Hey guys. I just picked up a 10 amp Variac for this repair and future projects. Any tips on bringing up an amp on a variac?
What to look for?
What increment levels/steps/time to bring up the amp?
Should I be looking at heat changes as well as voltage and current?
Was going to attach as many meters as possible to monitor the startup.
Luckily, this is a mono-block. Makes things a bit easier.
Thanks,
Vince
What to look for?
What increment levels/steps/time to bring up the amp?
Should I be looking at heat changes as well as voltage and current?
Was going to attach as many meters as possible to monitor the startup.
Luckily, this is a mono-block. Makes things a bit easier.
Thanks,
Vince
One of the cheap gadgets I recommend to use with the variac or even DBT is a killawatt meter. I bought one several years ago to check power draw on appliances and electronics in my house. A couple of years ago I pulled it out while powering up a diy project and I've been using it ever since. Every piece of electronics I work on now is plugged into it. Its very handy you can see at a glance VA, Watts, Amps, etc.
Bummer! Sorry to hear this happened.
Based on your description of the failure, I would suggest checking the diodes in the bridge rectifier for that channel and also the power supply filter caps. It could be the hum you are hearing is power supply related due to voltage surge from letting the magic smoke out. The fact that there is a hum makes me suspicious that the failure also took a bridge rectifier diode.
I would also suggest double checking the center pin of the outputs with reference to that bank's heat sink. Yeah, I know you're already checked it. I would check them again. Do it with the output stage disconnected and no power to the unit. You want it disconnected just to be sure you don't get a reading from someplace else. As you know, you should not have continuity from the FETs to the heat sink. I'm suspicious that you still might have a short.
Did you replace the source resistors for each output device? They are difficult to accurately measure and will almost always be toast if the outputs fried.
Finally, you might also have problems with the soft-start board due to the current surge. Depending on its design, the MOV or resistors could either be open or shorted and not passing enough current for the amp to properly start but that won't solve the hum issue.
Good luck,
Steve
Based on your description of the failure, I would suggest checking the diodes in the bridge rectifier for that channel and also the power supply filter caps. It could be the hum you are hearing is power supply related due to voltage surge from letting the magic smoke out. The fact that there is a hum makes me suspicious that the failure also took a bridge rectifier diode.
I would also suggest double checking the center pin of the outputs with reference to that bank's heat sink. Yeah, I know you're already checked it. I would check them again. Do it with the output stage disconnected and no power to the unit. You want it disconnected just to be sure you don't get a reading from someplace else. As you know, you should not have continuity from the FETs to the heat sink. I'm suspicious that you still might have a short.
Did you replace the source resistors for each output device? They are difficult to accurately measure and will almost always be toast if the outputs fried.
Finally, you might also have problems with the soft-start board due to the current surge. Depending on its design, the MOV or resistors could either be open or shorted and not passing enough current for the amp to properly start but that won't solve the hum issue.
Good luck,
Steve
Last edited:
At the moment, I'm 3/4 of the way to rebuilding the amp.
All the p - ch mosfets and diodes check out ok. still need to check n-ch side. All the source resistors and caps check out ok. Rebuild the front end on to new board and all components are ok. Checked jfets and bjt. all ok. The power supply is ok.
I'm guessing either output protection or softstart start is faulty. Still need to check all components on n-ch side.
Do that tomorrow.
Thanks
Vince
All the p - ch mosfets and diodes check out ok. still need to check n-ch side. All the source resistors and caps check out ok. Rebuild the front end on to new board and all components are ok. Checked jfets and bjt. all ok. The power supply is ok.
I'm guessing either output protection or softstart start is faulty. Still need to check all components on n-ch side.
Do that tomorrow.
Thanks
Vince
The amp is working. No hum. 🙂
Taking measurement of the bias from both sides. The n-ch side is off .035mV from the p-ch bank. The higher of the two readings is .301mV right now. The heat sinks are warm. Put the lid on it and letting it run for now. Its playing music just fine.
Anything I can do about the off set of the two channel banks? P3?
Thanks
Vince
Taking measurement of the bias from both sides. The n-ch side is off .035mV from the p-ch bank. The higher of the two readings is .301mV right now. The heat sinks are warm. Put the lid on it and letting it run for now. Its playing music just fine.
