Hiend Audio, DIY, Hi-Fi, Stereo, Electronics site, for lovers of high-fidelity music reproduction. High-End Vacuum Tubes. Silicon. Do-It-Yourself Audio Systems. Schematics: amplifiers, speakers, horns, CD, DAC. Circuits, Topologies, Acoustics, Cables, Speakers, Voltage Regulators, Equipment Upgrades, Modifications, Problems, Solutions, Tips, Tricks. ------ Dla pasjonatów Hiend Audio, najwyższej wierności odtwarzania muzyki na sprzęcie budowanym, modyfikowanym samemu. Wiedza: elektronika, układy lampowe i krzemowe, wzmacniacze, głośniki, kolumny, tuby, DAC, kable, DIY, porady, sztuczki, problemy, rozwiązania, tor audio.



Aug 13 2014

Print this Post

GM-70 dilemmas …

So. I have finally done it.

I have purchased from a fellow DIY fan a “complete set” of building blocks, which could in essence constitute a GM-70 based pair of monoblock amplifiers.

The set contains various transformers, chokes, interstages, capacitors, but what is more important, a set of GM-70 and 300B tubes. So, now comes the great dilemma: what should I build out of it ?

This is more or less how it looks like (one channel, totalling 30 kg):

GM stuff 1 GM stuff 2

Ok. So the challenge is as follows: The “originally intended” schematic, as below, is a DRD type of setup.  This is the schematic that has necessitated the components, as in the Bill Of Materials, as presented on the photos above. That is supposed to be built out of the components:


Yes, indeed, it is a workable design, but I do not like the sheer amount of waste, in terms of energy, that is happening over here. Just check out that set of high-power resistors in the cathodes of the power tube (GM-70), and those at the cathode of the driver tube (300B). Check out the need to “fiddle” with setting the correct value of auto-bias for the power tube and for the driver tube, by means of mundane trial and error, each time switching off, using the screw-driver to loosen the wiper screw, readjusting, firming the screw again, then switching on … A sheer nightmare !

Also check out how the 700V DC for the driver tube is supposed to be generated. It is supposed to be a  cut down version of the 1200 VDC, by means of a … a WHAT ?  Yes. You said it. By means of a Resistor Voltage divider.  Wow !  That is one hell of a voltage divider. Two 50 Watt 4k resistors, totalling with 100 Watts of resistor divider power (down the drain).

The problem with this given set of components is to come up with a “method” of how to alternatively use these components, and specifically (one monoblock set listed here):

1). A power supply toroidal transformer 1200VAC – 0VAC – 1200VAC,  800 VA.   If we rectify this and filter via a CLCLC, it would total up to something like 1680 V DC.  Too much for my tastes. Most probably, an inductor front loaded filter would be more down to Earth for this project ( LCLC )

2). Output Transformer: Ogonowski LO SE80-2:  Primary Impedance: 6k; Secondary impedances: 4R and 8R (both available). R_DC = 70 ohm; P_max = 80W,  I_max = 0,3A.
If we were to connect 8 ohm speakers to the 4 ohm taps, this would significantly increase the primary impedance of the transformer, up to 12k. This is most fortunate that the 4 ohm tap is available, as a much higher primary impedance shall be of benefit here, and indeed very much needed.

3). Anode Choke: L = 75H, R-DC = 1kohm; I_max = 20 mA.
4). Interstage: Ogonowski TM-1:  Primary: 3kohm / 3,64H;  Secondary: 1kohm  / 1,165H
5). Anode filter choke:  10H / 72R / 0,5 A
6). Some microwave caps: 1uF / 2100VDC and 0,92uF / 2100VDC
7). GM 70 Vacuum tube
8).  300B vacuum tube
9). steel chassis top plate
10). Steel caps / covers for the main transformers and chokes.

Ok. So that is basically it. One Monoblock listed. Two sets of the above are available. Any ideas how to use the stuff ?

Just to put this in perspective, here are some images of know GM-70 designs, ones that are already out there, operational, and providing both music as well as ambient illumination for their proud owners ….

