JTM45 Build

[pdf64]

Ha, oh well, let's wish your tech good luck then!

[Littlewyan]

Well he has seen the issue himself now and believes the output stage is solid. Although when he tested the output stage he put a 4.2V 1Khz signal into the PI to test it clean. I suggested putting a larger signal into it as obviously when playing the amp I'm not playing clean but he said its the standard way of testing.

Anyway I'm a bit concerned because he said he'll get it working even if he has to mod it. It shouldn't need modding though .

[pdf64]

Please could you re-iterate the exact issue (or point me to a post that does)?

[Littlewyan]

When playing with the volume on 10 and treble on 10 with everything else at 0 and you hit a big chord, the initial sound sounds like its struggling to get through but then the volume slowly comes back up and the rest of the note sounds ok. When monitoring voltages I found the PI goes iffy.

Tested by putting a 1Khz sine wave into the amp, set the controls to make it go wrong and the 82K anode swings up to 250-300V and the 100K anode goes down to 160-180V. Both were balanced at 200V to begin with.

[pdf64]

That seems to describe blocking distortion; if the bias supply output voltage was monitored, ie at the node formed by the 2x220k grid leaks, the magnitude of the bias voltage may be seen to increase significantly with the attack of the big chord and then reduce as the caps discharge. Effectively tracking the note envelope.
Has that blocking distortion mitigation been tried to see if it helps?
I have found that the most useful single change for this is to reduce the grid leak resistors value to 100k.
That acts to temper the gain and max available voltage swing from the LTP, and to discharge the 0.1uF coupling caps faster.
It's a common thing for guys that use a lot of overdrive (from pedals or pre-amp) to complain that when they play loud at a big gig, when they stamp on the overdrive the volume goes down.
It's generally due to a 0.1uF/220k RC coupling at the power tube control grids acting as described above.
The shift in LTP plate voltages reminds me of Merlin's advice to not to have overly high a value of cap at the 'other'/inverting LTP input grid, in order to avoid blocking distortion; suggest trying 22nF in the first instance, same as at non-inverting input http://www.valvewizard.co.uk/acltp.html

[xtian]

"Like" button for this PI advice, Pete.

[Littlewyan]

Hmm do you think that affect would still happen even if I lifted the coupling caps so they were no longer connected to the 220k grid leaks or the power tube grids? As I did try that for definite and it didn't fix it.

[pdf64]

If the coupling caps were lifted then the LTP plates would be unloaded, which is a change in conditions, and also the VB+ would be higher due to the power tubes not being cranked and sucking the saggy B+ down; so not sure if the test is really that valid.

Whereas loading the LTP plates with 100k grid leaks may well help, as previously described.
The previously linked Merlin page mentions blocking distortion at the LTP, and that reducing the 'other' input grid cap value below the traditional 0.1uF helps with that.
Additional blocking distortion mitigation at the LTP may be to add some 'grid stoppers.
I suspect that will definitely help, as the settings that bring on the issue involve connecting a massive output with low source impedance (ie the CF tone stack driver) via a cap to the LTP main input grid. The cap is bound to charge up, pulling the grid down / its plate up, thereby putting that triode towards cut-off. The LTP action will inevitably force it's partner triode towards saturation.
This seems to explain the conditions you describe?

In reality though, my feeling is that the issue is unlikely to be a problem, because the settings result in an obnoxious tone and the user would normally tweak the settings for a better tone.
If such a tone is required and those settings must be accommodated, then fitting grid stoppers (eg ~22k-100k) to the LTP should help.

So, to summarize, my hypothesis is that almost any closely similar circuit should exhibit similar performance artefacts in response to those control settings and massive overdrive.

[Reeltarded]

Wow, so all along it just needed some stoppers? Ya think?

(ref pgs 1-39)

[Littlewyan]

Miles I already tried grid stoppers remember? On all stages. 100K.

I know lifting the coupling caps is a change in conditions but the issue with the PI going funny still occurred with the coupling caps lifted. Which says to me they are not the issue.

Also it seems a bit odd that the issue won't resolve itself unless I bring the middle up to at least 5 and even then its still partly present. Turning the bass up doesn't change the issue at all. So if I wanted to scoop it by turning bass and treble to 10 and middle to 0 I still have the issue.

Added to which Joe on this forum has a JTM45 and he tried the exact same settings. Said the tone wasn't great but the amp still worked correctly. So it can definitely be done.

[pdf64]

OK, I'm glad I phrased it as being a hypthosis then!

Joe on this forum has a JTM45 and he tried the exact same settings. Said the tone wasn't great but the amp still worked correctly

Was the same input signal (freq & amplitude) being used?

Also was his the exact same circuit? Folks seem to make a mod or two and still refer to their amp as a clone ('mod blindness').

[rp]

thinking back, this is what happened to me 25 years ago, pre www, pre TAG. two in a row, knocked the wind out of me. talked to a tech about it, he showed me an open matchless and said I could try ptp as a good way to avoid wiring issues. put the wind back in my sails and has kept me going ever since. my 5G9 will be with a board, first I've done in 25 years - now I'm spooked! best I have some 100K grid stoppers on hand!

[Littlewyan]

Well no I mean he tested it with his guitar. Its pretty much the same circuit.

[pdf64]

Does your amp exhibit the problem with a regular guitar as the input?

[Littlewyan]

Yep. As I'm playing I can see the PI swing when I hit big chords or certain notes.