EL34 YGL3 -> 6L6 SSS

[martin manning]

Yes the CF driver power supply needs a ground reference. Tie the CT and the center of the [edit] 22uF caps to chassis ground so you get +200V at the anode pins and -200V at [edit] V- with no tube installed. 455V with no limiter sounds good, and confirms there are no glaring errors. Once you ground the CF supply, put the CF driver tube in and check the bias voltage range on the output tube pins 5. If that looks good, leave it at the most negative voltage. Then, shut down and install the power tubes. When you first power up with power tubes installed, use the limiter again, and be sure to connect a load at the output. If things still look good, take the limiter out and set the power tube bias.

[bcmatt]

Thanks Martin! Just when I begin to think I understand how amps work to some small degree, I prove I can't be trusted to follow a layout that may be missing a few wires that are probably obvious to everyone else.

(I did connect the cathode cap/resister to the reverb driver tube in the end too, but not without first scratching my head and double checking the schematic first, wondering why there was a grounded cap and resister going nowhere).

So, just to be clear, you don't actually mean grounding the centre of the 2 220uF caps, you mean the negative end of the pair, which is also the centre of the 2 20uF caps? I assume you just didn't tap the space bar hard enough between the 2 and the 20, but at this point, I'd take your word over the layout and the schematic if you actually mean it.

[martin manning]

Yes the center of the two 22uF caps in the CF driver supply. That and the small transformer's secondary CT should go to the screen filter ground. Looks like there is a missing ground in the M Hartman layout because I don't see a chassis ground from there in the layout snip I used to draw the CF driver supply.
http://ampgarage.com/forum/viewtopic.ph ... 588#361588

[bcmatt]

Ok, thanks.

So, without the CF in, I get +199V on the Anode, and -134V on the Cathode. It is -198V before the 100K cathode resisters though...

[martin manning]

I can't think of any reason there would be current flowing through the cathode resistors without the driver tube installed.

[bcmatt]

I am quite confused. I seem to lose 67 volts across those CF cathode resisters, whether there is a tube in or out and whether the light bulb limiter is used or not. What should the expected drop be with tubes in? But, even with the tubes out, aren't we are making our way back down to ground at the cathodes of the power tube sockets, through 4x 470K resisters basically in parallel? Now that we are not floating that CF supply above common ground anymore, isn't it going to be dropping that voltage as we make our way through the 100K and 470k resisters?

I am still able to get a range of bias on the power tube sockets (pin 5) between about -38V and -55V

With the 60W light bulb limiter, and a CF tube in I get about +95V on the Anodes, -95V at the negative Supply, but -28V on the Cathodes.

[martin manning]

Sorry Matt, I was looking at the D Root schematic which does not have the 470k to ground at each power tube grid as shown on the layout. What you are measuring is correct, so I think you can proceed to adding the power tubes.

[bcmatt]

Awesome.
So, I biased the power tubes. Since it is a single 1R resister I am passing through for measuring the current, I guess I need to do the whole amp power instead of the individual tubes.

So, 120W/460V = 0.261A x 55% dissapation = 0.143A

So I dialed it in to about 144mV across the single resister. Is this how you guys are biasing this? I've only ever measured tubes individually before.

Anyways, things are silent until I add in the phase inverter. Then I get some low frequency hum with volumes all the way down that will eventually start to motor-boat.

My power tube voltages are:

Plates: 460V
Cathodes: 144mV
Pin 5: -50V
Pin 4: 462V


CF:
Plates: 190V
Cathodes: -50V
Grids: -50V

PI:
1, 6: 457V
3, 8: 112V
2, 7: 68V (letting go of these pins with DMM set off motorboating)

[martin manning]

The power stage and CF driver look (and sound) good. You could go a little higher in idle current, maybe 170mA, which would be around 60% Pa max since ~5% of cathode current is screen current. The PI plates look high, being the same as the power tube plates, indicating that the PI triodes are not drawing any current. Is the PI grounded properly?

[bcmatt]

Of course, you are right again, Martin. Thanks!

As you can see on the MHartman layout, the ground for the PI goes to the FX Loop send, I'm assuming to ground to the chassis there, but perhaps my jacks aren't the grounded type. So I connected that jack to the the ground of the FX loop return and the Speaker Jacks, and that solved it. It's dead quiet now.

That got my PI plates down to 278V and 285V, and my grids down to 31V and 33V.

This amp is very clean and capable of very bright. The step filters are hugely helpful.... of course, with my 6 position switches, I got rid of the value between the highest two filtering values on the 7 position layout, and that is pretty close to where I want it for the high filter (on that missing second highest setting).

I'm using all 12ax7s but a 12at7 for the driver. I tried a 12ax7 there too, thinking it might introduce some magic, but the one I swapped in actually seemed duller and less bold. Maybe it couldn't handle the stink like the12at7.

I was playing with the trimpot at the PI and am not really catching on to what it is supposed to be adjusting. I'll have to play with this thing a lot more... but first I better document all my voltages now...
Thanks Again, Martin!

So what value of light bulb should I regularly play this thing with in my limiter? I'll screw my limiter to the top of the head cab.

[martin manning]

OK, sure; it's always fun to see one come alive. The CF driver circuit is designed for a 12AX7. A 12AT7 will bias way down in the mud of the anode curves unless you cut the 100k cathode resistors down to ~47k and recenter the bias adjustment. Maybe try some other 12AX7's in there and see if you find something you like. The PI trimmer is pretty subtle. You can read about setting it for an ODS in the files section, and I expect that will apply equally well in your case.

[bcmatt]

I found another 12ax7 that is sounding quite a bit better.
So, all tubes should be ax7s? Is there some other position that I was supposed to use it in?

Is this PI trimmer supposed to balance the plate voltages on the PI? I think I read somewhere last night that it is suggested to dial it so that one plate is about 6 volts above the other, but I can't remember which one...or where I found that...

I just realized playing last night that my TMB controls act really weird (in rock mode)... Is that a normal thing with these? Like dialing the bass control down actually chokes out almost the whole signal. This is in rock mode.

[martin manning]

M Hartman layout and other SSS schematics seem to use 12AX7 everywhere except for those with a reverb, where I expect there would be 12AT7 driving the reverb transformer. Tube types are generally not marked on the schematics I have collected, though.

PI trimmer adjustment procedures are discussed here: http://ampgarage.com/forum/viewtopic.php?t=18048

Since there is a 1k8 at the bottom of the bass pot, it looks like the Hartman layout uses the SSS 002 tone controls. Maybe check your wiring against that schematic.

[bcmatt]

Thanks Martin,
So, I'm quite embarrassed to admit it but full disclosure... after focussing most of my attention toward analyzing specifically how my pots and switches were wired several times against the layout and the different schematics... I eventually included the wires to the preamp board... and it was immediately obvious my mistake. I can't really blame the MHartman layout for this because it should be so obvious:
3 differently coloured wires come from three cap values on the board and go to each of the TMB pots. Logic would say to check the schematic as they sort of change colour coming out from under the board on the layout... or think about the fact that the bass should be the biggest value, the middle should middle value and the treble should the be the smallest value cap. Instead I just assumed that the colors stayed pretty much the same... red stays red, green stays green and purple must be blue. Rookie mistake.
I straightened them out just now and the tone controls work great now.

Next I will play with the PI trimmer... and perhaps look into wiring up a presence pot... mostly because I have an extra knob on the end still unused.

[bcmatt]

I messed a bit with the PI Trimmer and I probably cannot swear under oath that I am noticing real differences sonically, but for now I set the preamp side plate to be about 6.5V higher than the Power amp side since I don't have a 1000Hz tone generator.

I also have messed a bit with the filters. The Bass filter seems fine for me... good range of filtering, but I already dropped the largest cap value on the high filter to make room for the missing .001uF one that I skipped since I only have 6 steps instead of 7. I am finding myself switching only between the 150pF spot and the highest spot trying to find something in between because this amp is quite bright. I'm tempted to put in something in between like maybe a 68pF and bump all the other value around and also drop the .0051uF.... unless there is another component I should look at to shift all these into a more useable range.

I think I will order some more of the top-hat knobs and go silverface fender for the panel look.