Glowing screen grids

[renshen1957]

Well as usual your just running the crap out of your amps over there

Perhaps larger screen resistors or lower that B+

I have used larger screen resistor values on amps (2k2-2K7) on amps that are run full bore all the time and on smaller wattage amps that are invariably used as a screamer, however this will make the amp sound a little darker (unless compensated elsewhere in the preamp). That might be enough to drop the screen voltage.

Best Regards,

Steve

[John_P_WI]

I have used larger screen resistor values on amps (2k2-2K7) on amps that are run full bore all the time and on smaller wattage amps that are invariably used as a screamer, however this will make the amp sound a little darker (unless compensated elsewhere in the preamp). That might be enough to drop the screen voltage.

Best Regards,

Steve

+1

Also the larger screen resistor values can make the amp more touch sensitive and responsive.

Too large values can make things a little "squishy", experimentation is key.

John

[azatplayer]

I dropped a 1K in between the choke and the screen R's. Fixed things up ok, but gave the amp too much compression. Was kinda cool for a few minutes, but not what this amp is about.
I pulled it and have a 500R in there at the moment, not bad. Gonna try 200R next. I should pull the choke so i can read its dcr, but that requires lifting the power supply board..
But i have inadvertently discovered a fix for the harsh top it has.
I built the amp using the mod that others have used, to tighten the bottom in the fast mode, placing a lower value coupler after the second stage. Pete Thorns amp supposedly has a .0022 like the Express does.
I left the .1 in the gradual side as im using Danas switch which allows me to use 2 values. Gradual side has always been good, but the fast side has been had more treble. So i ended up with a compromise using the same values on both, .0068. This sounded pretty good apart from the high end brittleness.
Also kinda sounded a bit strangled. Like it needed a stent in an artery.
So i changed them back to both .1uf last night as spec, and man what a difference. Soooo much fuller and rounder, much much nicer.
It does tend to fart out when driven in the fast mode hard, but its way better than with the lower couplers.
I read that Pete Thorns amp has something else done to it, clearly you cant just do that cap.
I know from dicking with an express/liverpool, that dropping the 150K grid ref on the third stage to reduce gain, requires changing the coupler there too to keep the same response in the amp. This is the same spot, and i knew this when doing it, but listened to guys who have done it with apparently good results.
I think there must be a further mod that Friedman did on that amp.

[Roe]

I...Pete Thorns amp supposedly has a .0022 like the Express does.
I left the .1 in the gradual side as im using Danas switch which allows me to use 2 values. Gradual side has always been good, but the fast side has been had more treble. So i ended up with a compromise using the same values on both, .0068. This sounded pretty good apart from the high end brittleness.
Also kinda sounded a bit strangled. Like it needed a stent in an artery.
So i changed them back to both .1uf last night as spec, and man what a difference. Soooo much fuller and rounder, much much nicer.....

I'm using 15nf in slow mode and 3n3 in fast mode but this is a k60. the switch is after the caps. this is much easier on the switch, since you dont have hundreds of volts over it

[azatplayer]

Bloody funny business this amp building caper!
If ever " your milage may vary" was apt for a bunch of people, its here!!
Such subtle differences everywhere, and slight missed things, solder joints, differnet caps, types of resistors, names on iron. Gees. No two amps the same!
Would be a cool experiment to have guys build the same circuit and have a show, hear what the differences are, cos theyll be there!

[M Fowler]

Yes Don I agree the amps would have recognizable differences.

I built one K60 clone and the other I designed my own boards all the same component values but the coupling caps were different type.

The non clone was built into a Mojo Fender Bassman chassis and custom Lopo head cab. I can hear some difference but don't prefer one over the other though.

The next one will be a K50 which should not sound the same as the two K60s based on the obvious differences in components, lower OT primary, same PT and solid state rectified.

Krank them up and run those tubes hot

Mark

[Structo]

@ azatplayer

Interesting what you said about the harsh top end going away when you increased the resistor to the screens.

Did you measure your other B+ voltages after you added that resistor?

I believe that correlates with what Andy Le Blanc found when he was experimenting with it.

Supposedly Robben Ford's Dumble has a 300 ohm or so resistor instead of a choke.
I remember experimenting with that resistor in my D'lite but I later installed a choke.

Some Marshall amps had a floating resistor feeding the screen resistors on the tube sockets.

This JTM45 has a 1K feeding the 1K screens.

