PGA2310 preamp turn on mute

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hi. I'm trying to design PGA2310 based preamp and I guess last one problem left to solve is turn on/off thump. It's pretty heavy and if it will thump into 120w power amp it will be very loud or even can kill speakers I guess.

As far as I know the easiest way to do this is to use relays that will short output of PGA2310 to ground. But one of my requirements is no relays to make preamp completely silently operating.

I saw in some car audio head units that output mute is done using simple bipolar transistor. I've attached schematics of one of these solutions. I tried to do the same with 4n35 optocoupler lying around but with no luck, the thump is still pretty loud even though it's a bit softer than it was before. Honestly I don't understand why this head unit solution works at all because as I understand open bipolar transistor conducts current only in one direction thus eliminating only half of the signal and another half will still produce noise.

I'm also thinking of trying to use something like TLP227G or maybe analog switch capable of 5V single supply operation. (I'm using Arduino to control PGA2310).

What would you recommend to use for this case?
 

Attachments

  • Screen Shot 2017-11-08 at 12.07.41.png
    Screen Shot 2017-11-08 at 12.07.41.png
    104.6 KB · Views: 284
How does your power up sequence look like? You should first power up the arduino so it can set its pins to proper state before the PGA2310 gets the power. Otherwise if arduino pins are still in high-Z state and the PGA2310 gets the power then the mute is off and you have a turn on thump. Same on power down. Sequence the turn off such that the PGA2310 shuts down first and only then the arduino.
 
How does your power up sequence look like? You should first power up the arduino so it can set its pins to proper state before the PGA2310 gets the power. Otherwise if arduino pins are still in high-Z state and the PGA2310 gets the power then the mute is off and you have a turn on thump. Same on power down. Sequence the turn off such that the PGA2310 shuts down first and only then the arduino.

This is exactly the case. Arduino is powered first from always on +5v source, then after signal from remote it turns on +/-15V power supply that powers up PGA2310. I tried even putting MUTE pin of PGA2310 directly on ground making it always on but still there's thump during power on and even worse during power off. (Since during power off rails voltage lowers slowly because of PSU capacitors it results in prolonged noise in speakers).

I understand that some kind of mute circuit has to be built around but want to find some alternative to conventional mechanical relays. Something like PhotoMOS relays or transistors or ...
 
Administrator
Joined 2004
Paid Member
Mute transistors have very high E-B reverse breakdown voltages. A normal transistor will be from 5 ~ 7 VDC. I don't know if you looked carefully, but the transistors in the schematic are "digital transistors" that also have a resistor in the base circuit. You must use the exact part number shown as these transistors are specifically designed as muting transistors. The 2SC2878 was an early, well known muting transistor. A signal diode in series with the base might solve the reverse breakdown problem you have.

Personally, I'll take a distortion-free relay to ground mute over a transistor mute any day of the week, and twice on Sundays! You can get reed relays that are pretty quiet, and you do have to watch that the reed relay is designed for audio use (distortion problems). Although, that shouldn't be an issue shunting the signal to ground.

-Chris
 
IIRC, when in mute mode, the outputs of the PGA chips are essentially grounded through a 10k resistor. Additionally the device should default to muted at startup regardless of the mute pin state (admittedly I’m not as sure about that). Are you sure the thumps aren’t down stream of the PGA? Does the system really only thump with PGA in line?
 
Administrator
Joined 2004
Paid Member
Hi BJosephs,
when in mute mode, the outputs of the PGA chips are essentially grounded through a 10k resistor.
I would hardly call a 10K load, "grounded". Most signal sources will happily drive a 10K load forever and ever without much signal drop. This load is completely negligible from a signal point of view.

Now, if it were a resistance less than 10 ohms, that you could call grounded, not a great ground, but certainly signal loading and greatly reduced amplitude.

-Chris
 
PGA2310 datasheet says this about how muting works in it "Muting is accomplished by switching the input multiplexer to analog ground (AGNDR or AGNDL) with zero crossing enabled." So this means that it shorts inputs but not outputs when MUTE pin is active. I've verified that thump comes exactly from it and from nothing else in the circuit. By the way the circuit is (RF filter -> DG409 input switch -> OPA2134 buffer -> PGA2310).

I've checked schematics for Arcam CD player and indeed found muting circuit based on transistors. However as I can see muting transistor is driven by some kind of mute driver circuit that is built on transistors too and has it's own dual power supply. Why is this circuit needed? Maybe this is a stupid question but I'm novice to all of this.

Also, can you please briefly explain why diode in series with base might help?
 
Administrator
Joined 2004
Paid Member
Hi belyakove,
Well, to start, just because a circuit is designed a certain way doesn't mean it performs well. That's lesson #1.

Muting the output is always a good idea during start up and power down. Especially if the product is supposed to be a high quality one. That way you can be sure that even an induced noise will not make it out of your gear.

