How wild is the swing (Iq)? If it's going off the charts then I suspect it could be the Ge but in my case, it was only swinging because I had it on a lab supply and it was hitting the current limiter. On linear PSU, it simply drew enough to trip the fuse. One way to confirm the rest of the board is to switch out Ge for Si and remember to change R122.
Just to confirm, you haven't connected the jumper at this point have you? Still only dealing with the OS is that correct? (not FE + OS)
Here's a tip for testing with the Ge - don't run the legs through the holes, just bend them up and under so it's easy to desolder later. I had to amputate 2 of the 1943 because they didn't want to leave peacefully no matter how nicely I asked.
Stephen
Just to confirm, you haven't connected the jumper at this point have you? Still only dealing with the OS is that correct? (not FE + OS)
Here's a tip for testing with the Ge - don't run the legs through the holes, just bend them up and under so it's easy to desolder later. I had to amputate 2 of the 1943 because they didn't want to leave peacefully no matter how nicely I asked.
Stephen
Hi Stephen,
For the misbehaving channel I'm still at the initial OS bias stage. The front end has not been adjusted and the jumper is not in place.
The measurement across the 0.22R resistor goes from about 60mV to 80+mV. It might bounce around within a few mV for a while and then suddenly start dropping down to 60-ish mV. The P104 pot also seems to not be able to really lower the DC offset. The dc offset moves around a fair bit as the bias changes so I'm not sure how much effect P104 is having.
Thanks,
Dennis
For the misbehaving channel I'm still at the initial OS bias stage. The front end has not been adjusted and the jumper is not in place.
The measurement across the 0.22R resistor goes from about 60mV to 80+mV. It might bounce around within a few mV for a while and then suddenly start dropping down to 60-ish mV. The P104 pot also seems to not be able to really lower the DC offset. The dc offset moves around a fair bit as the bias changes so I'm not sure how much effect P104 is having.
Thanks,
Dennis
Well, the amp just blew a fuse. I was playing a signal on the 'good' channel while I tried again to adjust the other one when that happened. Now I have find more fuses and try figuring out which channel the fuse to blow.
Dennis
Dennis
Dennis , you're Old Dog, so I'll not ask for proper population of parts (position, value, orientation)
sole thing coming to my mind is that Ge is always first suspect
do you have spare one, or - use that one from already proven channel, to try second channel
sole thing coming to my mind is that Ge is always first suspect
do you have spare one, or - use that one from already proven channel, to try second channel
blown fuse - same thing happened to me with first Russkie I put in prototype - it behaved perfectly for setting, but gone south on first sign of signal, even without load
check, Ge, check IRFP also
if Ge is Dodo, replace IRF9510 too
maybe in this moment is best to try both channels with Si P polarity , to be sure that pcbs itself are operational
then, when pcbs are verified, start putting Ge in, and to lessen possible damage - put temporary fuses in rails
say that 2A is more than enough, including listening test with moderate power
check, Ge, check IRFP also
if Ge is Dodo, replace IRF9510 too
maybe in this moment is best to try both channels with Si P polarity , to be sure that pcbs itself are operational
then, when pcbs are verified, start putting Ge in, and to lessen possible damage - put temporary fuses in rails
say that 2A is more than enough, including listening test with moderate power
Thanks for the suggestions. I have some 2n1099 of different brands (along with the pair you had kindly sent). Stephen also gifted me a pair of the Si parts.
After a couple more fuses I can confirm the first channel is still good and plays music.
Before the fuse blew, I had reset the trimmers on the 'bad' channel and restarted the OS bias process. I think I
got as far as about 60mV across the 0.22 resistors when the fuse blew. I noticed that even at around 30 to 40mV
the measurement was wobbly so the Ge part probably was already showing problems...
FYI, I used a pair of GM Delco 2n1099 because I thought it would be amusing to have GM parts in an amplifier. 🙂
I hope to have time to take that channel apart tonight.
Cheers,
Dennis
After a couple more fuses I can confirm the first channel is still good and plays music.
Before the fuse blew, I had reset the trimmers on the 'bad' channel and restarted the OS bias process. I think I
got as far as about 60mV across the 0.22 resistors when the fuse blew. I noticed that even at around 30 to 40mV
the measurement was wobbly so the Ge part probably was already showing problems...
FYI, I used a pair of GM Delco 2n1099 because I thought it would be amusing to have GM parts in an amplifier. 🙂
I hope to have time to take that channel apart tonight.
Cheers,
Dennis
when dealing with proven constructions***,both channels in same case, I'm setting first channel with second disconnected and then I'm leaving first one connected while working same procedure with second; of course - battery of DMMs staying connected on first one, me having half an Eye on it
***Old SOul being exception - each Ge part needing simply more time to be verified ...... sorta reminding me of proper matchsticks proofing procedure
***Old SOul being exception - each Ge part needing simply more time to be verified ...... sorta reminding me of proper matchsticks proofing procedure
Are you building a T-Ford, esteemed Mr. Hui?
Hope you crack it. Or, I know you will 🙂
Regards,
Andy
Hah!
Perhaps we're seeing some early failures with these old Ge parts.
So far I'm read in this thread:
1) ZM had one (not 2n1099) bias up ok but failed with test signal
2) Stephen had one fail after 6 hours of use.
3) I had one fail during OS bias procedure.
I'm hoping that running continuously for a day will catch the early failures.
Any thoughts on testing these parts outside of an amp to weed out the weak parts?
Given ZM and Stephen's experiences a quick test will likely not do it.
Perhaps we're seeing some early failures with these old Ge parts.
So far I'm read in this thread:
1) ZM had one (not 2n1099) bias up ok but failed with test signal
2) Stephen had one fail after 6 hours of use.
3) I had one fail during OS bias procedure.
I'm hoping that running continuously for a day will catch the early failures.
Any thoughts on testing these parts outside of an amp to weed out the weak parts?
Given ZM and Stephen's experiences a quick test will likely not do it.
Perhaps you could mount them on some sinks and run them for an extended period either at or a bit over specs for the amp. Test to use case vs. test to mfr. rating or test to failure.
I realize that method won't take into account the failure mode of when a test signal is applied. I'm scratching my head over that one... wondering if it had something to do with it actually passing a signal vs. bad timing. Only ZM would know if it oscillated etc., and I did not read the specifics.
AC/Delco was known for making spark plugs.
I realize that method won't take into account the failure mode of when a test signal is applied. I'm scratching my head over that one... wondering if it had something to do with it actually passing a signal vs. bad timing. Only ZM would know if it oscillated etc., and I did not read the specifics.
AC/Delco was known for making spark plugs.
just put it in amp but install temporary 1A rail fuses
if biased, apply some mild signal, leave it for half a day
if biased, apply some mild signal, leave it for half a day
Yeah, I'll try to get some inline fuse holders.
The Delco transistor in this case certainly fulfilled the promise of its name. 😀
The Delco transistor in this case certainly fulfilled the promise of its name. 😀
Dennis
Hope it works out for you !!! I will be trying the same shortly with Delco parts - I have 5 or so to play with - also have some of the big Silicons if you need a pair
Best
Bob
Vikings, polite, smiling ...... ?
Timeo Danaos et dona ferentes..........
Sometimes this gang (zm andynør peppe + me) are just a naughty pack… tsstsstss
Hi Bob,
Thanks for your kind offer. Stephen gifted me some 2sa1943 when I saw him last month.
I am really looking forward to getting this working.
Have fun with your build.
Dennis
Thanks for your kind offer. Stephen gifted me some 2sa1943 when I saw him last month.
I am really looking forward to getting this working.
Have fun with your build.
Dennis