Bassman and selenium rectifiers

[paggerman]

Hi all, and happy new year to you all. Some time ago I acquired a power amp from an unknown organ. I started to trace the circuit and soon realised that it had been hacked with all sorts of power resistors soldered in series and left to flap around. It had a pair of 7519a power valves both of which test Ok. Knowing that these valves are rare, I am thinking of putting them up on eBay to partially fund a Bassman type build on the chassis and transformers. I have a choke lying around that I could use.

Does anyone have any suggestions for the model of Bassman to base my build on (accepting that I will need to make compromises according to the iron)? A single rectifier model will probably be easier to fit on the chassis.

Secondly, the Bassman diagrams I have looked at on Schematic Heaven show a selenium rectifier for the bias circuit. Can I simply replace this with a 1N4007 or is there a preferred substitute?

Very many thanks in advance

Jeff

[Steve Conner]

Yes, a 1N4007 will do fine.

There's a lot of "mojo" surrounding the Bassman. The 5F6A is the classic version that most people choose to clone.
Fender Bassman - Wikipedia, the free encyclopedia

[MWJB]

The 5F6A is a bunch of versions and nearly everyone clones a slightly different version ...of that version ;-). Each permutation has its fans, one thing I would recommend is using the 59 Bassman RI/6G12A Concert style presence control (4.7K load resistor, 25K pot), as the version shown on the original Fender 5F6A schem (& Marshall JTM, 5K pot, no fixed load resistor) is noisy in use.

Also build in to the design the 1.5K grid stoppers that connect to power tube grids (see Super Reverb AB763 for power tube socket wiring if using 6V6/6L6 style tubes). These are not shown on the original Fender schem.

Instead of a single 56K resistor for the bias circuit load, use a 50K cermet trim pot, wired as a variable resistor, in series with a 33K-39K fixed resistor to ground.

As Steve says the 1N4007 is a good choice for the bias diode, no one uses the selenium rectifiers nowadays (selenium is toxic & the selenium rectifiers produce a noxious smoke when they fail), the change will have no impact on tone.

A typical bassman runs 470vdc (+/- 10vdc) at 30-35mA per tube with a GZ34 rectifier. If your PT can put out that kind of voltage, wire the first filter cap after the rectifier as per the Super Reverb AB763 also (2x100uf 350v caps in series, each bypassed with a 220K 2W resistor).

[km6xz]

If you want to replicate the power supply of Bassman that used the selenium rectifiers be cautious of vintage rectifiers, they would work for years in operation but derated gradually if stored without current flowing through them. They also generate a little heat because they are losing about 20-25% of power going through them which is turned into heat. Their forward drop is about 1 volt per plate and the there will be a number of plates stacked to get the desired reverse voltage capability. The number of plates time about 25 volts will be the reverse rating of the part and the current will be based on the area of each plate. I remember building a monster power supply for a surplus ART-13 transmitter and I stacked a dozen selenium stacks from old TV sets to get the current and voltage rating needed for the transmitter's 813 and 811 anodes, 1,150 volts dc. I used that transmitter all through junior high and high school and that was the last time I built anything new with those large and inefficient rectifiers except for meter rectifiers that were tiny and are still used in a couple commercial vom's.

[MWJB]

I think we're talking at crossed purposes here, the 5F6A Bassman used a GZ34/5AR4 in a typical, centre tapped, full wave B+ supply. The selenium rectifier was just for the bias supply's negative dc voltage.

[km6xz]

Sorry, I was not implying that selenium rectifiers be used in the HT supply but how to identify one that would work, in this case, one with 3-4 plates for the bias supply. He would probably only find surplus or old parts that had no part numbers or specs so to make a suitable bias supply as a reproduction would require identifying one on sight.
I was just letting someone know a bit about them because not many people building today worked with them first hand when they were new stock parts. Few people under 45 have ever heard of them.

