Hi everybody.
I joined your forum a couple of days ago, following a Member's advice.
Very shortly, I will send some questions on a Valve amplifier using the GM-70s.
I have successful experiences with amps using 6C33s, 300B, EL34 and 2A3, but never dared to start with high voltage tubes.
Now it's the time. So I will ask few questions, hoping that some experienced builders will give me some useful advices of any kind, to make sure I do not repeat the mistakes they made at the beginning.
Thanks and talk to you very soon.
Ari Polisois
High Voltage GM-70 Amp
#4 High Voltage GM-70 Amp
Thanks.
The present guidelines (subject to change according to advices received), are :-
0. Power stage = two GM-70 valves in parallel (see end of post).
1. Anode Voltage = approx. 1 kV
2. Bias = - 97 V
3. Corresponding anode idle current = 100 mA
4. Internal resistance ( calculated ) = approx. 1700 ohms
5. Anode load = 7k
6. Driver stage valve = 6SN7GT ( two sections in parallel )
7. Power supply dc voltage for above = 420 V
8. Approx. plate resistance = 7700/2
9. Idle anode current = 8 mA
10. Bias of 6SN7GT = adjustable
11. Load resistor = 6k ( two 12k resistors in parallel)
12. Expected gain = 8 to 10x
13. Voltage amplifier = 1/2 ECC82
14. B+ = 420 V
15. Approx plate resistance = 7700 ohms
16. Idle anode current = 2 mA
17. Bias = variable ( average 0,9 V ).
18. Load resistor = 180k
19. Expected gain = 15-17x
Note : coupling between voltage amplifier and driver, as well as between driver and power tubes is direct (no caps ).
All the figures quoted are approximate and will be checked after building the amplifier.
I will draw a schematic as soon as possible and post it.
Regarding the power stage, the two GM-70 (per channel) will be paralleled undirectly as per figure attached, via a novel kind of output transformer ( designed as 4x4 SC-SCC ), that provides also for the self compensation of the magnetic flux in the core. Moreover, the GM-70s need not to be matched. A phase inverter is necessary for a proper operation of the OPT, otherwise the bass range will be reduced.
BR
Ari Polisois
The present guidelines (subject to change according to advices received), are :-
0. Power stage = two GM-70 valves in parallel (see end of post).
1. Anode Voltage = approx. 1 kV
2. Bias = - 97 V
3. Corresponding anode idle current = 100 mA
4. Internal resistance ( calculated ) = approx. 1700 ohms
5. Anode load = 7k
6. Driver stage valve = 6SN7GT ( two sections in parallel )
7. Power supply dc voltage for above = 420 V
8. Approx. plate resistance = 7700/2
9. Idle anode current = 8 mA
10. Bias of 6SN7GT = adjustable
11. Load resistor = 6k ( two 12k resistors in parallel)
12. Expected gain = 8 to 10x
13. Voltage amplifier = 1/2 ECC82
14. B+ = 420 V
15. Approx plate resistance = 7700 ohms
16. Idle anode current = 2 mA
17. Bias = variable ( average 0,9 V ).
18. Load resistor = 180k
19. Expected gain = 15-17x
Note : coupling between voltage amplifier and driver, as well as between driver and power tubes is direct (no caps ).
All the figures quoted are approximate and will be checked after building the amplifier.
I will draw a schematic as soon as possible and post it.
Regarding the power stage, the two GM-70 (per channel) will be paralleled undirectly as per figure attached, via a novel kind of output transformer ( designed as 4x4 SC-SCC ), that provides also for the self compensation of the magnetic flux in the core. Moreover, the GM-70s need not to be matched. A phase inverter is necessary for a proper operation of the OPT, otherwise the bass range will be reduced.
BR
Ari Polisois
- Attachments
-
- 4x4 - when the dc offset takes place.pdf
- (22.68 KiB) Downloaded 347 times
#5 High voltage GM-70 amp
Sorry, one file was missing.
Here it is.
Ari Polisois
Here it is.
Ari Polisois
- Attachments
-
- 4x4 OPT_different connections.pdf
- (21.86 KiB) Downloaded 252 times
#6
Personally I would look for something with a lower Ra and higher current than a single 6SN7 to drive a GM70, I found that they work with that sort of driver, but sound very slew rate limited.
I am trying to understand what you are suggesting with the transformer. Are you using one transformer to drive both the left and right outputs (R1+R2, L1+L2). Assuming you are canceling DC in the core by using two primary windings out of phase, how is the signal voltage induced? You mention a phase splitter to enable bass output, does that mean you are making use of leakage inductance to produce HF output? If you need a phase splitter, why not just use push pull? What advantage does this scheme provide over push pull (not that I understand how it works yet)?
I am trying to understand what you are suggesting with the transformer. Are you using one transformer to drive both the left and right outputs (R1+R2, L1+L2). Assuming you are canceling DC in the core by using two primary windings out of phase, how is the signal voltage induced? You mention a phase splitter to enable bass output, does that mean you are making use of leakage inductance to produce HF output? If you need a phase splitter, why not just use push pull? What advantage does this scheme provide over push pull (not that I understand how it works yet)?
