Oops!
me> scan of Cathode Phase Inversion, O H Schmitt, Journal of Scientific Instruments, Volume 15, Number 3, March 1938
This is not the 1938 paper but another from 1941. Still good stuff.
me> scan of Cathode Phase Inversion, O H Schmitt, Journal of Scientific Instruments, Volume 15, Number 3, March 1938
This is not the 1938 paper but another from 1941. Still good stuff.
In my 50+ years of making things with tubes I have seen two small signal tubes where the grid shorted to the plate.
One was a loose tube that had been banged around in a box for an untold number of years. It was toast when I plugged it into a DIY guitar amp resulting in a fried plate load resistor. The grid had cone completely loose inside the tube and had shorted to both the plate and cathode. This lesson taught me to use 1 or 2 watt plate load resistors. It just makes the amp more reliable when something unusual happens. More common than a shorted tube......An expert "amp modder" sticking the wrong tube in the socket! Yes, I have found an EL84 stuck in a 12AX7 socket.
The second was also in a guitar amp head that did sit on top of the speaker cabinet. It would emit a loud pop every once in a while after being on for some time. I had tried to fix it several times, but could not duplicate the problem on the bench. One day the owner brought me the amp saying that it had finally died. One of the 12AX7's was dead and replacing it restored operation with no further pops. Autopsy on the dead 12AX7 revealed several malformed grid wires and a few that were broken loose from the support rod. It was made that way. I have seen some poorly made tubes in my days, but what are the odds of finding TWO like this! Yes, a spare heater was included in each tube in case one burned out.......no word on just how to change it.
It would seem that an input tube shorted from grid to plate could fry the volume pot in the guitar, or make it become noisy from DC running through it.
One was a loose tube that had been banged around in a box for an untold number of years. It was toast when I plugged it into a DIY guitar amp resulting in a fried plate load resistor. The grid had cone completely loose inside the tube and had shorted to both the plate and cathode. This lesson taught me to use 1 or 2 watt plate load resistors. It just makes the amp more reliable when something unusual happens. More common than a shorted tube......An expert "amp modder" sticking the wrong tube in the socket! Yes, I have found an EL84 stuck in a 12AX7 socket.
The second was also in a guitar amp head that did sit on top of the speaker cabinet. It would emit a loud pop every once in a while after being on for some time. I had tried to fix it several times, but could not duplicate the problem on the bench. One day the owner brought me the amp saying that it had finally died. One of the 12AX7's was dead and replacing it restored operation with no further pops. Autopsy on the dead 12AX7 revealed several malformed grid wires and a few that were broken loose from the support rod. It was made that way. I have seen some poorly made tubes in my days, but what are the odds of finding TWO like this! Yes, a spare heater was included in each tube in case one burned out.......no word on just how to change it.
It would seem that an input tube shorted from grid to plate could fry the volume pot in the guitar, or make it become noisy from DC running through it.
Attachments
Last edited:
> fry the volume pot in the guitar
"Could", but not for-sure.
Take 300V, 100K plate, 100K pot. We get 1.5mA so 150V so 0.225 Watts. Such pots are usually rated 0.25W. So hot, but not sudden death.
Similar path-tracing for part-volume and 5K pickup shunt lead to smaller heat. Worst case could be 0.050W in less than 20% of the total pot element, which gets us to rating.
Yes, it would crackle real bad.
If YOU have seen this twice, I don't think I will live long enough to run into it.
"Could", but not for-sure.
Take 300V, 100K plate, 100K pot. We get 1.5mA so 150V so 0.225 Watts. Such pots are usually rated 0.25W. So hot, but not sudden death.
Similar path-tracing for part-volume and 5K pickup shunt lead to smaller heat. Worst case could be 0.050W in less than 20% of the total pot element, which gets us to rating.
Yes, it would crackle real bad.
If YOU have seen this twice, I don't think I will live long enough to run into it.
And I'll probably never see it again since I don't fix amps any more, but I do collect cheap tubes in cardboard boxes at hamfests if they are cheap enough. Still the odds are slim.
Correction: it would not crackle. P-G1 short makes tube gain slim to none, probably dead-zero. So it crackles but you can't hear it.
The effects of DC current through a pot wiper, especially the old carbon on phenolic versions are often permanent. There will be tiny pits of "deadness" that cause noisy or erratic operation even once the DC is gone. Sometimes this can be fixed with a pencil, but it won't last long.
I found this out many years ago when I really didn't know what I wad doing. I have no experience on modern conductive plastic versions, but I imagine that they could be fried with enough current.
Reading these two authors is a study in contrasts.
