PREAMP PLATE VOLTAGE EFFECTS

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Does the B+ to the plate of a 12AX7 affect it's "headroom", voltage swing, ability to amplify without clipping? Tube data sheet allow up to 250 VDC. I currently only have about 100 volts, going then thru the standard 100K load resistor. Am getting more clipping than I'd like. Don't want to change the 100K, but can lower the power supply dropping resistor to raise it. Should I increase it, then look at the cathode biasing? Thanks!
 
Is the clipping you refer to in this triode - or further down the chain?

Max headroom in this triode will be at approx half B+ (check the loadline). So increasing B+, and rebiasing so plate voltage moves to the new 1/2 B+, will increase headroom in this stage (my understanding - happy to be corrected)
 
Administrator
Joined 2004
Paid Member
You need to identify whether it is this stage that is clipping or when further up the signal chain. The change you propose would work, you may need to increase the value of the supply decoupling capacitor in order to provide comparable filtering of the B+ if you subsequently end up with a bit of buzz.

You may want to reduce the plate currently slightly to compensate for the higher current due to the increase in plate voltage, this will give you some additional headroom as well.

You aren't just over driving the first stage are you - about a volt in would do it.
 
Great comments and questions! Firstly, I make sure my SigGen is putting out at guitar level. Secondly, yes I am getting very mild clipping from the second gain stage (after the first gain pot). Have a 5E3 on a breadboard (and a couple related posts on this forum describing my exploits), so have added gain pots before the PI gain stage (pin2) and (most recently) another gain pot before the PI itself (pin7). Added a PI grid stopper pot, some help. But I DID boost the preamp tube (first 2 stages of gain) from 96-140 VDC, and it helped. This is all with a pair of 6Y6s (very different than the JJ 6V6s I started with), and have rewired and added components to accomodate it's requirements. Got it to work reasonably well (sounded better on guitar than it looked on the scope!). My goal was to gather data on these tubes for the future (can get 12W diss per tube with only 240 plate volts!). Then I will concentrate on getting the PI working correctly. Too many variables and I can't tell what's doing what. Thanks again.
 

Attachments

  • IMAG1524.jpg
    IMAG1524.jpg
    947.4 KB · Views: 423
  • IMAG1486.jpg
    IMAG1486.jpg
    71.1 KB · Views: 418
  • DESIGN #3.jpg
    DESIGN #3.jpg
    244.2 KB · Views: 410
100V B+ is way too low for 12AX7.
Yes it will clip early but you will run into problems with grid current before that (It will be noisy with an unpleasant tone)

Ideal anode voltage is in the range of 1/2 to 2/3 of B+ and in any case not less than say 140V for a low current stage (Less than 0.5mA) and not less than 180V for a higher current (approx 1mA) stage.
As a really rough rule of thumb that first "less than 0.5mA" will have a cathode resistor of 2K7 or higher, the 1mA would have a cathode resistor of around 1K5 or lower.

Lower the B+ dropping resistor to the stage to restore B+. With a typical 10uF power supply decoupling/bypass capacitor, you only need a couple volts drop across a B+ dropping resistor for it to do its job of decoupling the supply.

Cheers,
Ian

Your post above came in while I was typing, You may well have another problem, A B+ of 240V is marginal for a Cathodyne Phase Splitter.
 
Last edited:
I don't know why the preamp plates run so low, I copied the 5E3 schematic, and was using a PT that used to power a PP 6V6 organ amp, and was getting 355 VDC to the plates..... I wish the schematic had voltages on it (some do, this doesn't). Especially where the PI is concerned. But thanks for the tips. I built a Champ clone (my first amp - I'm still new to this), and pushed it's preamp 12AX7 plates to 200V. By "Lowering the dropping resistor" do you mean the 100K load resistor or the 22K on the B+ supply? And I don't understand "decoupling the supply".... Power supplies have filter caps (parallel to ground), and preamp stages have DC decoupling caps between them. Please explain? Thanks Jeff
 
Hi Jeff,
A "quick and dirty" reply - I hope it make sense to you.

I mean the 22K on the B+ supply should be reduced (try 2K2).

Decoupling Caps and Filter caps are the same thing we just think about the primary function slightly differently depending on where they are in the amp.

When an amplifying stage draws current from the B+ supply it is in the form of a steady state DC current (the idle current) with the signal current swinging either side of this.

