single ended quirk

[Andy Le Blanc]

I've been working on a SE amp to use 6l6/6550, kt88/90/100/120.

I've noticed a funny distortion with kt88/100 when the G2 voltage is set too closely the plate voltage.

Say within... 5v negative to the plate, it is not apparent when the G2 becomes more negative.

Its been consistent with both KT88 and KT100, and its not there with 6l6 types.

And its not there with P-P AB at similar voltages. It's a easy fix, just set the G2 lower...

But I was wondering why?

I inadvertently ran into G1 current with EL84 and EL34 vrs 6v6 and 6l6.
Again its easy to avoid once your aware of it, its natural to the tube type.
You can get away with some paraphase inverter designs with 6l6 but have to accommodate the design for EL34 for example.

I'm making a similar assumption here too. But a good explanation would nice, I haven't found one yet...

[Stevem]

Whats the plate voltage you have?

[martin manning]

Screen circuit oscillation? Do you have a screen resistor at the socket?

[Andy Le Blanc]

fender voltages... champ/princeton pt ... plate voltage is landing around 375 in the circuit...screen is -5v (370)

yes the screengrid resistor on the socket, I use a bleeder and the voltage is adjusted on the power rail.

6K load... everything is within typical operation, head scratcher.

and its consistent across the tube types I have on hand.

it not a problem to work with a different operating point.

Oscillation might be it, I'll review the old sheets for the tube type and triple check the recommended applications, might have missed something useful.

[statorvane]

6K load... everything is within typical operation, head scratcher.

Well for a KT88 in SE, 375 V on the plate, 42 W plate dissipation, I get a Z of around 3.3K. From Merlin's SE pages, Z=Va**2/Pa. You might try dropping your primary impedance and re-evaluate.

Also (I'm just getting back into SE after a 10 year hiatus), I have noticed the AX84 SEL. (6550 / KT88) runs the screens about 50 V less than the plates. This was a surprise as my only SE amp prior was a P1 eX. I am assuming these big octals were designed for ultra-linear operation with reduced screen voltages and may not sound their best when the screen voltage is close to the plate.

Hope this helps.

[Andy Le Blanc]

Doing some digging on the KT66/88 types... the salient statements I've found we're precautions
about the tubes high mutual conductance and parasitic oscillation, the values G1 screen resistor is typically much higher than I assumed.
The larger value 10k is recommended for the tube type, I think I have 1.5k or a 2.2k at the moment... I try that avenue next.

KT88 are a beefed up KT66, design around 1937 and used thru the war.
I don't think they were designed for ultra linear or distributed load circuit.
There is a feed back element with that design, and operation depends upon which % tap is used, the G2 voltages come with 10v of the plate in some of the quoted data.

I was encouraged to use the 6k load from the TAD KT88str sheet, and I had a 6k SE opt on hand...

I'm still digging, I would be nice if its simply the G1 screen resistor and parasitic oscillation., we'll see.

[martin manning]

Some KT88 history: http://www.vacuumtubevalley.com/Magazines/VTV19a.pdf

[Phil_S]

Some KT88 history: http://www.vacuumtubevalley.com/Magazines/VTV19a.pdf

Forget the articles (nice read anyway), check the ad at the bottom of page 25!

[NickC]

Some KT88 history: http://www.vacuumtubevalley.com/Magazines/VTV19a.pdf

Forget the articles (nice read anyway), check the ad at the bottom of page 25!

Unprecedented response to your touch ....

[matt h]

(deleted)