When I designed my own class ab amp I had to come up wit ha bias setting routine.
Turn bias pot until bias is zero volts.
Input a sine wave to input and monitor output to speaker with a speaker connected.
Turn up the bias slowly until crossover distortion disappears.
You can see if the bias is the problem by turning the bias down to zero volts and see if the heatsink gets hot with no input signal.
Be careful though as giving too much bias could blow your amp up.
Turn bias pot until bias is zero volts.
Input a sine wave to input and monitor output to speaker with a speaker connected.
Turn up the bias slowly until crossover distortion disappears.
You can see if the bias is the problem by turning the bias down to zero volts and see if the heatsink gets hot with no input signal.
Be careful though as giving too much bias could blow your amp up.
You need to find out which way the pot works to turn the bias down.
I use a signal generator and scope to set bias.
And to complicate matters further, there are TWO bias setting/sensing systems in series, one for each half of the output stage (Q14/Q15), TWO bootstrapped loads (R91/92/99/100) and so on.
The designer went all the way and beyond trying to make it ultra-symmetrical, overdrive in a certain way, etc.
Obviously it *does* work, and added complexity is not a factor with automated assembly, but when burnt it's a mess, even for experienced Techs.
R91, 92, 99 & 100
I pulled one leg of R91, 92, 99 & 100 and ohmed each. All four are within 1% of being right on the money.
As I stated yesterday, I had changed Q15 but didn't have a new one for Q14. Today I replaced Q14 with a TIP31C. Its way overkill but its an NPN. I turned the amp on and played it for a couple minutes on the clean channel. The temp of Q21/20 climbed up to around 115 deg. I switched to the drive channel and continued and it didn't take but about two minutes and 20/21 were about 150 deg. I turned off the amp because it shuts down around 157deg.
I removed Q20 and 21 and installed new ones just in case they had been damaged from overheating. Nothing changed. It still heats up just the same as before.
I pulled one leg of R91, 92, 99 & 100 and ohmed each. All four are within 1% of being right on the money.
As I stated yesterday, I had changed Q15 but didn't have a new one for Q14. Today I replaced Q14 with a TIP31C. Its way overkill but its an NPN. I turned the amp on and played it for a couple minutes on the clean channel. The temp of Q21/20 climbed up to around 115 deg. I switched to the drive channel and continued and it didn't take but about two minutes and 20/21 were about 150 deg. I turned off the amp because it shuts down around 157deg.
I removed Q20 and 21 and installed new ones just in case they had been damaged from overheating. Nothing changed. It still heats up just the same as before.
I'd be surprised if the power stage ever hit overdrive.
I worked with the FM65R (1x12", claimed 65w) for a couple of years with some very loud drummers, and found it stayed clean at very loud levels.
Problem was that my ears started distorting at the SPLs we were playing at, so the above may not be 100% accurate.
IMO, it'd be best simply to gut the existing amplifier, and replace with some 2nd hand PA amp. Sit it in the bottom of the cabinet or something.
Something like this. - I can vouch for that particular model being very good. If the asking price stays low, snap it up.
Chris
I worked with the FM65R (1x12", claimed 65w) for a couple of years with some very loud drummers, and found it stayed clean at very loud levels.
Problem was that my ears started distorting at the SPLs we were playing at, so the above may not be 100% accurate.
IMO, it'd be best simply to gut the existing amplifier, and replace with some 2nd hand PA amp. Sit it in the bottom of the cabinet or something.
Something like this. - I can vouch for that particular model being very good. If the asking price stays low, snap it up.
Chris
You hit 2 nails in the head.
1) I also think the circuit was designed to, say, "soft clip" or "bend the wave instead of chopping it" or something like that, and that might be the purpose of the added stuff .... which complicates repair, because "everybody is connected to everybody" in a DC coupled power amp.
So an error *anywhere* will throw output DC voltage, bias current or both out of whack.
Good luck trying to find it
2) I was also thinking, but didn't want to complicate badraven's life even further, that there is a "plan B", which I have used myself on some repairs, typically when the PCB is *destroyed* , as when it has a hole in it like somebody used a blowtorch.
That would be to pull everything to the right of C40 , meaning U6, Q6 to Q21, D16 to D39, all parts connected to them which are seen to be part of the power amp, clean the PCB, and add on the outside/bottom of the chassis a 100W/4 ohms generic Power Amp PCB, connected directly to the internal speakers.
It would only have 7 connections to the approppriate pads of the original PCB:
* * Input signal (from C40) and signal ground (from where R56+C41 ground)
* +42V (from top of C57)
* -42V (from bottom of C60)
* Spk+ from P1
* Spk- from P2
* New PCB ground to PSU ground , with a 1mm (#18) wire to the pad joining -C71 to + C72 (main filter caps.
You can mount the new pcb close to the chassis bottom with standoffs, with its own heat sink.
Trying to reuse the original one will complicate matters.
It won't even "look bad".
Almost forgot; you'll need 1 or 2 10mm - 3/8" holes in the chassis to pass those wires , deburr them and use some rubber grommet to avoid damaging wire insulation.
If you are fed up with that amp, it's one way to go.
Or make it even quicker, and mount inside some used cheap power amp that fits.
FWIW I often make such "substitutions", with good results.
Basically "all SS amps sound the same" .... or close enough.
And any power amp is better than a dead one
1) I also think the circuit was designed to, say, "soft clip" or "bend the wave instead of chopping it" or something like that, and that might be the purpose of the added stuff .... which complicates repair, because "everybody is connected to everybody" in a DC coupled power amp.
So an error *anywhere* will throw output DC voltage, bias current or both out of whack.
