DIY Audio Projects Forum
http://diyaudioprojects.com/Forum/

GM70 Triode Stereo Oddwatt SRPP / SIPP
http://diyaudioprojects.com/Forum/viewtopic.php?f=9&t=1957
Page 8 of 15

Author:  sampleaccurate [ 17 Jun 2010, 01:44 ]
Post subject:  Re: GM70 Triode Stereo Oddwatt SRPP / SIPP

Edcor responded to my inquiry today about hi-potting their transformers.

They use 1000 volts, I assume RMS. I believe Hammond uses 2000 volts or 3000 volts on some models.

I'm sticking with the Edcor. They also said that it should easily handle 1500 volts. If that's the case it should work. I'm putting a maximum of 1000 volts RMS across the primary and the B+ adds another 500 volts. We'll see. :xfingers:

Author:  azazello [ 17 Jun 2010, 07:41 ]
Post subject:  Re: GM70 Triode Stereo Oddwatt SRPP / SIPP

6E5P like driver! for GU50, GM70 and........
http://www.goodsoundclub.com/pdf/6E5P_Swedish.pdf

Author:  sampleaccurate [ 17 Jun 2010, 10:00 ]
Post subject:  Re: GM70 Triode Stereo Oddwatt SRPP / SIPP

azazello wrote:
6E5P like driver! for GU50, GM70 and........ http://www.goodsoundclub.com/pdf/6E5P_Swedish.pdf

What is the maximum allowable B+ voltage for the 6E5P. I have a bunch of them.

Author:  azazello [ 17 Jun 2010, 11:29 ]
Post subject:  Re: GM70 Triode Stereo Oddwatt SRPP / SIPP

There is all parameters and U&I diagram...
All power tubes have bias between 60 - 120 v!I
For filling input cap every power tube, 6E5P must have Ia = more than 20 ma. U nom = 150-190 v. I use 190 v.

Author:  sampleaccurate [ 17 Jun 2010, 16:23 ]
Post subject:  Re: GM70 Triode Stereo Oddwatt SRPP / SIPP

azazello wrote:
There is all parameters and U&I diagram...
All power tubes have bias between 60 - 120 v!I
For filling input cap every power tube, 6E5P must have Ia = more than 20 ma. U nom = 150-190 v. I use 190 v.

The grid of the GM70 will bias at about -120 volts in my design. The grid must swing from 0 volts to about -230 volts for full power output, a total of about 230 volts peak to peak.

How do you propose to obtain 230 volts output swing using a 190 volt B+ on the driver tube? It will certainly work, but at the expense of reduced output power. If I could use 300 volts to power the 6E5P I'd be in business.

I need a driver that can both supply enough current to "fill the input cap" (low impedance) AND supply 240 volts of output voltage swing. I may in fact need to use a follower of some sort to buffer the 6SN7 SRPP, but according to the simulations I've done in PSPICE the 6SN7 SRPP will drive the input cap, simulated by placing a cap on the output of the SRPP stage to ground. The thing IS a type of follower and has an inherently low output impedance.

The maximum average cathode current of the 6SN7 by coincidence is 20 mA. So I may be cutting it close with the 6SN7. However, the output impedance of an SRPP is much lower than a single triode (from what I understand), so it may well be able to provide adequate current.

I'm considering using a solid state LM317 / BUX85 CCS to replace the upper triode. Such a load represents a very high impedance to the bottom triode and has many beneficial effects, including a lower impedance output and much better rejection of power supply hum.


I don't see the use of a solid state CCS in the SRPP as being inconsistent with the design philosophy of the oddwatt and the use of a CCS in the SIPP stage.

Author:  azazello [ 17 Jun 2010, 17:23 ]
Post subject:  Re: GM70 Triode Stereo Oddwatt SRPP / SIPP

For GM70 working area is between 0 /when tube is open/ and -120v /when tube is closed/. For normal work
/it is easy and simply to explain and understand/,... input signal must be amplified some times in this area...., for practice, very easy is to divide aprox. 120 to p.e. 0,6 - 1 v
output signal from CD that will give You needed mu for choising needed pretube.
If You want use 6SN7, please, and You will see the result...in practice.../I sow a lot of schematic with GM70, but didn't see with only 1 driver pretube with low mu....20-30/..... The choice is Yours....!

Author:  sampleaccurate [ 17 Jun 2010, 21:10 ]
Post subject:  Re: GM70 Triode Stereo Oddwatt SRPP / SIPP

azazello wrote:
For GM70 working area is between 0 /when tube is open/ and -120v /when tube is closed

Not according to the plate curves. The bias point for the GM70 is -120 volts on the grid at idle in my circuit. The tube won't cutoff (close) until the grid reaches about -260 to -270 volts, but the bias point and load I chose will cause the grid to go positive on transient peaks before the tube cuts off on negative peaks, so it should always conduct unless overdriven hard.

A 120 volt peak to peak voltage on the grid (0 to -120 volts) will only give 1/2 the output voltage possible and 1/4 the power output possible from the GM70. The schematic in the Russian tube book that Magnus linked me to shows about a -100 volt +/- fixed bias on the grid. There's a -160 volt bias supply that's divided by a 33K and 20K to reduce the negative voltage from -160 to about -100. The resistor values are given as ranges so the bias point can be adjusted to suit your needs and the tube. I used the lowest values to calculate the -100 volt bias, but the bias can go lower than -112 volts if the 33K is increased to 47K.

