Bias probe/Tweed Twin 5e8-a weirdness...

Blueshound

Member
Messages
413
I recently acquired a Cox Tweed Twin 5e8-a. It’s a straight up repro, it’s like it was built right off of the old Fender 5e8-a service layout. When checking the bias with my Weber bias heads into my fluke dmm, I was getting really low numbers (10mA-20mA), regardless of whatever tubes I tried. This seemed weird since the amp sounded really good, not like you would expect an amp to sound at such low bias reading. It broke up like you would expect a tweed amp to do and had strong bass response.

On a hunch, I checked the bias as detailed on Rob Robinette’s website https://robrobinette.com/How_to_Bias_a_Tube_Amp.htm using the transformer resistance method. I was somewhat surprised to find the bias readings were now much more in line with what I would have expected based on how the amp sounded (anywhere from 26mA to 40mA).

Somewhat confused by this, I checked my Victoria Silversonic with the bias probes and then compared the numbers against the transformer resistance method. The numbers were almost identical.

What might account for this? I had never suspected there was any issue with my bias probes prior to this, and they seemed to work just fine on my Victoria.
 

pdf64

Member
Messages
7,281
So with the transformer resistance method, where exactly are you putting the meter probes?
With either method, how about if the tube in V4 is removed?
 

Blueshound

Member
Messages
413
When taking the output tube plate voltages one lead was on a chassis ground, the other was on pin 3 of whichever tube I was taking readings on. When checking the center tap voltage, I was connected onto the standby switch and chassis ground. When checking the transformer resistance, I was on the standby switch and pin 3 of whichever power tube I was measuring the resistance to. I didn’t take any measurements with V4 removed.
 

pdf64

Member
Messages
7,281
Would the correct location to measure the OT center tap be where the choke feeds into the two 16 mfd filter caps?
That should be right, according to the Fender layout; whatever it should be where the red wire from the OT terminates on the board.

if its anything like later fenders,no, to the cap (s) before the choke
It’s not, hence screwy results.
https://el34world.com/charts/Schematics/files/Fender/Fender_twin_5e8a_schem.pdf
 
Messages
1,332
the 5e8a choke must be a hefty unit if it's passing all that current through it, that is intriguing. That choke would have some kind of regulation effect for the output tube h.t current, not quite like a choke input, but something surely, again, this is quite intriguing, why did they do it I wonder?
 

zenas

Member
Messages
8,697
The placement of the choke is common on the other "e" tweed circuits like the 5e7 Bandmaster and pretty much identical Super and Pro circuits. (forget the model numbers now) Are others having trouble getting readings with their bias probes on those circuits? I don't use a probe so I'm curious.
 

pdf64

Member
Messages
7,281
Breaking news - tube amps can sound great despite their power tubes idling somewhat below some proscribed dissipation % :eek:
I suspect that the reading from the bias probe is fine, and that the result from the transformer resistance method will align with it, once the procedural error with the latter is corrected ;)
 
Last edited:

Blueshound

Member
Messages
413
Breaking news - tube amps can sound great despite their power tubes idling somewhat below some proscribed dissipation % :eek:
I suspect that the reading from the bias probe is fine, and that the result from the transformer resistance method will align with it, once the procedural error with the latter is corrected ;)
You are correct, sir. Your help is much appreciated!
 

Blue Strat

Senior Member
Messages
30,151
Further, the way to "fix" the amp so you'll get higher than 20% idle current with the power tubes is to decrease the value of the bias range resistor between the bias pot and ground (assuming that there's a bias pot). If not, decrease the value of the fixed bias resistor (56K?) by about 10%. This part it trial and error as I know of no way to calculate the correct value.
 

pdf64

Member
Messages
7,281
If going that route, bear in mind that amp's iron was specified for use with much lower rated 6L6G tubes. If it is an exact repro, I suggest that the current draw you set the bias for should not exceed the capabilities of a 6L6G. Just because 6L6GC (assuming that's what's fitted in the amp) will cope with much higher standing current doesn't mean that the amp's iron will too.
 

HotBluePlates

Member
Messages
10,297
the 5e8a choke must be a hefty unit if it's passing all that current through it, that is intriguing. That choke would have some kind of regulation effect for the output tube h.t current, not quite like a choke input, but something surely, again, this is quite intriguing, why did they do it I wonder?
The C-L-C filter makes the d.c. fairly ripple-free at the 2nd filter cap. Fender chose to supply both the plate & screen from the same power supply node. Which should be interesting for those who argue tube ratings to the death, as the schematic shows ~400vdc for plate & screen while the 6L6G data sheet shows that both the plate & screen are over-volted.

Later, Fender probably realized that the push-pull connection of the output transformer tends to cancel ripple. That means the plates could be fed from less-clean d.c. than the screens, and have slightly higher voltage (for a bit more room to make more clean output power). The choke probably also tends to prevent very large current changes, so it's harder to get all the output power possible by working deep Class AB (closer to Class B operation). So the tapping point for the plate voltage gets moved to filter cap before the choke.

The screens, however, are a bit more sensitive (there is amplification from screen to plate) and doesn't have the benefit of a hum-canceling connection. Best to have their supply voltage filtered by the choke.

... That choke would have some kind of regulation effect for ...
IMO, references to the "regulation effect of chokes" are very unhelpful as we tend to think in terms of the modern sense of the word "regulation." The classic definition was "voltage-change from no-load to full-load."

Chokes oppose a change of current, and less current-change in a circuit with resistance means less change of voltage drops around the circuit. The circuit seems to have "good regulation" because the voltages don't change much.

But the inductance of a choke falls to zero if current falls to zero. And if you saturate the core of the typical choke by drawing too much current through it, inductance again falls to zero. Absent some special chokes (swinging chokes or saturable reactors), a designer choosing to use a choke also commits themselves to something like a Class A power stage where they expect little change of current (to give the choke the best conditions in which to operate). But having given the choke its ideal environment, the designer also chose conditions less likely to have big current changes or changes in voltage-drop.

Is it the chicken or the egg?
 




Trending Topics

Top