Author Topic: Can sub-woofer amp flow & diagnostic procedure be explained  (Read 704 times)

0 Members and 1 Guest are viewing this topic.

Offline Mick B

  • Contributor
  • Posts: 13
  • Country: us
Can sub-woofer amp flow & diagnostic procedure be explained
« on: September 19, 2018, 05:56:20 am »
Hello, some information. This does not work the amp TDA7294V was replaced as were most electrolytic caps, transistors test OK. Thanks for any and all input. 
 

Offline oPossum

  • Frequent Contributor
  • **
  • Posts: 957
  • Country: us
  • The other white meat
Re: Can sub-woofer amp flow & diagnostic procedure be explained
« Reply #1 on: September 19, 2018, 06:23:02 am »
Measure voltages at pins 1, 4, 7, 8, 13 and 15 of the TDA7294, and pins 4 and 8 of the BA4558s.
 
The following users thanked this post: Mick B

Offline basinstreetdesign

  • Frequent Contributor
  • **
  • Posts: 342
  • Country: ca
Re: Can sub-woofer amp flow & diagnostic procedure be explained
« Reply #2 on: September 20, 2018, 02:34:31 am »
Hi Mick, hows the weather there?  I hope you weren't affected too much by the storm.  BTW, some time ago Mick liked a circuit description I made for him on a different product so he has asked me through PM-ing to do it again for this one.

So:
The schematic consists of two signal paths and a power supply.

First the power supply.  Mains comes in through the power connector, is fused for 2A and transformed down to some lower voltage from presumably 120 Vac even though the schematic says "230 Vac" (?).  Interestingly enough, there is no power switch so the unit is constantly drawing power.  The switch SW2, even though it says "OFF" and "ON" doesn't power anything OFF or ON just kills the signal before it gets near the power amp.  I don't know why the transformer is marked 120V / 60 Hz when the input is marked as 230 Vac/ 50 Hz.  Anyways, the bridge rectifier produces positive and negative unregulated DC rails of somewhere in the neighbourhood of +25-30 V and -25-30 V or so.  At least judging from the 35V caps C14 and C15 that's what it should be, no higher.  These rails supply the power IC only.  Transistors Q1 and Q3 together with zener diodes Z1 and ZD1 form voltage regulators for +/-15V to supply the rest of the circuit.  The voltage on each of those zeners sits at 15V and is supplied with a few (3?) mA of current through R29 or R30.  Then Q1 and Q3 are emitter-followers whose emitters are at Vzener - 0.7V with the entire circuit (minus power stage) as emitter load.

The signal path enters at left through either the LINE IN connectors or the HI INPUT connectors.  Either are to be used at one time, not both.  The LINE IN signals are to be the usual 1.0 - 2.0 V rms range whereas the HI INPUT signal level is intended to be closer to 10 - 23 V rms.  Interestingly the HI INPUTs are for balanced signals, that is, the audio is supplied as two complementary signals, one direct (L+, R+) and the other inverted (L-, R-) yet the two sides of each pair are differently terminated.  R+ and L+ are connected to ground through 2.2 K but the others, R- and L- are connected to ground through 100 Ohms.  This puts a very strange load on the source.  Anyways the two L+ and R+ are added together through the 220K resistors and passed to the input amp U1A.  If the LINE INs are used instead, they are added together through R1 and R2 and go to U1A.  U1A provides a voltage gain of about 5.5, due to the ratio of R4 and R5, to give a max signal level at pin 1 of U1A of 3.5 - 7V rms.  At this point the signal goes into two paths, one through U4A and the other through U1B.  The first path is one to create the mute signal for the power amp and the second is to further process the audio.

