Vintage Modern KT66 Plate Voltage

[Ukmarshall]

Hi,

New Vintage Modern 100W and already need advice, so first post!

I recently acquired a second hand 2466 head in excellent condition. It's from 2010 so has twin bias adjustment pots, and the valves appear to be stock branded Marshall's. I've had a few hours playing time and as a matter of course decided to check the bias. I have a JVM 410 head and have successfully biased the amp several times, so have a reasonable idea of what to do but have no experience with KT66's.

I have read a few posts and the FAQ on the Vintage Modern so knew the plate voltage for the KT66 is high on these amps. However, I have a few questions regarding plate voltage:

1) As a matter of interest does anyone know why the plate voltage is so high on the 2466?

2) I live in the UK, and in my local area, the mains voltage is high with a norm of around 245v and have measured voltages in excess of 250v at certain times. I checked the bias when the mains input was at 244v and measured a plate voltage of 543v and a bias at the test points of 86mv for the left pair and 88mv for the right. Now if my calculations are correct, this works out at 96% of max plate dissipation (25W for the KT66). My only experience is with EL34 valves, where 70% is considered the max of plate dissipation and I'm assuming this value is considered the max for KT66's. I use 65% for the EL34's in the JVM to allow for the possible high mains voltage.

Now to my problem. If I bias to the worst case scenario of 250v input voltage and for 65% of max plate dissipation, then to achieve the correct bias of 0.058A pushes the plate voltage up to 565v which seems a bit high to me.

So, in my case is it better to run at a high plate voltage to keep the max dissipation to 65%, or, run at a higher bias current giving a higher max dissipation (75%+) but a lower plate voltage? In either case I can't see it being possible to get the plate voltage much below 560v.

Any advice, comments will be much appreciated.

[mickeydg5]

The 2466 power supply has a higher voltage at the HT than most with a nominal of about 520VDC I would think.
The PT is matched for 230VAC wall supply though. That is why your voltages are so much higher. Your utility is giving 244 and 250VAC.

The only thing to do is either get a voltage regulator or a variac to keep the voltage between 230 and 240VAC.

As for biasing with voltages like that, 543VDC, I would set idle at 35mA (mV) per tube maximum. So that is 70mV measured on each side.

[Ukmarshall]

Thanks for the reply.

The mains supply voltage is giving me the problems. As you suggest, a voltage regulator is something I've tried in the past but have not had much success. The retail type of auto regulator tend to be too 'coarse' and have defined voltage steps which aren't particularly sensitive.

Biasing at 70mA a side gives a plate voltage of around 560v at an input voltage of 250v. So I'm still uncertain if this is a safe voltage or is it best to run a high bias current to reduce the plate voltage a little. I've read the Shuguang KT66 sound best run hot. The recommended settings I've read that Marshall use are plate voltage of 530v with a bias of 72ma-80ma which equates to 76% to 85% of max plate dissipation which is hot.

Does anyone know what the absolute maximum allowable plate voltage is for the Marshall KT 66?

[mickeydg5]

No sir. You calculate your bias number based on plated voltage. The bias does not give you 560VDC. The wall voltage will determine the plate voltage. Adjusting the bias will alter the plate voltage only slightly. If you have, meaning actually measured, 560VDC at the plate then set the idle bias current for about 34mV (mA) per tube or 68mV for a pair.

Like I had mentioned a normal 2466 situation at proper wall voltage may give 520, maybe 530VDC at the plate. So 530VDC on the plate and 80mA idle current would be about 73%. 72mA would be about 66%. You have to account for both plate and screen current when using the cathode resistor bias monitoring method. That is about 4 to 6mA for each screen.

That amplifier comes stock with little 3/4 watt 1 ohm bias monitor resistors. Those are begging to fail sooner or latter. A 100 watt amplifier like that should have at least 5 watt per side.

[Ukmarshall]

Thanks for your perseverance !

I'm missing something here obviously. I was basing my calculation on worst case, so with regards to screen current I had disregarded it. As the screen voltage is only a few volts less than the plate voltage I thought the current draw would be very low , so didn't take it into account.

With an input voltage of 250v (set on aVariac) I have set the bias to 68mA per side and measured the plate voltage at 562v and the screen at 559v. So do you think 562v is Ok?

If you don't mind, how did you calculate 530v plate at 80mA is 73% and at 72mA is 66%?

Thanks

[mickeydg5]

Well most of these Marshall's designs are based on 6CA7/EL34 type power tubes, 40 to 100 watt output with 100 watters being 4 tubes. Marshall runs all of their 5881/6L6GC, KT66, 6550 and KT88 amplifiers like this.
So it is easy to just use the EL34 plate wattage rating of 25 and then add some for the screen.
25 / 562 = .04448
.04448 * 70% = .03113
That is 31mA then add 4mA for screen and walla, 35mA maximum per tube. Cut it down 1mA for any expected wall voltage fluctuations.

25 / 530 = .04716
.04717 * 73% = .03443
about 34mA or 35mA + 5mA = 40mA; 80mA for two tubes

This is how most people relate to and calculate for these amplifiers. Its not an exact science but gives a working target number. Use your ears for the rest. If 32mA sounds as good as 35mA for both low and higher volumes, then use the lower number.

[mickeydg5]

Although not cheap, this or something like it will hold the supply at about 230VAC or 240VAC. (it says 120 on the site but it is actually 230/240 as per manual for this model)

http://www.furmansound.com/product.php? ... =P-1400ARE

There was also the discontinued AR-2306 model.

[Ukmarshall]

I think I see where my calculations differ. It was my understanding that if you use the valve wattage divided by the plate voltage times by 70% formula that this included the screen current as this calculates the total valve dissipation. In order to obtain the correct bias current for the plate on Pin 3 you would need to subtract the screen current!

In any case, it still seems I have a high plate voltage and I'm not sure if running the valves at this voltage is going to cause problems.

[mickeydg5]

Yes you would subtract for screen from my number. Look at any tube data sheet for a KT66 and there will be seen both a dissipation for plate and screen.

I would feel safer or better with a lower wall voltage.
BUT
Marshall designed it with these tubes and voltages in mind. Most electronics have to operate within a good bit of tolerance and your wall voltage is maybe up to 8% positive.