GM70 Tractor mk2 - oh my

[andrew ivimey]

I have had a couple of days to put together an amp for the eggfest. What a mess! I threatened the Tractor and the tractor Mk2 it is.

Anode choked 6S45PE into audionote copper into GM70 into the cheapest Hammond transformers that will still not breakdown under the rigours I put them through. 540volts HT, I could go more but a) can’t be bothered and b) want to show that you can get very loud without the higher tension. (Oh I might go higher; it’s easy enough, but I haven’t really got the time.

I started with AC heating on the GM70s but yikes! What a hum. Ignoring the hum though I realised why I really like AC heating; a) easy and b) It sounds so GOOD!
I went on to rectifying things. And the sound got hard. But the hum was totally gone.

Now, checking current and voltage the GM70s are using 20 volts and 3.3amps. You’d think that was fine wouldn’t you? Well the 160VA toroid I am using to provide the 20volts / 3 amps per valve would overheat and melt down, or so it feels, if I wasn’t using the very splendid and worthwhile computer fan with the blue LEDs. The LEDs make all the difference.

What a mess as I said, I want to hear what GM70s really sound like. Nick, you are bringing a proper GM70 amp, aren’t you? Will, if you have those coppers it would be lovely to hear them compared to the bog standard graphites I’ve got.

So far then, dead loud and interestingly revealing – oh and I have an appalling conductive hearing loss at the mo’ – whatever next!?

I haven’t got the time to do much more to this version of the Tractor but it would be nice to take it out to the fields, so to speak.

[NickG]

Well. I would hardly say I was planning a proper one, its a bit of a lash up out of stuff I have lying about (no wonder I am running out of room), should get built this weekend.

I will bring my pair of copper GM70's, I was planning on using a string of resistors to set the bias, so we can set it at something nice for the copper, and up the current for the graphite ones, If Paul brings his 1k5 power supply and James TX's, we should be able to see how they sound with the anodes glowing :-)

I found in the past, that at 600v it actually sounds very good, and loud enough.

[Paul Barker]

Great, could someone sim it please? Otherwise we'll never know how it will sound.

[andrew ivimey]

'sim'?

simulate a hearing loss? shove your fingers in your ears, dear boy! the fascinating thing is that I have never had this before and so many of the children who pass through my tender hands have, plus ear ache and pain and burst ear drums blah blah blah,so on to ENT, but the hearing loss and 'fullness' with low frequency tinnitus is what I have - fascinating! antibiotics and sudafed type stuff - just the job ( and a couple off days off work) !

or do you mean something else? as some would post LOL - whatever that means.

right now I'm having a rolling stones evening - nostalgia - through WE91
(6SJ7, angela variety, very pleasant but this single mains transformer runs badly hot too - what is it about me?

I'm just going to let this ******oopps melt down - turn up the volume; can't hear nuffink anyway!

Assuming Nick's lash up has decent output transformers then what with the coppers, this is what I want to know. Anyone got some excellent 5k or more output txs?

[Paul Barker]

Yes, Tributes but they are mounted on the 212 chassis, I could bring that I suppose. Am bringing the James also, those are OK, not ground breaking.

I meant sim your amp, don't you know it's hopeless if it doesn't look good on the sim? Pull yourself into the 21st Century you dinasuar.

Don't talk to me about hearing, it's been proven I'm listening for different stuff to most other people. Obviously I'll never make a living in audio, just as well I don't want to.

We build for ourselves, if other people like it too, it's nice, but it doesn't benefit us. We aren't here to impress, we are here to enjoy and marvel at what massively different hearing we all must have.

[Andrew]

Don't worry Andrew, Paul's got a "bee in his bonnet" about us folks who like to sit in front of a computer and use simultions to help us design/understand/improve or otherwise model a circuit.

Just poke 'im in the ribs back and see if he jumps....

cheers,

-- Andrew

[andrew ivimey]

How on earth could a computer simulation mean anything!? weird or bizarrrrre or what.

The amp gets hot, I hear what it puts out .....

