I happen to have samples of both GM70 which I have never used;
but for what it is worth -
I was reading the other day in a back issue of Radiotronics Magazine - published by AWA from the 1930's until mid 60's - that Nickel Anodes performed better in SOME respects to Graphite. But in saying this - Graphite absorbed ( being Porous like a Sponge ) the electron refection off the screen grid and hence gave a cleaner space charge in the tube. This made triodes and certain Beam Tetrodes better in performance quality. Graphite, as I've always believed, handles higher plate dissipation / power ( an looks more rugged ); BUT, Nickel had electron reflections off its shiny inner surface that some how neutralized, or reduced audio distortion- despite producing more reflected space charge.
I'd say copper would perform similar to nickel.
So it's horses for courses !
Power is not everything - just as having very low distortion and a certain type of mellow tone is not everything either.
So for example - In a guitar amp, I would use the graphite anodes and maybe - the nickel in a Hifi.
Last edited:
I have and have used both. I prefer the graphite plate version in the context of my two stage IT coupled GM70 SE 20W design.
The output stage employs fixed bias, and I am typically running at 110 - 120mA at 1kV on the plate. Primary Z is 7K.
I know others who prefer the copper, but while I don't remember the exact conditions I do recall that both plate current and plate voltage (<800V) were much lower, and output power was a little over half.
I've also noticed over three sets of coppers that they generally exhibit corona discharge phenomena and possibly leakage across internal insulators to a much greater extent than the graphites.
The graphite version sounds excellent and under my conditions of use is a more consistent performer with far fewer anxiety inspiring quirks. (A problem when plate voltage is 1kV and above) I should note that I have had no tube failures with any GM70 since I built the amplifiers and put them into use in late 2011.
The output stage employs fixed bias, and I am typically running at 110 - 120mA at 1kV on the plate. Primary Z is 7K.
I know others who prefer the copper, but while I don't remember the exact conditions I do recall that both plate current and plate voltage (<800V) were much lower, and output power was a little over half.
I've also noticed over three sets of coppers that they generally exhibit corona discharge phenomena and possibly leakage across internal insulators to a much greater extent than the graphites.
The graphite version sounds excellent and under my conditions of use is a more consistent performer with far fewer anxiety inspiring quirks. (A problem when plate voltage is 1kV and above) I should note that I have had no tube failures with any GM70 since I built the amplifiers and put them into use in late 2011.
I have both types. My amp runs at 1050v and 90mA, two stages, RC coupled. In this amp, the graphite plates sound more dynamic with strong bass, while the copper plates sound more refined with subtler nuances but softer bass than the graphite.
I prefer the sound of the copper plates, but graphites are also very good.
No corona problems here, but reliability problems with the coppers: two were bad and had to return to the seller, 3 others developed internal gas and rapidly ate their getter flashes... No such problems with my graphites.
I prefer the sound of the copper plates, but graphites are also very good.
No corona problems here, but reliability problems with the coppers: two were bad and had to return to the seller, 3 others developed internal gas and rapidly ate their getter flashes... No such problems with my graphites.
I use Penta Labs new production graphite plate 833C in my Midlife Crisis amps; I'm very happy with the sound, and in 9 months of use so far no issues, in fact zero drift in plate current to date, running 2300V B+ at 160mA. Non-graphite 833s such as the 833A probably wouldn't survive well since plate dissipation would be over their recommended max.
I had the misfortune of receiving four fairly rare Souz graphite plate GM70 from Ukraine that were improperly packed, internal structure and filaments completely destroyed. The seller made good and replaced them with comparables of another vintage and manufacture. I was really looking forward to trying those.
Most of my graphites seem to have been made by Reflektor.
The coppers were all made by different makers, but I am not sure who made any of them.
Most of my graphites seem to have been made by Reflektor.
The coppers were all made by different makers, but I am not sure who made any of them.
Pardon my ignorance- is the 833 a similar tube to the GM70?
General question- do the graphite plates in the GM70 use zirconium?
No, the 833 is totally unrelated to the GM70..
How would one tell whether there was zirconium in the graphite plate GM70? I have never seen mention of it in a bilingual or English language data sheet. Is there any tell tale clue in the appearance of the graphite?
I can say that the quality level of that plate structure is quite good based on appearance. (Have not pulled one apart so far, but level of finish is high)
I run mine very near the max rating and replace them around 20% beyond the rated life time of 1000 hours. Don't want to find out what happens if that tungsten filament breaks and touches something it shouldn't. Over that time they are parametrically stable and only need a minor bias adjustment or two. They are still amazingly inexpensive, and pair of them in push pull is good for up to 500Wrms out. I run SE class A1 only and get about 20W, A2 would up the ante to between 40 and 50W..
How would one tell whether there was zirconium in the graphite plate GM70? I have never seen mention of it in a bilingual or English language data sheet. Is there any tell tale clue in the appearance of the graphite?
I can say that the quality level of that plate structure is quite good based on appearance. (Have not pulled one apart so far, but level of finish is high)
I run mine very near the max rating and replace them around 20% beyond the rated life time of 1000 hours. Don't want to find out what happens if that tungsten filament breaks and touches something it shouldn't. Over that time they are parametrically stable and only need a minor bias adjustment or two. They are still amazingly inexpensive, and pair of them in push pull is good for up to 500Wrms out. I run SE class A1 only and get about 20W, A2 would up the ante to between 40 and 50W..
None at all, I apply all voltages simultaneously - never had any problems doing so. I can confirm the getter is traditional, I had forgotten that zirconium is used for gettering in some graphite plate tubes as I haven't designed anything that uses them. Also they do not run with visible color on the plate even at well beyond rated dissipation.
I choose it because it is fairly easy to drive (mu around 6.7) and quite linear. I had originally planned on using something else but supplies of that tube were unobtainium at the time.
These are triodes..
http://www.mif.pg.gda.pl/homepages/frank/sheets/018/g/GM70.pdf
http://www.mif.pg.gda.pl/homepages/frank/sheets/084/g/GM70.pdf (Note there is an error in the pinning shown on this data sheet)
http://www.jacmusic.com/nos/GM70/index.htm (Partially disassembled copper plate version can be seen here)
http://www.jacmusic.com/nos/GM70/index.htm
http://www.mif.pg.gda.pl/homepages/frank/sheets/018/g/GM70.pdf
http://www.mif.pg.gda.pl/homepages/frank/sheets/084/g/GM70.pdf (Note there is an error in the pinning shown on this data sheet)
http://www.jacmusic.com/nos/GM70/index.htm (Partially disassembled copper plate version can be seen here)
http://www.jacmusic.com/nos/GM70/index.htm
Tried some 300B mesh and carbon, and had similar results.
But apart from these I not know, just tried these two in exact same situation.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.