PSU design help !! (DIY 6L6 SE with choke -debugging)

[FatGaz]

Hi, All !

I've been re-building a very old DIY project amp recently and happend to find a choke that I could use to help reduce the noise in a single ended circuit. So I've built her up with a preamp using a paralleled 12ax7 1st stage and a second stage going into a cathode follower before the tone stack. Debugged a few issues, and making progress....

I've built it up and powered up for the first time. no bangs / sparks (ditched the DC heaters for the preamp -doh!)

I have one 6L6 going into a hammond 125Ese OT (wired for 5K). the b+ voltage is hopefully OK for SE use.

I have discovered that my PSU design* is wrong(?) and need help as I'm a bit out of my depth !

I have a slightly larger voltage on the screen than on the plate, I think this is BAD ???

the circuit is as follows, 'BR' from bridge recto. E is common/ground line; equals sign is a cap; [resistor value]
{measured voltage} - UK mains with a 330 ohm cathode bias giving +17 volts (not sure what this should be)
ignore the dots, they are for spacing; dashes are connections


...{305V}.....................{298}..................... ...{297}..............{262}...............{253}
..................................B1(OT)................ .....B2(screen)..........b3 (pre)...........b4(pre)


BR--|----[100 ohms 7W]----|--------{choke}----|-----[10k 2w]----|----[10k 2w]----|
......|...............................|................. ...........|.........................|.................. ....|
......=47u..........................=47u................ ......=47u...................=47u.................=47u
......|...............................|................. ...........|.........................|.................. .....|
E---|----------------------|--------------------|------------------|-----------------|

the voltage measured at the tube socket (via the OT and a 1K screen res) = 291V plate, 293 screen.


Problem is I'm dropping more voltage across the OT to the plate than I am across the choke to the screen.
it's not much, screen is about 2 volts higher than the plate (at idle of course, which should be full power for se)


basically I had tried the above because I had read that in an SE amp, you get mains hum in the output tube produced more from the screen supply than from the plate supply, so I put the choke there. it's after the main plate supply, so doesn't have to cope with much current. (I think the choke is rated around 10-20H and 50-75mA).


SO what should I change ??? I'm not 100% sure the choke can cope with the current for the 6L6 as well.
goal is that i'd like to minimise the noise.

HELP !

Gareth.


* I kind of guessed at this without really understanding it - had trouble finding any example schematics from other SE amps to help guide me or for me to copy.

[Bruce / Mission Amps]

2v is really nothing to be concerned with.
You could lower the power tube's idle current a little by lowering the DC voltage to the screen though.
1000 ohm screen resistor in this set up is obviously not enough resistance to do what you want.... that's reall all that is "wrong".
Measure the DC resistance of the OT and use a 2w to 5w resistor much larger then that.
Install it after the choke, but in series with the wire that feeds the screens... not the rest of the B+ rail.
You could just put it right on the tube socket lug 6 to 4, as I think you're doing with the 1K but make it a lot bigger ... simple.
The screen grid and the rest of the preamp stages are not drawing as much current through the total DC resistance of the choke (in the B+ rail) then the power tube's plate is through the DC resistance of the output transformer.... basic Ohm's Law math there.
example: 2ma screen at idle dropping say 5vdc.... = 2500 ohms ... not 1000 ohms. etc etc etc

[FatGaz]

Hi Bruce,

OK, plate draws most current, SO it's no supprise it drops more voltage than the choke - I can see that now thanks (doh, it was obvious too)

I'm not quite following what you're saying about changing the circuit. sorry.


At the moment I have a 1K (two 2K 2watt parallel) from the PSU B2 node directly to the screen pin 4.

Are you saying I should put a resistance in series with the choke before B2 ? (i guess this will cascade down to the preamp; bad?)

OR are you saying I should just make that 1K larger so the screen voltage drops more ?

Also, do the rest of the voltages look OK ? I'm not 100% sure about biasing a 6L6 se.

Thanks !!

Gareth.

[FatGaz]

OK folkes, a couple more measurements.

without unsoldering any wires, the OT measures 98 ohms and the choke measures 156 ohms.

also, the cathode resistor is 244 ohms. maybe a bit small ?, should I try closer to 330 ohms ? not too sure what current to aim for with my plate voltage.

what should I try next as a screen resistor ?
should i get a ceramic (box) one or make do with parallel metal film 2w ones again ?

I'll need to buy more resistor supplies tomorrow. (maplin uk) I guess a couple of 5Ks at least for 2.5kohms.

all advice greatfully received !

cheers,

Gareth.

