EEVblog Electronics Community Forum
Electronics => Projects, Designs, and Technical Stuff => Topic started by: dark_hawk on October 10, 2021, 01:45:15 pm
-
Hi,
This is schematic is used to make an audio amplifier/ power board go into standby mode when there is no sound detected for an x period of time. Reason for this is that the unit has very bad thermal design and the power board gets really hot even if there is no sound being produced.
All credits of the design go to the original designer TGAH08 whom I can’t find anywhere and his website is down for 3 years now.
(https://i.imgur.com/Hz4eGBY.jpg)
His description of the circuit:
The auto standby circuit is built around the LM324 quad operational amplifier.
It has six wires to connect, +9v Red, -8v White, ground Black, standby control in Grey and out Yellow (the existing pcb track linking these two points, needs cutting) and an audio pickup Black screened.
The audio signal is picked up from the output of the opamp feeding the sub channel amplifiers. I used this signal as all the seven audio inputs bass signals are summed together and are sent to the bass along with the dedicated bass signal. This way if any of the eight channels have an audio signal the circuit will keep the unit switched on.
The audio input is buffered by the first opamp so as not to load the audio signal in anyway. The second opamp has a gain of about 200 and boosts the signal to a usable level. The third opamp is an active rectifier with a peak hold, this gives a DC voltage out when audio is present, and this holds TR1 on which keeps capacitor C5 discharged.
With no audio signal, TR1 switches off and C5 starts to charge up via R11. With the values given the CR time is 484 second or 8mins (CR time is to 0.7 of supply voltage), this voltage is connected to one input of the fourth opamp. This opamp is set up as a comparator and it’s other input is set at about 0.7 of supply voltage by R7 and R8, about 6.2v.
If at any time an audio signal is picked up, TR1 will be turned on and rapidly discharge C5 via R13 (100R) and the timing period will restart. If C5 reaches 6.2v the fourth opamps output will switch high and pull the control line to 5v via R4 and D1, this will switch the sub to standby in an orderly way. At any time an audio signal will ‘wake up’ the sub by switching this line back to 0v.
If the sub is switched to standby using the control pod the I/O control ics output will switch to +5v and this will, via R3 and D2, switch on TR1 discharging C5 and the timer will be held at zero. This will ensure an instant switch on the next time the standby button is used to switch the sub on.
What I want to do is use this design to force the unit into standby when not in use, but I don’t want it to wake the unit up when sound is detected again. I want to switch it on manually.
How do I do that?
Thanks
-
Add feedback around amplifier 4 to make it latch.
-
Thanks for the reply.
I'm sorry, but how do I do that?
-
How? Add a resistor from pin 1 (output) to pin 3 (non-inverting input). That will provide positive feedback. You will need to experiment with the value, but 10K or 47K would be a good starting point and work down from there.
A NC, PB switch in series with this resistor should provide the manual reset you want.
PS: I am only speaking to your question, below. I did not look for other ways to provide the action you want and there may be a better one.
Thanks for the reply.
I'm sorry, but how do I do that?
-
I'm sorry, but how do I do that?
I offer a different approach to this problem. Instead of latching the fourth opamp, I used its output signal to block the 'wake up' feature so any audio signal arriving later will not reset the fourth opamp. A more detailed description follows:
If at any time an audio signal is picked up, TR1 will be turned on and rapidly discharge C5 via R13 (100R) and the timing period will restart. If C5 reaches 6.2v the fourth opamps output will switch high and pull the control line to 5v via R4 and D1, this will switch the sub to standby in an orderly way. Also the "wake up" function via audio signal is disabled(bypassed to GND) using TR2 via R18 & R19. (The resistor R18 is there to enable user to reset the logic using the standby switch on the sub.) Any audio signal after this timeout period will not ‘wake up’ the sub anymore. The output of the fourth Opamp will remains high.
If the sub is then manually switched to standby using the control pod the I/O control ics output will switch to +5v and this will, via R3 and D2, switch on TR1 discharging C5. This will force the fourth Opamp output to go low and remove the blocking of audio 'wake up' feature by TR2. Therefore, if the sub goes into standby via timeout, user can reset the logic by switching the sub manually to standby and then back to service.
You can use any small signal NPN transistor for TR2.
-
Thank you very much for the answer and for taking the time to review it.