The audio signal continues through U1B with a voltage gain there of unity.  Switch SW1A allows the polarity of the signal at that point to be inverted or not presumably to match the polarity of the L and R signals at the main speakers.  It's unity gain either way.  U2A and discretes, C2, C3, R11 and R13 form a 2nd-order HIGH pass filter at about 7 - 9 Hz.  The sub then produces no noise below this frequency.  The op-amp U2B and all the discretes between it and U2A form a 2nd order LOW pass filter somewhere around 50 - 80 Hz.  No noise above this frequency is to be heard.  Both of these filters are unity gain for in-band frequencies.  The two adjustments VR1A and VR1B (same shaft?) adjust the cutoff frequency of the LPF.  The next stage, U3A, is a straight voltage gain of about 1.5 - 2.5 adjustable by VR2A.  The final stage at U3B is a last 2nd order LPF at about 250 Hz just to make sure that no high frequencies get past that point.  From there the signal hits the famous ON/AUTO/OFF switch where it is  or is not shorted to ground.  Resistor R22 is there just to save U3Bs ass from being shorted directly to ground, and then the signal goes to the power amp, U5.

The second signal path starting again at U1A pin 1 derives the MUTE and STANDBY signals for the power amp.  This path goes to U4A which applies a very large voltage gain by virtue of the ratio R31/R38 = 48.  This saturates its output with rail-to-rail signal swing (at full power).  Diode D2 half-wave recifies this and applies it to the base of Q2.  The two transistors Q2 and Q4 form another high-gain circuit.  As long as there is signal at all, Q4 keeps C13 charged up to a significant voltage, at least several volts.  When the signal abates and dies to nothing.  C13 may discharge through R35 over the next several seconds.  C13 and R35 have a time constant of about 1 second.  This voltage is fed to U4B which compares it to 0.7V derived from the forward voltage of diode D1.  The switch SW2B, part of the famous "ON/AUTO/OFF" switch clamps the detector ON in its upper position, allows it to work under control of the signal in its middle position or clamps it off in its lower position.  Thus at the output of U4B is both the MUTE and STANDBY signals for the power amp derived.  U4B also drives an LED through R43 to indicate that [EDIT: MUTE is active MUTE and STANDBY are not active and the power amp is enabled].  Network R44 and C9 smooth out some of the faster transients (0.22 msec) that may remain in the MUTE signal.

The last section to be considered is the power amp.  It is a one-chip affair that is basically a large op-amp.  It receives a
signal of max amplitude of 7-10 Vrms from U3B through C20.  R47 and C20 makes sure that the signal is readjusted to ground bias and it goes into pin 3 of the chip.  Resistors R48 and R49 set its voltage gain at ( R48 + R49 ) / R49 = 23.  C21 and R49 set the lowest frequency that will be amplified as 1 / ( 2 x pi x C21 x R49) = 7.5 Hz.  The networks with R51, R52, C23, C22 and D3 set the response time of the chip to the MUTE signal generated by U4B.  The voltages at pin 9
and 10 of the power amp must be above 1.5V for the amp to come on.  It has one output pin, pin 14.  From this output pin there is, besides the connector to the speaker, is a R-C network C10 and R46.  This AC load is known as a Zobel network and this load must be there to keep the chip U3 from bursting into oscillation which it may do if it were not there.

Now to the subject of troubleshooting:
If you get no sound out whatsoever, then first make sure of the basics, so like oPossum has said check the supply rails, both of them at all chips.  You wouldn't, by any chance have a unit wired for 230 VAc, would you?  So, with the switch in ON position, first make sure that the voltage at pins 9 and 10 of the power amp are HIGH (above 1.5v) to allow it to work.  Then apply some signal and then start at one end of the signal path and check the output of each op-amp in the audio signal path to see if you get signal there.  This narrows it down to a single audio stage.

Hope this helps.
« Last Edit: September 22, 2018, 04:11:46 am by basinstreetdesign »
STAND BACK!  I'm going to try SCIENCE!
 
The following users thanked this post: Mick B

Offline Mick B

  • Contributor
  • Posts: 13
  • Country: us
Re: Can sub-woofer amp flow & diagnostic procedure be explained
« Reply #3 on: September 21, 2018, 02:44:26 pm »
Hello again, the storm was only a drizzle by the time it got here thanks for asking. The schematic is for two amps I have the PSW10 and I believe the other was for 250v @50hz model. Also note C14 & C15 are  4700uf @50v everything else appears to be correct. Its going to take me sometime to digest all the information you have provided. I can't thank you enough. First order of business a better understanding (2nd order HPF) as I move through the signal. I will no doubt have questions. Once again thank you for your support as I endeavor to understand these things.
  oPossum if you read this thank you also, I will get around to what you suggest.  :-+
 

Offline Mick B

  • Contributor
  • Posts: 13
  • Country: us
Re: Can sub-woofer amp flow & diagnostic procedure be explained
« Reply #4 on: October 02, 2018, 08:58:01 pm »
Hello again,I hope everyone is doing well.I have done lots of reading and got familiar with LPF & HPF as well as the Zobel networks. The math is killing  me, one of my bigest regets is having to learn math @ 65 that I blew-off when I was a teenager, this is by far the hardest and the thing that slows me down the most. As Dave says "For you young players", Learn the math now while you still have a memory. Enough preaching.     
 Following oPossum and basinstreetdesign advice, I have used a sine-wave of 160Hz @ 1.5v followed the signal from UA1 to UA5
1. I have a good variable signal up to 7.76 pk-pk on pin 3 of U5
2. I have +14.65v at pin 8 and -14.37v at pin 4 on all the op-amps   
3. I have -35.6v on pins 15 & 8 and +35.7v on pins 13 & 7 of U5 NOTE:C14 & C15 are 50v
4. Pin 9 with switch in on position +11v in off position -.8v
5. Pin 10 switch in on position +9.91 in off position .85v
6  Voltage drops on ground pins 1 & 4 on pin 1 found 00.3v drop and on pin 4 a 0v drop
7 attached is a picture with the scope leads attached to speaker, playing Allman Brothers
tried this at different input levels with more or less the same results.
This amp seems to work BUT is SERIOUSLY DISTORTED AND requires a lot of output gain to acheve that I think my new TDA7294 is TOAST possibly from a 16v cap at c24 instead of 50v or my heat-sink and fan weren't enough for the TDA7294
Any other ideas? 
Thanks Mick
 

 

Offline basinstreetdesign

  • Frequent Contributor
  • **
  • Posts: 342
  • Country: ca
Re: Can sub-woofer amp flow & diagnostic procedure be explained
« Reply #5 on: October 07, 2018, 01:50:54 am »
Your scope shot seems to show 60Hz power spikes every 16 msec.  This shouldn't happen.  Did you just measure the rails with a meter or did you scope them?  It looks like the +/-35V is corrupted by large power ripple and it is being passed on to the speaker output.  The voltage ripple on the +/-35V rails shouldn't be larger than about 5V max.  If that is so then the +/-15V may also be similarly corrupted.  If so, then it would similarly corrupt all of the op-amp stages.  It should have no ripple at all.  Come to think of it, it looks like the bridge rectifier may be toast and causing those power spikes.

Also did you put in the sine wave and then follow it to each op-amp output on scope?

Putting a 16V cap at C24 by mistake might burn out that cap under full power but would not damage U5 or any other part of the amp.  Once replaced, the amp should be fine.

BTW: If I were you I wouldn't sweat the math for any of this for the sake of this amp.  Especially the Zobel network.  Maybe when you have free time you could look up the theory for those filters and see if the component values make sense but that would be purely recreational.
« Last Edit: October 07, 2018, 01:52:40 am by basinstreetdesign »
STAND BACK!  I'm going to try SCIENCE!
 

Offline Mick B

  • Contributor
  • Posts: 13
  • Country: us
Re: Can sub-woofer amp flow & diagnostic procedure be explained
« Reply #6 on: January 11, 2019, 03:01:51 am »
 Hello again, hope your holidays were enjoyable. It took this long to get back to you because I moved my shop. For the good news, thanks to you and my persistence my sub-woofer sounds like it did when it was new. What I found was the bracket that holds the TDA7294 up against the heat-sink was touching the pins, I re-drilled the heat-sink. I want to thank you again, your experience was invaluable to me and my understanding of these various circuits. Hope I can prevail on you in the future should something interesting show up.
 Thanks Mick   
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf