Hoffman Amplifiers Tube Amplifier Forum
Amp Stuff => Tube Amp Building - Tweaks - Repairs => Topic started by: stevefmt on March 30, 2025, 06:45:02 am
-
I am tweaking a 5879 pentode gain stage in my current amp project preamp section. It is preceeded by a 12AX7 lashed up as a bypassed high-gain first triode, 1M gain pot, then unbypassed second triode to a 1M interstage divider whose attenuated leg drives the pentode grid. Pentode load is a tone stack with ~1M total Z.
The yellow trace is the grid input signal. The purple trace is after the plate output coupling cap and across the tone stack. Progressively left-to-right you see what is happening as I increase drive (via the gain pot). Note that as the grid starts to round off and clip on the +peaks, the -peaks on the pentode plate are not symmetrically clipping, but are narrowing and kinking.
I have not seen this on any of those scope trace diagnosis web pages or books I have. So what does this diagnose? Too much grid-current? Screen compression? Blocking distortion?
Steve
-
need a schematic, I flunked English visualization class
then unbypassed second triode to a 1M interstage divider whose attenuated leg drives the pentode grid.
Is the I.S. driver an inductor(transformer), another tube?
-
Tube before interstage divider. See attached.
-
The presence control looks weird. How was it set for those scope shots? I guess at a low resistance, hence a treble boost.
How about if you just set it up as a regular pentode stage, screen grid fully bypassed, how does the overdriven wave look then?
-
... The purple trace is ... across the tone stack. ...
The strange shape is due to the reactive parts (caps) in the tone stack. What you thought was a clean-sine will also look strange when measured across a tone stack.
The yellow trace is the grid input signal. ... Note that as the grid starts to round off and clip on the +peaks, the -peaks on the pentode plate are not symmetrically clipping, but are narrowing and kinking. ...
The pentode is being driven into grid current, flat-topping the Yellow Trace.
By the 3rd picture, grid-drive is 11.44v peak-to-peak (!!!) or >>4.5v peak, and your schematic shows the bias across the pentode's cathode resistor is about 1v. You're slamming the bloody hell out of the 5879.
Positive-going swing of the Yellow Trace flat-tops because the driving signal's peak exceeds the pentode's bias, causing grid-current of the pentode.
Pentode grid goes from being ∞Ω to maybe few-kΩ, and this very-low-load clamps the output of the 12AX7 in this direction.
____________________
Inspecting the Purple Trace to understand the pentode's harmonic-contribution is pointless, given the way the reactive load will mangle the waveform, because it changes the relative-levels & phase of the different frequencies present. (As soon as the wave stops being a "sine wave" then there's more than 1 frequency present).
Twist any tone control knob, and the waveforms will change.
If you wanted to test/tweak/audition the "preamp up to the pentode" then you would run it into a pure resistive load for your scope-checks & listening, and wait to add a tone stack later.
-
You're slamming the bloody hell out of the 5879.
beat me to it
:laugh:
-
I felt that tube's pain. :icon_biggrin:
-
I've seen a few instances of grid-current distortion in some of my builds, usually into the PI.
I've solved it by either: 1) subbing in a lower gain tube for the preceding 12AX7, or 2) decreasing the value of the grid leak resistor, or both.
But I haven't tried increasing the bias of the receiving tube. Would that work? Seems to me it may not be enough.
-
It's an overdrive preamp, so per se, heavy grid current and cut off clipping are normal operation.
The trick is tweaking it so that it's pleasing and useful.
-
4.5v peak, and your schematic shows the bias across the pentode's cathode resistor is about 1v.
use your scope post 1 meg gain pot and set it for ~~ 2VAC, use your input device to "adjust gain", evaluate sound, tone, happy factor...
-
Thanks for the helpful replies. I will sub in a 1M load for the tone stack. I knew I was overdriving the pentode, but there is little info on just how much clipping is in the useful/desirable range. I will play with the interstage divider to see where I can tame it down toward a 2V max swing and safe grid territory.
