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 22 2014

Print this Post

GM-70 push pull musings ..

gm70 pp v 2.0

Has anybody out there tried a PUSH – PULL version of an amp, based on a pair of two GM-70 tubes yet ?

These here are some preliminary musings on a Push Pull GM-70 setup, which would provide the benefit of net standing DC current flowing through the respective halves of a PP primary adjustable to net ZERO.

I foresee Pentode Constant(*) Current Drains, and each is finely adjustable and trimmable via respective pots in each branch of the cathode of such pentode based CCD.

As long as I keep the currents flowing (and displaying on the respective meters) at the same value, the net DC current component flowing through the primary shall be zero.  No magnetizing of the core. The core may be of a slightly smaller size, with no air gap. Possibly higher inductance. Possibly better bass response.

These CCD’s with their high dynamic impedance also serve the purpose of “lifting” the voltage, or rather “decoupling” the cathodes of the GM-70, so that these cathodes can freely adjust with their voltages according to the programmed currents, and to the current momentaneous voltages of the grids of the GM-70 tubes.  The grids of the tubes are directly connected to the output of a SRPP driver stage, which is capable of both fairly high voltage swings, but also, what is more important, of draining significant grid current, in situations where we drive the grids of the GM-70 into positive voltage territory.

Please note that there are ultrapath capacitors between the HV B+ and the respective cathode of GM-70. This basically closes up the AC signal path around the GM-70, her respective half of the primary winding, and the Ultra-Path capacitor.

The CCD’s are actually not constant, as they are “modified” / “modulated” by a fraction of the output voltage from the respective half of the secondary winding, via appropriate resistive voltage divider, so as to create negative feedback loops around the GM-70′s, the CCD’s, and output transformer.

Alternatively, I was thinking about how to do a “true” current feedback path, binding the secondary windings respective half with the cathode current of the GM-70 to ground, but this seems to be in conflict and mutually exclusive with the concept of the Constant Current Drain. If you have any ideas on how to “marry” / join the CCD concept with the Negative Current Feedback concept – please, please, please DO let me know.

Another, wider NRB loop, also dual path symmetrical, goes back to the cathodes of the respective SRPP driver branches.

Please note that there is a galvanic connection and a Direct Drive of the GM70 grid from the output of the SRPP.  Some may say that this is a very risky business … What shall happen if the lower tube of the SRPP fails ?   The response is fairly simple: the power tube will melt, … or at least WOULD melt if it were not for the CCD, which keeps the cathode current in rein.

I hope that such a solution shall provide a reasonable degree of failsafe protection, and at the same time the benefit of direct galvanic coupling of driver stage and output tube.

The 1/2 height voltage of the SRPP shall be 1/2 of SRPP B+, i.e. 0.5 * 600V = 300V DC.

The cathode of the GM-70, driven by the CCD, shall by itself adjust its voltage so as to match the cathode current of the GM-70, as “programmed” by the CCD.

A “third” negative feedback loop, this time a “global” one, goes straight back to the cathode of the very  input tube.

An alternative input stage would basically also possible, for example one with a DC coupled LTP stage, accepting two independent paths of a global NFB.   This solution would call for the -250V DC power supply line as a nice drain for the LTP tail resistor.

Zasilanie 600 1680 poprawione

A basic version of the power supply and filters, providing the needed voltages of 1700 V and 600V.

Alas, After sleeping over this upper power supply module, I came up with the observation and “unease” that actually, during the positive sine wave, only the top toroid is actually conducting (LOOK AT THE DIRECTION OF THE DIODES !!).   During the negative going sine half-cycle, the same applies to the lower toroid. This can not be.  We need to “balance” them, so that each has something to do during each of the half cycles.  It is actually VERY important to think about which diode is directed in what direction, at which half cycle does it conduct and to what toroid is it connected to.

After some musings, I came up with this lower schematic, which exemplifies the workings of my thought process:  The green color represents a positive half sinewave cycle, or, simply speaking, the “magnetic field pumping current upwards”.   The red color represents the negative half sinewave, or the “magnetic field pumping current downwards”.

Direction of the arrow suggest from which toroid is the current sourced/drained from (upper toroid vs. lower toroid).

When you look at the paired arrows at the filter caps, you shall notice that they resemble a pair, which represents a full sinewave cycle.  The first of the pair represents the first part of the cycle, the second of the pair represents the second part of the cycle.

With the connections as depicted below, we now see that each toroid has now something to do within each of the half cycles.   No half cycle renders any toroid without a workload.

Please also take note of the GROUND SHIFT.

I actually reverted back to the idea of connecting the ground to the center taps of both transformers. This allows me to work with lower voltages (referenced to ground), but at the same time, to have the same high voltage available for the output tube (if you look at the difference between +600V and =600V – it still constitutes 1200V …).

{{ To keep a long story short: John, many kind thanks for the TIP ! }}

Odpowiednie fazy 2

This last option would potentially also provide -250V for powering a Long Tail Pair resistor, if we wish to change the input circuit to an input LTP with two completely symmetrical global NFB paths.

This one could have an issue with PS ripple, though.   Right hand side “alternative” tabs show voltages relative to a “shifted” ground. The place where we “attach” the ground pin to the power supply is somewhat arbitrary.  But there are certain PS ripple implications that go along with that.

As in the case of any inductor front loaded filter, these branches will obviously need to be loaded with some minimum working current so as to prevent an above average output voltage from those filters.

Permanent link to this article: http://hiend-audio.com/2014/08/22/gm-70-push-pull-musings/