This GM70 amp cost very little (as it is all scrap box stuff ... later ... I want to improve things. maybe GM70 (and all those russian valves aren't so bad after all - and that's where I could, but don't want to, spend money ...

whisper ... potential? not bad eh oooerrr - maybe - que faire ??? no room - I'll chuck out some stuff - honest

hm, change of vinyl time, strange my balance is way off, nearly broke an
rb250 let alone the 103 that is precariously tracking Between the Buttons - oh nostalgia, I could weep.

Turn it up!

[Andrew]

Hi Andrew,

You been on pseudoephedrine again? ;)

cheers,

-- Andrew

[andrew ivimey]

I wish ...........
......
......
.......
......

just being off work for two days ..... and I like my job

WE91 continues not to melt down, tho' the mains transfomer is too hot to touch - pity if it does go adn takes other bits with it. My Czech 32bs are safely boxed on the shelf; EH gold grids can take their chance.... Carmina Banana isn't over yet.

[andrew ivimey]

later

so much for melt down, the fuse blew

next time I don't use a fuse


so I have moved on to T-amp - hm boxed in and cagey - oh I suppose its all right really but I'd rather barbecue mains tranformers.

[NickG]

Stand back, there is no stopping him now...

[Andrew]

Hi Andrew,

I wonder if there's some DC on your mains? Apart from the high volume - and I heartily approve of course - something's got to be causing all your mains TXs to overheat.

cheers,

-- Andrew

[andrew ivimey]

I connected my maplin cheap digimultimeter to 20volts DC and applied to the mains - there is a regular fluctuation of up to 3.2 volts every half a second or so.

disconnected swampy filter - same result.

Now could this be true - whence cometh, then?

or is it cheapness maplin meter? - the other cheapmaplin does it too.

D.C. on the mains in not healthy!! if true, what do I do?

[Global]

Hi Andrew

http://www.diyparadise.com/dablok.html

Sounds like you may need to use 2 sets of 3 diodes to tame your problem.

You can build heavier-duty variations on this to go in a mains supply box but I have used this successfully to stop buzzing in a torroidal transformer on a 'per amplifier' basis.

Cheers

Mark

[andrew ivimey]

Mark, Hi, thanks, I remember discussion about this a while ago. Looks like I will have to try it.

But where does the D.C. come from?

Back to work ....laters!

[Global]

Hi Andrew

Ther may be anumber of reasons for DC on the mains. I have heard it said (haven't done any experiments) that earth points in a fairly close area can differ in potential by some tens of millivolts so, depending where Neutral is bonded to Earth, you might have a PD.

Could be c*** of course!

Mark

[Shane]

Can we extract and find a use for that DC? After all, it's free...

[andrew ivimey]

Diyparadise, 'Dabloc' is in place - let's wait and see.

Anyone else out there use them???

[NickG]

I did at one point, and found it was a improvement. I blew it up with the 211 and never rebuilt it, maybe I should.

[andrew ivimey]

DaBlok don't help - more diodes, less capacitance????? dunno

Back to the computer fan ....

sounds pretty good, though.

[Paul Barker]

There is usually a potential between neutral and ground, but has that anything to do with a dc potential between phase and neutral? If so how? If not, where does any dc potential come from?

Is not the best way to blockdc in our circumstances a 3kw gapped centre tapped isolation transformer? I don't think I'd prat around with diodes.

[andrew ivimey]

OTOneH, I agree with an isolation transformer.

OTOH, they are expensive and bulky.

You've got one haven't you! ;)

[Paul Barker]

I have a small 1kw centre tapped isolation transformer but it isn't gapped and won't handle dc any better than your transformers, though it will block the dc for you. The problem is that the dc causes the transformer performance to sag during the cycle which is audible, you must have a transformer upstream which is so designed to handle dc to smooth out this problem. I guess the lucky people are the ones with a transformer perched on a pole outside their listening room window.

I keep asking Ali Tait to get me one of these so I can invert it and make a gm100 power supply, but he is most of the time dealing with the larger underground transformers further up the chain of voltages. I forget how many thousand volts he deals with 40,000 or something. I think the pole mounted transformers just go from 10,000 to 230 (LOL we all know it's only called 230 so we can say we conform to EU standard, nothing has actually changed it's 240 as it always was, trouble is nobody told the Chinease)

[G.Axiom]

Hi again.
Perhaps I should think droppin in in the next Eggfest. :D
ROTFL
This really is a small world.
Just saw this thread and now, look at whats lying in my garage.
It's a transmitter PS from a disbanded ship radio station, in good working condition. Salvaged it, it was goin to garbage. Ran it up and it puts out 3,3 kV. Looks as a good PS-candidate for that GU81M lying on a CD box on the top of the choke. Considering B+1,5 kV, 0,5A per channel. Thats a good, quite conservative op point to GU81M.
I think I will try toroidal power trans as output trans, this is intended to feed those subs in my active 2-way system, so no worry about HF roll-of.

Or, there is a sister ship which will disband its radio station in the next docking in the fall. Perhaps I should go PP-monoblocks:D

[James D]

Oh that is very nice!

Whats its input 48V dc?

James

[G.Axiom]

Oh that is very nice!

Whats its input 48V dc?

James

I think it is 12,6v, 11A for filaments.

You see, I really hope I can keep me from starting this project. This _thing_ has been lying there for two months now and I am trying to forget it but I cant. I think you know... puttin together 150w SE, with less than 1000 Euros. I just think those power transformers as outputs...
Have a good local source for the tubes, so no worry about that.


Edit: oops, which input? It has 220, 230 and 240v ac primaries, if you are asking that? Just plug it in the nearest wall inlet and pray:D

[andrew ivimey]

I think there are several of us feeling a little envious.

How does one get hold of such things???

[Paul Barker]

Keep a weather eye on ebay, Ht transformers capable of 1 amp 3kv they come up for under a ton once in a while. They alone are a two man lift and that would be strong men. Buyer collects of course.

[G.Axiom]

I think there are several of us feeling a little envious.

How does one get hold of such things???

Well, I could easily let it change hands, if somebody really is interested. Besides, I think I most probabaly could arrange free shipment to Felixstowe, Hull or Immingham.
You see, I got three kids and wife....:D , and I am running my +100dB-sensitive system now with 6AS7-se:s to subs and VT25 to mids, with marvellous results. I really dont need any 100+watt monster, but still I am afraid that one of thous days I start this project, you know, just because if something is possible to do it usually becames done also:D that's humans nature, I think.
It is some 25 years old, OTOH it really is built as "best money can buy"; as marine radio equipment usually is. That is solely visible brands are suchs as RCA, Siemens and so on. Trans are "Gaston of Norway". EI -core and oilpaper insulation.
I have no idea of its real worth, I could change it to something interesting.

Edit: as I have a good local source I most probably could supply some GU81M:s as well. And James D: mains has 220,230, and 240 inputs.

[Martin213]

Go on Paul,you know you want it really :p ;) :D

[andrew ivimey]

I have started to take this Tractor thing seriously .... maybe that's not a good sign.

The blue leds (and the fan) keep the toroid from getting meltdownable. DaBloc didn't help but I was using 10k caps and only one set of diodes
- pour on more?

An isolating transformulatormarble might helpymould the heathingable proleblemmo.

I am on my third set of output transformers;

The Hammond 125FSEs were where I started and I tried 5k and 10k. The latter was too quiet (only 540 volts B+)

I liked the sound - well it worked didn't it!

Eastern Audio (remember them?) 5K sounded different - less bass, more well maybe mid.

But I think the 'bass' from these two transformers was thuddability rather than clarity.

So I went on to AE C cores but only 3.5K / 80ma (should be all right.. sweat breaks out - these were expensive for an amp such as Tractor and were very happily ending a WE91clone until the well specc'd <on paper> 'eastern Audio' mans transformer shorted out on the primaries).

interesting life I lead, n'est-ce pas

The sound is so very much more clear with C core - all those tinkly things and bits that are added to the mix, were oh so clear and 3 dimensionallywhole. But much less bass. It was there sure, but less clout in the balance.

