It works well enough for cvanc according to post #24, but if you want to make it more symmetrical, how about two 2N3904s in antiparallel, each having its own base resistor? Of course the signal still has to be small enough not to cause emitter-base breakthrough, as discussed earlier in this thread.
An audio guru once told me the voltage dependent capacitance of muting transistors was the main reason not to use them. Is that true?
the Rohm mute npn has an CB capacitance of 8pf when the base is 10V negative to the collector. this capacitance increases to 10pf when the voltage changes 6V, so 2pf modulation with 6V change. If you drive the circuit with a relative low impedance this 2pf effect is minimal. But then you need a very low ohmic mute resistor, or lots of current into the mute transistor. turning it off only needs the negative bias of the base, if there is no negative supply available you can create one using any available clock . for example the I2S clock of your dac.
When using standard transistors I meant as special muting transistors did not exist then. The capacitance an sich was not the problem but their value depending on the signal voltage level was. That audio guru has meanwhile passed but as he was pretty often right. There were a few tricks to minimize the effect like using the emitter at the signal side etc. (I forgot those) but best was to not use them at all. I have done so ever since when possible.
Last edited:
The difference with the old choppertransistors is the that modern mute transistors have very high gain. Both share the high Vbe reverse voltage. Already 40 years ago panasonic had special mute transistors.
This is for an 8 track reel to reel recorder (Fostex R8). All started when there was distortion at the output on one of the tracks if the input was high enough (peak levels). At normal levels it was clean and I did my last recording using this track carefully and recording only at low levels
Unfortunately 2N3904 doesn’t work. It does some muting and the output is undistorted but not dead silent muting when powering off the device
Thanks for the smt suggestions. It’s a great idea especially with the adapter. 2SD2704K with the adapter even matches 2SC2878‘s pinout
Here is the schematic,
I’m open to suggestions but I think I’ll go with one of the smt equivalent transistors, smt muting transistors are very cheap (even the adapters) or I’ll order a few 2SC2878A or B from eBay. I suppose A or B (low or high hfe) doesn’t matter, right?
Unfortunately 2N3904 doesn’t work. It does some muting and the output is undistorted but not dead silent muting when powering off the device
Thanks for the smt suggestions. It’s a great idea especially with the adapter. 2SD2704K with the adapter even matches 2SC2878‘s pinout
Here is the schematic,
I’m open to suggestions but I think I’ll go with one of the smt equivalent transistors, smt muting transistors are very cheap (even the adapters) or I’ll order a few 2SC2878A or B from eBay. I suppose A or B (low or high hfe) doesn’t matter, right?
Last edited:
The suggested muting transistors have a lower on resistance due to the high Hfe. The negative excursion of the signal when not muted does activate the collector base diode, and the forward diode in the base line is pulled in reverse blocking. Room for distortion. A fix would need a negative bias rail, that does not seem available is this design
If node POWER_MUTE is pulled to ground during normal operation, then voltages more negative than about -1 V to -1.2 V (diode forward voltage drop plus transistor base-collector forward voltage drop) will unintendedly activate the mute transistor. That's practically independent of the type of transistor, although the resulting distortion may be worse with some types than with others.
No idea if this is an issue, as I neither know what drives POWER_MUTE nor what the peak signal level is. There is an annotation about -10dBv, probably misspelled -10 dBV, but that is presumably a nominal RMS level.
No idea if this is an issue, as I neither know what drives POWER_MUTE nor what the peak signal level is. There is an annotation about -10dBv, probably misspelled -10 dBV, but that is presumably a nominal RMS level.
yes of course, thanks for that
unfortunately there’s no way of a negative bias rail. The only way is to take all those muting transistors out of the circuit. If the R2R was mine I would do it. Then powering off the devices from speakers to source
No negative rail anywhere does put a damper using regular transistors as shunt switches. JFETs too.
Hazy thinking but would a P Channel JFET work? R26 on the output is the vital component as it ground references the output. I tried a simulation of it and it would all depend on the FET characteristic for both on resistance and also gate voltage properties. If the control voltage is only +5v it might not be doable but if higher it could be.
An open branch will also do, there is no need for a negative rail. Even if POWER_MUTE is switched to ground, that may work fine if the signal levels are small enough.
Without information about the driving circuit and the peak signal level, it's all just a shot in the dark.
Without information about the driving circuit and the peak signal level, it's all just a shot in the dark.
Do you have any experience if higher resistor values are sufficient for pulling the Base to the negative rail?