If there is any AC impedance in the power supply line (and there always is at least some small impedance) then that changing signal current can generate a signal voltage on the power supply line. The preamp decoupling/filter cap is there to eleiminate or at least minimize this signal voltage on the B+.

If you run 2 stages with the same signal phase of a single B+ power supply "node", then you will usually have stability problems because this signal current induced signal on the B+ line from one of these forms POSITIVE feedback into the other stage connected to the same B+ node.

You can, however, run two stages with opposite signal phase from the same B+ node because then the signal induced variation from one of them is NEGATIVE feedback to the other.

The problem comes when you have more than 2 stages. For a 3rd stage you need to generate another power supply "node" with a series resistor plus a "decoupling" cap to 0V to generate a new "decoupled" power supply node. That is, the 2 B+ nodes are decoupled from each other.

In addition the phase splitter has both phases present and so it should always have its own dedicated power supply "node"

The signal current induced "signal" on the B+ at one "node" is divide down by the series resistor element and the AC impedance of the"bypass / decoupling / filter" capacitor to 0V to produce a new "decoupled" power supply node.

Cheers,
Ian
 
Last edited:
The dropping resistor is the 22k resistor on the B+ supply. Decoupling the supply is the filter cap for that node which is 16uf. Between the preamp stages the caps are usually called blocking caps because they block the DC to the next stage, or they are also called coupling caps not decoupling caps because they couple the different stages together.

With the schematic and the known B+ you can use ohms law to figure out all of your voltages around the circuit. I also recommend downloading Duncan PSU designer.

Give me a minute and I will work out each section of the stock 5E3 circuit and it's quiescent voltages.
 
Last edited:
You guys are great thanks! A lot to digest, "opposite signal phase from the same B+ node"..."generate a new "decoupled" power supply node", but I'll re-read it a bunch of times, look at schematics, and download Duncan PSU designer. Would LTspice help with providing voltage information for the 5E3 circuit?
 
I will start by current draw and go from there. Each preamp tube should be pulling about 2mA each.

With B+ voltage at 390 I get 180v at the plates of V1 and 1.5v at the cathodes. V2 gain stage 180 at the plate and 1.37 at the cathode. V2 PI stage 222v at the plate, 51v at the cathode, and 19v at the grid. The bias for the PI is -1.5v grid to cathode.

6V6 cathodes at 22v and plates at 374v to ground which gives me a plate voltage of 352. Cathodes share a 250R common resistor. Screen voltage is 352 as well and pulling about 2mA each. Plates pulling about 42mA each.
 
Awesome-finally some numbers to compare to mine! I'm in the ballpark, but will start tweaking to see what changes. My breadboard has two 6Y6s in it right now, so plate and screen are way low, but will finish what I want to do with them and go back to 6V6s and see what I can do. Thanks!
 
I'm scratching my head....put 6V6s back in, restored proper wiring and resistances (as per 5E3), and replaced 5Y3 rect with a GZ34 for 10 more volts. Get 365V B+, but only 292V and 248V for B+2 and B+3... Pulled 6V6s, get 407, 357 & 300 for B+ 1,2,3. Amp works, 6V6 plate current 42 mA ea...gonna check for bad solder joint or something....
 
Thank you Arctic Breeze (and FMB and Jazbo). Pointing me in the right direction. My breadboard now stays clean up to about 75% overall gain. Found my 5K B+ dropping resistor (to screen) was 7.0K.... but dropped it to 3.47K, screen went from 300 to 330V (screen current 11 mA), dropped "22K" (I was using a 16K) to 10.9K, raised the 3 gain stg plates from 103 to 132V (current to the 4 triodes 5.4 mA). the 3 B+s are now 369/330/270, my 6V6s are 362/330/20.6 (Vp,Vs,Vk). I know the dropping resistor are way below 5E3 values, but needed to get the voltages up, and the currents seem fine (no shorting going on anywhere). My PI is 202/70 (plate/cathode) - right where Mockingbird calc'd it at. Why do you say I have 6 triodes? I have 3 gain stages, 1 PI stage, and two 6V6s....(4 tubes ttl). Oh, and a 1M pot in front of PI pin7 (replacing the 5E3 1M resistor) is set at 750K (big stopper). I'm liking where it's at now... maybe can now start tweaking the bias (scope shows assymetric clipping - want to play with that). Thanks again!
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.