• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

GM-70 design approach

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I have been brainstorming a gm 70 design but have been getting stumped in a few areas. This is my first attempt at a high voltage design so I am unsure about a few things regarding the power supply design.

Filter chokes that have a dcwv of 1kV+ seem hard to come by. How do some people go about obtaining these chokes? Is there a way to use a lower test choke and decrease the potential that it sees?

I am planning on making it a SE parallel design using the hammond 1642SE 5K 75W OT. I am thinking of using a 6sn7 type as an input tube followed by an 807 to drive the gm-70’s. I have yet to start drawing any schematics on paper, but I wanted to run this past a few of you and see if anything really stands out as a large problem with this design approach. After looking at the 807s datasheet, do believe it will drive the gm 70 well. Is this a correct assumption? Thanks!
 
Administrator
Joined 2004
Paid Member
I placed the chokes in the CLCLC plate supply to the GM70 on the negative side of the supply in my amps. See old ARRL Radio Amateur Hand Books for useful high voltage supply design details. Since the only the ripple voltage and DC IR drop appear across the choke there is not much stress on the insulation and standard good quality chokes can be used.

I would strongly suggest you check out Electra-Print for your OPTs, they will not be unreasonably expensive and they are quite good from my direct experience. (An extremely expensive GM70 high end amp uses Electra-Print OPTs) I had jack design a pair with 7K primary rated for 150mA and 40W, with a single copper 8 ohm secondary.. In class A1 at 1kV/120mA I am getting a little over 20W, driver stage is an IT coupled triode connected D3A running at 20mA.. (Note: 25W in A1 with 5K primary is possible, I went for higher DF and lower distortion)

This amplifier replaced a 300B SE amplifier I designed 13yrs ago, the goal was to achieve at least 5dB greater output than that amplifier, with considerably better linearity at my usual listening levels. I choose to go with a two stage amplifier because I am finding in my old age that fewer stages sound better. This amp clobbers the old one in terms of resolution and dynamic behavior with my high efficiency 3 way speaker system. (Onken/Ultraflex based)
 

Attachments

  • DSCN0044R.JPG
    DSCN0044R.JPG
    214.8 KB · Views: 2,245
Administrator
Joined 2004
Paid Member
The D3A in triode has a measured mu of 80 and can swing absolutely obscene amounts of voltage as long as the driven grid current is not too many mA.. I need about 180Vpp swing to drive the GM70 to full output in A1.. Open circuit swing is in excess of 300Vpp at a couple of % thd on a 200V rail.. (Varies somewhat from sample to sample)

The D3A is coupled to the output stage with a 1:1 IT, gain is more than sufficient for the target output power. The driver stage was not intended to furnish appreciable grid current but is capable of furnishing at least 4mA so some very limited operation into A2 is possible. Given an speaker efficiency of over 100dB and >20W per channel it gets loud.
 
Administrator
Joined 2004
Paid Member
Are you using D3A as pentode or triode?

Nvm: you said triode... but have you tried pentode?

Pentode is not a good mix with IT, and I am looking for the low source impedance I get with triode connection and no feedback local or otherwise. I prefer the distortion spectrum in triode..

I do have an SE amp with 6J7 pentode driver into a 6V6, (triode or tetrode mode) it is not one of my favorite amps.
 
... After looking at the 807s datasheet, do believe it will drive the gm 70 well. Is this a correct assumption? Thanks!

Believe me, one of best drivers for GM-70 is 807 (in triode connection).
Just try to find russian analog G-807 (Г-807) from ebay for example.
For input tube you can use russian 6N1P-EV (6Н1П-ЕВ).
Very good double triode, long life military production, works very well at anode current about 1mA.
Ua about 6mA for 807 is good enough.
(now somebody will start with comments "aaah, at 6mA linearity is not good, blah blah blah...")

An advice for chokes...
Just buy such chinese winding machine (again ebay) and use reel and core from old trafos,
you just have to find the proper size.
I'm using the same chinese winding sh*t even to wind output transformers for GM-70 :D
 
Last edited:
Thanks!

Thanks everyone for the advice, I will look at some old radio books and get a clearer image on the HV supplies.

I looked up the russian 807 on ebay and could not believe the low price on these! How does the sound compare to Sylvania or RCA?

Also no one commented on the thought of running them in parallel to get more power, are there any complications that may occur by doing this?
 
Thanks everyone for the advice, I will look at some old radio books and get a clearer image on the HV supplies.

I looked up the russian 807 on ebay and could not believe the low price on these! How does the sound compare to Sylvania or RCA?

Also no one commented on the thought of running them in parallel to get more power, are there any complications that may occur by doing this?

First what to parallel? Driver or output tube? It's not very clear...