The diode in series with a muting transistor prevents reverse breakdown. Look at the situation again. Specialised muting transistors are designed with higher than normal reverse breakdown of the E-B junction. Normal transistors will generally not work for that function because they only have a 5 to 7 volt reverse breakdown, and they may leak before that. By putting a silicon signal diode in series with the base, you are preventing it from breaking down. This setup may cause other problems, I don't know. However it might work and allow you to use a normal, high gain BJT for muting. I would simply use a transistor designed for that purpose if I felt the need to use a transistor here. Another part that might work well would be a J-Fet. Again, I have seen them used, but never used on myself.

To my way of thinking, you have gone to the trouble of extracting the best performance from a circuit that you can. Why on earth would you then compromise it in the muting circuit?? Using a relay of some kind would tend to eliminate all influences of a muting circuit during normal circuit operation. BJT semiconductor junctions have a way of breaking down or becoming leaky under reverse bias conditions.

-Chris
 
Hi BJosephs,

I would hardly call a 10K load, "grounded". Most signal sources will happily drive a 10K load forever and ever without much signal drop. This load is completely negligible from a signal point of view.

Now, if it were a resistance less than 10 ohms, that you could call grounded, not a great ground, but certainly signal loading and greatly reduced amplitude.

-Chris

We have a misunderstanding, but it's my fault because I was looking at the wrong datasheet (PGA2311) anyway. From said datasheet:

"MUTE disconnects the internal buffer amplifiers from the output pins and terminates AOUTL and AOUTR with 10-kΩ resistors to ground."

So, the internal amplifiers are not driving a 10k load, obviously this would accomplish nothing with regards to muting. The output pins are disconnected from the internal amplifiers and tied to ground through a 10k resistor, so no signal at all is present. My bad because I though all the PGAs accomplished muting the same way.
 
Last edited:
Hi belyakove,
Well, to start, just because a circuit is designed a certain way doesn't mean it performs well. That's lesson #1.

Muting the output is always a good idea during start up and power down. Especially if the product is supposed to be a high quality one. That way you can be sure that even an induced noise will not make it out of your gear.

The diode in series with a muting transistor prevents reverse breakdown. Look at the situation again. Specialised muting transistors are designed with higher than normal reverse breakdown of the E-B junction. Normal transistors will generally not work for that function because they only have a 5 to 7 volt reverse breakdown, and they may leak before that. By putting a silicon signal diode in series with the base, you are preventing it from breaking down. This setup may cause other problems, I don't know. However it might work and allow you to use a normal, high gain BJT for muting. I would simply use a transistor designed for that purpose if I felt the need to use a transistor here. Another part that might work well would be a J-Fet. Again, I have seen them used, but never used on myself.

To my way of thinking, you have gone to the trouble of extracting the best performance from a circuit that you can. Why on earth would you then compromise it in the muting circuit?? Using a relay of some kind would tend to eliminate all influences of a muting circuit during normal circuit operation. BJT semiconductor junctions have a way of breaking down or becoming leaky under reverse bias conditions.

-Chris

Ok, you've convinced me, I'll use reed relays for this purpose. :) However I still have questions. I've seen in other circuits that resistors are used before and after connection to mute relay. Also I've seen that 10uF electrolytic capacitor is used between opamp output and muting point for dc decoupling. So the questions is what values of these resistors should be? And also if this electrolytic can be omitted? As I understand having electrolytics in signal chain is not very good thing.
 
We have a misunderstanding, but it's my fault because I was looking at the wrong datasheet (PGA2311) anyway. From said datasheet:

"MUTE disconnects the internal buffer amplifiers from the output pins and terminates AOUTL and AOUTR with 10-kΩ resistors to ground."

So, the internal amplifiers are not driving a 10k load, obviously this would accomplish nothing with regards to muting. The output pins are disconnected from the internal amplifiers and tied to ground through a 10k resistor, so no signal at all is present. My bad because I though all the PGAs accomplished muting the same way.

Checked 2311 datasheet and it indeed has different MUTE and Power on/off behaviour than 2310. Seems from description that 2311 doesn't have this problem. Pity that 2310 is not the same.

Hi BJosephs,
Well, that makes more sense. The 10K resistor is there so that any charge on an output capacitor doesn't change. That would make a pop noise.

-Chris

Maybe my question is stupid but if output would be left just unconnected why would that affect output capacitor charge?
 
Check easy stuff like: do you still get thump with pre-amp inputs shorted to ground?
Do you have any DC offset at output? Have you a buffer on the output and this is causing thump?
According to data sheet, gain is at zero on power up until volume setting is sent by processor.
So activating mute on power up does nothing.
My PGA2320 pre-amp signals power amps on from standby (5v trigger from Arduino pin) and no thumps.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.