Author Topic: Auto-Standby circuit doesnt work as expected  (Read 6858 times)

0 Members and 2 Guests are viewing this topic.

Offline MichaelaudioTopic starter

  • Regular Contributor
  • *
  • Posts: 212
  • Country: de
Re: Auto-Standby circuit doesnt work as expected
« Reply #25 on: August 05, 2020, 05:10:44 pm »
So I found out that when I short R69 the Voltage on Pin 5 ic9 goes back to zero when its been over7.7V , so Q2 fully closed then. Same thing when I remove R70. When I remove R69 and look at Pin 1 ic9 with the scope when Pin 7 gets a high, Pin 1 is also getting also high for also short time(its pulsing a little bit)
« Last Edit: August 05, 2020, 05:31:35 pm by Michaelaudio »
 

Offline MichaelaudioTopic starter

  • Regular Contributor
  • *
  • Posts: 212
  • Country: de
Re: Auto-Standby circuit doesnt work as expected
« Reply #26 on: August 06, 2020, 06:48:50 pm »
So it seems no one can help my but anyway, when I remove R64 and R65 it works. Now I havent any more ideas (C1 and c15 arent the problem either)
 

Online Ian.M

  • Super Contributor
  • ***
  • Posts: 13218
Re: Auto-Standby circuit doesnt work as expected
« Reply #27 on: August 06, 2020, 07:39:29 pm »
Am I correct that R64, R65 carry the signals going to the activity detector circuit round IC9A?

If so, it would seem that cutting power to VCC causes a glitch affecting those signals, which causes a false activity detection so turns the power back on.

You need to modify it to either lock out the activity detector for a few seconds after the timer cuts power, or to slow down the speed with which Q2 can reset the timer so it needs sustained activity, not just a btief glitch.  Inserting a resistor in series with Q2 collector may do the job - I'd start by trying 1K.
 

Offline MichaelaudioTopic starter

  • Regular Contributor
  • *
  • Posts: 212
  • Country: de
Re: Auto-Standby circuit doesnt work as expected
« Reply #28 on: August 06, 2020, 08:08:52 pm »
Yep.
It actually doesnt turn off at all, the voltage of Pin 7 IC9 switsches so quickly that it doesnt turn off Q4. Looked at VCC while Pin 7 is pulsing - nothing to see, looked at the inputs  (r64/R65 going straight to the RCA connector) same thing - nothing. And I dont see why a 1k to the collector would change something, there is already a 2M in series isnt it?
 

Offline Renate

  • Super Contributor
  • ***
  • Posts: 1460
  • Country: us
Re: Auto-Standby circuit doesnt work as expected
« Reply #29 on: August 06, 2020, 08:15:26 pm »
I'm late to the party, but 2 M ohm and an electrolytic?
You like living dangerously.
A long analog RC delay is so Flintstones.

Even back in the day, they would use an RC oscillator and count 16384 or more pulses for long delays.

The whole approach of an extremely large time constant RC network with an electrolytic cap is irretrievably flawed

A quick glance at the TPA3116 shows that there is no disable pin (or did I miss some multi-pin usage?)
Use a chip that has both mute and disable.
Intercepting power to the chip is a poor approach.

(I just retrofitted my guitar amp. It has both mute and disable. I have an ATTiny85 running power sequencing.)
 
The following users thanked this post: Ian.M

Offline MichaelaudioTopic starter

  • Regular Contributor
  • *
  • Posts: 212
  • Country: de
Re: Auto-Standby circuit doesnt work as expected
« Reply #30 on: August 06, 2020, 08:42:55 pm »
I know I know. But I already have the pcb and its working fine except for this problem, so I dont wanna make a new board and new schematic, i wanna fix the actual problem (spent way too many hours on this project to give up)
 

Offline tooki

  • Super Contributor
  • ***
  • Posts: 13157
  • Country: ch
Re: Auto-Standby circuit doesnt work as expected
« Reply #31 on: August 06, 2020, 11:28:35 pm »
I'm late to the party, but 2 M ohm and an electrolytic?
You like living dangerously.
A long analog RC delay is so Flintstones.

Even back in the day, they would use an RC oscillator and count 16384 or more pulses for long delays.

The whole approach of an extremely large time constant RC network with an electrolytic cap is irretrievably flawed

A quick glance at the TPA3116 shows that there is no disable pin (or did I miss some multi-pin usage?)
Use a chip that has both mute and disable.
Intercepting power to the chip is a poor approach.
You missed it. There’s a mute pin and a shutdown (SDZ) pin. One of the op-amps in the full schematic (which he inexplicably refuses to post in this thread, despite posting it in another thread already) is part of another circuit (drawing power from an internal power rail despite the datasheet explicitly saying not to do this) to control the mute pin.
« Last Edit: August 06, 2020, 11:31:14 pm by tooki »
 

Offline Renate

  • Super Contributor
  • ***
  • Posts: 1460
  • Country: us
Re: Auto-Standby circuit doesnt work as expected
« Reply #32 on: August 07, 2020, 01:43:49 am »
You missed it.
I did!
If you look at the block diagram (which is what I did) the innocuous "SDZ" in the upper left corner doesn't really seem to go anywhere.

