Bloozmaster HRM in a '67 Bassman chassis, 4x6L6GA

[David Root]

Just finished wiring it up, except the heaters. The oddball position is the buffered loop tube, no room in line with the other noval tubes. It is fed 12V DC from the board to its left. Did not want to run heater wires under the main board.

HRM tonestack brought out to the faceplate. Ampdoc 5V relays powered off the 5V PT winding.

V1 is a Mazda 12AX7S, labeled Dario Miniwatt. V2 is an Amperex 7025, the PI is a longplate GE 12AX7. V4 is a Tungsol long plate 12AX7, will have to change that to a 12AT7 before I crank her up. Forgot about that.

Two bias pots, Magnetic Components 4 ohm Twin OT, PT is a Mercury Magnetics low voltage tweed Bassman 310-0-310, perfect for 400V or so plate/screen. Heyboer choke. PT may be slightly under powered for 4 x 6L6GA, but if I can get 50W or so that will be OK.

The 6L6GAs are JAN Sylania from the '40s and early '50s.

I wanted the 4-tube tone in 50-60W rather than 100W. It is based on Henry's Bluesmaster schematic, with separate MV for clean and OD.

OOops! I wired the power grids wrong. HUA, there is a lot of it going around.

[Structo]

Hi David,

Looking good!

Ahh, re-read your post and I see you spotted your mistake.

[M Fowler]

David,

Nice amp, it takes a while to wire one of these Dumble style amps doesn't it. Did you use mostly 22 awg wire?

Mark

[Bob Simpson]

What's in the noval socket north of the PI trimmer?

Bob

[jelle]

Looking good. Question....the layout of the OD plate and cathode wires is different...intentional? The cathode wire of OD1 usually goes with OD2's plate and cathode wire to the board.

That amp was a lot of work, wasn't it? Came out nice!

[Drumslinger]

nice job David!

[David Root]

Mark, it is a mix of 22, changing to 20 at the PI and the loop buffer. I have ordered a bunch of 20 ga from Steve at Apex Jr. All stranded silver plated teflon insulation.

It does take a long time, about 3 months worth of weekends.

Bob, that is the FX loop buffer tube. I put a '62 Tungsol 12AX7 in it and took the pic and posted before I realized I need a 12AT7 in that spot. Duh!

Jelle, no, I forgot about that! Will correct it tomorrow. (I start Xmas vacation tomorrow). This one took longer than usual as I had a lot of drilling on the chassis and drilling a steel chassis is tougher than aluminum, even though I have a 16" drill press.

The four 6L6GAs are actually two pairs that are about as far apart as you could get I think. One pair is about 21mA each, (soft) the other about 48 mA, (hard) on my MaxiMatcher at 400V and -48V bias.

If I change bias to -60V the hard tube pair drops to about 30mA, so that is why I put in two bias pots. I think I will need -45 or so on the soft pair and about -60 on the hard pair.

I do have a pair of Tungsolski 5881s which match the soft pair of 6L6GAs quite closely. Given that if I do that I will have two push pull pairs with one of each tube in it. How do you think that would work out? Might even get some power section overdrive?

[David Root]

Corrected and finished wiring today. Has too much bass, but I addressed that in the Bluesmaster too much bass thread.

Another issue. Since I'm using 6L6GAs the plate/screens are at 400 V at 57% Pa, which is about right I think. I reduced the Screens-> PI dropping resistor to 470R/3W to keep the PI close to 400v, actually 398. However, no FET, so the V1 and V2 B+ are about 25 and 40V too high respectively. I have the standard Bluesmaster 15K, 10K dropping resistors PI->V2->V1.

Rather than plug in the standard 150K to ground to simulate the FET current draw, I calculate that if I replace the 15k dropper with 24K I should be just about right on with the required V2 and V1 voltages. This is pretty close to the standard ODS 22K anyway, which has the desirable plate voltage lowering effect recently discussed at length in another thread.