Whenever an honest man discovers that he's mistaken, he will either cease to be mistaken or he will cease to be honest.
#7 High voltage GM-70 amp
Dear Nick,
What valve would you suggest to replace the 6SN7GT ?
The two sections in parallel will have about 3k9 Ri and, combined with the load ( 6k ), would give quite a small value facing the GM-70 grid.
I could not calculate the Miller effect, not knowing the anode/grid interelectrode capacitance of the GM-70, but I feel that the low values above should not restrict excessively the HF range.
If you have in mind another valve to improve the driver performance, it would be useful if the pins would concern the same electrodes, so that I could find, by trial and error, the differences in behaviour.
Regarding the Steve Bench's matrix, I could not get the drawing and description by clicking on the page you mentioned. If possible, kindly post it to this thread.
As for the OPT, you need two of them, one per channel and, as mentioned, you also need two valves per channel.
The best feature of the layout you see in the pictures is the self compensation ( one of them explains when and how it takes place).
In two words, the anode current of the first valve crosses the upper primary of the left leg of the magnetic circuit and then the lower primary of the right leg of the circuit, thus creating two fields. The same occurs with the seconda valve, with the other primaries. The DC fields are opposed, causing the cancellation of the flux generated by the idle current.
The AC fields, having opposed phases, due to the splitter or inverter, will not fight each other and the swing will develop freely.
As regards the connections of the secondaries, many options are possible, as you can imagine.
The calculations show that, even with different idle currents of V1 and V2, the offset is achieved, with the proper connections of the windings. Therefore, no matching of the valves is necessary.
This is different from a push pull transformer, where you have to provide for a separate balancing device of the idle currents. The 4x4 takes care of that automatically.
The boosting effect of the bass output takes place as in a push-pull transformer, because the push and pull windings do not interfere with each other. The half primaries of each valve are in series and the inductance is four times that of a single prmary (supposing the number of turns is equal, all over the transformer)
Practically, the size of a 4x4 is not far from that of a classic PP OPT, but the advantage of self compensation is a plus.
Of course, more practical and reliable details can be obtained when the whole amplifier, with this kind of transformer, will be completed and tested. Luckily, I can rely to several friends, some of which I consider "golden ears" , to express an objective opinion regarding the sound. But, in any case, I do not see why it should be worse than that with a standard OPT.
What valve would you suggest to replace the 6SN7GT ?
The two sections in parallel will have about 3k9 Ri and, combined with the load ( 6k ), would give quite a small value facing the GM-70 grid.
I could not calculate the Miller effect, not knowing the anode/grid interelectrode capacitance of the GM-70, but I feel that the low values above should not restrict excessively the HF range.
If you have in mind another valve to improve the driver performance, it would be useful if the pins would concern the same electrodes, so that I could find, by trial and error, the differences in behaviour.
Regarding the Steve Bench's matrix, I could not get the drawing and description by clicking on the page you mentioned. If possible, kindly post it to this thread.
As for the OPT, you need two of them, one per channel and, as mentioned, you also need two valves per channel.
The best feature of the layout you see in the pictures is the self compensation ( one of them explains when and how it takes place).
In two words, the anode current of the first valve crosses the upper primary of the left leg of the magnetic circuit and then the lower primary of the right leg of the circuit, thus creating two fields. The same occurs with the seconda valve, with the other primaries. The DC fields are opposed, causing the cancellation of the flux generated by the idle current.
The AC fields, having opposed phases, due to the splitter or inverter, will not fight each other and the swing will develop freely.
As regards the connections of the secondaries, many options are possible, as you can imagine.
The calculations show that, even with different idle currents of V1 and V2, the offset is achieved, with the proper connections of the windings. Therefore, no matching of the valves is necessary.
This is different from a push pull transformer, where you have to provide for a separate balancing device of the idle currents. The 4x4 takes care of that automatically.
The boosting effect of the bass output takes place as in a push-pull transformer, because the push and pull windings do not interfere with each other. The half primaries of each valve are in series and the inductance is four times that of a single prmary (supposing the number of turns is equal, all over the transformer)
Practically, the size of a 4x4 is not far from that of a classic PP OPT, but the advantage of self compensation is a plus.
Of course, more practical and reliable details can be obtained when the whole amplifier, with this kind of transformer, will be completed and tested. Luckily, I can rely to several friends, some of which I consider "golden ears" , to express an objective opinion regarding the sound. But, in any case, I do not see why it should be worse than that with a standard OPT.
#8
You can get more numbers here
http://www.tubes.ru/techinfo/ModulatorTubes/gm-70.html
The issues with driver stage are more from what I have heard using different driver stages, I find a 300b or 2a3 makes a good driver. el84 or el34 should work well as well. you need to consider that grid current starts before you get to 0v, so its helpfull having a low impedance driver.
