Sovtek Heaters

[Wolfie26]

I wanted to run my 6SN7GT tubes off of constant current sources or constant voltage sources. So I built 2 circuits using LM317's to see what happens.
The circuits are for running just one tube by it's self.

Constant Current:
Sovtek-1 6SN7GT 5.552vdc at 600ma. (3.33w)
Sovtek-2 6SN7GT 5.573vdc at 600ma. (3.34w)
Sovtek-3 6SN7GT 5.632vdc at 600ma. (3.37w)
Raytheon 6SN7GT 6.344vdc at 600ma. (3.8w)

Constant Voltage:
Sovtek-1 6SNTGT 6.306vdc at 650ma (4.09w)
Sovtek-2 6SN7GT 6.305vdc at 650ma. (4.09w)
Sovtek-3 6SN7GT 6.304vdc at 640ma. (4.03w)
Raytheon 6SN7GT 6.300vdc at 600ma. (3.78w)

Sovtek-1 was purchased in the 90's the other 2 in the last couple of weeks.
Sovtek specs for the heaters: 6.9v max to 5.7v min with the current being 600ma +60ma or -50ma.

So I am undervolting using a constant current source on the Sovtek tubes.
It seems to me that these Sovtek tubes need to be run in constant voltage only or you will under power them.

Sovtek data sheets always reference to 6.3 so I'm thinking constant voltage without trimming to 3.78 watts is the way to go..

Any suggestions on what to do in a situation like this..?

 

[analog_sa]

Always though the rule was simple: adjust your current source until the voltage across the heater gets to nominal. Don't rely on current figures taken out of datasheets, it's the voltage that matters.

 

[50AE]

I also suspect heater resistance differences can be a highly suspected culprit for tube mismatching. Had better selection with some Soviet 6P45S after equalizing the heater VA products.

 

[leadbelly]

It's not a "situation". Either regulating current or voltage is fine. That level of deviation is normal in tubes.

 

[HollowState]

The proper cathode operating temperature of all indirectly heated receiving tubes is determined by the filament voltage. Not by the current drawn. Current specifications are needed for connecting like values in series or assuring adequate power when paralleling. This also applies to directly heated cathodes with very few exceptions. Those being a tiny handful of large industrial tubes. And unless you live in a third-world area where incoming power service wildly fluctuates, filament regulation of 6SN7 types is futile and unnecessary.

 

[audiowize]

Hollow state is correct, regulate the voltage if anything, not the current. Some regulators will allow for current to be limited, so you could limit the inrush current to about twice the quiescent current, but not much beyond that.

 

[wiseoldtech]

Self regulation by using a competant transformer with sufficient rating is all you need.
Don't go nuts trying to figure into things that are really meaningless.
Tubes, by design, are forgiving creatures, and a bit of voltage higher or lower isn't worth fretting about.
For instance, the "design center" for a 6xxx tube's filament is 6.3 volts, but anything from 6.0 to 6.7 isn't going to hurt things.

 

[nigelwright7557]

i once accidentally connect 12ax7 to 30 volts dc for a few seconds.
it got a bit bright but survived.
i tend to just use a 6v3ac transformer.
the transformer also acts as a current limiter on power up.

 

[mondogenerator]

I did the same to 1Zh29b, used a 2x1.5V cell with the filaments in parallel instead of series connection.

It survived, pretty much unscathed

 

[Kay Pirinha]

6SN7GT's were basically designed for TV service. The heaters explicitly were allowed to be fed either in parallel with nominally 6.3 V AC or DC or in a series string with 0.6 A nominally. There was even the 6SN7GTA version with controlled warm up time.
What is sold as 6SN7GT's nowadays are re-labeled Russian 6P8S's, whose data sheet doesn't even mention TV use and/or series heater feeding. Hence, I'd also prefer to apply constant heater voltage on them.
Best regards!

 

[Wolfie26]

Thanks everyone for all the feedback. I was surprised at the horrible performance of the constant current driver with the Sovtek tubes. Worked fine with the American NOS tube. but under powers the poor Sovtek's. Notice how similar the Sovtek tubes are in watts under constant voltage. Most transformers are designed for 115 or 117vac so they tend to put out more then 6.3vac hence my trying out some 317's as current and voltage sources. I'm going with the voltage sources.

 

[TG]

What is sold as 6SN7GT's nowadays are re-labeled Russian 6P8S's

6N8S rather
6P8S is an extremely old and rare output pentode.

 

[petertub]

6SN7GT's were basically designed for TV service. The heaters explicitly were allowed to be fed either in parallel with nominally 6.3 V AC or DC or in a series string with 0.6 A nominally. There was even the 6SN7GTA version with controlled warm up time.
What is sold as 6SN7GT's nowadays are re-labeled Russian 6P8S's, whose data sheet doesn't even mention TV use and/or series heater feeding. Hence, I'd also prefer to apply constant heater voltage on them.
Best regards!

Maybe russian 6SN7GT are relabeled 6P8S. JJ sells the real stuff.

 

[kodabmx]

I run 6N8S two at a time from 12V in series. In fact, I run all my 6.3V tubes in series from ~12VDC. The only exception being when there is only one tube. Then I use a buck converter for it from 12V.

My monoblocs power the tube heaters in series from 11.6VDC since it was easier to use 18VCT and full wave than 24VCT and a regulator

 

[Kay Pirinha]

6N8S rather
6P8S is an extremely old and rare output pentode.

Of course ! Pardon me for my typo, please.

Best regards!

 

[JMFahey]

I was surprised at the horrible performance of the constant current driver with the Sovtek tubes. Worked fine with the American NOS tube. but under powers the poor Sovtek's.

Your fault, not theirs.
Why apply constant current when standard safe to operate any filament is constant voltage?

The reason being that filament metal has POSITIVE resistance behaviour, so it will self regulate on voltage input and vary wildly with current source.

The idea behind series filments was only to allow for *cheap*, emphasis on cheap, large count tube devices such as TVs to *hopefully* reach 117V or thereabouts and save on a hefty winding and some iron weight, a compromise in the best case.

That said, even your improperly fed Russian tubes WILL work.

Don´t blame them.

 

[Wolfie26]

The Sovtek tubes are marked 6SN7GT / 6H8C. The 6H8C is printed in Gold along with 9007. They seem to work really well. I have one swinging 150 volt sine waves cleanly up past 20khz as a phase splitter.

Because of their terrible tolerance on the heater (+60ma to -50ma.) I now know not to run them constant current

 

[gerrittube]

Hi Wolfie26,

That’s a lot of swing! Do your use the 6sn7 as a Long Tailed Pair or as Concertina? I would like to see your diagram.

Regards, Gerrit

 

[Wolfie26]

Oops, that's 50vpp or 100 volt sine wave. I copied a Glass Audio design for a pp 6550 amp by a Japanese designer. It was a concertina type. Was able to push it higher. I did modify the design a bit. Will go have a look.

 

[kodabmx]

Mine will swing that kind of voltage, but with the help of a 6F12P as seen here:

If you build this, make R2 2k. Also, less power but less distortion can be had by switching the OPT for VPT12-whatever.

EDIT: And in fairness, here is my PSU for it so anyone who like to can build it.