I've tried it using another BC182L and it does not seem to work. I'm monitoring C5 to know if there is an audio signal being detected, and the cap keeps charging like it's not picking anything up although I'm feeding it a signal. Correct me if I'm wrong, but I think R18 is blocking the signal from opamp 3 to TR1. Voltage at the base of TR2 with an audio signal and the unit turned on is 0.6V. Also when switching the unit on after it was off, the cap resumes charging from the point it was at before switching off. Unfortunately I don't have a scope to look at the signals.
The circuit works fine without the modifications, only change I made was lower the gain from 200 to about 85 as it was picking up all sort of noise. Also decreased R13 to 68ohm so as when there is only voice audio being played it wouldn't pick up the signal well, so having a lower ohm resistor will help discharge the cap further down with the short detection periods of voice.
-
Also decreased R13 to 68ohm so as when there is only voice audio being played it wouldn't pick up the signal well, so having a lower ohm resistor will help discharge the cap further down with the short detection periods of voice.
What you have said indicated that values of R13 and R15 are very critical to the present operation of the circuit. In that case, I would propose to disable the audio signal in another way which does not affect the function of R13 and R15.
-
Nice work, there are better alternatives for lm324 on the market, it is mostly used for voltage comparators, signal filters and VCO's.
-
Also decreased R13 to 68ohm so as when there is only voice audio being played it wouldn't pick up the signal well, so having a lower ohm resistor will help discharge the cap further down with the short detection periods of voice.
What you have said indicated that values of R13 and R15 are very critical to the present operation of the circuit. In that case, I would propose to disable the audio signal in another way which does not affect the function of R13 and R15.
That's very smart. Thank you.
Now it works, picks up the audio signal fine when there is audio, and after it switches to standby it bypasses opamp 2 signal to ground and does not pick up any further signal.
One problem though, When you play audio, and stop it there is a sound that comes out of the speakers like a static (or a discharging of something?) that lasts for 2 seconds and then disappears. This sound does not happen with the unmodified design (maybe happens once when you switch the unit on but does not repeat after that). I just removed the jumper connecting pin 5 to the collector of TR2 and the sound disappeared. Should R18 be increased to maybe 100K? What exactly is the purpose of R18?
Edit: I changed R18 to 20K and that static sound seems to disappear. But at the same time I also changed C1 to a nichicon FG audio cap. Don't know which eliminated that static sound but it's gone.
Thanks again.
Edit2: That static sound is there when playing low volume sounds even after removing TR2 and the modification. If I remove the shielded audio in cable it's gone. And with that cable removed if I touch the probes of the multi meter to ground and the negative of C1 the static sound comes back and this time continuous and not with low volume audio. I think something is leaking from this board to the pre amp board of the sub. But where? And how do I prevent it?
-
Nice work, there are better alternatives for lm324 on the market, it is mostly used for voltage comparators, signal filters and VCO's.
It's easy to find and seems to be doing the job fine. To be honest It's the first time for me using op-amps and this one seems like a good place to start.
-
Edit2: That static sound is there when playing low volume sounds even after removing TR2 and the modification. If I remove the shielded audio in cable it's gone. And with that cable removed if I touch the probes of the multi meter to ground and the negative of C1 the static sound comes back and this time continuous and not with low volume audio. I think something is leaking from this board to the pre amp board of the sub. But where? And how do I prevent it?
The symptom looks like problem with ground loop between different parts of the amplifier system.
What is the power supply arrangement of your LM324 device? Does it use a transformer and connects to mains for power? If yes, is it housed inside a metal box and has earth connection to 0V?
-
LM324 is powered from a capacitor on the pre amp board. The main power board of the unit is a SMPS not a transformer. And the box is wood but during testing everything is hanging in the air. And there is no earth connection unfortunately.
Please note that this static is not there when the small board is not connected, or even when the audio in cable is disconnected. It's something from the small board.
How do I prevent that ground loop or "isolate?" the small board?
(https://i.imgur.com/rDq9HT2.jpg)
(https://i.imgur.com/hRBsWjh.jpg)
-
You dont ned the buffer. Make a non invering amp with Av of 200.
-
I think I found the problem but I don't know how to fix it. That sound being picked up is the "sound" of TR1 being switched on and off with the signal from op amp 3. When I removed TR1 and connected the board, the sound disappears. Now how would I prevent that switching on/off "sound/signal" from being leaked back to the pre amp?
-
You dont ned the buffer. Make a non invering amp with Av of 200.
Thanks for the reply.
