Marshall 1962 Bluesbreaker: new valves & biasing - is everything ok?

[Aran]

Hi folks!

this is a question about my current situation, not BB biasing in general

I've just got a 2007 BB Ri, the PCB version. It is fully stock and came with 5881 power valves. I decided to go for KT66s (NOS Gold Lions) as the stock 5881s was to damn smooth
The plate voltage (on pin 3, with KT66s mounted) reads 385V. Kinda small?
I set the KTs at 60%, so 25 / 385 x 0.60 = 0.03896, just 0.039. Kinda high?

So, is it correct? I found the plate voltage with KTs kinda low (it was 407V with 5881s), so the bias is higher. I can't go lower that 37mV with onboard bias pot.

My main concern is both KTs become physically very hot after like 10 min of working - I didn't check the temperature but I can feel they produce more heat than 5881s pair. Additionally, the plate in both KTs become slightly red. Well, it's nothing like typical, glowing "red plate", but if I watch carefully, the plate become a bit orange/red (didn't saw it with 5881s).

Is it like 66s do that or something's wrong?

Opinions?

 

[Jon Snell]

This amplifier is designed for 6L6GC, NOT KT66s.
The bias, HT Anode voltage is totally wrong and the KT66s wil fail very quickly, if they are in fact true KT66s.

 

[Aran]

Ok, but putting KT66 valves into Bluesbreakers is kinda popular mod and I didn't heard off bigger problems than bulbs size matter.

 

[Pete Farrington]

For a given set of anode and screen grid voltages and anode current, KT66 typically require a greater magnitude of negative bias voltage applying to their control grid than other 6L6 family valve types.
To facilitate this, R28 should probably be reduced a bit, eg to 180k or 150k.

https://el34world.com/charts/Schematics/files/Marshall/Marshall_bluesbreaker_reissue_45w_1962.pdf

 

[Aran]

Thank you! I'll swap the resistor and we will see the result.

Some news I think I should mention. I've just put in stock 5881 that was with this amp to check how they were biased (I didn't do it yesterday). The biasing pot was on minimum. After 10 min of warming up I checked the plate voltage (400V) and then the bias (38mV). Huh? 1mV more than with KT66s I put the biasing pot where it was "originally", around 3/4 of rotation. The plate voltage increased to 407V (exactly as it was yesterday, before valves swap and rebiasing) which caused both valves to reach 42mV. Well, the BB was biased kinda hot Around 75% if bulbs are 23W max dissipation, right?

This is my first Bluesbreaker Is it normal?

I'll check the voltage with KT66s pushed to 42mV.

 

[Pete Farrington]

When checking voltages in an amp, always check the mains voltage being fed to it.
It doesn’t make sense that a change in bias increased anode voltage and current.

 

[neikeel]

I have to partly disagree with Jon.
The only issue with KT66s in these is their physical size (and yes Gold Lion KT66 are true KT66).
They are quite capable of coping with 500v on the anode if correctly biased under the 70% idle maximum.
The Dagnall D2505 and Drake 784-389 OT are both in the 8k primary impedance range similar to the original Drake -103 so that is not an issue
The chief concern (and where I might agree with Jon) would be the extra heater current required from the PT. I have not heard of any failures in that respect, but am prepared to be furnished with multiple examples if that is indeed the case. The rectified output of the RI is a bit lower than the original Drake -55 (450vdc) and an anode voltage of 400v is typical when set up correctly in correct range.

 

[Aran]

Thank you guys for all responds. I did a small testing today with different biasing, checking pin 3 voltage as well.

37mV = 385V
42mV = 399V
43mV = 400V
44mV = 402V

Every reading is around 70%, considering the valves are really 25V max dissipation and this proportion "naturally" stays all the time when spinning the pot.

I can't go lower than 37mV with onboard biasing pot. I didn't go further than 44mV. Having 37mV bias both valves are red plating. It wasn't that super-duper heavy glow of anode, but yes, it was a bit red/orange. The "higher" the bias I set, the lesser valves are red plating. With 44mV it was barely visible but still a bit of slightly red point if you really look hard.

 

[neikeel]

Those figures are the converse of what I would expect - the higher the bias current the hotter the valve runs and the measured anode voltage typically drops.

 

[Aran]

The OT is Dagnall TXOP 00003.

So, what should I do to run KTs safely in my BB? As I know, even slight red plating isn't good. I fired up the amp with 44mV bias. The sound was "proper". I mean no oversaturation, oscillations etc. It sounded nice, totally different than with 5881s, more old

There is no amp tech around where I live to do the maintenance so I must do all the things myself which in fact is kinda cool, challenging, but cool I also would like to learn as much as possible in case of valve amp repairing/modding etc,that's why this thread and all my questions I think it's natural progress from pedals modding/building

 

[Browneyesound]

Those figures are the converse of what I would expect - the higher the bias current the hotter the valve runs and the measured anode voltage typically drops.

Yes, something doesn’t seem right. If you are observing higher anode voltage at higher tube current, and you see redplating when the current is lower, something is amiss.

I would double check everything connected to the tubes and the bias supply.

 

[neikeel]

So with 44mA bias current (bias probe or resistor between pin 8 and ground?)
What is your pin 3 voltage ?
If it is 400v or lower then you are good to go.

 

[Pete Farrington]

There seems to be a parallel thread at TGP.
The mV figures seem to have actually been misunderstood, ie rather than representing 37mA cathode current, 37mV was -37V at the output of the bias supply.

 

[neikeel]

Ah that makes more sense.
Hence me asking what he was measuring and where.