PS. I attach some pics of the build - chassis & cabinet from a skip (with new top), and some other parts are 'salvage' too..




note above the addition of a regulated 9v supply for effects pedal(s) <- added after aborting the dc heater supply (didn't work, probably because of the in-rush current of cold heater fillaments ?)

I hope the speaker still fits ( and it doesn't pickup any hum from the PT...) its a 70's g12h

yes, the layout is crap, but I'm hoping to get away with it...

[FatGaz]

more info, found choke specs - 10H nominal (5 to 10) & 100mA max.

so maybe, the choke would be ok for the whole amp ? (i've read something about large ripple voltage input causing problems though ?)

Also, found more 6L6 datasheets, but this has confused me more than enlightened. basically they tend to give the example of:

class A - 300V plate 200(!!)V screen ~55mA plate (5mA screen) possibly 220 ohm cath res ? -12/-14 V bias.

So, if i end up with say 290V plate and 280-270 screen, do i still go for about 50-55mA and ~12-14 V bias (currently at 17V).
{or is this saying that i must reduce the screen a lot}
or do i just measure and check dissipation for plate & screen, then just give it a go ?

bit bifuddled.

Gareth.


(PS, i have an issue to debug also, the first stage (parallel triodes) plate voltage is way too low, 80V ! it's a shared 220K plate resistor with ~250V on it - not sure what is causing this to be so out of whack. cathode is 0.8V. have also checked this with the other tubes pulled. )

[loudthud]

If you look at the plate and screen voltages with a scope, you will see that the plate has prehaps 10V peak to peak of ripple on it that goes above and below the screen voltage. The screen is nice smooth DC with virtually no ripple. This is perfectly normal, nothing to worry about.

[FatGaz]

She lives !!!!

cathode current @ 72mA, screen now has a 2.5k resistor so is a few volts below the plate, looks like a happy 6L6, nothing glowing that should not. thanks for the help Bruce !

amazed, totally amazed that my tone stack works fine 1st time; I was expecting to find a mistake or two there, especially as it is point to point between the pots.

sacrificial speaker has survived the first test strum. I can hear no hum from it !!! hiss, but no hum - wow...

I have one query which is the plate voltage on the first stage seem too low ? it's a 12ax7 that is wired with the two triodes parallel, sharing a 1.5K/25uF cathode and sharing a 220K plate resistor. the B+ is 256V, BUT is only 78V on the plate ???? does this seem reasonable ???

any ideas ?

Gareth.

Just going to get the 1*12 cab to plug in and annoy the neighbours.

[FatGaz]

err, ok, maybe there is a bit of hum with a proper speaker... not terrible. but just a little bit more than I'd like. I'll need to investigate further.

plus points, I have a switch that changes the slope resistor (and maybe one of the caps) in the tone stack - really useful !
gives a warm jazz or a more mid-ish rock setting. Also having an on/off/on bright switch for 2 different caps is very useful too.
It can get fairly twangy clean. Pity the cheap PT can only give me 275V on the plate of the 6L6.

I like the 9V supply for my pedals - really handy. (might buy myself a reverb pedal for this amp.)

I can hear the preamp farting out a bit when it's dimed using the hot input jack, I guess it might be too high a level into the second gain stage's grid ? it's not the best though-out design.

I have two switches for the cathodes of the two gain stages; difficult to tell a difference between settings, one halves the 1st stage cathode resistor, the other switches the 2nd stage cathode between a 2.2 and a 25uF capacitor - i can just about hear this when it's overdriven. I think I might alter the 1st stage switch to change the capacitance also. maybe between 25u and 1u or 0.68u.

Gareth.


(good gig on bbc radio 3 at the moment - monday night is jazz night)

[loudthud]

I have one query which is the plate voltage on the first stage seem too low ? it's a 12ax7 that is wired with the two triodes parallel, sharing a 1.5K/25uF cathode and sharing a 220K plate resistor. the B+ is 256V, BUT is only 78V on the plate ???? does this seem reasonable ???

any ideas ?

78V is a little low. Got scratchy pots on your guitar? A low plate voltage causes grid current to flow making a small negative voltage on the grid.

Edit: here's something you won't find anywhere. Curve tracer photos of a 12AX7 in parallel. Tube was a NOS GE. Load lines are 220K, 100K, and 47K with 300V B+. The -1.5V bias line looks pretty good for each load. Calculated bias resistors are 2.142K (220K), 1.2K (100K) and 714 ohm (47K).