I guess the simple answer is try it and use what sounds best to your ears. I know for myself a 6sn7 on its own doesn't do that good a job of driveing a 300b, the gm70 is a much harder thing to drive.
If you just search for "steve bench matrix amp" you will find a link to Steves pages, and there is a link on there to the matrix.
I won't clame to understand the transformer idea, but that may just be a problem in my understanding, but if you have a phase splitter and DC cancalation in the output transformer, it does sound very similar to a push pull output satage. Anyway, I look forward to seeing the final circuit.
http://www.tubes.ru/techinfo/ModulatorTubes/gm-70.html
The issues with driver stage are more from what I have heard using different driver stages, I find a 300b or 2a3 makes a good driver. el84 or el34 should work well as well. you need to consider that grid current starts before you get to 0v, so its helpfull having a low impedance driver.
I guess the simple answer is try it and use what sounds best to your ears. I know for myself a 6sn7 on its own doesn't do that good a job of driveing a 300b, the gm70 is a much harder thing to drive.
If you just search for "steve bench matrix amp" you will find a link to Steves pages, and there is a link on there to the matrix.
I won't clame to understand the transformer idea, but that may just be a problem in my understanding, but if you have a phase splitter and DC cancalation in the output transformer, it does sound very similar to a push pull output satage. Anyway, I look forward to seeing the final circuit.
Whenever an honest man discovers that he's mistaken, he will either cease to be mistaken or he will cease to be honest.
#9 Hi Voltage GM-70 amp.
Daer Nick,
Thanks for your suggestions.
I found some more information on the GM-70. The input capacitance (5-11) looks rather high, but not excessively. We'll see in practice, as you say.
Regarding Steve Bench matrix amp, I had a look at it but could not fully understand how it works. Anyway, I could see that, through quite a complex circuit layout and several additional components, Steve could get rid of the unwanted cross-talk between the two channels, building up in the same iron core. Steve also included manual adjustments that, IMHO, look critical.
A simpler ( but not as drastic ) result, has been obtained by Eng. G. Mariani (of GRAAF, Italy) and myself with the Split Core - Stereo Common Circuit transformer, in which you find two primaries (one for the left and one for the right channel) on the same core, with their closely coupled secondaries. These two primaries are physically apart ( unlike most of the push-pull transformers' half primaries) and, moreover, a "flux escape" is inserted inside every primary bobbin (consisting in 5-6 turns of magnetic band, 0.35 mm. thick), to divert some of the magnetic lines, thus reducing the coupling ratio between the primaries of the left and right channels. The crosstalk is limited to the lower frequencies and its impact on the stereophonic effect is acceptable.
However, in the case of the 4x4, I mentioned that I would use one output transformer per channel. Therefore, no cross-talk can possibly take place in the OPTs.
One side that I will have to check (and this can only be done with the amplifier built) is whether the two "paralleled" GM-70s, in each channel, disturb each other, causing oscillation or intermodulation. Also, needless to add, the quality of the sound.
I will try to get at this point as quickly as possible, but, to be realistic, it will take at least another couple of months, considering the family and work commitments.
In the meantime, I will be grateful for more advices on the matter.
I would like to add that I would not be against trying also the simpler SC-SCC transformer (described above), that I have used successfully with some other valves, in case of major problems with the 4X4.
Ari Polisois
Thanks for your suggestions.
I found some more information on the GM-70. The input capacitance (5-11) looks rather high, but not excessively. We'll see in practice, as you say.
Regarding Steve Bench matrix amp, I had a look at it but could not fully understand how it works. Anyway, I could see that, through quite a complex circuit layout and several additional components, Steve could get rid of the unwanted cross-talk between the two channels, building up in the same iron core. Steve also included manual adjustments that, IMHO, look critical.
A simpler ( but not as drastic ) result, has been obtained by Eng. G. Mariani (of GRAAF, Italy) and myself with the Split Core - Stereo Common Circuit transformer, in which you find two primaries (one for the left and one for the right channel) on the same core, with their closely coupled secondaries. These two primaries are physically apart ( unlike most of the push-pull transformers' half primaries) and, moreover, a "flux escape" is inserted inside every primary bobbin (consisting in 5-6 turns of magnetic band, 0.35 mm. thick), to divert some of the magnetic lines, thus reducing the coupling ratio between the primaries of the left and right channels. The crosstalk is limited to the lower frequencies and its impact on the stereophonic effect is acceptable.
However, in the case of the 4x4, I mentioned that I would use one output transformer per channel. Therefore, no cross-talk can possibly take place in the OPTs.
One side that I will have to check (and this can only be done with the amplifier built) is whether the two "paralleled" GM-70s, in each channel, disturb each other, causing oscillation or intermodulation. Also, needless to add, the quality of the sound.
I will try to get at this point as quickly as possible, but, to be realistic, it will take at least another couple of months, considering the family and work commitments.
In the meantime, I will be grateful for more advices on the matter.
I would like to add that I would not be against trying also the simpler SC-SCC transformer (described above), that I have used successfully with some other valves, in case of major problems with the 4X4.
Ari Polisois