This amp seems to be haunting me.
First time I attempted to build it I ran into problems and ended up selling everything DIY.
This time I was given some soldered boards and attempted to finish it.
I have built mono blocks and everything fired up without any smoke or such.
I raised the bias to 400mV across one of the 0.68R emitter resistors. I have two pairs of 15003/15004 on each channel.
I shorted the signal input to ground and made the adjustments without a speaker or load connected, just a DMM set to mV across the output terminals. I turned the other pot so I got around 50mV DC and proceeded to connect a cheap car speaker to test the amp.
The speaker started to squeal (I wore ear protection) and the DC on the output showed almost 25V and then the fuse blew.
I replaced the fuse and connected a 8R 50W resistor instead of the speaker and tried again.
Bias still at 400mV and the DC across the output was a few hundred mV.
I dialed it down and tried with the speaker again.
Still a squeal but no high DC.
I have a 2x18v of 200VA transformer and 2x68.000uF per channel and a star ground between the caps.
To this point the following is connected:
Two of the wires from the transformer.
-Driver board ground.
-Main board ground.
-Negative speaker post.
-And one wire to chassi ground/safety earth.
Disconnecting the last wire doesn't change anything.
Where the hell do I start looking?
I am getting tired of this amp 😭
First time I attempted to build it I ran into problems and ended up selling everything DIY.
This time I was given some soldered boards and attempted to finish it.
I have built mono blocks and everything fired up without any smoke or such.
I raised the bias to 400mV across one of the 0.68R emitter resistors. I have two pairs of 15003/15004 on each channel.
I shorted the signal input to ground and made the adjustments without a speaker or load connected, just a DMM set to mV across the output terminals. I turned the other pot so I got around 50mV DC and proceeded to connect a cheap car speaker to test the amp.
The speaker started to squeal (I wore ear protection) and the DC on the output showed almost 25V and then the fuse blew.
I replaced the fuse and connected a 8R 50W resistor instead of the speaker and tried again.
Bias still at 400mV and the DC across the output was a few hundred mV.
I dialed it down and tried with the speaker again.
Still a squeal but no high DC.
I have a 2x18v of 200VA transformer and 2x68.000uF per channel and a star ground between the caps.
To this point the following is connected:
Two of the wires from the transformer.
-Driver board ground.
-Main board ground.
-Negative speaker post.
-And one wire to chassi ground/safety earth.
Disconnecting the last wire doesn't change anything.
Where the hell do I start looking?
I am getting tired of this amp 😭
is there a Zobel network on the output (usually 10r in series with 100n)?This amp seems to be haunting me.
First time I attempted to build it I ran into problems and ended up selling everything DIY.
This time I was given some soldered boards and attempted to finish it.
I have built mono blocks and everything fired up without any smoke or such.
I raised the bias to 400mV across one of the 0.68R emitter resistors. I have two pairs of 15003/15004 on each channel.
I shorted the signal input to ground and made the adjustments without a speaker or load connected, just a DMM set to mV across the output terminals. I turned the other pot so I got around 50mV DC and proceeded to connect a cheap car speaker to test the amp.
The speaker started to squeal (I wore ear protection) and the DC on the output showed almost 25V and then the fuse blew.
I replaced the fuse and connected a 8R 50W resistor instead of the speaker and tried again.
Bias still at 400mV and the DC across the output was a few hundred mV.
I dialed it down and tried with the speaker again.
Still a squeal but no high DC.
I have a 2x18v of 200VA transformer and 2x68.000uF per channel and a star ground between the caps.
To this point the following is connected:
Two of the wires from the transformer.
-Driver board ground.
-Main board ground.
-Negative speaker post.
-And one wire to chassi ground/safety earth.
Disconnecting the last wire doesn't change anything.
Where the hell do I start looking?
I am getting tired of this amp 😭
No, but nevermind.is there a Zobel network on the output (usually 10r in series with 100n)?
The other channel is dead anyway. Can't even raise the bias above 0.
Since I haven't soldered the boards anything can be wrong.
I think I'll just put it on the shelf for a rainy day. I still have the chassis, cooling and psu anyway.

I had a problem on my KSA-50 build that was due to the grounding of the input and board...are you working with Pinkmouse's boards?
This amp is well worth the time...still one of my favorite amps....compared it to a real KSA-250s and still preferred my DIY build...
This amp is well worth the time...still one of my favorite amps....compared it to a real KSA-250s and still preferred my DIY build...