Good luck trying to find it
2) I was also thinking, but didn't want to complicate badraven's life even further, that there is a "plan B", which I have used myself on some repairs, typically when the PCB is *destroyed* , as when it has a hole in it like somebody used a blowtorch.
That would be to pull everything to the right of C40 , meaning U6, Q6 to Q21, D16 to D39, all parts connected to them which are seen to be part of the power amp, clean the PCB, and add on the outside/bottom of the chassis a 100W/4 ohms generic Power Amp PCB, connected directly to the internal speakers.
It would only have 7 connections to the approppriate pads of the original PCB:
* * Input signal (from C40) and signal ground (from where R56+C41 ground)
* +42V (from top of C57)
* -42V (from bottom of C60)
* Spk+ from P1
* Spk- from P2
* New PCB ground to PSU ground , with a 1mm (#18) wire to the pad joining -C71 to + C72 (main filter caps.
You can mount the new pcb close to the chassis bottom with standoffs, with its own heat sink.
Trying to reuse the original one will complicate matters.
It won't even "look bad".
Almost forgot; you'll need 1 or 2 10mm - 3/8" holes in the chassis to pass those wires , deburr them and use some rubber grommet to avoid damaging wire insulation.
If you are fed up with that amp, it's one way to go.
Or make it even quicker, and mount inside some used cheap power amp that fits.
FWIW I often make such "substitutions", with good results.
Basically "all SS amps sound the same" .... or close enough.
And any power amp is better than a dead one
Last edited:
JMFahey actually what you say about removing everything to the right of C40 makes sense since the drive and such is all created in the preamp on this amp. So are you suggesting I build a power amp like you describe or are you suggesting I purchase a power amp? Do you have something particular in mind? Thanks!
BR
BR
I'm not an electronic pro here, so I can only sit back and watch and learn like you. I have knowledge about certain amps as I've dealt with them before, but this isn't one of them.
I'm sure you will get to the bottom of this. You started with a busted amp, broke it further blowing fuses and all, then over the course of a hundred posts were guided into getting the thing (almost) working. You are close to finishing.
if you go the route and get a seperate power amp, may I suggest a Class D board?
Something like this.
http://www.parts-express.com/pe/showdetl.cfm?partnumber=320-303
You can just use one channel.
I'm sure you will get to the bottom of this. You started with a busted amp, broke it further blowing fuses and all, then over the course of a hundred posts were guided into getting the thing (almost) working. You are close to finishing.
if you go the route and get a seperate power amp, may I suggest a Class D board?
Something like this.
http://www.parts-express.com/pe/showdetl.cfm?partnumber=320-303
You can just use one channel.
Last edited:
If the heat-sink is getting to hot then the output transistor are passing too much current.
The first few stages of the amp are about voltage amplification so shouldn't affect it.
Then the bias stage gives the correct bias voltage to the output transistors.
You could try changing r95 and r96 to 3k3 and see if the problem goes away.
This would decrease the bias voltage so you might get some crossover distortion but it would prove where the problem is or isn't.
The first few stages of the amp are about voltage amplification so shouldn't affect it.
Then the bias stage gives the correct bias voltage to the output transistors.
You could try changing r95 and r96 to 3k3 and see if the problem goes away.
This would decrease the bias voltage so you might get some crossover distortion but it would prove where the problem is or isn't.
Nigel
I have been sitting here for the last little bit thinking the same thing. I thought, if I change the right reisistor(s) I could cool the output transistors off and eve if it were to drop my power who cares, this amp has plenty of power. I will give this a try. Thanks!
I have been sitting here for the last little bit thinking the same thing. I thought, if I change the right reisistor(s) I could cool the output transistors off and eve if it were to drop my power who cares, this amp has plenty of power. I will give this a try. Thanks!
Last edited:
ouimetnick
Thanks for the advice. If I can't correct this heating problem soon I think I will try this. Thanks!
BR
Thanks for the advice. If I can't correct this heating problem soon I think I will try this. Thanks!
BR
I'm not an electronic pro here, so I can only sit back and watch and learn like you. I have knowledge about certain amps as I've dealt with them before, but this isn't one of them.
I'm sure you will get to the bottom of this. You started with a busted amp, broke it further blowing fuses and all, then over the course of a hundred posts were guided into getting the thing (almost) working. You are close to finishing.
if you go the route and get a seperate power amp, may I suggest a Class D board?
Something like this.
2x100W TDA7498 Class-D Amplifier Board 320-303
You can just use one channel.
Just for kicks, connect the Power amp out to a (borrowed) amp similar to the one suggested by Chris in post #229 and hook it to your speakers.
If you like what you hear ...... now you know.
As of adding a new Power Amp PCB, for practical reasons please avoid Class D, ChipAmps, Class A, Zen, etc.
We are not "upgrading" or "modernizing" anything here, we are backtracking to a classic, simple, *well* proven design to be able to go onstage with this amp.
The least "surprises", the best.
Said new PCB should be able to supply around 100W into 4 ohms, when fed +/- 42V , to be able to use the current PSU.
Here simplest is best.
If you like what you hear ...... now you know.
As of adding a new Power Amp PCB, for practical reasons please avoid Class D, ChipAmps, Class A, Zen, etc.
We are not "upgrading" or "modernizing" anything here, we are backtracking to a classic, simple, *well* proven design to be able to go onstage with this amp.
The least "surprises", the best.
Said new PCB should be able to supply around 100W into 4 ohms, when fed +/- 42V , to be able to use the current PSU.
Here simplest is best.
PREamp out!!
Using the power amp output in its current faulty state could destroy any subsequent equipment.
I think that's what you meant, but I feel we should be clear on this point.
Chris
PS - to the OP, have a look on eBay for Samson Servo. There's a few bargains to be had, if you can get one to fit nicely in the cabinet.
Last edited: