volume control IC is the best option, something like the pga2311. But this is an ancient chip, can you recommend something newer and more decent?
- Status
- Not open for further replies.