I think I do need it. The audio in cable is connected to the output of the opamp for the subwoofer channel on the pre amp board. I think connecting directly to the gain op-amp will load the signal to the amplifier ics being fed by this op-amp. Please remember I didn't design this circuit and to be honest I don't have the knowledge to do so, but the one who made it thought it was necessary, plus, do you think the buffer is is the reason why the transistor switching noise is reaching the pre-amp board?
-
-
-
With the switching noise audible, try connecting an 25V electrolytic capacitor of 100uF or more, with short legs, across pin4 and pin11 of LM324 by hand, i.e. its positive and negative supply pins to see if it has any effect on the switching noise.
If by luck this eliminated the noise, you can make the connection permanent. Then insert resistor of 100R each into the incoming supply connections before it reaches the above electrolytic capacitor. i.e. instead of direct positive and negative power to LM324 pin4 and pin11, they each passes through a 100R first.
-
The Cap at the power input of LM324 does not help.
But adding a 100 ohm resistor at the small board input does make a difference. That switching noise is lessened a lot, it's still there but very faint. Even one 100 ohm resistor at the ground input makes a big difference.
I experimented with adding several values of inductors at the input but it does not make any difference.
Also, if powering the small board from a 9V battery (positive and negative while leaving the ground and the -8V connected to the preamp board) eliminates that noise completely.
Edit: Adding that 100R resistor at the ground had a side effect though. Voltage is now fluctuating at all the positive input of LM324! And this somehow makes TR1 turn on randomly when there is no input. A good solution I think would be to power the small board through a DC-DC isolation module, but I can't find those locally.
-
Edit: Adding that 100R resistor at the ground had a side effect though. Voltage is now fluctuating at all the positive input of LM324! And this somehow makes TR1 turn on randomly when there is no input. A good solution I think would be to power the small board through a DC-DC isolation module, but I can't find those locally.
You should not add any resistor in the ground connection wire, only to the +9V and the -8V wires.
If adding 100R has some effect, you can safely increase the resistor value to 470R to see if you can still hear the noise.
-
Edit: Adding that 100R resistor at the ground had a side effect though. Voltage is now fluctuating at all the positive input of LM324! And this somehow makes TR1 turn on randomly when there is no input. A good solution I think would be to power the small board through a DC-DC isolation module, but I can't find those locally.
You should not add any resistor in the ground connection wire, only to the +9V and the -8V wires.
If adding 100R has some effect, you can safely increase the resistor value to 470R to see if you can still hear the noise.
Adding a 100R to the +9V and -8V has no effect, the switching sound is still present.
Where do I put the 470R resistor? You said that I shouldn't add it to ground!
-
Your should add the 100R ~ 470R resistor as shown in the image.
-
Your should add the 100R ~ 470R resistor as shown in the image.
That's exactly how I put them. And they have no effect on the switching noise.
-
That's exactly how I put them. And they have no effect on the switching noise.
Try adding one resistor in the +9V line and one 100uF capacitor from +9V to Gound wire instead to see if there are any improvement?
-
That seems to have solved it.
Adding a 100uF cap between +9V and ground seems to have solved it, even without the resistor. I could've swore that I tried adding a capacitor at the input at some point and it didn't fix it then.
I increased the 100uF to a 270uF. Hope that's OK.
Thank you very much for you help.
If it's not too much trouble, could you please explain why I should not add a resistor to the ground? and What is the purpose of R18 in your modification?
-
If it's not too much trouble, could you please explain why I should not add a resistor to the ground? and What is the purpose of R18 in your modification?
I'm glad you've finally got the circuit working in the manner you wanted.
Ground is a common reference point for the rest of the circuit and we wanted its potential to be as steady as possible. If you add resistor in series with the ground line on part of your circuit, then any current flow in this ground resistor will create a potential difference between the two parts of the ground lines inter-connected by this resistor. As you have found out, this will have serious side effects like making your TR1 turns ON randomly without audio present. So normally we seldom place resistors between the ground lines of different part of the circuit. This is difficult to explain clearly, hopefully someone can provide you with more informative comment on this subject.
The purpose of R18 is to make TR2 more stable in regard to noise triggering. Without R18 TR2 will turn on if the output of the opamp ever goes higher than 0.7V. By adding this resistor, TR2 would require at least 1.4V of steady opamp output for it to turn on. So its purpose I would say is to increase the noise immunity of TR2 against noise.