Holy crap, it works!

[lmolter]

I'm very excited. I finished wiring my Bassman this weekend and got the tubes today. I had already verified that the heaters were wired correctly, and I brought the amp up with a 5U4 that I had laying around. I was initially concerned because all DC voltages were high, and there was talk here and there that the Mercury ToneClone PT put out too high a high voltage. Senior members told me that everything would be fine after all the tubes were installed and there was load on the B+. Right on.

Anyway, the 5881 plates sat at 460v and with the bias all the way down, they drew 35ma each. That's 70% of the rated 23W plate dissapation. And since nothing was smoking, I plugged in the guitar and... beautiful sound (with a little hum). No motorboating, no squeals, no hiss, no dead air (radio jargon for silence). And aside from transposing the Normal and Bright channels on the first preamp, all controls seemed to function normally.

Unfortunately, I don't have a 4 X 10 cabinet yet, so my load was a Radio Shack PA speaker converted to 2ohms via resistors. So I imagine the sound will be much better and LOUDER with real a real 2ohm inductive load.

I still want to put the scope on it and look at the PI and power stages just for my own education. And I have to eliminate the minor hum.

In closing (for now), I just want to thank everyone who helped me along the way by answering my most basic and confusing questions. I plan to post a "What I've Learned" piece here after I get the beast tamed. I'm close, I think, but the scope may show things that I can't discern by ear. Oh, I'll post some pictures, too.

Thanks to all,
Larry

[lmolter]

I went through and measured all voltages, and all were within 10% (high) of those marked on the fender schematic. Wait before you blast me... I know from other threads here that those voltages are inaccurate, but I don't know which ones can be trusted and which ones cannot. In my measurements, all measurements were high, but the PI grids measured 13v and 14.2v in comparison to the schematic (I know...) that showed 22v and 23v. Any thoughts?

I also had some snap, crackle, and pop sounds after the amp got warmed up. These went away when I swapped the PI with the previous stage (V3 with V2). Other than that and the minor hum, I'm still pleased.

Larry

[zhyla]

Ah, it's always a good feeling when you first hear guitar sounds come out of a newly wired amp, congratulations. I'm in a similar stage with a Bassman as well. What a great amp.

[tubeswell]

the PI grids measured 13v and 14.2v in comparison to the schematic (I know...) that showed 22v and 23v. Any thoughts?

Larry

Hi Larry

Do you mean the output tube grids measuring -ve voltages? or the phase inverter grids (which you might get an AC voltage swing on with a signal going through, but if the PI is cathode biased, then you shouldn't get any quiescent DC voltage at all).

If you do mean the -ve voltages on the output tube grids, then measuring the negative grid voltage is not a reliable method for setting bias (and more-so if the -ve bias voltage is lower than what the schematic shows it oughta be). Instead, you need to calculate the output tube bias by either determining the tube current or the plate current and multiplying it by the plate-to-cathode voltage (which is the same as the Plate to ground voltage on a fixed bias amp) to get the power dissipated in the tube- i.e. Amps*Volts=Watts

Tube current is less risky to manage, however will give you the plate and screen current combined, and you can deduct about 5mA for the screen current in each tube.

For this method you can put a 1R 1% (say 0.6W metal film) resistor between each output tube cathode and the ground - these can be permanently wired in. Then when you read the mV across each resistor (between the cathode and the ground) you will get the equivalent in mA (because 1V/1R=1A). From there you can work out the approximate plate dissipation (in Watts). You should be aiming for about 70% of the output tube's max rated dissipation at idle for Class AB1 for each tube.

[lmolter]

Hi Tubeswell,

No, I meant the PI grids. The output tubes are fine and are drawing about 35mA (with a plate voltage of 460V). The 5F6A schematic shows the PI as cathode-biased through a 470ohm and 1M divider with 34V at the cathodes (mine is 44), and 22V and 23V at the grids (with mine at 12V and 13V). Off all my measured voltages, these two are the only ones that were not higher than the ones noted on the schematic. I was just wondering if this is 'normal'?

Thanks for the reply,
Larry

[MWJB]

Please confirm that you have a 12AX7 in the PI?

You have the 10K tail resistor? What is the voltage at the junction of the 470ohm cathode & 10K tail?

What are you getting at the PI plates?

I would expect voltage at the grids to be up, rather than down.

[tubeswell]

Ah yes I see now. The 10k creates a voltage which is watered down by the 1M grid load resistors (but still less than what is at the cathode through the 470R), but leaving the grids at a relatively high DC voltage, hence the need for the blocking caps in front of the grids. If the 10k was bigger you create more voltage, and if it was smaller - less. Thanks MWJB

[DirtyGeorge]

As suggested by MWJB you should measure the voltage at the 10K/470ohm junction. Since there should not be any current flow through the 1M at idle the voltage at that junction should match the grid voltage.