Unrelated: Looking back at my notes, my guitar amp uses a TDA7296, 60 W, class AB.
Interestingly it uses max 65 mA quiescent current, compared to the TDA3116 at 50 mA (which is class D).
« Last Edit: August 07, 2020, 02:06:12 am by Renate »
 

Offline MichaelaudioTopic starter

  • Regular Contributor
  • *
  • Posts: 212
  • Country: de
Re: Auto-Standby circuit doesnt work as expected
« Reply #33 on: August 07, 2020, 06:19:48 am »
@tooki  i wouldnt do it again  but the cap on gvdd is charged with 70uA for like 3sec. So shouldnt be a problem. Everything works just fine except for the problem im trying to get help with.
 

Offline Renate

  • Super Contributor
  • ***
  • Posts: 1460
  • Country: us
Re: Auto-Standby circuit doesnt work as expected
« Reply #34 on: August 07, 2020, 12:39:27 pm »
Why don't you remove the big P MOSFET, bypass it, invert the sense of your gate drive circuit and make it an open collector into the FAULTZ & SDZ junction.
 

Offline MichaelaudioTopic starter

  • Regular Contributor
  • *
  • Posts: 212
  • Country: de
Re: Auto-Standby circuit doesnt work as expected
« Reply #35 on: August 07, 2020, 01:01:49 pm »
Bc I want to cut the Power for everything not just the tpa's
 

Offline Renate

  • Super Contributor
  • ***
  • Posts: 1460
  • Country: us
Re: Auto-Standby circuit doesnt work as expected
« Reply #36 on: August 07, 2020, 01:34:25 pm »
Bc I want to cut the Power for everything not just the tpa's
Aw, c'mon, how much current is the other stuff taking?
What other stuff? If you power it down how can you tell when the audio goes live again?
 

Offline MichaelaudioTopic starter

  • Regular Contributor
  • *
  • Posts: 212
  • Country: de
Re: Auto-Standby circuit doesnt work as expected
« Reply #37 on: August 08, 2020, 08:45:09 pm »
As soon as a signal is applied to the inputs
 

Offline Renate

  • Super Contributor
  • ***
  • Posts: 1460
  • Country: us
Re: Auto-Standby circuit doesnt work as expected
« Reply #38 on: August 08, 2020, 09:16:59 pm »
Trivia: This thing has a Bluetooth module, but you can't use it unless you feed the line inputs to deactivate your muting circuit.

The TDA's you can turn off with the SDZ.
Your detect circuit has to stay powered all the time.
You are only saving the power consumption on those middle opamps if you switch power.
That's only 1.4 mA/op amp.

So you really want to brute force switch the power to that middle section?
Without using a reliable shutdown/mute timing?

 

Offline MichaelaudioTopic starter

  • Regular Contributor
  • *
  • Posts: 212
  • Country: de
Re: Auto-Standby circuit doesnt work as expected
« Reply #39 on: August 09, 2020, 12:19:24 pm »
I realized that too, so I will make that it doesnt turn off when bt is deleted.
I want it to work yes, I already changed connector nsfw on the pcb, spent hours of work just to get rid of some Problems (and 3 weeks to make schematic and layout) and now that everything works beside the stby circuit i want it to work no matter what, so yea I dont wanna redesign it or make it work another way than expected. I guess thats understandable.
 

Offline MichaelaudioTopic starter

  • Regular Contributor
  • *
  • Posts: 212
  • Country: de
Re: Auto-Standby circuit doesnt work as expected
« Reply #40 on: August 15, 2020, 04:15:44 pm »
I found it out. It was the switching ic! (SN3257-Q1/TSSOP)
Does anyone know if a pin compatible switching ic exist? 
 

Offline tooki

  • Super Contributor
  • ***
  • Posts: 13157
  • Country: ch
Re: Auto-Standby circuit doesnt work as expected
« Reply #41 on: August 15, 2020, 09:34:41 pm »
I realized that too, so I will make that it doesnt turn off when bt is deleted.
I want it to work yes, I already changed connector nsfw on the pcb, spent hours of work just to get rid of some Problems (and 3 weeks to make schematic and layout) and now that everything works beside the stby circuit i want it to work no matter what, so yea I dont wanna redesign it or make it work another way than expected. I guess thats understandable.
mmmm.... no. If I discover my approach was wrong, I’d rather scrap it and start again with the right approach that will work well. Anything else is putting lipstick on a pig.
 
The following users thanked this post: Renate

Online Ian.M

  • Super Contributor
  • ***
  • Posts: 13218
Re: Auto-Standby circuit doesnt work as expected
« Reply #42 on: August 15, 2020, 09:58:21 pm »
mmmm.... no. If I discover my approach was wrong, I’d rather scrap it and start again with the right approach that will work well. Anything else is putting lipstick on a pig.
There's a rather more earthy phrase that comes to mind: 'gilding a turd'.   Mind you, if you are willing to spend enough money that can literally be done.  First freeze dry it, next resin impregnate it and fully cure, then plate it with electroless copper to build up a conductive layer, and follow on by electrolytic copper to the desired thickness.  Finally send it out for the ENIG plating process we are all familiar with from PCB production.  At the end of the day, however gilded, its still a turd, useless for anything except a scatological paperweight or a modern art exhibit.
 
The following users thanked this post: tooki


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf