GM70 into A2

All the parts have to be put together at some point. Have fun!
Post Reply
PET-240
Posts: 23
Joined: Sun Jan 29, 2012 1:30 pm
Location: BrisVegas, Austalia

GM70 into A2

Post by PET-240 »

Gudday All,

Quite new here, have been emailing Dave and Steve regarding chokes ond more recently assistance looking into building a driver section capable of driving a GM70 into A2.
Steve's email suggested to start with what I require and what im capable of, so we can all find out what the point is.
My background is machine electrical control systems, I am currently involved with the repair and maintenance of RF welders which run from 4kW- 20kW, generally around the 8kV B+ mark. I have not and could not design one of these. I can work with relays and plc's, I have not the skills to do electronic circuit design. I have a gift of fixing machines, most of which I have no prior knowledge of prior to standing in front of them, some of which have no documentation. I also seem to remember useless bits of information that get correlated in my head and actually come together and be useful. I also have a good mechanical mind, can draw pictures in my head, am really handsome, very modest and and little need to be proven wrong. Be nice if life was a fairy tale.....
What I would like, I am in the process of building a preamp not of my design, I am going to ask the designer man what I can share here and ask him to come look and join in. He has advised that it won't be capable of driving a GM70 into A2. Lampizator has a design for the two halves of a 6N6P to drive a 70, I would like to avoid using an IT(bet you it ends up with one) because I've never heard one, not cause they bad, hence the preamp. I have both 6N6P's and 6H30's coming to play with, I don't know if they are capable of driving a 70 into A2, I don't know how to correctly pick operating points and so on, I am doing reading as much as possible, but I learn better doing. I will ask stupid(to you knowledgable folk) questions, but I feel the dumbest question is the one you don't ask.
Why A2? Mainly because I can, also because my speakers- Lenehan Audio ML1 References, Oz grown boxes(yes boxes), also have some Hawthorne 10"PSI's OB's and a bunch of the Audience A3 drivers to build an array......sometime.
I also want to wipe the smile off a few guys here that reckon SET or SEP is only for horns etc. I reckon a 70 in A2 will kick the bum of the the ML1's quite nicely. I have also read that two tubes on parallel can add some smear, the ML1's will reveal that, they are proper built speakers.
See here, http://www.lenehanaudio.com.au/models.php , they have a big brother and so on but the site is being revisited currently.
I want an amp that brings the emotion etc but also the power to rock when asked without sweating it out.
I hope that's enough info, I ask you gents for your assistance please!

Thanks Muchly,

Drew[/url]
Last edited by PET-240 on Sun Feb 19, 2012 12:38 pm, edited 1 time in total.
PET-240
Posts: 23
Joined: Sun Jan 29, 2012 1:30 pm
Location: BrisVegas, Austalia

Datasheet

Post by PET-240 »

TerryG
Posts: 3
Joined: Mon Feb 20, 2012 6:23 pm

Post by TerryG »

Hi D!!!,

Well I am here, now what? Just kidding....

I will first address why I fell the 6N6P or the 6H30P will not drive the GM70. First off lets qualify that, it will drive the GM70 into A2, but the problem isn't if it can provide enough current, but its voltage swing.

The only way I can figure to drive the GM70 all the way to clipping would be to do the 6N6P with a driver stage before it so it could have enough voltage swing to take advantage of all the power and into A2 where you want it to operate. (I do know we may get some nay-sayers here against A2, but everyone is entitled to their opinions) Anyway what you need is a tube that can provide the grid current, when the signal goes into the positive region. Of course you will have the tube (GM70) biased a little negative we are assuming.

When I did the calculations the GM70 looks like it will draw about 10ma of current when well into A2. But then that is just math on paper so you better be prepared for a little more. Rather than try to make a preamp to drive the GM70 why not just build a gain stage drive circuit for the GM70 and drive the entire amp with almost any preamp?

There are other issues here that in my opinion need to be addressed first, like the B+ of the GM70 as this will effect your bias point, which in turn will tell you what you need specifically to drive the tube. When I was working on my plans for a GM70 I favored higher voltages over 1Kv. Which it seems you would be at home working with.


T
PET-240
Posts: 23
Joined: Sun Jan 29, 2012 1:30 pm
Location: BrisVegas, Austalia

Driver Stage

Post by PET-240 »

Hey T,

Terry is the guy who designed the pre I'm building, for you who may be wondering why we know each other.
Ok, been looking about, could I use a D3a to amplify the drive into a 6N6P to drive the 70?
D3a has gain of 77, know it is a pentode, not sure if g3 is internally connected or not. The other thought is the 6C45>6N6P>GM70
This will take awhile as I want to "design" it, but those of you with much more experience than I tell me where I'm going wrong, cap sizing etc to me would be guessing at best.
If I can put the drive stages to bed as to what is suitable, then figure loadlines etc, the powersupply caps are worrying me once we hit 1-1.2kV, I have ceramics aplenty, but no filtering caps, this will be something to keep in mind.
Thanks Terry,

This will be fun........and mind bending.....but tax deductible for me.
The joys.
sbench
Posts: 296
Joined: Fri Jan 28, 2005 3:45 pm

Post by sbench »

Hi Drew and all,

For the GM70, one operating point appropriate for A2 operation looks like it could be 1250 volts and 100mA, loaded into 12.5k load. This would require about -128 volts on the grid and about 300 volt peak to peak signal (Limits of -278 volts and +22 volts). Into 12500 ohms, that would produce a power of (2210 to 210 volts = 2000vp-p=707vrma into 12.5k) or about 40 watts.

Without the A2 excursion, that same load line produces 29 watts.

300 volt p-p drive, (106vPMS and perhaps 1 or 2 watts implies a relatively healthy driver. If you assume a few tenths volt sensitivity, you will need a gain in the driver stages of perhaps 300.

Does anyone have any idea whether a GM-70 is likely to sound better with a 12.5k or a 10k load line?

Dave, 10k or 12.5k or 15k at 1000v, 1250v or 1500v would be the likely op points I would think. How's the xformer look at those points?

And who is that wierd Steve character? I thought he died a few years ago. :shock:

Warmest Regards,
Stephie <3
dave slagle
Posts: 2085
Joined: Sat Jan 22, 2005 3:54 am
Location: NYC
Contact:

Post by dave slagle »

Hey all,

Well... 1250V scares me. 100ma at 40W output is no small task particularly considering the peaks will be over 2KV.

My suggestion is a bit of a cop out but I would biamp with a line level crossover. This way the 10 octaves or so of bandwidth can be centered on each driver rather than trying to do a full range amp and encroaching on the 20-20Khz ends.

(getting 12-13 octaves in a smaller more same 8W transformer is possible, but I would expect getting 10 out of this beast to be a challenge.)

dave
Get Your Fix
www.hifiheroin.com
sbench
Posts: 296
Joined: Fri Jan 28, 2005 3:45 pm

Post by sbench »

Awww, over 2kV of output? :wink:

What Spice land shows is 6SN7 as a cathode follower: +300 volts on the plate, -127v on the cathode and a 30k resistor to -350 volts. The -127v couples to the GM-70. Adjust the bias by pot on the grid of the 6SN7. The gain stage is a cascode 6N1P/6N1P stage, with a 150k resistor to the GM70's 1275 volt supply, witha 374 ohm resistor in the bottom cathode, partially bypassed (47 ohm degeneration). This provides 50 watts at clip from a 14k to 8 ohm transformer. At 30 watts, distortion is only a few tents percent. Sensitivity = .5vrms producing 50 watts.

You do need +1275V, a +300v and a -350 volt supply in addition to the filaments.

Tube complement as described is 1 6N1p, 1/2 6SN7 and 1 GM70 per channel.

Spice file attached.

Enjoy,
Stephie <3
Attachments
driver_6sn7.asc
(4.37 KiB) Downloaded 140 times
PET-240
Posts: 23
Joined: Sun Jan 29, 2012 1:30 pm
Location: BrisVegas, Austalia

Awwww Crikey

Post by PET-240 »

G'day,

Just reading that, and that you get around 25W(?) in A1, I'm starting to see the cost/difficulty/benefit ratio is waaaaaaay outta balance. I don't mind doing things because I can, but this possibly pointless.
Stephie, may I ask what is max possible power in A1 please? I read all your loadlines papers last night, want to print out the curves and go doodling, really want to understand this more, having you folks watching my back is inspiring and a great safety net.
Appreciate the input so far, must say my butt puckered at Stephie's and Dave's responses above.

Cheers,

D
sbench
Posts: 296
Joined: Fri Jan 28, 2005 3:45 pm

Post by sbench »

For A1 operation, it looks like 30 to 35 watts into 10k is about the limit. The 50 watts I described above does include transformer losses etc, so from a practical viewpoint you are probably comparing a 30 watt A1 to a 50 watt A2 amplifier. For that added power, you end up with a 3 stage vs 2 stage driver and more complex power supply requirements.

For A1, a transformer coupling to the GM-70 is practical and a simple bias supply.

Whether it's worth it or now? Well, that is an individual call.For some people the complexity is not a barrier; for others, simplicity rules.

Warmest Regards,
Stephie
dave slagle
Posts: 2085
Joined: Sat Jan 22, 2005 3:54 am
Location: NYC
Contact:

Post by dave slagle »

You may also want to take note that while the graphite plate GM70's are rated at 125W, the MUCH better sounding copper plate ones will not handle that dissipation.

dave
Get Your Fix
www.hifiheroin.com
sbench
Posts: 296
Joined: Fri Jan 28, 2005 3:45 pm

Post by sbench »

That's an excellent point. I've never physically met a GM-70, so the sounds are only from second hand info. Also, it is now ALWAYS true that the "hottest" operating point corresponds to the best sounding one.

The data above is only from a theory view... I've built 845 amps, but not GM-70 ones.

Warmest Regards,
Stephie
PET-240
Posts: 23
Joined: Sun Jan 29, 2012 1:30 pm
Location: BrisVegas, Austalia

Post by PET-240 »

Yup.......right.

Methinkin on all this, what if we dropped the input sensitivity back to 1.5v for 50W?
This is a curiosity question, would that not simplify the driver stage?
To clarify Stephie, I mentioned the 6CA7 as a possible driver for the 70? This is one of those things that I just threw out there.
I always figure simplicity is best, but can someone describe the sound differences between A1 and A2 please?

Thanks

D
sbench
Posts: 296
Joined: Fri Jan 28, 2005 3:45 pm

Post by sbench »

I have made A2 amps from 841s and 809s. I find them quite good, but not quite as warm as A1. What I find in the 841 amp is a very detailed and accurate sound. Pleasant to listen to, but somehow darker and more sterile sounding than A1 amps.

The real key in those amplifiers is to make the transition between no grid current (A1) and grid current draw invisible. That is tough because the impedance seen at the grid is very rapidly changing during that transition point.

My 801 push pull "harmonic equalizer" amp (shown in these pages somewhere) actually moves between A1, AB1 and AB2 modes of operation. If you listen carefully, you can tell which regime you're listening to, but it is very very difficult to do so.

Warmest Regards,
Stephie
PET-240
Posts: 23
Joined: Sun Jan 29, 2012 1:30 pm
Location: BrisVegas, Austalia

Post by PET-240 »

G'day Stephie,

Ok then, thoughts on running two 70's in parallel and staying SE? I have heard that it can add some smear if the two tubes being used arent matched perfectly.

True-false? Your thoughts please? I haves feeling you have built more amps than I own tubes.........

Thanks,

D
sbench
Posts: 296
Joined: Fri Jan 28, 2005 3:45 pm

Post by sbench »

G'day Drew,

Some people are not bothered by paralleling tubes, others find the smearing you mention. I personally find that if you match the tubes AT SEVERAL OPERATING POINTS in the range you are expecting to operate, then they behave themselves and you get no smearing.

The problem with matching only at one point is that the tubes are guaranteed to be similar ONLY at that voltage and current. It is very enlightening to manually match a set of tubes, and then change the operating points. You often find that those tubes you so carefully matched become mismatched at other points. Not always, but often. You CAN find tubes out of a reasonable sized batch that DO behave very similarly and those are the ones you use.

If you match to a few percent over the expected range, you're safe enough, at least to my ears.

What's a good range? For the GM-70 thus envisioned, the quiescent point is -127v grid, 1250 volts, 100mA. Now, you might find 6 candidates that all bias at 1250v and 100mA with -124 volts on the grid, and 4 candidates that all bias -128 volts at 1250v and 100mA. I'd then try 1000 volts and 125mA, and check what bias is required. Sort the ones out that STILL maintain match. Then try 500v and 200mA and do the same. Choose those from eiither of your original "sets" that are STILL matched. Those are the ones you use.

For stereo, it is good to have BOTH channels bias similarly. Why? because the gains will more likely match, and the impedances etc. Thus, you will not get images changing around on you.

Warmest Regards,
Stephie
PET-240
Posts: 23
Joined: Sun Jan 29, 2012 1:30 pm
Location: BrisVegas, Austalia

Post by PET-240 »

Stephie,

That harmonic equaliser, Crikey!

Found the thread, read it, can it be done with the 70's? Is it easier to do a push pull this way than trying to drive the 70 hard in SE, or is it the same thing having two tubes into A2 then AB?
Just trying to get my head around it, told you I'd ask silly questions!
Though it seems you get more amp for the effort in a way, is the P-P iron easier etc?
Would you do it a different way now with more experience?

Thanks,

Drew.
sbench
Posts: 296
Joined: Fri Jan 28, 2005 3:45 pm

Post by sbench »

P-P makes things easier in a number of ways, since even order products cancel. DC balances so the iron doesn't have to be as large for a given power level, and it is WAY more efficient. You get a lot more power, sometimes 3x as much as two paralleled SE tubes. Many people do not consider even the best P-P to sound as good as a nice SE design.

I think I mentioned somewhere that of all the amps I have built, I like my matrix 813 the best. Three amps stand out:
1. a SE 801 amp. So sweet and pure a sound. Terribly under powered for my needs, but a sweet, delicate sound.
2. an 845 SE amp. Nice sound.
3. a matrix 813 amp. I still consider this to be my best sounding amp. I built it in too small a container... I'll rebuild it in some way when I get a chance, but I don't think I'll touch the design at all.

Stephie <3
sbench
Posts: 296
Joined: Fri Jan 28, 2005 3:45 pm

Post by sbench »

A quick model up shows the GM-70 operating PP Class AB1 (not driven into grid current) can provide about 150 watts. 8kCT to 8 ohm output transformer, 1500 volt anode, -170 volts bias for about 48 mA per tube quiescent current.

The 170 volt bias suggests 340v p-p drive (120v RMS) are needed on each grid. This could be accomplished using a 1:1:1 driver transformer. You would get about 0.5 to 1 volt sensitivity from 1/2 6SN7 driving 1/2 6SN7 driving a suitable trnsformer.

Thus the tube complement per channel would be 1 6SN7 and 2 GM-70.

Warmest Regards,
Stephie
PET-240
Posts: 23
Joined: Sun Jan 29, 2012 1:30 pm
Location: BrisVegas, Austalia

Post by PET-240 »

G'day Stephie,

Amazes me that you can do that so quickly, next job for me, read about the matrix amp on grey gum. It is about sound, the ability to hit dynamics easily etc.

I shall return....... Triumphant sounds loud...doesn't it?
PET-240
Posts: 23
Joined: Sun Jan 29, 2012 1:30 pm
Location: BrisVegas, Austalia

Post by PET-240 »

Hey Stephie,

That matrix does sound interesting, the 813 as you have it would be most interesting, though the output tranny is something I'm glad you figured out as that is a serious bunch of wires! Though if wound specifically for the 70, would be simpler?
Going to have a look about on the output of the 813, get more of an idea as to it's capabilities, get an idea about what the 70 would do here.

Audio........such a simple hobby.
TerryG
Posts: 3
Joined: Mon Feb 20, 2012 6:23 pm

Post by TerryG »

dave slagle wrote:You may also want to take note that while the graphite plate GM70's are rated at 125W, the MUCH better sounding copper plate ones will not handle that dissipation.

dave
All,

Regarding the issue of wattage capacity and the GM70, a very reputable source (dare I say who) told me that he ramped the GM70 way up to 1500v, and up to about 200ma and the tubes didn't even break a sweat. This same source says that the currently available "made up" datasheets were for audiophiles, and are very conservative compared to what the tubes can actually do. These are transmitter tubes and usually they are operated such that the plates are intended to glow during operation and last several thousands of hours doing so.

It would be interesting to see how much SET wattage you could get from this tube, a lot depends on the driving circuit of course. Personally I like interstage transformers, to my ears they sound better than anything else. I struggled with plate chokes for years, they sound good too, but for some reason are much more problematic.

I have yet to experiment with GM70's personally but just looking at them (holding one in my hand right now) they do seem to be very hefty, with very thick cathode filaments. My intention was to use a 10K primary with an 8 ohm or 16 ohms secondary depending on the speakers I use. I intend to see what they can do comfortably, without acting as a place heater in my home, although I don't know if you could really avoid this. The thing about single ended amps , and particularly A2, is that to my ears it sure sound like live music to me, very lively dynamic sound. Problem with the GM70 is your not going to hear any A2 until you get to louder levels. Might be a hair rasing experience though.

Stephie,
How low in impedance does the grid drop do you think when in A2? It didn't seem to drop below 10K if I recall correctly when I was putting them on paper, meaning it shouldn't be to hard to drive if you have the voltage swing. There seems to be many tubes that could drive them, but your looking at, at least two gain stages before the power stage.

Cordially,
Terry
PET-240
Posts: 23
Joined: Sun Jan 29, 2012 1:30 pm
Location: BrisVegas, Austalia

Post by PET-240 »

Heya T,

Good points those, the SE thing still interests me, have heard the 6SN7's can have a bit of a hard edge to them, no idea about cathode followers and their sound characteristics, I take it Stephie's usings them because of their capacities in regards to voltage swing etc.
In regards to supply voltage, if supplied with 1200V, how does the supply reach 2kv? I can see the method of how the output trans supplies the ac waveform, just trying to grasp how the input voltage varies so much. Is it the P-P input voltage is say 300v, gain of 7, so the plate voltage will swing to 2100v? Simplified I realise!

Got me most curious this.......
sbench
Posts: 296
Joined: Fri Jan 28, 2005 3:45 pm

Post by sbench »

Stephie,
How low in impedance does the grid drop do you think when in A2? It didn't seem to drop below 10K if I recall correctly when I was putting them on paper, meaning it shouldn't be to hard to drive if you have the voltage swing. There seems to be many tubes that could drive them, but your looking at, at least two gain stages before the power stage.
Since I do not own any GM-70's I have to rely on the data sheet as available information. For the GM-70, what I have in the "triode_nh.txt" model file is simply 1k and a diode. That's because I had no additional info in order to build an accurate grid model. If someone has curves on grid current vs grid and plate voltage, I can make the model more accurate. Right now, the accurate grid models are associated with devices that have grid current info. (833, 801, that sort of transmitting tube).

Re being able to overdrive beyond "audiophile" limits. If the curves are reasonably accurate, you can put any voltage and current you like into a Spice model, and the tube will respond accordingly. I tried to stay within the "published" limits, although we girls occasionally like to have things pushed a little bit further. (O my, did I really say that?) :oops:

Stephie <3
TerryG
Posts: 3
Joined: Mon Feb 20, 2012 6:23 pm

Post by TerryG »

I have a pair of these muscle tubes, but haven't really put them through their paces yet. With these larger tubes the filiment current is unreal.

BTW, if you are going to use a cathode follower to push the GM70 into A2 you better have a pretty big coupling capacitor, at least 10uF I figure.

All I can say is go for it Drew!!

Terry
PET-240
Posts: 23
Joined: Sun Jan 29, 2012 1:30 pm
Location: BrisVegas, Austalia

Some light reading......

Post by PET-240 »

G'day again all,

Much reading and learning going on in my head.

Gonna add some links if I can get the iPhone to let me, of some other builds, you guys may see something I don't, I know you will, but I'm curious.......must have 9 lives?
This is a series, as yet incomplete at ultra fi by Peter Sikkens I think?
http://ultra-fi.blogspot.com.au/2011/01 ... 1.html?m=1
http://ultra-fi.blogspot.com.au/2011/03 ... 2.html?m=1
http://ultra-fi.blogspot.com.au/2011/05 ... 3.html?m=1
http://ultra-fi.blogspot.com/2011/08/co ... m70-4.html
http://ultra-fi.blogspot.com/2011/12/co ... m70-5.html
http://ultra-fi.blogspot.com/2011/12/co ... m70-6.html

And this one elsewhere,
http://www.audiodesignguide.com/Fiat/GM70d.html

I am hearing what you folks are saying, the Coppatone is talking A2 and 40W at a different operating point, not saying its right or we are, just a few more thoughts possibly won't hurt. Makes some interesting thoughts here and jc adds a couple of cents too. Prob pay to have a geezer at em all.

Dave, I realise it's early days, but can you please look over the last link, it talks about specs for their trannies. Also, in regards to the core, if we end up SE, would we be able to change bobbins if we had to but re-use the same core? Just wondering if we changed the current spec as per Terry's suggestion above re: published operating points and curves, if we spec the trans for up to 150mA, or is this where we can play with gaps as required, or use the cancellation winding to best effect? I guess that would muck with primary impedance and loading? Or is that just from the secondary speaker loading?

Stephie, after reading all your Matrix info......twice mebbe three times, would it be an advantage going that way or the way we are looking? Just I know you have both types of amps, guess depends one way or the other.

Terry, page 6(edit:,nope page 5) shows what I think is his close to final schematic, where would the cap you mention be? He seems to have eradicated caps in the signal path.

Thankyou Linesman, Thankyou ballboys,

Drew.
IslandPink
Posts: 56
Joined: Thu May 12, 2005 8:01 am
Location: Denbigh, North Wales

ACF2 ?

Post by IslandPink »

My friend Nick ( G ) recently built 211 and 100TH ( Eimac ) amps with the J.Ross McDonald ACF-2 circuit as driver.
http://jrossmacdonald.com/jrmpubs/040Ca ... llower.pdf
That should work quite nicely, although you'd need something like 35-40mA in the right-hand section to handle the Miller cap of the GM70 .
They do sound nice though when driven properly - especially at 1800V and 120mA as we once tried ( briefly ) - I think that one used something like a 2A3 or 300B as driver though .

Mark
PET-240
Posts: 23
Joined: Sun Jan 29, 2012 1:30 pm
Location: BrisVegas, Austalia

Post by PET-240 »

G'day Mark,

Can you advise as to the remainder of the circuit please?

Trans imp, power supply etc?

Thanks!

Drew.
Post Reply