Bassman 135 question

[Smokebreak]

I fixed up one of these the other day and am a bit confused about the bias feed. I understand the concept of the output tube matching pot, conversion, and all that jazz. This amp already had a bias pot sketchily installed, albeit in the right spot, and kept the matching pot.
Anyways, the bias feeds are 47K and 68K. Why is that? I'm not even questioning why they are relatively low, but more why they can only be as close as 11K apart.
The amp hummed really badly unless I had the pot right at 50% rotation, and it was silent. The WingedC 6L6s were mismatched, by modern standards, I think 10mA apart or so IIRC haha. Customer didn't want a new quad if he didn't have to, so the matching pot came in very very useful.

On an entirely different note, what had been done to this amp was new to me.
The big sockets had been replaced at some point. Instead of removing all of the connections from the original tube pins, all 8 pins of each of the 4 sockets had been clipped at the base, and then very shittily cold soldered to the backs of the new socket pins. Some poor wires were holding on for dear life, and some had already lost.

[martin manning]

Good question. I'd have thought maybe an alternate way to balance the PI output (note the plate load resistors are both 47k), but the 47k and 68k bias resistors are offset in the wrong direction for that. Maybe the values are swapped in the schematic? Are they positioned that way in the amp?

[Malcolm Irving]

The DC resistance of one side of the OT is usually higher than the other side (due to that coil having the ‘outside track’). For a given brand of transformer the difference is consistent. Maybe this was a way to give the higher resistance side a less negative bias voltage? Two perfectly matched tubes would then DC balance nicely at the mid position of the balance pot.

[Stevem]

Is the amp being used for Bass, or clean guitar only?

[Stevem]

Fender found production of these amps to go faster with just the bias balance pot and a hum balance pot in the fil string!

[martin manning]

The DC resistance of one side of the OT is usually higher than the other side (due to that coil having the ‘outside track’). For a given brand of transformer the difference is consistent. Maybe this was a way to give the higher resistance side a less negative bias voltage? Two perfectly matched tubes would then DC balance nicely at the mid position of the balance pot.

I don't think there is enough grid current at idle to make any difference there. A quick Google didn't yield any explanation. Maybe this is something that came from the Hi-Fi world?

I see the UL super reverb has the same 47k and 68k, but they are on opposite sides of the PI. That leads me to believe the reason is about balancing the PI, maybe addressing bias excursion under overdrive conditions too, and the Bassman 135 schematic is wrong.

[Malcolm Irving]

... I don't think there is enough grid current at idle to make any difference there....

Yes, I agree.

[Firestorm]

I'm sure Martin is right about the Bassman 135 schematic being in error. All of the other UL schematics show the 47K/68K flipped. Interestingly, it's only the UL versions have that 21K difference: in roughly the same period with 2x6L6, the grid returns are both 100K; with 4x6L6, they're both 68K; with 6x6L6, they're both 33K. So something about Fender's version of the distributed load/UL circuit "requires" that balance correction.

[Smokebreak]

I think you've got it , Martin. I did verify that the resistors on the board were 47K and 68K but I didn't trace them out unfortunately. Steve, this amp was for a bass player, but I played guitar through it for a while and it sounded great.
Something else of interest was that the 1st stage bypass caps on the pre cathodes is 5u for the bass channel and 25u for the guitar channel. At first that seemed backwards(at least the bass side) but once I thought about it, I guess the bass guitar already brings those lower frequencies to the table in the first place, so there's more of an importance to keep things tight, thus the lower cap there.

[rooster]

Ed Jahns brought this to Fender when he was in charge of design there. (Just prior to Paul Rivera era.) He was a hi-fi guy and he changed many things that contributed to a more powerful, 'cleaner' amp. He also added some things that dealt with high frequency oscillation within the typical (previous to his tenure) Fender BF circuit. No, you can't typically hear what he fixed, but your dog and the OT might disagree with you.

A brilliant man on so many levels, he is not remembered well in Fender's history. The worst problem he created might be his use of PTs that delivered too high a plate voltage. Witness the UL SFBassman 135, the 70 watt UL SFProR/SR and the 120 watt UL SFTR. The Philips STR 6L6 tubes of the day made all of this happen and unfortunately many tubes manufactured today do not do well with these higher voltages. Haha, do you have a gig that needs a 120 watt TR? Anyway, there are ways to alter them, but it's really not worth the effort. Nor will you ever get the amp into the sweet spot of it's BF predecessors.

Eh, my .02.

[gingertube]

This twigged my interest because of the age old debate about whether the phase inverter should be perfectly balanced or should have some "deliberate" imbalance to promote some extra 2nd harmonic distortion for a "warm" tone.

So I did a quick calc to see what would have given perfect PI balance I this amp.

I assumed that the 47K and the 68K to the bias pot are indeed shown backward. The input side of the PI needs the smaller total load .

So total effective load on the cathode driven side (to V6 ,, V7) is 47K parallel 68K = 27K8
Then calculated the effective load for the input side for perfect balance. That turned out to be 26K6
47K parallel 62K is spot on.

So for the resistors to the bias pot you may wish to try 68K for the V6,V7 feed and 62K for the V4,V5 feed.

It may just be my preferred sound but every time I have "balanced" the phase inverter on any amp I have preferred it that way,

Cheers,
Ian

[Stevem]

How many here have played around with balancing both the AC and DC functions of a LTPI?

[Stevem]

How many here have played around with balancing both the AC and DC functions of a LTPI?

[Stevem]

How many here have played around with balancing both the AC and DC functions of a LTPI?