Now I have 20 mfd (solen) ultrapath caps, ultrapathing away and 'Wish You Were Here' scratching away. There is plenty of bass now but it might be what it is supposed to sound like. - not sure/ dunno.

Oh and I have a nice chromed audionote choke for next - this should make a difference.

And then a 5U4GTB as rectifier rather than the diodes which are so easy, cheap, dull looking. Valve rectumification means fewer volts. But there is a hollow looking, rather forlorn space on the baseboard, right there next to the 'eastern audio' - oh no, not again! mains transformer with spare windings a plenty - could have used 211s. dead easy.

Decisions, decisions - no precision

The Ae sounds very different from the other two, it does confuse the issue.

laters

[Martin213]

Andrew,Nick,Paul,anyone :o

Could someone please tell me what a GM70 is,and explain,in layman's terms,the difference between a GM70,845,and 211 :confused:

Many thanks,in anticipation :o

Cheers

Martin

[andrew ivimey]

GM70s are direct heated triode (DHT) transmitter valves from the old soviet union. The last thing I heard about them was that they were invented as pulse transmitters to send morse signals down the length of the trans siberian railway. As that is quite a long way, er they needed to be powerful. GM70s have one draw back. The heaters take 20volts at 3amps. You can't use ac, which is very unfortunate. B+ can be quite happy at 1500volts. They have enormous potential for power at those hair raising levels that Paul and Nick seem completely at home with.

The ordinary and bog standard graphite anode GM70 costs less than £10 each on eBay. The copper anodes are not quite a rare as hen's teeth and are reputed (by e.g. Paul and Nick) to be a superior valve. I have been surprised to find there is a difference in graphite ones too.

211s as NOS GE VT4Cs are beautiful, silky smooth, less power and happy with about 1100 volts tho' Philip @ Bluebell has a cunning way of running them at 1500volts, too. F@#~in* awesome! They are very easy to drive and only use 10volts for heaters - this can be easily achieved using AC (as I do in my ongaku - which is superb!)

845s look like 211s and are somewhere in between 211 and GM70 in character. But getting NOS (or old good, but used, is much more difficult) I bet they are superb, but have never had the pleasure.

Shuaguang do both 211 and 845s. We have had some pleasure from Shuaguang 845s, haven't we! They are very cheap and some would say, sound very cheap too.

But you really can't beat the best DHT single ended. full stop.

Sure, each DHT valve has its signature, that's why tears come to some peoples' eyes over 300bs. 2A3 and 45s have their place, but at the top, in terms of power and excitement (or is it anxiety) the transmitter triodes have it; and they can drive modern difficult speakers.

(We are hoping for great things from a powerful GM70 setup from Nick to drive the notorious ESL57s of Colin. My speakers at home are such an easy load and 93dB sensitive they make rubbish sound not bad so Am happy to give this Tractor an airing).

I have probably missed quite a bit out, but building transmitter SE amps is like electric lego; dead easy.

Getting it right is something else :rolleyes: :D ;)

[Martin213]

Thank's Andrew,
I think i understand :o

[Martin213]

Found This:

www.yarland.com/en/products/66s.htm

Makes interesting reading :rolleyes:

[pre65]

Hi-see the ornec site

http://www.ornec.com/product/GB/100163/

see they have cut the price,but out of stock !

Philip

[NickG]

INteresting that one site says 9w at 5%, which I can believe, 20W max output. And the other site 23W RMS, which I dont.

[Martin213]

Still sounds gorgeous Nick :)

Where do you think the problem liath? :rolleyes:

[NickG]

Well, I think the problems above lie in the people selling them thinking that 9w doesn't sound much, so making up a reduculus figure to try and sell them. Nothing wrong with 9w, lower than a 845 can do, but not in itself a problem.

I think the fundimental design of the amps fine, just cost cutting means that where we would start to think about a pair of 845 transformers at £200, the entire amp is that much.

I think the obvious thing a DIY amp would do diferently (if I remember what people found before) is using transformers with bigger core's and with a higher load impedance.

But if it sounds good, it can't be half bad.

[Global]

Mine outputs 16w at onset of visible clipping. This is about what you would expect from the HT and circuit used.

