Simms watts 200 question

[catalin gramada]

Hello. Got a Simms watts 200 to restore, 4xkt88 in UL mode supplied at ca. 650V arround. What is wondering is its poor 12ax7 driver into big 220k grid leaks. Thinking more to add a pair of 12au7/12bh7 cathode followers in between and to reduce common grid leaks to 100k or less to have better drive capabilities. What are You thinking about, please? Thx.

[Helmholtz]

Why redesign? Does the amp exhibit a problem?

[catalin gramada]

Why redesign? Does the amp exhibit a problem?

It seems unsafe with those big value resistors. 12ax7 seems weak to can push the power tubes into grid current territory. For the moment it not work properly as it is, have to do a fresh recap. but the reponse in bass is weak for actual standards.

[catalin gramada]

Thinking more like this.


Supplied at 600v seems within the limit. With 33 k on the plates got 240v iddling and max 500v max negative swing for 0 mA current to 100 k load. 100k grid resistors also should be safe enough for a pair of kt88 biased for less than 35w dissipation , as data sheet states ( 220k per tube biased less than 35 w-fixed bias op) Biased at -10v got 300V pk-pk voltage swing. Should be safe enough I think.

[Helmholtz]

Some thoughts:

- Where did you find the info that the 12BH7 could stand plate voltages of 600V at signal peaks?
- Additional tube will reverse signal phase.
- Additional loop gain may cause stability problems and might require redesign of NFB circuit.

To me it looks like overkill.

[catalin gramada]

Is not 600v on the plate. It will be supplied at 600v , almost same as screens for max excursion. The plates will stay at 240v around at idling and max excursion negative peak will be 500V for 0 mA with 100k load. This is within the absolute admissible voltages.

[Helmholtz]

Is not 600v on the plate. It will be supplied at 600v , almost same as screens for max excursion. The plates will stay at 240v around at idling and max excursion positive peak will be 500V for 0 mA with 100k load. This is within the absolute admissible voltages.

So your screen supply drops by around 90V (taking into account the cathode voltage of the 12BH7)? What about transients as supply sag takes some time?
I couldn't find a 500V limit either.

[nickb]

Why redesign? Does the amp exhibit a problem?

It's been working for 50 years so I don't see the justification for a redesign unless you are seeing thermal runaway issues with the KT88s. Seems better to solve the known issues which are

(1) Not working right. What exactly is/are the problem(s)?
(2) Lack of bass. Have you measured the frequency response of the just the power amp? Just the -3db LF and HF points relative to 1Khz into a resistive load will do. -3dB is where the voltage is about 0.7 of the 0db value.

[catalin gramada]

I just looking what Marshall did in their 200w heads. 12AU7 driver supplied at 620V with 310V on plates at iddling,biased at 15v around

[Helmholtz]

I can't see PI supply voltages from the Marshall schematic. It's a UL design not requiring a separate screen supply. And yes, the ECC82 has a 550V Vp limit at zero current.
But I dunno, is the 200W Marshall actually a reliable, time tested design?

FYI, the Marshall model 2000 (250W, 6x6550) and model 2001 (375W, 8x6550) both used a "standard" ECC83 PI.

[catalin gramada]

UL yes, like mine. Voltages in the chart attached for 200w

[Helmholtz]

UL yes, like mine. Voltages in the chart attached for 200w[ATTACH=CONFIG]57300[/ATTACH]

Yes - no PI supply voltage identifiable from the Marshall chart.
But I would estimate something like just below 600V at idle.

[catalin gramada]

C'mon... with 15V over 1.5k R cathode meant 10mA / 2=5mA/ tubeŚ47k=235v + 310on plate=545 V.. I think...almost 600...not quiet but close

[Helmholtz]

C'mon... with 15V over 1.5k R cathode meant 10mA / 2=5mA/ tubeŚ47k=235v + 310on plate=545 V.. I think...almost 600...not quiet but close

Yes, up to 550V at idle would be safe for an ECC82.

[catalin gramada]

And still 12bh7 at 600v, 33k plates/100k load, great swing, much better



It will take only 5-6 volts aroud on input to get power tubes full drive