JTM45 Build

M Fowler · Post by **M Fowler** » Fri Aug 07, 2015 3:47 pm

Sounds like the main filter caps drain when the guitar sends a large signal, too much sag.

Littlewyan · Post by **Littlewyan** » Fri Aug 07, 2015 4:36 pm

Nah tried replacing the mains filter cap, same issue. Also don't forget the amp works fine with everything flat out. Just going to give the tech a call to see how hes getting on.

pdf64 · Post by **pdf64** » Fri Aug 07, 2015 10:30 pm

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

Does it actually cause a problem you can hear?
And if it did, is it conceivable that those settings would be selected?
In the same way that a Mesa Boogie will oscillate if everything is turned up in lead mode.

Littlewyan · Post by **Littlewyan** » Fri Aug 07, 2015 10:58 pm

Yeah you can hear the issue straight away, thats how I found out about it!

Well basically I'm limited if I leave it. I can't turn the middle below 5 or the treble above 5. Pretty crappy tone controls with limitations like that. Plus I may switch to the 33k/500pF Tonestack in future, however that atm just makes the issue much worse.

My tech txt me tonight saying hes done a few mods but needs to do more testing before commenting further. Hopefully he hasn't changed the circuit as thats not really a fix in my eyes.

I'm starting to wonder if its something like a resistor has too high inductance for some reason and is causing the problem. Everything else has been replaced!

pdf64 · Post by **pdf64** » Sat Aug 08, 2015 12:31 pm

What is the static VB+?
Did you used 10k or an 8k2 dropper to the PI node?
I'm wondering whether the output from your CF tone stack driver is particularly strong for some reason.
From the info I'm fairly sure that the issue is blocking distortion at the LTP, and so it should be able to be brought under control by the normal tweaks, eg reducing cap values, grid stoppers.
Have you tried it with various guitar / pickup types, and if so is it a problem with some rather than others?
Or using a clean boost type pedal?
I don't get this issue with my JTM45 type amp, but then it's by no means a clone.
Also I run the presence at max, in addition to the bigger bright caps.

Littlewyan · Post by **Littlewyan** » Sat Aug 08, 2015 12:34 pm

The HT is 480VDC.

I've used 2 x 8K2 Droppers for the PI.

Its not the guitar as the guitar is fine on my Express amp. There is a boost switch on the guitar which makes the issue worse.

pdf64 · Post by **pdf64** » Sat Aug 08, 2015 3:04 pm

Even though it should sag lots under load, the VB+ looks on the high side of things. And the 8k2 droppers will tend to keep the downstream nodes even further to the high side too.
A clean boost will tend to send the input level right up, eg ~4Vpk / 2.8Vrms.
As described, as I see it the root cause is a massive signal from the CF low impedance source slamming the LTP PI.
Your big input signal and high B+ will tend to corroborate with that.

The Express was designed 'expressly' to be overdriven, ie transition in and out of clipping gracefully without any unpleasent artifacts; blocking distortion has been well catered for, eg no CF driving into RC coupling with long time constants.

Whereas the Bassman etc was probably designed to be used clean, and that the circuit almost always sounds glorious overdriven is likely a happy coincidence.
With a bit of tweaking to mitigate blocking distortion etc, hopefully we can get rid of the 'almost'.

Let's see what your tech comes up with, but if it's not to your liking, there are options to explore, eg 10k droppers, reduced LTP grid cap values.

BTW which choke type did you use, eg the 20H (resistance ~500 ohms) or the 3H (resistance~100 ohms)? My experience is that plenty of resistance, eg >=500 ohms, in the supply to the choke node is generally beneficial, in regard of tone, response and tube life when overdriven.
That seems to align with Trainwreck designs.

Littlewyan · Post by **Littlewyan** » Sat Aug 08, 2015 3:41 pm

The choke is a 3H.

I'm pretty sure its not the high B+ as my old 1987xl ran with 470VDC on the plates and I never had this issue. Plus this amp originally only had 1 8K2 dropper, I added the other later on and it didn't make any difference.

I still haven't heard from the tech so I'm assuming the mods didn't work.

Don't forget I'm sure I recreated the issue by driving the PI by itself with my oscillator. The CF was not connected at that point. I know I definitely recreated the issue by driving the PI directly from the mixer resistors completely bypassing the V2 circuit altogether.

pdf64 · Post by **pdf64** » Sun Aug 09, 2015 10:17 am

OK, trying to rationalise this in the context of my blocking distortion hypothesis, I think that the 1987 has bigger dropper resistors in the B+ string, and a higher resistance choke. So despite the Si rectifiers and high static VB+, the 1987's static VB+ down the line may be lower than yours.
With high signals, the 1987's EL34 screen grids may draw rather more current, hence increased voltage drop across the choke, compared to a JTM45's 3H choke and KT66, so further adding to the discrepancy in pre-amp VB+.
So the high VB+ line of your amp will facilitate a high max potential signal voltage out of the 1st stage, as the loading is low, thereby pushing the LTP into blocking distortion, even without the enhanced drive capability of the V2 CF stage.

Littlewyan · Post by **Littlewyan** » Sun Aug 09, 2015 10:37 am

I've also got EL34s, plus with a valve recto the sag will be much greater. The B+ drops to 400VDC when you start playing. Also the 1987 uses the same choke, 3H 105R. I thought blocking distortion is mainly affected by lower frequencies?

Reeltarded · Post by **Reeltarded** » Sun Aug 09, 2015 3:09 pm

Amplitude

Littlewyan · Post by **Littlewyan** » Mon Aug 10, 2015 8:06 am

Well then its less likely to be blocking distortion as I can fix it by turning the middle control up which makes the signal much larger.

Littlewyan · Post by **Littlewyan** » Mon Aug 10, 2015 8:50 am

Just rang my tech. He said he modded the AC Supply but didn't say what he did, moved the feedback wire from the 8Ohm Tap to the 16Ohm tap and put a capacitor over the feedback resistor. None of which it should need. Hes just modifying the circuit by doing that, not fixing a circuit that should work. But he said let him test it first so who knows, see what happens. I think hes found the issue is in the PI now and not earlier in the circuit.

pdf64 · Post by **pdf64** » Mon Aug 10, 2015 4:53 pm

Littlewyan wrote:Well then its less likely to be blocking distortion as I can fix it by turning the middle control up which makes the signal much larger.

My experience is that blocking distortion may be worse at some settings rather than others, and it's not intuitive. Whatever, lot's of treble can make it more prominent.
Don't know why - maybe something to do with the phase advance that goes with treble boost, or the harmonic content against time of the ADSR note envelope?
The fact you've noted a shift in dc conditions when overdriven indicates that there's a bias shift; if that get significant then it's described as blocking distortion.

Littlewyan · Post by **Littlewyan** » Mon Aug 10, 2015 10:45 pm

Got the amp back and it has the same issue. If it was blocking distortion I would have seen it on the scope which I'm sure I haven't.

Tube Amp Archive

JTM45 Build

Re: JTM45 Build

Re: JTM45 Build

Re: JTM45 Build

Re: JTM45 Build

Re: JTM45 Build

Re: JTM45 Build

Re: JTM45 Build

Re: JTM45 Build

Re: JTM45 Build

Re: JTM45 Build

Re: JTM45 Build

Re: JTM45 Build

Re: JTM45 Build

Re: JTM45 Build

Re: JTM45 Build