GM ready 2

GM ready 1


Here are some preliminary ideas that I came up with, but this is just in the works, subject to change …

GM70 v.2

This is a rough approximation of the concept:

a). Incandescent (i.e. low-tech) Light bulbs (230V/60W) as cathode resistors for the power tubes. YES ! Ordinary, rude and crude, light bulbs.  Power PTC devices. Positive Temperature Coefficient devices. With some thermal inertia, at that. How come nobody uses these ordinary light bulbs in hiend-audio power amps ? Just think about it.  If you are concerned about the filament’s impedance jumping around as a function of temperature as a function of the bass guitar within a rythmic rock-and-roll piece of music, then OK, just to be on the safe side, BYPASS the light bulbs with an electrolytic capacitor and a foil capacitor. Done.

b). The Alan Kimmel, CAMUS Style “Jumping Jack Flash”,  jumping iron cores of both the interstage transformer, as well as the output transformer.  Check out the capacitors providing AC coupling of the CORE of the transformer with the anode and it’s AC voltage swing.  Moreover, check out the resistors (10 Megohms) connecting the jumping core with the ground, providing a DC reference, and yet “insulating” the iron core from the ground.   The Core is hence free to “jump around” together with the music. This essentially renders any parasitic capacitances between the primary winding and the core as harmless. But:  BEWARE: do not TOUCH. The chassis of the Transformer Core is isolated from ground and hence possesses deadly voltage swings on it. Do not touch, or else thou shalt be jumping in electrocution shock, in sync with the bass guitar.

c). A DRD type of choke connection, providing DC anode voltage to the anode of the input stage tube.

d). Jack Elliano’s “Ultrapath” style capacitors, closing the AC signal loops in both of the output stage, as well as in the driver stage.  Each such loop consists of just three elements: The Tube, the transformer, and the Capacitor.   This way, we kind of like “eliminated the capacitors of the Power Supply Filter from the “signal path” – as now the signal has a much “shorter” route to travel upon.

B.T.W.:  you can read about the ULTRAPATH concept HERE:




Here are some basic specs and anode characteristics of the GM-70 tube …



GM-70 Anode GM-70 Specs

A nice point to start off with would be, say 1200 VDC at 100 mA quiescent current, at -120V bias (give or take).

Assuming that we use the 4 ohms tap to drive an 8 ohms speaker, we essentially “have” a 12K primary winding. That would place us with a maximum swing from say -20V bias to -220V.

At  -20V bias, we would be somewhere in the region of: 100mA + 700V/12k = 100mA + 58 mA = 158mA, at an anode voltage of 500V.

At -220V bias, we would be somewhere in the region of: 100mA – 700V/12k = 42mA, at an anode voltage of c.a. 1850V.   (Actually, there is a nonlinearity involved; from the graph it would rather seem that the current is this point, as extrapolated by the 12k impedance line, would rather be in the whereabouts of 30mA).   These are obviously just rough approximations as taken from a haphazard graph.  Fine tuning will be necessary anyway.

So, assuming a choke loaded rectifier and LCLC filter, the 1200 V DC standing quiescent voltage seems to be “workable”.

But, obviously, for the driver tubes, I would need a much much lower anode voltage than that.

But how do I “get” a lower voltage, to power the driver stage, to start with, … withOUT those crazy 2 x 50 Watt resistors to hammer down the 1200V down to 700V DC ??

GM70 v.2 zasilacz

Here is a concept that I came up with.

Actually, the concept “sucks”, because it implies a need to place the the power supplies for both monoblocks into a joint (“integrated”) housing of a joint (both channels) power supply unit.

So that would imply that I shall be stuck with THREE boxes:  The integrated power supply box, the LEFT monoblock and the RIGHT monoblock.   What is more freightening still, is that from the PS box, I shall need to route DC power supply cables to the monoblocks. One of those routes shall be 1200V DC.  Errr … How do I route 1200 VDC across the carpet of my living room ?

OK, I can live with that. So the idea now boils down to the following: How to “reduce” the voltages of the power supply transformers and “create” a lower voltage, such as the 600V DC that would be required for the driver stage?  Check out the drawing above. Check out that the two toroidal transformers, from the primary side, are actually connected in series … So each of them receives just one HALF of the mains voltage. Hopefully, having a more or less (?) equal split of the mains voltage between them.

But fret not, the secondaries are fully symmetrically loaded, so that a “self-leveling” act takes place on the primary side. Each of the primaries has upon it exactly one half of the mains voltage. If it were not to be so, then the one with the “more than half” voltage would consecutively produce a bit higher secondary voltage. This in turn, this “higher” secondary voltage, would cause the rectifier path associated therewith to work “harder”, to effectively push “more” current, and hence this secondary would be loaded “harder”. But this means that a more heavily loaded secondary will pull down the impedance on the primary side of that half. A self balancing act indeed.

OK, but this is all but a vague concept.  My request for your kind inputs and feedback here.

Any and all corrections / ideas or concepts, especially from folks who actually have successfully completed such a similar beast, would be kindly appreciated.




Permanent link to this article: http://hiend-audio.com/2014/08/13/gm-70-dilemmas/