KT88 JMP

[Tage]

an associate was asking me about using KT88's in his JMP. I'm sure this is possible as some were built with 6550's, although this one is a EL34(british?) model. so if I change the bias supply and grid leak/stoppers and add some screen grid resistors, 1k 5 watts, will a swap just need a bias adjustment afterwards? I'm sure the output transformer will not be optimized for KT88's, but I still think this will result in some sort of power increase right?

Just another question that should be pretty easy for anyone to once again clear up for me...

[greekie]

The functionality of a tube is much better understood if it is thought of, as it's british word - a valve. A KT88 can be though of as a bigger valve than a EL34 with a wider diameter. But if the power supply isn't able to make use of the bigger valve, power will remain more or less the same.

What would be the purpose of using KT88's?

Jake

[Tage]

I can't think of a better reason for using them other than a minor tonal variation. He is hoping for more headroom or a bit cleaner sound. I told him he should probably buy a different amp then, I assumed that was all I would be told here too. I suppose there will be less power amp distortion as the tubes will be running no were near there maximum dissipation, or is that a wrong assumption? the power supply gives about 480V to the plates, it's a mercury transformer rated for 520V the output transformer is a Mercury as well it's "optimized for KT88/6550/6L6". Also no worries this thing is a clone-ish, no original stuff being harmed.

[MWJB]

Fitting 6550 or KT88 to a 6L6, or a EL34 amp may 20-30% more power, but dynamics will tighten up, the tubes will distort less.

If the amp runs EL34 now, then it should already have 1K 5W screen grid resistors & 5.6K grid stoppers. 6550 or KT88 only draw marginally more heater current than EL34, so they can be considered a drop in replacement (as can 6L6), requiring just a bias adjustment. Be aware that bear trap tube retainers should not be used, spring cradle retainers will prevent the metal base rings in some 6550/KT88 grounding & killing the tube.

[Steve Conner]

Well, putting KT88s in it won't turn it into a 100 watt amp. But if it has beefy oversized Mercury transformers, you might be pleasantly surprised. Maybe it would do 70 or even 80.

Besides the extra wattage, which isn't really that noticeable on stage, the tubes will change the sound. EL34s generate a lot of distortion that gives them a dirty, aggressive bite even when they're not cranked up full. But KT88s are more linear and hi-fi sounding, which many people appreciate in a high gain master volume amp. Maybe it even gives the impression of more headroom than the small increase of measured power.

They should be drop-in replacements as noted by MWJB. And as he says watch out for that metal base, in an ideal world the amp maker would have chosen the pinout so that it ended up grounded, but it isn't an ideal world.

[Tage]

excellent stuff, so if it has spring type reatainors I should be good to go? my brothers JCM800 had those on it and they didn't contact the base of the 6550's as I recall.

I was fairly certain it would not be a problem, the only other thing is I was looking at the SED data sheet for KT88's and noticed that they are only reccomended to have a input grid resistance of 220k if they are in a fixed bias configuration. I know that lots of amps use only 100k grid leaks and 1k input grid stoppers for 6550/kt88's is that the reason they use them? not a critical question but I'm kinda curious.

Whoops just looked at the JJ one and it says 100k for a setup Wa+g2>35Watts, so my question still stands, are 220k grid leaks the amp has for use with kt88's too big?

[pdf64]

My buddy has a Park 75, which is basically a JMP with KT88, supposedly does 75 watts. It's a while since I last checked it over, so not sure of the OT primary impedance.
It would be best to lower the grid leak resistors from 220k; I think that the 6550 marshalls used 150k. Pete.

[dai h.]

Personally I would be careful not to bias it hotter since my understanding is that Marshall biased their 6550 amps like EL34s(i.e. 25W dissipation)--this is with stress to the B+ winding in mind (may not be made to handle the extra current draw). Also, I'd probably use the lower grid load/bias return, then maybe increase the coupling cap(s) size for (low freq.) compensation (bigger cap = less low freq. loss) as necessary.

[Ronsonic]

I can't think of a better reason for using them other than a minor tonal variation. He is hoping for more headroom or a bit cleaner sound. I told him he should probably buy a different amp then, I assumed that was all I would be told here too. I suppose there will be less power amp distortion as the tubes will be running no were near there maximum dissipation, or is that a wrong assumption? the power supply gives about 480V to the plates, it's a mercury transformer rated for 520V the output transformer is a Mercury as well it's "optimized for KT88/6550/6L6". Also no worries this thing is a clone-ish, no original stuff being harmed.

There is definitely more clean headroom to be had from KT88 or 6550s in Marshall type amps. Not more total power, that's pretty well limited by the rest of the amp. If he's playing metal then this gets even better with a tighter bottom end that won't mush out when fed the sludge. For classic rock, blues and other stuff I prefer EL34s, but that's just a matter of taste.

The difference won't be huge, but it is there and a guitarist will feel it.

[Alex R]

If the slightly higher heater current is a problem, which it probably isn't, you'll see a lower heater voltage with the KTs.

[Tage]

Yeah I emailed the guy from Mercury and he said they make those PT and OT's to handle the increased heater draw and current load on the HT winding so it should be ok to go for a slightly higher dissipation I think. I already upped the coupling caps to 100nf with a PPIMV, I think that should suffice. Now I also lowered the grid leaks to 100 ohms and the control grid stoppers to 1.5k

Thanks for the feedback guys

[EFK]

On the MKII JMP made for the US market w/ 6550s, 82K grid R were used. Don't know if it's relevant, just mentioning it. I think Pete's suggestion above of 150K would have been a better fit, maybe so many wouldn't have converted over to EL34.