Yes, Pinkmouse boards.I had a problem on my KSA-50 build that was due to the grounding of the input and board...are you working with Pinkmouse's boards?
This amp is well worth the time...still one of my favorite amps....compared it to a real KSA-250s and still preferred my DIY build...
Your Input grounds going to the boards, correct?
EDIT - found my posting from years ago - sounds like it may be your issue...
https://www.diyaudio.com/community/threads/krell-ksa-50-pcb.31077/page-378#post-1430591
"I did not have the signal ground attached to the Starground."
EDIT - found my posting from years ago - sounds like it may be your issue...
https://www.diyaudio.com/community/threads/krell-ksa-50-pcb.31077/page-378#post-1430591
"I did not have the signal ground attached to the Starground."
Correct.Your Input grounds going to the boards, correct?
EDIT - found my posting from years ago - sounds like it may be your issue...
https://www.diyaudio.com/community/threads/krell-ksa-50-pcb.31077/page-378#post-1430591
"I did not have the signal ground attached to the Starground."
I'll check that on the working channels as soon as a couple of beers leave my system. Signal ground to star ground.
If I get one channel working, that will make it more interesting to troubleshoot the other.
Your Input grounds going to the boards, correct?
EDIT - found my posting from years ago - sounds like it may be your issue...
That did it!
There are three ground points with a line drawn below it. I took this as all three are on the same ground which they weren't.
One extra wire to star ground and it plays music. There is a large DC surge on turn on but it dips to 50mV rather quick.
There is still a large ground humm I'll have to find the source of but that is a minor thing, I hope.
Then there is the dead channel. Psu works but I can't get below 500mV DC to the speakers and I have no voltage over the driver emittors, so no bias voltage at all.
that will give you peace of mind....First channel quiet and working.
Other channel still dead. I measured some of the resistors stated on the wiki and basically no voltages match.
I guess that some or all of the transistors are dead.
The easiest fix would be to just replace all of the ones on the driver board.
Good deal! Before you go trashing possibly good transistors, make sure you have the + / - PSU voltages and power grounds (corrected to signal ground) on the input board, if all good, trace those same voltages to the other driver and both output boards....Then there is the dead channel. Psu works but I can't get below 500mV DC to the speakers and I have no voltage over the driver emittors, so no bias voltage at all.
If all ok, check legs of output transistors for shorts....
Publications made many run away from EF1 in search for speed as class AB amplifiers reigned. In many situations the EF1 isn't as bad as the picture painted and especially for heavily biased class AB amplifiers. And do we really need that speedup capacitor really?😉
Corrected to signal ground?Good deal! Before you go trashing possibly good transistors, make sure you have the + / - PSU voltages and power grounds (corrected to signal ground) on the input board, if all good, trace those same voltages to the other driver and both output boards....
If all ok, check legs of output transistors for shorts....
I just ran dual wires 🙂 .
Otherwise I built the Amps side by side so all wiring are identical between them, not counting the boards themselves which I didn't solder myself.
And by the looks of the soldering, a short is not a far fetched idea.
But new transistors for the main board cost me around $10 so it will probably save me both time and frustration to just desolder the ones on the board and remove the solder blobs at the same time while replacing them.
If any are ok I keep them as spares.
Replaced the transistors and now both channels work. I'll just have to run them for a while and adjust bias and such 🙂
18v 200va is too small.
I use a single 24v 300va and with 3 devices per rail with massive fan cooled heatsink.
https://www.diyaudio.com/community/threads/krell-ksa-50-pcb.31077/page-468
I use a single 24v 300va and with 3 devices per rail with massive fan cooled heatsink.
https://www.diyaudio.com/community/threads/krell-ksa-50-pcb.31077/page-468
Small, yes. But it works 🙂18v 200va is too small.
I use a single 24v 300va and with 3 devices per rail with massive fan cooled heatsink.
https://www.diyaudio.com/community/threads/krell-ksa-50-pcb.31077/page-468
Yes, but not much power in Class A before it becomes an AB amp (depending on load impedance).Small, yes. But it works 🙂
Yes.Yes, but not much power in Class A before it becomes an AB amp (depending on load impedance).
I have a history of failures with this amp and I had the present transformers on a shelf so I wanted to get it working before I made any investments.
When it comes to power I have a small room in the basement and easy driven speakers. So I don't need that much power.
But I'll keep my eyes open for a pair of larger ones.
- Home
- Amplifiers
- Solid State
- Krell KSA 50 PCB