-
the 2V isn't a "goal" per say, it's a starting point to see if you're in the right ballgame
-
More benchtime this morning on this. I put a ~500k pot in for the interstage divider to play with pentode grid drive (the original 1M in my first post being way too touchy). With the guitar plugged in (guitar vol. max), I played around with the gain control and the interstage divider pot until I got a range of desirable clean vs overdrive tones. With the divider wiper at 1285Ω above ground, it was clean at 9o'clock gain, even with some heavy pick attack. At 12o'clock, into some nice bluesy distortion. At 3o'clock it was crossing into some rock/metal crunch. At full clockwise, a hard strum powerchord has an interesting sustained 'swirling' character.
Liking that sound (for now), I then subbed a 1M resistor for the tone stack, and disconnected from the PI/PA stage, I scoped across the 1M load and various points in the preamp as I put signal generator tones through it. I wanted some numbers to what I was hearing. Obviously my pentode grid drive is much lower now. My max grid swing with full cw gain and 300mV instrument jack signal at 1khz was 2.85v.
So then, with the interstage divider as it is now with the wiper so close to ground, it appears I have way more gain out the second stage triode than I really need. So I could set that stage up for lower gain, or perhaps even try a 12AY7 or 12AU7 there.
By the way, this project is just one big experiment. I have long wanted to try many of Blencowe's preamp book ideas and learn more doing so. I'm using his medium-gain preamp design, but trying a pentode for the third stage. I also have a James tone stack, his effects loop idea, and then an LTP-PI, Lar-Mar MV, and p-p 6AQ5's. Fun Fun Fun!
-
pretty much all my builds have an AU7 in front, low B+ and just enough gain to drive a TS with the 2nd triode acting as recovery for TS loss plus a little left over.
the other configuration is the 1st triode is a gain stage, the 2nd a cathode follower, which has no gain but can drive loads better into the next section.
hunt up the Valve Wizards page and read up on DCCF (DC coupled Cathode follower), once I started using that "section" I never looked back.
-
Here;
How to design valve guitar amplifiers (https://www.valvewizard.co.uk/)
-
Thanks for the helpful replies. I will sub in a 1M load for the tone stack. I knew I was overdriving the pentode, but there is little info on just how much clipping is in the useful/desirable range. ...
You may want to check out the book Guitar Amplifier Overdrive (https://www.amazon.com/Guitar-Amplifier-Overdrive-Ulrich-Neumann/dp/132959665X).
... I knew I was overdriving the pentode ... I will play with the interstage divider to see where I can tame it down toward a 2V max swing and safe grid territory.
Your goal is distortion, so overdriving the 5879 is not "a problem." In fact, it's your goal.
Your question was, "What am I looking at," which is why I answered the way I did.
- In itself, "grid current" is not necessarily bad, especially when you want distortion.
- The thing to know is that when the peak of the drive-signal equals/exceeds the bias voltage, the tube will definitely draw grid current.
- When a tube draws grid current, its input impedance drops severely, which usually alters the performance of the stage before (clamping or "clipping" the output).
The points above will help you make sense of the scope-shots. The question of "how much drive is good" or sonic-merits of a circuit is completely subjective, and depends on whether it makes a sound you like. You can attempt to discover some correlation between "sounds good" and "waveform on a scope" but that's mostly up to you to discover. There are some resources out there to explain how to interpret scope waveforms for square-wave & sine-wave testing, but those won't explain "good-sounding guitar amp dirt" via a scope display.
-
... the other configuration is the 1st triode is a gain stage, the 2nd a cathode follower, which has no gain but can drive loads better into the next section...
6G6 Bassman bass channel style :) https://el34world.com/charts/Schematics/files/Fender/Fender_bassman_6g6_schem.pdf
... By the way, this project is just one big experiment. ...
Kudos on trying this stuff out!
Rather than do it on a 'novel' design, it may be a good idea to investigate a standard overdrive preamp circuit first, eg 2204 etc, to get a feel for things?