redplating in power section - baffeled could use some help

[spinfresh]

Hello, and thanks for some feedback in advance.

here is what I have going on,

So far I have build 3 of these amps and they all have the same issue,

each amp uses different tubes, different transformers and different B+ voltage, other than that everything else is the same.

1. el34 - 465v at tube, and 3.2k OT
2. 6v6 - 426 at tube, and 6.8k ot
3. el84 - 440v at tube and 6.8k ot

also they are all fixed bias with the correct amount of ma's for b+ voltage

each of the amps have 8 diode rectifier with .001 caps jumping each diode, and each amp has a center tapped transformer.

the power supply consists of 100uf/100k to ground x2 similar to a marshal

the preamp tubes go through another 32uf 22k x2 for filtering

each amp uses 2 output tubes

so in my shop the amps are stable and work perfectly however when I take them to practice the power tubes red plate, (regardless of if the preamp has tubes in it or not)

the line voltage at my shop is 117v and it has new wiring and in practice space has 124vac and old hacked up wiring.

any ideas? -- thanks in advance.

BTW, the lead dress is close to perfect, so are the heater wiring, all inputs are shielded, I have tried 100 ohm 470 ohm, and 1k grid stoppers, as well as removed all of the preamp tubes and they still redplate with the preamp tubes removed... also in my shop the bias holds solid no fluctuation.

P

[spinfresh]

here is the schematic for the powersupply - the bridge is not grounded due to the CT on the transformer, other than that it is more or less identical.

[Blackburn]

You're using a bridge rectifier with a center tapped PT? Are you grounding the CT or floating it and grounding the bridge?

[spinfresh]

I am grounding the center taps on all 3 amp and floating the bridge

[spinfresh]

If I ground the bridge I get too much b+ since my transformers are 270 0 270. And the schematic with the non ct transformer calls for a 190 190 power trans.

[ToneMerc]

Unless i'm looking at this wrong I would expect it to redplate the way the bias is wired to the standby switch. When in standby mode, your bias supply is open.

TM

[tubeswell]

Put your standby switch between the rectifier and the reservoir cap. Seems to me that the way your schematic shows the wiring, you don't have any bias unless the standby is flipped on, but 1/2 of your rectifier is still supplying current pulses to the the filter section, and the heaters are on - so yes the output tubes will red plate.

[spinfresh]

Unless i'm looking at this wrong I would expect it to redplate the way the bias is wired to the standby switch. When in standby mode, your bias supply is open.

TM

the way I see it and the way it seems to work is the standby switch turns off both the bias and the b+ 2 of the amps has the bias pulled from the B+ windings, and the other uses a dedicated bias tap on the transformer. both redplate the same

[LOUDthud]

The bias supply in the schematic is not appropriate if the center tap of the HV winding is grounded. That circuit is intended for a bridge rectifier arrangement where the minus side of the bridge is grounded. It tries to force the peak voltage of the entire winding on the 64uF cap. Take a look at the Trainwreck Express schematic for a proper circuit.

[Phil_S]

Unless i'm looking at this wrong I would expect it to redplate the way the bias is wired to the standby switch. When in standby mode, your bias supply is open.
TM

the way I see it and the way it seems to work is the standby switch turns off both the bias and the b+ 2 of the amps has the bias pulled from the B+ windings, and the other uses a dedicated bias tap on the transformer. both redplate the same

The evidence is overwhelming. You have runaway bias voltage when under full power. Dontcha think maybe the logic has a flaw? BTW, you don't need a standby switch. Simplify the whole thing be eliminating that switch. (You can search here for discussions on the topic. I'm not wanting to expand this thread on something that's been covered here extensively. Or check Merlin's site for a good article.)

[Littlewyan]

So to summarize:

1. Ditch the Bridge Rectifier and go with a Full Wave Rectifier like the TW Express has.

2. Either eliminate the standby switch or at least put the bias supply before it.

Edit: That bias supply is very....different. Never seen anything like it. Where did you find that?

[Banzini]

Edit: That bias supply is very....different. Never seen anything like it. Where did you find that?

It looks like the bias supply from the Vox AC50.

[gui_tarzan]

I'm confused. Isn't the bias supposed to be negative DC? As drawn this looks like it would be positive to the power tubes.

I'll rephrase, it looks like the diode is in the wrong place. Shouldn't the diode be inline with the path going to the tubes? In this pic it goes from the supply to the ground.

[martin manning]

I'm confused. Isn't the bias supposed to be negative DC? As drawn this looks like it would be positive to the power tubes.

The output from the 4uF series 4uF (=2uF) is AC voltage. The diode dumps positive voltage to ground, while negative voltage charges the 68uF cap through the 47k. As the cap charges the mid-point of the waveform is pushed down to the DC bias voltage.

The way it's set up presently, with the CT of the transformer grounded and the bridge ungrounded, it's operating as a FW rectifier.

[JazzGuitarGimp]

The positive half-cycles are shunted to ground through the diode, leaving only the negative half-cycles to charge the bias supply filter cap (the 64uF, 64V lytic), which, BTW is shown correctly with its positive terminal at GND.

As long as the negative side of the FWB is left floating, the pair of diodes on the back (negative) end of the bridge have no effect on the power supply.

This circuit should work, cutting both B+ and bias simultaneously via the standby switch, but I tend to be a "suspenders AND belt" kind of guy when it comes to design. So I would be more inclined to not interrupt the bias supply with the standby switch. I would also use a more conventional bias supply circuit.