Cant bias 6L6 amp

[Stratz]

Hi all. I just removed the cathode bias setup on an old Bogen CHA33 and installed a 10K bias pot, 100uf/100v cap, diode etc:
I've done this many times with 100% success but this time no matter what I do I cant get the bias below ~ 85%.

My AC feed is 400v and I have my decade resistance box in place of the dropping resistor but no matter what resistor value I choose it wont balance out. If I have say 35mA then I'll have 490v on the plates. If I change the resistor setting on the decade box to a higher or lower value I get the same problem.
It's always either too much plate voltage or too much grid current, or vice versa

Has anybody run into this problem before?
Thanks.
Rob.

[Enzo]

Changing "the" resistor? Please put up a schematic of exactly what you built. In any amp, B+ climbs as current sinks. SOme of that is due to winding resistance in the transformers. Some is not. Perhaps instead of thinking about "the" resistor, you might think about your whole divider. 400vAC? Fine, is it STAYING 400v or is it adjusting itself as well?

[pdf64]

Due to HT sag, as the power tube bias is cooled off (ie control grid more negative) the HT voltage will rise.
It's normal.
Don't justify excessive plate dissipation by thinking that the VHT will get too high.
The worst thing is to have both excessive.
1st off what is its heater voltage eg 6.3Vac?

[Stratz]

Enzo. I used the Fender Princeton AA964 bias arrangement with the 100k dropping R but with a 10k pot/15k resistor rather than the 27k fixed bias resistor (see attached schematic)

pdf64, it uses 6.3vac filament heaters.

I've used this same circuit on many amps and I can always get it to balance out to 65% to 80% PD. I dont understand why this amp is giving me a hard time. I cant post the Bogen cHA 20 schematic because I downloaded it from Radio Museum and I dont think they allow re-posting of their files but the Princeton schematic is below..

Thanks very much for the replies.
Rob.

[Stratz]

Bogen CHA-33 Schematic

I found the schematic (on this forum no less). It has a few mods to it but none that effect the phase inverter, power section or power supply. Heres a direct LINK which is clearer to read.

Rob.

[Tom Phillips]

I don't expect that you added the Princeton tremolo circuit. Correct?

Tell us what range of bias Voltage your circuit achieves at pin 5 of the power tubes. You can measure that with the power tubes removed if you are concerned with stressing the tubes.

[Stratz]

I don't expect that you added the Princeton tremolo circuit. Correct?

Tell us what range of bias Voltage your circuit achieves at pin 5 of the power tubes. You can measure that with the power tubes removed if you are concerned with stressing the tubes.

No tremolo Tom I'm just using the bias circuit. I've also used the Bassman AA864 bias circuit before as well.

Off the top of my head I can get anywhere from -8v to -60v depending upon which resistor I choose on the decade box. The problem is that any resistor value that choose (within the proper mA or PV range) will not balance between PV and grid current. Turn the pot one way the PV goes doen (normal) but the grid current goes up (normal also) I just cant find a spot below ~85% PD.
I've tried 6L6GB, 6L6CG, 5881's. Currently I have a brand new set of JJ 6L6GC's installed because they're pretty tough and I dont want to ruin any of my good tubes.

Thanks for the reply.
Rob.

[pdf64]

Sorry, it's the actual measured heater voltage I was hoping for.
Without the nominal circuit voltages that have now been supplied, that's the only indicator of whether the PT primary (and hence all its secondary outputs) is being supplied with its intended mains voltage.

...any resistor value that choose (within the proper mA or PV range) will not balance between PV and grid current. Turn the pot one way the PV goes doen (normal) but the grid current goes up (normal also)...

I think you may mean cathode or plate current, rather than grid current?

...Off the top of my head I can get anywhere from -8v to -60v depending upon which resistor I choose on the decade box...

Need an actual measurement of the grid voltage; leaking caps etc could be screwing it up.

[Tom Phillips]

The Bogen CHA-33 does use very high plate voltage on the 6L6s. I think that you may need a more bias voltage. I.e. -65 V or such. This can easily be the case with modern 6L6s with their all over the map tolerances. Also would be good to make sure you are really achieving as much as -60V at the power tube grid pins.

[Stratz]

Sorry, it's the actual measured heater voltage I was hoping for.
Without the nominal circuit voltages that have now been supplied, that's the only indicator of whether the PT primary (and hence all its secondary outputs) is being supplied with its intended mains voltage.


I think you may mean cathode or plate current, rather than grid current?


Need an actual measurement of the grid voltage; leaking caps etc could be screwing it up.

Sorry, Heaters measure ~6.4vac, screen voltage was about 80v less than the plate voltage when it was cathode biased. I dont know if the screen voltage changes with the bias
All coupling caps are brand new if that helps.

[Stratz]

The Bogen CHA-33 does use very high plate voltage on the 6L6s. I think that you may need a more bias voltage. I.e. -65 V or such. This can easily be the case with modern 6L6s with their all over the map tolerances. Also would be good to make sure you are really achieving as much as -60V at the power tube grid pins.

I had my DMM hooked up to the grid while trying to bias it Tom. Depending upon the dropping resistor used I saw between -8 and -60 vdc

[Dave H]

I used the Fender Princeton AA964 bias arrangement with the 100k dropping R but with a 10k pot/15k resistor rather than the 27k fixed bias resistor

Princeton only has to bias 6V6s. For 6L6s with a 100k dropping resistor try a 20k pot/27k resistor.

[g1]

It sounds to me like you need to reduce the 100K feeding the diode.

[Tom Phillips]

To cut to the chase it would be really nice if you would mark up the Bogen schematic with the circuit and component values and voltage readings that you actually have. It doesn't have to be pretty but it could save a few thousand words AND get us to the answer. I.e. One modified amp - one circuit vice two schematics for circuits that are not completely what you have.

[Stratz]

To cut to the chase it would be really nice if you would mark up the Bogen schematic with the circuit and component values and voltage readings that you actually have. It doesn't have to be pretty but it could save a few thousand words AND get us to the answer. I.e. One modified amp - one circuit vice two schematics for circuits that are not completely what you have.

Next time Tom, it's a very good idea.
I finally figured out how to grid bias it. The model I have is from 1956 and it has two 5Y3 rectifier tubes. I've seen later models with a single 5U4 rectifier from the factory. I removed the 5Y3's and rewired one socket for a 5U4GB. It worked great.

I have 460 volts on the plates and 46 mA to the grids which gives me 71% plate dissipation. 460 PV may sound a bit high but the schematic calls for 480v (in 1956 6L6's were probably $0.50) so rather than use a nice NOS set of 6L6 or 5881 power tubes I put in a new set of JJ 6L6GC's's. I'm pretty sure they can stand up to the punishment.

It sounds very nice. ike a Blackface Bandmaster or a Super Reverb if I use my pedal.

I need to make a few labels for the face plate. Input, Vol, Treble. Middle, Bass, Stby, On/Off and it'll be finished.

Thanks so much to all that have helped me so many times. You guys are the greatest.... I really appreciate your help.
Thanks again.
Rob.