Seeking 18W enlightenment

[Lectroid]

About a year ago I finished an 18W amp from by as circuit by tubenit. After initially getting it running, I documented the voltages at that time on the schematic below.  Those numbers on the schematic are a year old.

I re-measured the voltages yesterday:
A=334
B=323
C=288
D=271

   Plate      Cathode   Rp      Rk
------------------------------------------------------------------------
V1b:   90      2.87       100K       1.5K
V1a:   82      2.85       100K       1.5K

V2b:   312      12.06    Reverb   1.5K
V2a:   82      2.85       100K      1.5K

V3b:   194      1.77       82K      (1.5K eff/shared 820R)
V3a:   82      2.85       100K      (1.5K eff/shared 820R)

V4:   328      11.53   -      213R
V5:   328      11.53   -      213R


Here's my question.  When I compare the original voltage numbers from a year ago (schematic) with the numbers I got yesterday, they are close to what I recorded a year ago.  Overall, I'm happy with it.

EXCEPT--on the first stage tube V1.  V1b's plate voltage is 90 volts, or 38 volts lower than it was a year ago.  Likewise V1a's plate voltage is 82, or 40 volts below what it was a year ago.  See the schematic above for the original numbers. 

Why would the voltages for the V1 first stage now be so far below the voltages I recorded when the amp was just built?  All the other tubes' voltages are back close to those original values.   I've rolled a few tubes through there but voltages don't change with that.  I'm probably missing something obvious.  Can anyone point it anything else can I test?

[acheld]

That is hard (for me) to understand.   While your plate voltage is lower, your cathode voltage is higher, which points to the tube, I guess.   It is always possible the initially recorded voltages were wrong, but your cathode voltages are otherwise in line, so maybe not likely.  I guess you could check the 5K dropping resistor between C and D, but that would not account for what you're seeing.

I'm stumped; hopefully not chumped.

[SEL49]

That is hard (for me) to understand.   While your plate voltage is lower, your cathode voltage is higher, which points to the tube, I guess.

That's normal. The plate voltage is lower because of higher current through the tube causing more voltage to be dropped across the plate resistor and that leaves less voltage to be dropped across the tube. Higher current means the cathode voltage will be higher.

Those resistor values are normally used on a 12AX7, not a 12AU7.

[ac427v]

SEL49 wins the prize this morning.
I missed that V1 is a 12AU7 on the schematic.

[Lectroid]

Tubenit's original circuit called for a 12A_7 in the first socket so I first tried a 12AX7 in the first stage.   The goal was good cleans, with a smooth onset of overdrive.  But at first it was too hot, too much overdrive too soon on the dial.So I swapped in a 12AU7 for the 12AX7.  That lowered the amount of overdrive available and that helped balance the amp's sound.

My original thought was that the anode voltages on those triodes were too low.  But if the resistor values are wrong, that could explain why my voltages were low?

What would be correct resistor values for a 12AU7 in that first stage? 

I looked for example schematics but didn't find one that used a 12AU7.  Anybody know of one to look at?

[acheld]

Most folk just use standard values for 12ax7's.  Most often that's what I would do -- but, in your case the plate voltage is low enough that I'd recalculate using rich kuhnel's tube calculator over at ampbooks.com.

[Lectroid]

… I'd recalculate using rich kuhnel's tube calculator over at ampbooks.com.

I’ve been playing with those calculators, just tweaking values to see how the results change. But I don’t understand the theory well enough. Would I change the cathode  resistor or the anode  resistor?

[pdf64]

If you're looking for centre biased, the manufacturer's info has a table with suitable values according to various scenarios / goals, eg p3 of https://frank.pocnet.net/sheets/093/1/12AU7A.pdf

[Lectroid]

@pdf,
I have seen those tables, thanks. They seem a little conservative for guitar amps center bias to me but I’m no expert. 

Anyway, I still would love to know  the answer to my last question:

… I don’t understand the theory well enough. Would I change the cathode  resistor or the anode  resistor

@acheld suggested changing the anode and/or cathode resistor but didn’t say more.

I can lower the anode resistor value to get a higher plate voltage.  But that would raise the cathode current and the bias voltage, right? At that point, would I lower the tube’s cathode resistor to bring the bias voltage back down to what it was? (Roughly center-biased?)

I think that was what @acheld meant but I’m hoping someone can confirm that for me.

[acheld]

What I meant was what I actually do (when it matters, only, cause it's a pain), and that is to draw load lines and try to find a good bias point.  Change parts values on the bench, and try again. 

I use Kuhnel's calculators as a starting point; I wish I understood how to use the calculators to their fullest extent.    Seems to me PDF's suggestion will be easier to use (that table is great!).

[Jennings]

Would I change the cathode  resistor or the anode  resistor?



Either and/or both can be changed to get to your desired point. The datasheet for the valve should give you guidance on max and suggested values. Personally if it’s just a little more or less juice you’re after I’d often just up or lower the cathode resistor value a notch or two so long as it doesn’t throw anything out of limits.

[Merlin]

Would I change the cathode  resistor or the anode  resistor?

Leave the anode resistor as 100k (unless you think you still have too much gain in which case you could reduce it). Adjust the cathode resistor for your preferred tone, there is no 'correct' value in a guitar preamp.

[Lectroid]

Leave the anode resistor as 100k (unless you think you still have too much gain in which case you could reduce it). Adjust the cathode resistor for your preferred tone, there is no 'correct' value in a guitar preamp.

@Merlin,

Sanity check?  When I read your: "Leave the anode resistor as 100k (unless you think you still have too much gain in which case you could reduce it.)", I think by "it" you meant gain.  If so, my guess is I'd  ower the value of Ra to reduce the triode's voltage swing, and that should that lower the gain.  Is that right?

[Merlin]

Yes, a smaller anode resistor will reduce gain (it will reduce maximum output swing too, but maximum output swing isn't really important since it is nearly always more than enough to drive [or overdrive] the next stage)

[Lectroid]

Understood.  Many thanks for your comments.  I've learned a lot about biasing for clean vs. dirty that I had not known.


 

[mresistor]

Have you tried a 12AY7?

[Lectroid]

Yes that’s what I have in there now.   Once I got the B+ issue fixed, I found a 12AY7 gives me a better range of cleans before it start to break up.