Here are the plate curves and something close to my loadline - I changed it to reduce the plate dissipation at idle since this was drawn but it illustrates my point. IGNORE the red stars. The bias point at idle is -120 volts on the grid, 1200 volts on the plate, and 100 mA plate current. The grid voltage can swing from 0 volts to -240 volts without saturation or cutoff respectively.

Does anybody out there have an opinion on this? Azazello seems confident, but I think he's wrong. If he's right I need to do some serious reading and rethink this entire circuit.

Image

Author:  sampleaccurate [ 17 Jun 2010, 21:49 ]
Post subject:  Re: GM70 Triode Stereo Oddwatt SRPP / SIPP

azazello wrote:
divide aprox. 120 to p.e. 0,6 - 1 v
output signal from CD that will give

p.e.? (Professional Engineer? - yes, I am registered in the state of Nevada as a P.E.)
CD? (compact disc?) - I prefer SACD
:confused:
Could you write the words out? I don't understand your abbreviations.

Author:  gofar99 [ 17 Jun 2010, 22:29 ]
Post subject:  Re: GM70 Triode Stereo Oddwatt SRPP / SIPP

Hi, Using a SS load on a SRPP is possible, but I suggest that the 317 is not the one to use. The maximum voltage on it is 37 for the standard one and 57 for the HV version. I would try an IXCP 10M45S instead. It can handle 450 volts and 2-100 ma depending on your not exceeding the max dissipation. SRPPs have been used in many designs to drive difficult loads. The advantage of low output impedance and large voltage swings lends itself to those uses. The greatest drawback is that you must consider the heater to cathode breakdown voltage rating. If you use a dual triode, then you need to account for both cathodes. If you use a completely floating arrangement there is a tendency to introduce noise in the circuit if you ground reference it then the upper cathode sees about half the B+. On new tubes the rating is often 200 V fine, but as they age I have found that really 100 is closer to actual. I use a "bias" voltage of about 1/3 the B+ applied to the stage. With say 80 volts bias you can safely go to 240 B+. This is usually fine to drive most tubes as a well designed SRPP can easily reach 80% of that in output. Depending on the tubes used it can be quite linear throughout the whole range as well. You may also want to check out Mu follower stages. They are similar and can be designed to be non-symetrical and in some cases get around the H-C problem.

Good listening
Bruce

Author:  sampleaccurate [ 17 Jun 2010, 23:05 ]
Post subject:  Re: GM70 Triode Stereo Oddwatt SRPP / SIPP

gofar99 wrote:
Hi, Using a SS load on a SRPP is possible, but I suggest that the 317 is not the one to use. The maximum voltage on it is 37 for the standard one and 57 for the HV version. I would try an IXCP 10M45S instead. It can handle 450 volts and 2-100 ma depending on your not exceeding the max dissipation. SRPPs have been used in many designs to drive difficult loads. The advantage of low output impedance and large voltage swings lends itself to those uses. The greatest drawback is that you must consider the heater to cathode breakdown voltage rating. If you use a dual triode, then you need to account for both cathodes. If you use a completely floating arrangement there is a tendency to introduce noise in the circuit if you ground reference it then the upper cathode sees about half the B+. On new tubes the rating is often 200 V fine, but as they age I have found that really 100 is closer to actual. I use a "bias" voltage of about 1/3 the B+ applied to the stage. With say 80 volts bias you can safely go to 240 B+. This is usually fine to drive most tubes as a well designed SRPP can easily reach 80% of that in output. Depending on the tubes used it can be quite linear throughout the whole range as well. You may also want to check out Mu follower stages. They are similar and can be designed to be non-symetrical and in some cases get around the H-C problem.

Those are some good points. I've taken most of that into consideration.

My CCS will be an LM317 in combination with a BUX85 high voltage transistor also capable of 450 volts output (see the GM70 circuit below). The 5 volt DC source using the zener and the B+ will be changed to a 10 volt DC source from a small transformer - the resistors required are too big and hot. The LM317 will only see about 9 volts at all times - the transistor provides the high voltage and protects the LM317. I've got the combination of the two on one channel of my oddwatt and it works perfectly. Comparing the two channels with test instruments and my ears reveals no difference whatsoever in performance. So I should be good to go with that.

As far as the heater to cathode voltage, the 6SN7 has great specs in this department - one of the reasons I chose the tube. The heater can swing 300 volts relative to the cathode according to the Tung-Sol datasheet, 200 volts below and 100 volts above. It can also handle high B+ voltages for a large output voltage swing. If I set the bias properly on the filament as you suggest it should be OK with one tube - the two triodes sharing one heater biased at a fraction of the supply voltage. 1/3 sounds about right. I'll calculate the voltages and set the bias right in the "middle". I didn't really consider the aging of the tubes though. That's a point I overlooked. I think I'll make sure to leave a margin of safety there. IF I can't accomplish the job with a dual triode I can always use two tubes and leave one of the triodes connected such that it is cut off. Then I can bias each heater separately. That is, if I don't use a CCS to replace the top triode which is what I'm leaning towards right now.

I agree, the linearity of the SRPP on my PSPICE simulations is great, both with a tube or a SS CCS. Nice circuit.

Image

Page 8 of 15 All times are UTC - 6 hours [ DST ]
Powered by phpBB® Forum Software © phpBB Group
http://www.phpbb.com/