[paggerman]

Thanks - and an observation

Sorry, I was not implying that selenium rectifiers be used in the HT supply but how to identify one that would work, in this case, one with 3-4 plates for the bias supply. He would probably only find surplus or old parts that had no part numbers or specs so to make a suitable bias supply as a reproduction would require identifying one on sight.
I was just letting someone know a bit about them because not many people building today worked with them first hand when they were new stock parts. Few people under 45 have ever heard of them.

Yep. 43 years old and had never heard of selenium rectifiers til I started amp building two years ago. Many thanks for your reply. Your posts are consistently valuable to me.

[paggerman]

Great tips.

The 5F6A is a bunch of versions and nearly everyone clones a slightly different version ...of that version ;-). Each permutation has its fans, one thing I would recommend is using the 59 Bassman RI/6G12A Concert style presence control (4.7K load resistor, 25K pot), as the version shown on the original Fender 5F6A schem (& Marshall JTM, 5K pot, no fixed load resistor) is noisy in use.

Also build in to the design the 1.5K grid stoppers that connect to power tube grids (see Super Reverb AB763 for power tube socket wiring if using 6V6/6L6 style tubes). These are not shown on the original Fender schem.

Instead of a single 56K resistor for the bias circuit load, use a 50K cermet trim pot, wired as a variable resistor, in series with a 33K-39K fixed resistor to ground.

As Steve says the 1N4007 is a good choice for the bias diode, no one uses the selenium rectifiers nowadays (selenium is toxic & the selenium rectifiers produce a noxious smoke when they fail), the change will have no impact on tone.

A typical bassman runs 470vdc (+/- 10vdc) at 30-35mA per tube with a GZ34 rectifier. If your PT can put out that kind of voltage, wire the first filter cap after the rectifier as per the Super Reverb AB763 also (2x100uf 350v caps in series, each bypassed with a 220K 2W resistor).

Great info. Once I'm back home and settled down I'll start drawing up a design that incorporates these features. It would appear that not all Bassmans are created equally. Thank you.

[pdf64]

Another minor mod I find beneficial for 5F6A type amps (when used for guitar) is to wire the mid control as a variable resistor, ie link the wiper and hot track end. It provides more bass cut at the CCW end of the bass control, without losing the ability to get any of the 'pre-mod' settings.
Admittedly this doesn't have the reliability / stability improvements that the others mods do, it's purely an 'extending user control' type of thing.

[nashvillebill]

This is a selenium bridge rectifier from an old Grundig radio. Ain't those blobs of melted selenium lovely?

[paggerman]

Thanks for all the advice thus far. One more question(for now), if you would: having previously built only cathode bias amps I am soaking up as much as I can about fixed bias before embarking on this project. I have looked at several Bassman schematics over at Schematic Heaven but on none of them is there anything to adjust the bias. The 'how to' at aikenamps.com talks about looking for trim pots. Is this what MWJB refers to in his tips (50kcermet)? If so, why only one? My interpretation of the talk about bias was to ensure that both valves in a pp configuration had the same current running through them. Therefore is it not necessary to adjust each valve individually to obtain an optimised 'target' current?

Was bias in the original Bassman designs simply ignored to cut production costs or was there some other voodoo involved?

Back home in 10 days, can't wait for my fix of solder fumes.

Many thanks

[MWJB]

Yes, the 50K cermet pot as part of the bias voltage divider load is the bias trim pot.

Buy matched tubes (+/-5mA, any competant tube vendor can do this) and you only need one bias trim pot...I find balance/individual pots are a bit of a bind, as the idle currents see-saw, turning a 30 second job into a few minutes. At 30-35mA per tube you might not notice a difference of 2 or 3mA...sometimes you might get away with, or even prefer a mismatch of a little over 5mA (tames highs, some feels it "warms" the tone")? My preference is for a dead match, but if I don't have that to hand, I have a listen & if I can't hear anything "wrong" then, as far as I am concerned, there isn't anything "wrong".

Fender gradually reduced the idle current over the production run of the 4x10" bassmans (& even colder come the 6G# amps), as long as they didn't have tubes fail under warranty they were happy. Bias wasn't "ignored", it was just set to "satisfactory production tolerances" (as it is on many, many mass produced amps made today) and by negative voltage, rather than by mA per tube.