Anything I can do about the off set of the two channel banks? P3?
Thanks
Vince
Attachments
Totally normal. The Vgs of the N-channel devices is different than the P-channel.
Nope. No need to change it. Just get zero DC offset and be happy.
😀 😀 😀
That's cool. 😀. But, which do I follow? Just the positive side or the side that's higher?
Thanks.
Vince
Thanks.
Vince
I don't understand your question...
Bias until your heatsinks are quite warm after 30min. The difference in the N and P channel is to be ignored. They will still be in the same ballpark.
DC offset is measured at the speaker terminals.
Bias until your heatsinks are quite warm after 30min. The difference in the N and P channel is to be ignored. They will still be in the same ballpark.
DC offset is measured at the speaker terminals.
Just mean what side do I take the bias reading from? What source resistor, n ch or p ch side? Where did you measure from? I know about DC offset... It's 10mV.
Doesn't mater. Any source resistor is fine.
If DC offset is zero or close (and 10mV is fine) the Bias through the rest of the devices will be close.
Once set, then check drop across all the device's resistors just to make sure you don't have any mosfets hogging current.
If DC offset is zero or close (and 10mV is fine) the Bias through the rest of the devices will be close.
Once set, then check drop across all the device's resistors just to make sure you don't have any mosfets hogging current.
Unfortunately this amp was built with sandcast resistors that fit perfectly on the board. Radial resistor leads are hidden. Can only test from test points on output boards.
Thanks Jim, and everyone!
Vince
Thanks Jim, and everyone!
Vince
If your shooting for a certain bias mv number then just average the P & N mv's to get the number you desire.
That's interesting. So in the event that both banks are not equally biased, one can take an average. This is the first amp I worked on with parallel output FETs, so this is new to me.
The reason I was concerned in the first place is because of the diodes in the output stage need to be in a certain temp range so the current doesn't runaway. The article says .400v is where this may happen. This is why I stopped at .300v for the higher bias reading of the two banks.
BTW- the mosfets Vgs measured tolerance was .03v max. For example, one mosfet would be 3.86v and another 3.89v. Some were exactly the same Vgs.
Thanks,
Vince
If I'm working on a commercial amp and it calls for a specific mv drop across the output resistors then I'll usually average the P & N readings. When I'm working on DIY Class A gear the heatsink temp is usually what limits my output bias setting. Of course there's other things to worry about, but most folks will hit a high heatsink temp wall and and have to stop.
So, now the repaired mono-block is noticeably louder than the other mono-block. The idss of the jfets is between 8-10mA, more like between 8-9mA idss. The feedback resistors are the original resistors. They are impossible to confuse, because they are heat sinked and were measured (~110 Ohm) and dropped back in the circuit.
The bias was set to .350v. Can raising and lowering the bias increase gain? I suppose it can to some degree. Maybe if it's higher, it can drive the load better/easier? I'm not sure if Scott took an average of the banks, like discussed above. Maybe the bias reading was .350v and the other bank is .400v or higher, possibly creating an imbalance. Possible?
Any ideas where this extra gain came from?
Thanks,
Vince
The bias was set to .350v. Can raising and lowering the bias increase gain? I suppose it can to some degree. Maybe if it's higher, it can drive the load better/easier? I'm not sure if Scott took an average of the banks, like discussed above. Maybe the bias reading was .350v and the other bank is .400v or higher, possibly creating an imbalance. Possible?
Any ideas where this extra gain came from?
Thanks,
Vince
You might want to consider that the quieter one is the problematic one and that the one you just repaired is fine. I would take a look at the quieter amp and check that the resistor values, etc. are correct in that amp.
I thought about that. All things being equal, the changes were made on the broken channel. However, the problem that precipitated breaking the amp was due to incorrectly trying to adjust the amp to get a warmer sound. But, what I'm saying doesn't prove anything. Also, all the parts in the faulty amp were reused except the jfets, bjt and mosfets. All resistors are the same.
Last edited:
recheck value of resistors in Jfet sources , same as feedback resistors
all that , taking in account that replacement Jfets are genuine
all that , taking in account that replacement Jfets are genuine
- Status
- Not open for further replies.