After a few investigations the situation is as follows:
On the right channel I can set bias and it is as we speak approx 400mV but on this channel the DC-offset can't be lowered below 200mV, that's the minimum value I have been able go get.
On the left channel it it no reaction after Q105 and Q106. It seems like they couldn't take the abuse. What to do?
Edit:
And the outputtransistors get up på 60 degrees quite fast at 640mV bias.
On the right channel I can set bias and it is as we speak approx 400mV but on this channel the DC-offset can't be lowered below 200mV, that's the minimum value I have been able go get.
On the left channel it it no reaction after Q105 and Q106. It seems like they couldn't take the abuse. What to do?
Edit:
And the outputtransistors get up på 60 degrees quite fast at 640mV bias.
Well, I thought that atleast one channel is as it should be so I set the bias and the offset as good as I could and carried a old centerspeaker down from the attic. I plugged it in and tested with my mp3-player and what do you think happend? The offset jumped up from below 200mV to a couple of volts and blew the speaker. Or it just sounded like a small gassleak and went dead. When I set the bias just before I did it with the input shorted and no load on the output.
This was the channel which I didn't mix up the caps...
Anyone in Sweden who wants to take over the project for a modest sum?
This was the channel which I didn't mix up the caps...
Anyone in Sweden who wants to take over the project for a modest sum?
input cap?
Do you have an input cap? if there is any DC on the output of the mp3 player it will be amplified and then present on the output of the ksa50...
Sounds like you might have killed some transistors, you probably need to remove and test them...
Also make sure your diodes, zener and ordinary are installed with the correct polarity or the front end voltages will be weird...which will make all the voltages weird...
Stuart
Do you have an input cap? if there is any DC on the output of the mp3 player it will be amplified and then present on the output of the ksa50...
Sounds like you might have killed some transistors, you probably need to remove and test them...
Also make sure your diodes, zener and ordinary are installed with the correct polarity or the front end voltages will be weird...which will make all the voltages weird...
Stuart
No inputcap but the mp3 is the same player I tested my friends Krell with (and same boards, components etc..) so nothing wrong there.
I just checked and the diodes are installed the correct way.
yeah, I did check some of the transistors on the left channel and they seem to be broken but that doesn't explain the right one, the one that busted the speaker.
voltages...
IIRC in the wiki there is a table of voltages that should be present at certain points around the board, assuming you can leave your amps turned on for a while you might measure and then compare the voltages. If you list them here people may be able to help you diagnose the problem...
Stuart
IIRC in the wiki there is a table of voltages that should be present at certain points around the board, assuming you can leave your amps turned on for a while you might measure and then compare the voltages. If you list them here people may be able to help you diagnose the problem...
Stuart
No, all as specified in you BOM. Or atleast all components are from the groupbuy a while back.
Accually, I'm placing adds and going through my stuff as we speak. If I only can get someone to take over this project and let me have back some part of what I've spent then I'll probably get rid of everything down to my last capacitor, import a chineese integrated amp and settle with it. Perhaps take up some project in a year or so but as of today, I've had it.
Sorry for sounding so glum guys
yet another observation. When the amp is turned on theres a DC-surge of about 3V across the speakerposts that lasts for a second or two before it is lowered. When I then switch it of ir is again raised but this time to above 15DVC which slowly discharges!!!
These aren't very acurate measurement since they where made without a load.
What can be wrong?
These aren't very acurate measurement since they where made without a load.
What can be wrong?
kmj said:
And because of this, are fairly meaningless. if you have no load, things float around all over the place on start up and shut down.
Kmj, I know you want to just get it working, but aimlessly playing around won't help. The only way to troubleshoot is to work systematically. Print out a schematic, and go through every single resistor and cap, checking values. As each one is done, mark it with a permanant marker or dab of correction fluid.
Once that's all checked, then pull all the trannies and test them. At this stage you can test without the output boards connected, and the bias turned right down and I recommend you do so. The Krell behaves perfectly like this, and it avoids damaging expensive output devices.
kmj said:
Perhaps take up some project in a year or so but as of today, I've had it.
Sorry for sounding so glum guys
Hi KMJ, I understand you pain. I've given up on DIY for the last year after my project (an Aleph-X) created smoke instead of sound. Like your amp the thing is virtually done and just needs troubleshooting. You put all this effort in and when it does not work its very depressing. I'm just glad I've got some other things that turned out ok.
Your case layout is excellent. I would suggest
- disconnecting the output stage and driver boards and then testing the PS and delay circuit to make sure you get the right voltages and the right delay. Put a power resistor across the caps to bleed out the voltage and so you can see that the thing can provide some current.
- Next fire up one driver board and you should be able to measure the DC voltages to see if they match the Wiki; then check the other one. Your case layout has the driver boards quite accessible, should be a minimum of disassembly. If you have a scope you can feed some test signals into the driver board and observe the output-- the driver board works without having the output stage connected.
-Then connect the output stage and see if it does anything funny.
I suspect that you may have wired the output boards incorrectly, so check your devices and the pinouts.
Hi, I'll do that, eventually anyway
Removing the outputstage seems like a necessary thing since the fan I usw doesn't manage to cool the sink even with only one channel running. It doesn't take more than two minutes before the controller gives of a warning for temeratures abouve 60 degrees. But I have a 230V industrial fan which moves alot more air to install, this however renders the fancontroll useless for all things exept monitoring the temperature. The PSU and softstart works as it should, stable 37,5VDC are supplied to all connections.
Strange that this heatsink can't handle even one channel thou....
That's just it, this was only the latest one of a longer bad streak. But I'll sort it out sooner or later.
Then these values doesn't depend on the bias? That will make it easier.
Oh, would there do as a dummyload or is it cutting it a bit close?
50W 6ohm resistors
heatsink temps...
KMJ,
While you are debugging I wouldn't worry about the heatsink temps, or trying to get everything set up to maximum class A levels. The "boring plan", (which I used) is to start with every thing working with no outputs, it makes for a fine headphone amp...
Then make sure you have the preset for minimum output current and attach the outputs. Then slowly (ie 10% per day) increase the idle current, until the heatsink reaches 55c or the current reaches the 1.9A required for 50w class A...since the amp works fine with the idle current set low, it is no big deal to use the amp while adjusting...
Checking everything out before attaching the outputs can save a lot of grief...and speakers.
50w 6 ohm resistors sound perfect for a dummy load, you won't be able to perform full power tests for too long, but resistors are pretty tough and can stand overloads for a while, dunking them in a bucket of water on the end of long, long leads makes them virtually indestructible...
HTH
Stuart
KMJ,
While you are debugging I wouldn't worry about the heatsink temps, or trying to get everything set up to maximum class A levels. The "boring plan", (which I used) is to start with every thing working with no outputs, it makes for a fine headphone amp...
Then make sure you have the preset for minimum output current and attach the outputs. Then slowly (ie 10% per day) increase the idle current, until the heatsink reaches 55c or the current reaches the 1.9A required for 50w class A...since the amp works fine with the idle current set low, it is no big deal to use the amp while adjusting...
Checking everything out before attaching the outputs can save a lot of grief...and speakers.
50w 6 ohm resistors sound perfect for a dummy load, you won't be able to perform full power tests for too long, but resistors are pretty tough and can stand overloads for a while, dunking them in a bucket of water on the end of long, long leads makes them virtually indestructible...
HTH
Stuart