In the standard ODS however, with either the FET or substitute 150K to ground, the V1 dropper is 2K2. In the Bluesmaster it is 10K.

Is 24K /10K possibly too much of a good thing in this respect, or was Mae West right when she said that "Too much of a good thing can be wonderful"?

[Bob-I]

Rather than plug in the standard 150K to ground to simulate the FET current draw, I calculate that if I replace the 15k dropper with 24K I should be just about right on with the required V2 and V1 voltages. This is pretty close to the standard ODS 22K anyway, which has the desirable plate voltage lowering effect recently discussed at length in another thread.

I'll be interested in the results of that. There's a major difference in the 2 methods of reaching the desired voltage. My guess is the FET resistor at the end will not increase the draw with more input where the dropping string resistor will react to the increased draw of V1 and V2 by dropping the voltage. I haven't compared the 2 methods but I'd expect that the FET resistor will not produce as much voltage difference on V1 and V2 as the larger rail resistor.

Watching for more info.

[David Root]

Posting the completed, corrected gutshot.

Bob-I, sounds good with 24K/10K dropping resistors on V2 and V1 and no 150K to ground subbing for the FET circuit. Voltages seem a little high on V1 but it sounds good.

I have to lower the 20K to the loop send stage, B+ is 309 instead of 345, and probably tinker with the 2K2 to the return side.

Happy Christmas to all!

UPDATE DEC 27. Decided to leave well alone on the loop voltages. they looked OK on TubeCad.

I decided that taking the V2 to V1 dropping resistor from 10K to 27K would possibly be too much of a good thing, so I tacked in large resistors on the V1 node to ground until I got V1A and B to 175 and 182 V plate. This was a 270K, basically 1mA draw to ground in place of the FET draw, and the rest from the 24K V2 node dropper. Sounds very good to me.

I don't have snubbers on V2A or V2B. What might they typically be? I don't think they need to be very big.

[David Root]

Well I now have the following DC values:

B+ 401, 400, 397, 310, 290 with send loop 304 and return 310V

V1A, 182/1.45
RCA '50s long blackplate 12AX7, perfect balance
V1B, 188/1.55

V2A, 206/1.54
'60s GE long grayplate 12AX7, decent balance
V2B, 211/1.45

PI, 259/259/36 '60s GE long grayplate 12AX7, perfect balance

Loop (12AT7) Send cathode 33V, return 150/2.43 '85 Mullard CV4024

400V on plates and screens of 6L6GAs, cathode current 28mA each.

I don't know if the 24K instead of 15K dropping resistor and/or the 270K (1.0 mA) FET partial sub to ground are doing anything different, I haven't really cranked it yet, but it sounds wonderful in clean and OD.

OD entrance trimmer and HRM TMB pots set per Scott's recommendations, trimmer 20K to ground, bass on 2-3, mids set to 4.5K to ground, treble about 40% up (4-5).

I also changed my strings. to a brand I had not used before, Pyramid, round core medium w/wound G. Stunning strings! Huge tone and sustain! 011 to 050 set.

[martin manning]

David, what are you using to check the balance of your pre-amp tubes?

[David Root]

Martin,
Vacuum Tube Valley twin triode "characterizer" as they call it. It shows gain on a relative scale, and DC voltage balance as null to unspecified difference.

So if you get two 100 readings on switching between triodes 1 and two, for example, you have a perfect gain match, and if DC balance shows on that meter the meter needle and is within the specified area that is +/- 0.5V, Dead center is perfect balance, needle does not move if you have that.

It is pre-calibrated. I have had it for several years now and not had it recalibrated, but it doesn't get moved around like a DMM does.

[martin manning]

I was afraid of that... I saw that device a while ago, but it is not currently in production.

[David Root]

No it isn't, for some reason. I think there are several folks on this forum that have one.

What is your opinion of it? They were not cheap, I think I paid $695 for mine.

The null DC balance is fine, but I would have preferred to have one that reads out actual transconductance and actual current as well, but those seem to come a lot higher $$.