Second... about russian military production.
One of most important advantages is very high vacuum achieved in their
military production (especially in 60's-70's) due to very high military requirements
(don't forget, in US all of military production came from private companies and low production costs was the most important thing there).
In USSR the most important thing was quality of production, not costs.
All military factories was owned by their government (communist party).
Current russian production is not as good as old one.
A very good example is Sovtek production of 2A3 and 300B (also known as Electro Harmonix).
 
Last edited:
Very sorry! I was really tired last night and obviously I did not clarify. I was referring to running the GM-70 outputs in parallel.

Very interesting what you said about the Russian production....I have experience with some of the old USSR teflon caps and you sure are right about them not cutting corners:D

Looks like I will have to order some of these USSR 807s. I have noticed that there are brown and black base versions. Are there any differences between these? Maybe production years?
 
The colour of the base is different mainly because of the time:

Ulyanovsk factory production (old logo), black base:

G-807-807-1625-5933-beam-tetrode-tubes-lot-of-4-partpic.jpg


Again Ulyanovsk factory production (new logo), brown and black base:

An externally hosted image should be here but it was not working when we last tested it.


GM70 in parallel is something useless except that case,
when you're using speakers with extremely low sensitivity
(or electric heater instead of speaker).
Tube SE amp (as a conception) also includes high sensitivity full range driver,
mounted (very often) in a BL horn enclosure.
Anyway, 20-25 watts of output power is enough
even for speakers with sensitivity about 82-84dB/W/m.
 
Last edited:
Administrator
Joined 2004
Paid Member
I can attest that the >20W I get out a single copper plate GM70 is more than sufficient with my speaker system, and frankly if you need more power than this you might want to consider something other than a GM70 in A1..

You could for example IT couple your triode connected 807 to the GM70 run it fairly hard (say 50mA or so) and you might get far enough into A2 to get 40 - 50W out a single GM70. I'd probably do a 2:1 step down or similar (3:1?) that should give you plenty of grid current capability. Note mine is designed primarily for A1and I have not tried this yet.. (Won't with these amps as I am very pleased with their overall performance)
 
Well in most situations I would not be concerned about wattage. I used to own some horn speakers and remember being happy with a 6 watt SET. Unfortunately my goal with this project is to power some large infinity's... Before you think I am crazy, I have tried all kinds of power amps on these speakers. To my amazement, they are very versatile when it comes to the variety of amplifiers that make them sing. They sound excellent with a 40 watt el34 amp and I have found that 50-75 watts on these is more than I really need. So my thought is if I could get that out of a SET it would be pretty cool.

...ok and I just may want to build a GM-70:D

My guess is that A2 will have higher distortion than A1 operation, correct? Would the B+ have to go up also?

Thank you guys for the guidance, I really appreciate it!!
 
Last edited:
Thanks everyone for the advice, I will look at some old radio books and get a clearer image on the HV supplies.

I looked up the russian 807 on ebay and could not believe the low price on these! How does the sound compare to Sylvania or RCA?

Also no one commented on the thought of running them in parallel to get more power, are there any complications that may occur by doing this?

hi
this tubes make your amp very dark sounding.
 
... I am thinking of using a 6sn7 type as an input tube followed by an 807 to drive the gm-70’s....
If you want to consider avoiding the use of a tube with a Top Cap (807), the EL34 in triode mode can swing a lot of voltage with relatively low distortion. 801A is another possibility if you want to try DHT.

For maximum power output on GM70 (and 845) the anode voltage should be as high as possible and bias it as hot as possible, to run close to maximum anode dissipation.
 
Administrator
Joined 2004
Paid Member
<snip>
...ok and I just may want to build a GM-70:D

My guess is that A2 will have higher distortion than A1 operation, correct? Would the B+ have to go up also?

Thank you guys for the guidance, I really appreciate it!!

Same B+ and since it is class A the average power consumption does not change. (The instantaneous power consumption will increase and low supply impedance is important.)

Results will depend almost entirely on how linearly the driver stage can provide the required grid current. The GM70 curves in the grid current region still look very good.
 
class A1 or A2?

Hi

Do you intend to drive the GM70 in A2 or A1?

You can have a look how I managed this on my TB3/1000 150W SE amp. I am using an EL34 simply because I like this tube very much.

ward's 150W SE amp blog part 3

By the way, we have a GM70 output transformer in our range. // version can be made with very good spec that are comparable to the ones of the single GM70.

http://www.monolithmagnetics.com/images/stories/datasheet_S-15.pdf

We also produce inductors , low capacitance versions if needed on a custom basis. The low capacitance versions (each layer isolated individually) will accept very high voltages with no problem.

You can indeed put them in the zero line as Kevin indictaed.

I wish you succes with your project.

grz, //WDC
www.monolithmagnetics.com
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.