Hello to all.
I decided to build the mentioned amp, becouse the one other that I got blew.
The schematic can be seen here
So I fallowed the construction guide and assembled the amp. Everything seemed to be ok. As I vas setting the bias I noticed, that the current was increasing with the temperature increase of T6,7,9 and 10 (they are mounted on common heat sink). I'm worried about this becouse of thermal runaway. When the temperature of this common heat sink settled the bias was steady, but heat sink was extremely hot to the touch. I've let the amp run about 1 hour. Everything seemed to be ok, except the small common heat sink teperature. So I decided to test the amp with some loudspeaker. I connected everything and there was some noise comming out of speaker (not humm). So I disconnected the input (I tought maybe it's the faulty cable) and then it happened. As soon as I disconnected the input I noticed smoke and the noise went crazy (I almost crapped myself
). So I instantly turned the amp off.
The resistor R20 lost his smoke. I measured every transistor and fet! They are all ok. Just the R20 blew!
I checked and rechecked every connection and whole PCB assembly before I wired the thing together. All was OK!
I was running the amp on +- 81V!
I can't find the error, so maybe someone can help me!?
I have no idea what have gone wrong!
I decided to build the mentioned amp, becouse the one other that I got blew.
The schematic can be seen here
So I fallowed the construction guide and assembled the amp. Everything seemed to be ok. As I vas setting the bias I noticed, that the current was increasing with the temperature increase of T6,7,9 and 10 (they are mounted on common heat sink). I'm worried about this becouse of thermal runaway. When the temperature of this common heat sink settled the bias was steady, but heat sink was extremely hot to the touch. I've let the amp run about 1 hour. Everything seemed to be ok, except the small common heat sink teperature. So I decided to test the amp with some loudspeaker. I connected everything and there was some noise comming out of speaker (not humm). So I disconnected the input (I tought maybe it's the faulty cable) and then it happened. As soon as I disconnected the input I noticed smoke and the noise went crazy (I almost crapped myself
). So I instantly turned the amp off. The resistor R20 lost his smoke. I measured every transistor and fet! They are all ok. Just the R20 blew!
I checked and rechecked every connection and whole PCB assembly before I wired the thing together. All was OK!
I was running the amp on +- 81V!
I can't find the error, so maybe someone can help me!?
I have no idea what have gone wrong!
I have a very similar circuit running on 50V rails and its stable. However VAS and its current source are not mounted to a heatsink and the drivers doesent have a heatsink.
The drivers get hot but not burning hot.
Its possible that 81V rails are just too much since it was designed for 75V rails.
The drivers get hot but not burning hot.
Its possible that 81V rails are just too much since it was designed for 75V rails.
You say that T6,7,9 and 10 were on a common heatsink but for the temperature compensation of the bias to work properly, it is also important that T8 is mounted on the MAIN heatsink with the output transistors. This is just a suggestion as you doesn´t say anything about T8 in your first post.
Also, are you sure that your amp really does have an adequate heatsink for use with higher voltage rails. Rails of a higher voltage will increase dissipation in the output transistors and may result in the idling temperature being quite a bit higher than expected.
As Tekko says, it may well just be that +/-81V is pushing it too much for an amp designed to work on +/-75V. Remember with +/-81V rails, it is a total difference rail to rail of around 162V which certainly isn´t a low voltage.
Also, are you sure that your amp really does have an adequate heatsink for use with higher voltage rails. Rails of a higher voltage will increase dissipation in the output transistors and may result in the idling temperature being quite a bit higher than expected.
As Tekko says, it may well just be that +/-81V is pushing it too much for an amp designed to work on +/-75V. Remember with +/-81V rails, it is a total difference rail to rail of around 162V which certainly isn´t a low voltage.
The layout is not so good.
Decouple the power supply to the differential amplifier with 100nF.
Ok, you try to help me, but I dont know how to do that
. There are 100nF capacitors from positive end negative rail to GND just to the right of small common heat sink.
The layout is not so good.
Decouple the power supply to the differential amplifier with 100nF.
Hi,
Already done with two 100nF caps (close to driver heatsink).
Quasi's DIY Audio Site
No problem,
I draw it it on the layout.
You should never ever forget that every wire is a antenna or transmitter.
The red one is the antenna. From the capacitor to the end. I draw the position for the additional capacitors in green.
And the blue line is a loop, not bad but also not good and it can not be changed.
http://hostarea.de/show.php/134426_layout.jpg.html
I hope you are not green-red-blind
I draw it it on the layout.
You should never ever forget that every wire is a antenna or transmitter.
The red one is the antenna. From the capacitor to the end. I draw the position for the additional capacitors in green.
And the blue line is a loop, not bad but also not good and it can not be changed.
http://hostarea.de/show.php/134426_layout.jpg.html
I hope you are not green-red-blind
moschfet,
Ain't your two (?) green ground wires forming one big loop?
Edit:
Although, personally I would have designed the layout differently. Reversing direktion of the powerstage for a compact ground with small loops and feedback pickup in the right place. I also don't understand Quasi's reasons running groundlink in parallell with the output trace. What could be the reason for doing so?
Ain't your two (?) green ground wires forming one big loop?
Edit:
Although, personally I would have designed the layout differently. Reversing direktion of the powerstage for a compact ground with small loops and feedback pickup in the right place. I also don't understand Quasi's reasons running groundlink in parallell with the output trace. What could be the reason for doing so?
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.