[pdf64]

It's a common 1k that then splits into individual 470 ohm resistors for each tube screen grid.
That common 1k makes a big difference; if a bypass cap is connected to that node, the large signal waveform shape changes noticably. and the tone seems to go 'harder'.
My feeling is that the unusual screen grid arrangement is a significant component of the JTM45 tonal character.

Large individual screen grid resistors cause significant local negative feedback, ie an inverted version of the main signal appears there, thereby reducing gain and output. As noted, they also contributes to a compression effect, due to screen grid current not being linearly proportional to plate current, rather it ramps up at high signal levels / overdrive.

[Structo]

Yes but that method is one that the guys at the Metro forum were using on other amps as well when they noticed their screens glowing.

Not sure what mean about negative feedback through the screen resistor, can you elaborate on that a little more?

I always thought going too big on the screen grid resistors caused the bass to sound mushier, or looser.

[azatplayer]

Just tried a set of my mates KT77's in there, have removed the series screen R, and hit the thing pretty hard, he yells out (as hes watching the tubes) for me to stop. He saw the screens getting hit pretty hard and as theyre his tubes...
I tried a set of JJ KT66's in there, wow, home run! So much more detail than the GT EL34's i was used to thru the breaking in of the amp. EVen the Svets i was running, the VA KT's, the Winged C 6L6's, none can compareto the sweetness and chime and bottom authority of the JJ 66's. Very detailed.
The GT's however when run real hard sound explosive.
Im so glad i finally got this amp sorted, been arduous to say the least, its near made it to the shelf for parts! What a journey.
Now just gotta wait for my new Larry Mann headshell, and Im golden!

[eddie25]

I like those KT77's, but I'm skeptical of the 800V rating for the screens. I have an amp I built with KT88's and almost 500v screens and I'm tempted to try them in that amp.

Did you change impedance on your amp when switching to KT66's? or did you leave it the same between tube changes?

I've been getting a little frustrated lately that most ready to buy Marshall trannies give you at least 475+ screens and that most tubes are rated for 450. But, most the amps I've built like this are still going with original tubes, years later and sound good.

[Gaz]

I've gotta say that I've never had more failures than with the JJ KT77s in non-master Marshall circuits. I'll never use them again, and I attribute the failures to a misaligned screen grid. I've even had 5 watt screen resistors open up with those tubes. Whenever I've replaced them with E34Ls or 6L6s, I never had the same problem recur. Buyer beware!

[Roe]

I...
I've been getting a little frustrated lately that most ready to buy Marshall trannies give you at least 475+ screens and that most tubes are rated for 450. ....

yes, the 100watter typically have high voltages. but sag in the power supply helps greatly

[pdf64]

Not sure what mean about negative feedback through the screen resistor, can you elaborate on that a little more?

I always thought going too big on the screen grid resistors caused the bass to sound mushier, or looser.

The screen grid is a grid - the voltage there will affect the operating point of the tube, Vac there will modulate plate current.
But also screen current is related to plate current, at least that's how it works out in guitar amp circuits.
The net effect of the above 2 things is that a version the input grid signal will appear at the screen grid.
The screen grid signal will be inverted compared to the input grid signal.
The larger the screen grid resistor is, the larger the signal will be.
The value of the screen grid signal will have a roughly exponential relationship to the input grid signal, at low signal levels it will be insignificant but at high signal levels its significance will increase - the screen grid resistor value will be a factor as to how significant it will be, in terms of tube operation.
If significant, there will be a local NFB effect because the screen grid signal is in opposition to the input grid signal, which serves to reduce the gain and plate current of the tube.
That may be perceived as a mushier bass, because bass notes require power and the power output is being reduced (by the effect of the large screen grid resistors). Also I guess that the output impedance may be increased, which would also tend to increase bass mushiness, but I've not worked out how that can be (easily) quantified yet.
Pete.

[azatplayer]

The Concorde has a 4K8 primary which kinda sits in the middle of a lot of tubes, 6L6's, EL34's, KT66's have all sounded great, but very different.
I had a set of JJ KT77's years ago after they were released, in a Crate combo.
They lasted about a week in that amp before the halo fell off one tube..
The seller would not take responsibility cos it was in a combo, no warranty implied. I assured my mate that the screens glowing was in line with what happens on these things, he still wanted them out
Fair enough!
Im gonna unpick the choke and measure its resistance, then work around placing series resistance up to about 500R combined i think.
When i dropped in that 1K a few days ago, it was from the screen node directly to the screen resistors. Ive seen a few amps that do this, but my logic is telling me series resistance with the choke may have less compression affect.
Gotta pull the PS board tho ;(