If you actually have 12v at the junction but 44v at the cathode then you are dropping 32v across the 470ohm. In that case your 470ohm is either a much higher value or you have way too much current though the PI.

If the voltages do not match then either you made a mistake wiring the circuit or something (oscillation?) is screwing the reading.

DG

[lmolter]

Ok, at the 10K/1M junction, I read 42 V. The cathodes are at 44V and the plates at 259 and 243. Everything but the grids is in tolerance. Ok... I just ran to the garage because I had a hunch that the 470 was a 4700, but no, it was 470. However, just for kicks, replaced the 12AX7EH with a plain 12AX7A and the voltages came up to 23 volts. Maybe there was something wrong with the tube? It was new in the box.

???

I'm putting the scope on it tonight to uncover more surprises.

Larry

[lmolter]

I put the scope on and fed in a 10mv, 1KHz tone on the Normal input. I traced it through each stage, looking for oscillations and oddly-shaped sine waves. Alas, none were found. Clean sine waves throughout. At the power stage, I hung a probe on each plate, a/c coupled them, and read about 40V p-p at the OT windings. The sine wave again was clean on both channels, with an undulating low frequency component (probably the 120Hz hum).

In a class AB1 push-pull, I thought each tube was biased slightly non-linear, that is, one half of the signal will be slightly less than the other half (from the same tube). In my case, each tube is putting out a symmetric sine wave, which leads me to think that they are biased in the middle of their linear region, sort of like a class A would be. If I'm blowing smoke here, please let me know and explain what I _should_ be seeing. I'm just having a hard time believing the amp is working as well as it is for a first-timer. Granted, I've been building electronic stuff since I was 10, got a degree in electronic engineering (but have been doing software for the last 27 years), and took my time wiring it up. I dunno.

Larry

[DirtyGeorge]

Each tube in a class AB P-P output stage will conduct 360 deg for small input signals. The nonlinearity will only occur when the input swing is sufficiently negative to drive the tube towards cutoff.

I'm happy to hear the PI grid voltage readings came into line with the other tube. I don't remember the details of the mechanism but I think a gassy tube can cause such symptoms. Ionized gas pulling electrons from the grid or some such.

DG

[lmolter]

You're right. I thought about it overnight (yes, I had dreams about sine waves) and I was slightly mistaken.

About the tube swap... I went to the local music store today and bought 2 new 12AX7 Groove Tubes. Guess what? The voltages went back down. Oh, well, the amp sounds good except for the hum, so I'll leave them in until I can figure out if something is amiss. Back to the hum... It's there without any input, and all inputs are shorting with no plug in them. What's really curious is that pin 2 (grid) of the voltage follower stage is very microphonic. This is the input from the two volume pots. If I tap the wire, I hear the tap in the speaker cleanly. If I rub the wire with my finger, I hear a rubbing sound in the speaker. And the hum increases, too, if I touch it. What the heck is that all about?

Larry

[MWJB]

Grid wires are usually pretty microphonic, shielding them is a good idea.

Pull the PI tube still hum?

Did you ground the preamp & PI filters to the input jack ground? You don't really want them grounded to trhe same place as mains & screen filters.

[lmolter]

Yes, the grounds are separate. I have a wire from the brass plate and a common ground from the eyelet board. These wires attach to the same ground lug separate from the other grounds. Is there a potential for a ground loop here?

As far as grounding the grid wire to the 2nd stage - that wire comes from under the eyelet board. It's going to mean unsoldering every connection to the board to get it out. If I snip it flush to the board and run a new wire from the 2-270K's, will the stub of the old wire act like an antenna to hum? Ick. Maybe I'll try rerouting it and shortening it up. I left a little extra on the wire in case of mistakes.

Larry

[MWJB]

"Yes, the grounds are separate. I have a wire from the brass plate and a common ground from the eyelet board. These wires attach to the same ground lug separate from the other grounds. Is there a potential for a ground loop here?" Whoa! You're losing me here.

Do you have a seperate wire from each ground point on the circuit board, going to either the relative ground point along the brass plate, or seperate wires from each point on the circuit board running to a common ground point on the brass plate near the input jacks? You shjould not have all the ground points "daisy chained" on the board, then running via one wire to a single ground point on the brass plate.

If shielding the grid wire for the second stage, mount the 270Ks directly on the middle lugs of the volume pots, free ends twisted together. Run the centre conductor of the shielding from the junction of the 2 resistors straight to pin 2 V3, don't thread behind the board (clip wires from volt pots to the 270Ks on the board, cut old exposed wire to pin 2 close to the board & leave wire behind board & 270Ks on board in place). Ground sheild to either one of the volume pot grounds. Shrink tube the free end of the outer jacket so that it can't slip down and ground out pin 2.