https://cheapestmall.ru/product/164868944780
i found this one if anyone else is interested, for 135 euro this looks quite awesome, not necessarly a big fan of the oversized case but still, 135 for such a device seems quite cheap
i found this one if anyone else is interested, for 135 euro this looks quite awesome, not necessarly a big fan of the oversized case but still, 135 for such a device seems quite cheap
I used the Muses72323, and as I understand it's a chip with a passive resistor ladder that is controlled by some logic in the same package.
The Muses sounds significantly better than both my other relay based attenuators that I mention in my above link. And the relay attenuators sounded way better than the blue Alps pot I used previously...
do you suspect the opamp buffers to be accountable for the better sound?
----
i was looking at the schiit freya plus
it has:
1. passive relay attenuator
2. (i think) passive relay attenuator + input and output buffer
3. tubes
(4. solid state tubes)
now a switchable tube buffer would be really awesome, but i kinda wanna avoid messing around with 300V, but i was thinking one could add opamp buffers for "some coloration" and it comes with the benefit of linear output/input impedance, which seems one of the big cons of volume attenuators, tho i think your usual 6bit relay attenuator also has some impedance matching resistors
opamp buffers could be switchable, so i can in the end switch between completely passive and opamp buffers
while opamp rolling i was always quite happy with the OPA1656, so probably would go with this one
This is probably due to the fact that FPGAs can switch 1000 times faster than MCUs (software is extremly slow). But that doesn't matter for click-free relay attenuators. The trick is the software. And of course you have to understand something about audio electronics.
Oh, sure. Gate logic is much faster than software. No argument there. But relays are slow devices. Even 'fast' relays take 5-10 ms to fully switch, so the speed of the logic (or MCU for that matter) is basically wasted.
I don't know this for a fact, but I'm willing to bet that Benchmark used an FPGA so they could make the attenuator a stand-alone module that they could reuse in other products. I seem to recall that the HPA-4 is built that way. The "secret sauce" is in the sequencing of the relays and/or in the place in the circuit where the switching is done. Then, of course, the marketing department caught word that an FPGA and an algorithm were used and turned those into marketing talking points. Can't blame them. It's how they operate. 🙂
Both Doug Self and Bob Cordell have chapters on click/pop-free switching in their respective line level audio books. Worthwhile reads.
Tom
I listened to a lot of different resisters and I liked the Vishay/Dale RN65 or RN70s. The large resistors have lower distortion. There is an industrial grade CCF which are less expensive and almost the same good sound. Caddock is a more expensive option through I can’t hear the difference from the RN65. The ultimate is the Vishay Nude foil precision resister, which are very expensive, but are extremely transparent sounding.
So now resistors, being a passive component has distortion, what kind of distortion would you say. Is there any technical literature or research paper on the subject?
My conclusion from these two books is that electronic switching always produces THD. Relays are absolutely THD-free. That way you are on the safe side. One disadvantage is that they are very slow.
I only posted about it in this thread:
It's been a long time 🙂
SSD1306 display, encoder or potentimeter control. Or it can be controlled by external MCU, the module can be switched to I2C slave mode.
1db steps, max attenuation is -63dB.
Later I'm going to create a separate thread for it (need to finalize the firmware, recently I received some IR receivers and remote to play with).
Yep, that's what I did there.
The MCU is ch32v003 and there is a driver chip (similar to ULN2003, I do not remember exact part number)
You might want to read Groner and Putzeys' article on distortion in relays then. It's in Linear Audio vol. 13. Not all relays are created equal. Reed relays can be particularly bad.
I haven't had issues with relays designed for small signal switching. They tend to have gold (or gold alloy) plating on the switch points. The Kemet EC2-series is an example. They're not cheap, but they are good.
Tom
Tom, yes this could be a problem since the interior is a glass tubular read relay with the coil assembly wound around the contacts so you would introduce a spike into the signal path when switching. I did not think of that, you just reminded me . Maybe this is what produces the zipper sound that someone mentioned earlier. Maybe I am chasing up ghosts and it is inaudible.
See linear audio article on distortion in resistors. I reference the article in my thread listed above.
The Relaixed passive system from Jos von Eijndhoven is excellent - 4 inputs, easily customisable impedance, can be stacked for balanced, remote control, simple MCU programming, 64 steps, and all open source (gerbers available for personal use). Also has a built in soft start for your power amp if you choose to use it. Can be powered from an onboard SMPS or go linear if you want. Proven solution, no clicks etc between steps.
https://www.vaneijndhoven.net/jos/relaixedpassive/
https://www.vaneijndhoven.net/jos/relaixedpassive/
I did a linear 4066 volume controller 35 years ago. At the time I did not put a log ladder in but rather used a 4040 counter and logic to log the clock speed. There was no sound quality as it did 100db s/n and 0.01% thd.
I am using a 45 year old !! Alps 100k tapped pot used in the original sx-780. It is still static free and flawless.
I am using a 45 year old !! Alps 100k tapped pot used in the original sx-780. It is still static free and flawless.
Here is the code I used to avoid clicks:
JavaScript:
void update() {
tmp = new & old;
// set going LOW bits (attenuated stages)
if (tmp != old) {
relay(tmp);
delay(ms);
}
// set going HIGH bits cascaded LSB to MSB:
if (tmp != new) {
int i, b;
for (i = 0; i < 6; i++) { // 6 relay stages
b = 1 << i;
if (((new & b) == b) && ((tmp & b) != b)) {
tmp += b;
relay(tmp);
delay(ms);
}
}
}
// or set going HIGH bits at once:
// if (tmp != new) {relay(new);}
old = new;
}