Bias selenium rectifier replacement

[lo-fi]

Hi,
I convinced myself to get rid of the old selenium rectifier in the bias circuit of an old Dukane 1U460 PA. It was still working but I know they're famous for failing after so many years.
So I installed a IN4007 silicon diode in place, also replacing the two old 20/150 caps with 22/450 double checking proper polarity (+ to ground).
The diode cathode (marked side) was soldered on the - side according to the schematic.
When I switched on the amp (no power tubes for safety reasons) I measured a totally different bias voltage: -26.80V (selenium, tubes installed) vs. +9V (IN4007, no tubes). It looks like the bias circuit is now far from specs.

Did I assume it was too easy? Can you see any big mistake in what I did? At this point I guess the original resistors values do not match the new scenario with IN4007 diode.
Any tip?
Thanks.

[Jazz P Bass]

The marked band of the 1N4007 is the Cathode.
That should connect to the transformer wire.

Dukane 1U460.pdf

[Helmholtz]

Hi,
I convinced myself to get rid of the old selenium rectifier in the bias circuit of an old Dukane 1U460 PA. It was still working but I know they're famous for failing after so many years.
So I installed a IN4007 silicon diode in place, also replacing the two old 20/150 caps with 22/450 double checking proper polarity (+ to ground).
The diode cathode (marked side) was soldered on the - side according to the schematic.
When I switched on the amp (no power tubes for safety reasons) I measured a totally different bias voltage: -26.80V (selenium, tubes installed) vs. +9V (IN4007, no tubes). It looks like the bias circuit is now far from specs.

Did I assume it was too easy? Can you see any big mistake in what I did? At this point I guess the original resistors values do not match the new scenario with IN4007 diode.
Any tip?
Thanks.

Positive transformer current flows from diode anode to cathode (marked side). This positive current charges the connected side of the cap to a positive voltage.
You may have damaged the bias filter caps by reverse voltage.

[lo-fi]

Thank you! I supposed the minus on the schematic should match the negative side of the diode (cathode). That's why I soldered it reverted.
It was not that difficult for the schematic to mislead my ignorance!
Now voltage is back in the proper range even though the declared value produced a quite cold bias for that class AB amp: -33V ==> 14.8mA (measured using transformer shunt method on a 5881 pin 3).
Before removing the selenium rectifier I set bias so that plate current was 32.5mA.

Thanks again for your kind support.

[lo-fi]

Unfortunately it's not over. Soon after having checked (with no power tubes) that bias voltage was correct I tried to reinstall the two 5881s.
As soon as I turned on the amp I noticed a blue glow on the 5U4GB rectifier and heard hum on the speaker. The meter started reading current on power tube's plate but the value seemed wrong since the very begunning. I turned it off immediately.
I fear the bias diode initial wrong mounting could have damaged something in the power supply circuit. The amp powers on but some major problem occurred.

[patlaw]

Are you using a light bulb limiter? If not, you might want to.

[lo-fi]

No, I'm not. I don't have it.
I thought it was an easy task but my stupid mistake has made it probably way complex.

I removed both rectifier and power tubes. The amp powers on and voltages on the PT secondary side are correct, including heaters. Fuse won't blow. Probably the old TungSol 5U4GB was disgusted by my ignorance and preferred retiring after so many years!
Or maybe one or more filter caps failed.
Bias voltage is correct and stable with the IN4007 so I guess my mistake did not damage bias electrolytic caps.
It's not easy for me to troubleshoot all those can caps. I have a spare NOS 5U4GB here but I'm not confident to plug it at this point.

(Needless to say English is not my native language. I apologize for everything that does not make too much sense)

[J M Fahey]

I was answering to an old question, sorry.

[lo-fi]

Yes, I know. I fixed the reverse diode mistake soon after Jazz P Bass advice and the bias circuit is now working fine.
I fear that during the short initial test with wrong diode positioning some other component of the amp was damaged. Hopefully it's just the rectifier tube but at the moment I don't know.
It's getting late on this side of the Atlantic. Will resume the job tomorrow.
Thanks everybody!

[mozz]

Are the filter caps original?

[J M Fahey]

It's getting late on this side of the Atlantic.

Which is? ..........
Please fill your location data.

[lo-fi]

Are the filter caps original?

Yes, they're still the original filter caps. I know I should have already replaced them all but since the amp was working fine (no hum nor noise) I started from the old selenium rectifier.
I guess the three cans will have to be replaced. Quite an hassle to figure out the layout assuming I'll have to use single caps to place somewhere in the chassis.

Anyway, the selenium rectifier replacement was completed. All the rest is a different matter, probably off-topic according to thread's subject.
Lesson learned: never start modding an old amp before you are not totally confident about what you're doing (never messed with bias rectifiers in the past). I supposed I knew what the schematic meant. I was wrong and now I'll medidate about it in the long and boring time required by troubleshooting the consequences of my superficiality.

Thanks, folks!

Oh well, I'm from Italy. Been causing troubles to tube amps since late '90s (I recall the early days of Ted Weber's Bullettin Board. Maybe someone from back then is still around here!)

[patlaw]

Everyone here has done something as ill advised as you did. (We usually say "stupid", but wrong decisions are not always stupid.) You will learn more from fixing it than you would had everything gone as planned. You learn the most from your mistakes. In other words, everyone one of us has "been there, done that."

[pdf64]

The key point is that you minimised the damage by verifying your work with the tubes out of the amp; it could have been a lot worse else.
We all have to make assumptions that we're not certain of occasionally.
Rather than meditate, I suggest you build yourself a light bulb limiter (and make sure to use it at the initial power up of new builds or repairs)

[g1]

Aside from the actual bias filter caps (C5, both halves), and maybe rectifier tube or power tubes, I don't think you would have caused any other collateral damage.
If the big filters are bad, they probably were anyway.

edit: looks like you didn't run the power tubes without bias, so you won't have damaged them either.