Sunn 200S 5AR4 B+ issues

[Mars Amp Repair]

Hey folks,
I did see another post regarding a dude who bought on of these and someone had modified the power supply to accept a solid state rectifier. My experience with the one I have is that even with the original 5AR4 rectifier, the B+ on the filter caps exceeds & almost equals the rating of the caps at 525Vdc. I assume this is because even with the original rated 117Vac, the B+ was already fairly high to begin with so with the solid state rectifiers it definitly exceeds the rated 525Vdc on the filter caps.
Anyway, this customer was using a solid state replacement for the 5AR4 without any mods to the filters, and the filter cap bulged out & spewed its guts under the chassis.
Long story longer, I did the similar thing the previous 200s text did, but from the Fender amps, I knew about the dropping resistors to even out the voltage to each cap that is in series.
NOW, the issue I'm having is that the 5AR4 is intermittently arcing inside when taken off of stby. 5AR4 is the rect the amp originally called for). The spec for the 5AR4 as far as input filter is 60uf max. The equivalent value of the input filter cap is 40uf, so I'm well below that (from 2ea 82uf in series with 220K dropping resistors). I also did the same with the 2nd B+ node filters with equivalent value at 22uf. The line current draw is not excessive.
If I use a 5U4 rect, there is no issue, except the increased sag owing to its specs.
I've already told this customer that given the need to readjust the idle wattage bias on the power tubes for the tube rectifier vs the solid state use, that he needs to decide which one he wants to use & I'll set the amp up for that. I doubt he'd be interested in having me add a switch to the bias ckt to change from tube to SS rectifier and would probably forget to switch it anyway!
Anyone have any ideas as to why this arrangement would cause the 5AR4 rect to behave this way? I calculate the extra current draw from the divider resistors is negligable.
I'm kinda stumped here.
Thanx, glen

[g1]

You have tried several 5AR4's, and not all of the same brand?

[Helmholtz]

Schematic?

[Mars Amp Repair]

You have tried several 5AR4's, and not all of the same brand?

I'll try that. Thanx, glen

[Mars Amp Repair]

Here ya go....thanx

[Tom Phillips]

Might be worthwhile to put the standby switch after the choke if your modified reservoir cap circuit can withstand the resulting unloaded voltage when the amp is in standby mode.

[Helmholtz]

Might be worthwhile to put the standby switch after the choke if your modified reservoir cap circuit can withstand the resulting unloaded voltage when the amp is in standby mode.

While that would avoid hot-switching the rectifier, it is generally not a good idea to wire a switch in series with an inductor.
Opening the switch will generate a large kick-back voltage and arcing between the switch contacts.

I'd consider splitting the reservoir cap into two, like 22µ before and another 22 after the switch.

http://www.valvewizard.co.uk/standby.ht

[Mars Amp Repair]

While that would avoid hot-switching the rectifier, it is generally not a good idea to wire a switch in series with an inductor.
Opening the switch will generate a large kick-back voltage and arcing between the switch contacts.

I'd consider splitting the reservoir cap into two, like 22µ before and another 22 after the switch.

http://www.valvewizard.co.uk/standby.ht

Now that's some really useful information. since it's easiest to just add the trickle charge resistor across the stanby switch, I guess I'll try that first. If unsuccessful for whatever reason, I I'll try moving the standby switch to the other side of the reservoir cap as I have the rating of the reservoir cap combo rated to take 900Vdc and the max the B+ ever gets at the first node w/o load is about 540VDC.
That's a great article.
The only part I have an issue with is that perhaps running the tubes w/no B+ does technically (according to the article) cause issues with 'Poisoning' the Cathode barium coating, the tubes running at idle bias is also going to wear the tubes out prematurely. I kind of cringe at the idea of possibly leaving an amp running at idle bias w/unattended with all that heat being emanated. I guess the question I would pose is whether the cathode poisoning or the running at idle wattage for extended periods will shorten the life of the tube more.
My take: Safety first. Put amp in stby w/unattended eg: when the band takes a break for safety and longevity reasons.
Your thoughts?
Thanx Helmholtz for all this great info. I'll follow up with what worked. glen

[Helmholtz]

Now that's some really useful information. since it's easiest to just add the trickle charge resistor across the stanby switch, I guess I'll try that first. If unsuccessful for whatever reason, ....

Drawback is that the amp will not be completely silent.
Resistor should rated for at least 600V.

I I'll try moving the standby switch to the other side of the reservoir cap ...

Good for the rectifier but bad for the switch as said earlier. A HV snubber cap across the switch should be added.
An example, assuming L = 5H, C = 100nF, I = 0.1A (total idle current) will result in a surge voltage around 700V across the cap when opening the switch.
I don't like this wiring.
A flyback diode across the choke might help.

The only part I have an issue with is that perhaps running the tubes w/no B+ does technically (according to the article) cause issues with 'Poisoning' the Cathode barium coating, the tubes running at idle bias is also going to wear the tubes out prematurely.

Yes, that's a common risk with all standby solutions that disconnect the HT.
But things will not be worse than with the original wiring.

[Mars Amp Repair]

The final solution as to inform the customer to NOT use the solid state rectitfier 5AR4 sub. I was going to modify the bias ckt to switch to allow him to select between SS & tube rect, but determined that the elevated B+ was just not good for any part of this amp.
As for the fuse blowing with the series'd caps, I determined that the reservoir cap really could not be larger than abt 30uf as designed, so used 2 x 47uf Caps in series. The 7uf lower than the orig 30uf made no difference in the power output, so I settled for that.
I also took the advice of G1 on the forum and got a couple of other brand rectifiers & found the Sovtek version seemed to hold up, however I had already changed the reservoir cap to the 23.5uf equiv, so I really can't say or certain.
NOTE: Given the amp was designed for 117Vac, the B+ still measures over 510VDC which is still more than the 10% buffer I believe good design practices require.
BTW, not being certain where JJ or Sovtek source their rectifiers, I just looked at the internal structure to pick something different.
Thanx all for you input on this matter.
I'm happy, cust happy, all is well with the universe! glen