Power transformer gets very hot after one minute idle

[elushi]

I got a problem with my amp. It's not a known brand, because it's from an old Argentinian maker. It’s a very very simple amp: SS rectifier, a 12AX7 in the pre-amp, a 12AX7 in the phase inverter and two EL84 in a push-pull output stage.
The problem is in the power transformer
I wanted to install a stand by switch. So I used the old power switch in the front as the new standby switch and I drilled a hole in the back of the chassis and I installed what it is now the power switch. I also put the whole power supply circuit with the filter caps and the rectifier diodes in a board and took of some resistors from the circuit. What I got now is in red. The resistors I took off are in light blue.

The thing is that when I turned it on, it all worked just fine, no issues. The stand by switch also worked. But what was strange (at least to me) is that when I hit the strings too hard in the guitar, now the amp got mute. No more sound coming out of the speaker. A few days later the amp didn't work anymore. The notes are at low volume and like dirty distorted. This is because only 87 VDC, more or less, are getting to the plates of the power tubes and a little less, like 70VDC to the pre-amp and PI tubes. The power transformer is motor boating (it also makes the whole chassis to vibrate) and gets very hot after a minute or two. It’s not a very strong vibration but certainly it's stronger than before.
No matter if I open the standby switch or even disconnect all the tubes, still gets hot. I supposed the primary winding was shorted, so I disconnected the secondary winding, took a measure and I got almost 600VAC between the two leads of the secondary. Don't know the PT specs but I guess it's 300-0-300 so I guess the measure is correct.
With the secondary winding disconnected the PT doesn't gets hot nor vibrates.
I took also a measure at the 6.3V winding. With the secondary connected it drops down to 5.7VAC. With the secondary disconnected it goes up to 7.1VAC
And I also took a measure to the current. Seems like the poor tranny is drawing about 0.4A! I think i got the answer to the vibration
One of the rectifier diodes was dead because I took it out from the pcb, I measured it and I got 5.5k ohms between anode and cathode and vice versa! I replaced it, turn the amp on and the new diode is now dead too. 10k between anode and cathode and vice versa.
As I am a very newbie, first of all, can anyone explain to me how the he-- the original power supply circuit was rectifying the ac current with those 1Mohm resistors in parallel with the diodes!?
Second, any idea why my amp is dying? Or dead?
Regards

PS: The amp was doing normally before I mess it up. Moreover: I took off those 1Mohm resistors like 5 months ago and the amp was working without them.

[Steve Conner]

The circuit is getting near to the 1kV PIV rating of the diodes. Adding a standby switch made it worse, because the diodes now see the unloaded output of the transformer.

Try two 1N5408s in series for each diode position.

[elushi]

Hi Steve. Thanks for your soon answer! It's my first post.
So you say that (with luck!) the primary winding is not shorted. Just the diodes are blowing up every time I open the standby switch? So that's why the diodes were doing good before I installed the switch. I'm going to try what you said. But anyway why does the diodes see such a high voltage?

[Enzo]

Becuase by adding the standby switch you have removed all the load from the high voltage supply. Normally when you turn it on the circuits are all there and the thing has to come up to its natural voltage. With the standby switch open, there is no such load.

Also, those 10 ohm resistors you removed were there to limit inrush current through the rectifiers. When you power up a discharged capacitor, at first the charging current looks like a short circuit to the power supply. The 10 ohms limits that surge current to make life less stressful to the diodes. Why did you remove them?

And this too: when you have two rectifiers like that, if one of them shorts, look at where the remaining one winds up - it is then directly across the windiing. The shorted one looks like a wire after all. This is stressful to the "good" diode. So the rule is when one of those diodes shorts, replace them both. The remaining one may test good, but it is stressed.

[elushi]

Try two 1N5408s in series for each diode position.

Thank you very much Steve. It worked!

When you power up a discharged capacitor, at first the charging current looks like a short circuit to the power supply

Oook, Now I get it.

Why did you removed them?

The amp is from 1970. The old resistors where so corroded (or something like that) that in the process of soldering the power supply circuit to the board, they both broke into peaces. I replaced them with two 2W metal resitors pair. But both blew up, obviusly. I'll place two 20 ohms 5W instead.

replace them both. The remaining one may test good, but it is stressed

Yes. I figured out that. So the four diodes are new.

Steve, Enzo. Thank you. You always are bringing some help to whoever needs it.