For example, in the section on preamp pentodes, Kuehnel points out that you can squeeze high voltage gain out of a small-signal pentode, and even suggests that this is one of the advantages of a pentode. But he never even mentions the severe and frequent microphony problems that seem to have plagued almost every manufacturer who attempted to use a small signal-pentode in the input stage.
(Merlin) Blencowe's approach to pentode preamps, by contrast, suggests keeping voltage gain down to help control microphony. Merlin also discusses the changes in harmonic spectrum (and associated changes in guitar tone) that occur with variations in anode load. Kuehnel doesn't even mention guitar tone and/or the influence of operating point and load impedances on it.
I have Merlin's preamp book, and used to own Kuehnel's "Guitar Amplifier Power Amps". The same pattern occurs throughout both books; for the most part, Kuehnel's approach reads like a standard audio engineering textbook, focused, at least by implication, on linear, small-signal, Hi-Fi operation, studied using elegant, clean, textbook mathematics.
Merlin's book, on the other hand, deals with all the weird and wonderful things that make valve guitar amplification so far removed from valve Hi-Fi design - everything from horribly-designed cathode followers that turn out to sound good, to grid current flow and the host of problems that come with it.
I entered the area of valve guitar amp design with a solid understanding of traditional (linear, Hi-Fi) solid-state audio electronics designs and principles. Starting from that point, for me, at least, Merlin's book was pure gold, full of information I hadn't and couldn't find anywhere else.
Kuehnel's book, on the other hand, was quite interesting as an abstract intellectual exercise, but pretty much useless in terms of actually understanding the assorted weirdnesses that go into valve guitar amp design.
Obviously this is just my opinion. You may, or may not, share it.
-Gnobuddy
So far, when trying to understand circuits, I've been taking the approach of figuring out the most ideal behavior, and then trying to understand the deviations from that. So hopefully both books will be useful to me!
This sounds really cool! I also have no idea where to find the first edition... Did he just take it out of subsequent versions? 🙁
And thanks to everyone for the information and links so far.
This sounds really cool! I also have no idea where to find the first edition... Did he just take it out of subsequent versions? 🙁
And thanks to everyone for the information and links so far.
I agree with Gnobuddy about Merlin. I interacted with him quite a lot in the past and IMO he definitely had the right approach for guitar amps, although I tend to think some of his ideas are a little to esoteric at times to be useful in real life guitar amps. Still fun to look at, though. 🙂
Sent from my LG G2 with CM13 using Tapatalk
Sent from my LG G2 with CM13 using Tapatalk
> Kuehnel... useless ... weirdnesses
The 3rd edition addresses overdrive. I agree he is not teaching a gut-level flavor of amplifier insight. I think his works give an outline which may, in time, be fleshed-in by other writers.
Apologies-- it seems the "1st edition" didn't really happen. The Cramer's Rule matrix analysis was in the 2nd. I gave my copy to a friend. I may have to get it back-- it sells for $40 now! (Try ABE.com ISBN 0976982218) If really stuck yet broke, try writing to the boss (Richard) at PentodePress.com Mention "PRR"- I think he knows the handle and we had some words about analysis techniques (it may be my fault he took out the matrices).
The 3rd edition addresses overdrive. I agree he is not teaching a gut-level flavor of amplifier insight. I think his works give an outline which may, in time, be fleshed-in by other writers.
Apologies-- it seems the "1st edition" didn't really happen. The Cramer's Rule matrix analysis was in the 2nd. I gave my copy to a friend. I may have to get it back-- it sells for $40 now! (Try ABE.com ISBN 0976982218) If really stuck yet broke, try writing to the boss (Richard) at PentodePress.com Mention "PRR"- I think he knows the handle and we had some words about analysis techniques (it may be my fault he took out the matrices).
Last edited:
I was taught Cramers Rule in high school, and have taught it to my own math students as recently as a few months ago.
For two equations in two unknowns, it's quick and effective. But, IMO, Cramers Rule is tedious, slow, and error-prone for larger sets of equations, even slightly larger; 3 equations in 3 unknowns is a lot more cumbersome than a 2x2 problem.
For anyone who's curious, here is a nicely done, very clear explanation of how to use Cramers Rule to solve two unknowns connected by two equations:
Determinants and Cramer's Rule for 2x2 Systems 1 | Coolmath.com
The days when you had to do this sort of thing entirely by hand are gone, though. If you shell out the $15 USD or so for a Sharp EL-W516x calculator, it will solve 2x2 and 3x3 sets of equations with numerical coefficients for you at the push of a few buttons!
-Gnobuddy
Microsoft Mathematics can do that kind of thing easily and you can download it for free!
- Status
- Not open for further replies.