I got this nice little amp to fix. It fizzed. I checked the tube : OK … then I saw R39 resistor was burnt.
I spiced the FET PI, replaced the burnt resistor and compare voltage : good !
(I used IRF150 model instead of IRF830 …)
A few days later, my customer came back : his amp had the same problem, again !
R39 burnt !
I found this topic on the Blackstar forum (very helpful, thanks a lot, guyz !) : http://www.blackstaramps.com/forum/viewtopic.php?f=32&t=2135&sid=6267625e63d87d9553d419c17f908d83
Trace broken, to rebuild …
Once TR3 is removed, I could check it (see : http://www.decibelcar.com/menuelectro/93-fettest.html , very helpful, thanks !) : TR3 OK
R39 and C46 removed, let’s check if there is a short somewhere :
We should get an infinite resistor between Source and Drain holes … and we read 50kΩ !
It seemed this amp had a bad PCB, so I drilled hole (0.7mm) to isolate TR3 pins, as suggested in the Blackstar forum :
and checked :
I removed R38, TR2 and C45 : no short there !
Soldered new 22k resistor, TR3, C46 … amp to the test !
Tips : Measure voltage across 4,7ohm resistor on cathode (=R33, common 12BH7 cathode resistor). Adjust to mV reading of approximately 46mV. Adjust balance pot for least amount of hum from speaker
12BH7 voltages :
- Anode : 470V (!)
- Grid 1 : -29V
- Cathode : 45mV
Hope this helps
I got another Blackstar HT5 to check. In this one, there is a new designed PCB (I didn’t notice the “R39 bug”) … but how to bias this new version ?
Here is how I did (I’m very sorry, the photos are of poor quality) :
PR1 : bias, PR2 : balance
To check balance, measure voltage between blue wire and ground, and compare to orange wire and ground (blue and orange wires go to 12BH7 grids 1, triode #1 and triode #2).
To bias the amp, measure between R33 (red arrow) and ground (or voltage drop across the resistor R33), you should read 46mV (470V at the anode) : bias PR1 pot !