It sounds good as standard and much better with all the mods already described. In order to get power at the expense of distortion, the O/P TX impedance is quite low and the sound is improved by using 4 ohm tap with 8 ohm speakers.

If you are going to buy one, remember that the HT rectification is DANGEROUS and MUST be upgraded as described elsewhere.

Edit - here! http://www.world-designs.co.uk/forum/showthread.php?t=32


Mark

[andrew ivimey]

To hell with those chinese jobbies. They are crap - start your own thread! ;)

Now, according to my maths which (is highly suspect) 3.5k should just be acceptable given that I am using only 540 volts but fabulously clear and detailed as it was, there was distortion - and where does this come from, I wonder.

So I have gone back to the fabulously cheap and cheerful Hammond 125FSEs and have them connected so I can play around with 5k and 10k. So much more coming my way. Must have a serious listen.

At 5K 'Anthem of the Sun' HDCD sounds lovely (the AE 3.5k c cores can go back into the mix for 300bs etc; hey, what's the point!)

Perhaps 10k will sound better. (later; it actually sounds the same but a couple of dB quieter but then I can then whack up the B+, 1kV here I come, but that wasn't the point.

5K is the starting point, not 3.5k and I wish I had some really decent 5K+ trafos.

More anon, but let's have no more of those chinese junks (on this thread).

But then maybe this is boring. Those who know what I am talking about, have the T-shirt already. What I am trying to get across is,

THIS IS EASIER THAN PP AND IT SOUNDS SO MUCH BETTER.

Even with this (faintly daft) lash up the huge potential is there and audible.

[NickG]

Yep, I have just been listening to the output stage I built this weekend. They do sound very good. I am tempted to say that they have much more of the sonic signature of 845's than 211's, thats not to say thats bad thing, its just a thing.

I have been using the Cu ones, I had forgotten what a nice colour of light they emit. I have a couple of things in ther amp I need higher power version so before the amp cooks itself. But unless something bad happens it will be ok for Eggfest.

Despite the reduced power, I think 10k is actually a nice load for the gm70, you can get loads of voltage swing given 1kv or so.

I think it should be interesting to see how they drive the statics.

I had forgotten how much heat they emit though!!!

[Paul Barker]

Yes last time we tried gm70 together we thought it was OK sounding, different to 211, but we prefered smoothnes of 211.

I think we should try really cooking the gm70 like that bloke on Yahoo GM70 group. Can you remember his operating point?

We should be able to achieve it with my 1k5 supply.

That would return Eggborough to it's roots.

[NickG]

Yep, that was my thought Paul. Your James TX's have the current capacity to do that if we have a variable 1k5v supply.

Should be able to find the op point from the archive, I will take a look.

Edit: Found it 1250v@175ma

[steve s]

Now, according to my maths which (is highly suspect) 3.5k should just be acceptable given that I am using only 540 volts but fabulously clear and detailed as it was, there was distortion - and where does this come from, I wonder.

.

Andrew, did you change the cathode resistors.. the grid volts would drop with the voltage... then it would distort as you turn it up..

I am having some fun getting my transitter valves going correctly.. I too could do with 1k power supply... its probably the easy option..

steve

[andrew ivimey]

Hi Steve, thanks for this suggestion. I have (not very methodically) done what you suggest, including cutting the cathode resistance down to 1 ohm and variously upwards to 1kOhm. I should have measured voltages and noted what I was doing. Distortion is directly affected by this, you are right.

Right now, bearing in mind that B+ is 540volts and the output trafo primary is 5k or 10k, I have splitter resistors on the heater of 22ohms with cathode resistor of 560ohms and cathode bi-pass cap of 47mfd. It feels right. Of course the 22ohm resistors even bolted to aluminium and 25watt rated, they get bloomin' HOT. A gentle breeze from an undervoltaged computer fan takes the heat right out though!

Using 10k tap only drops the output by a couple of dB, maybe 3 at the most.

It would be so easy to just whack up the potential to the 900s, oh and maybe a little more, and use 10k with no thought to other possibilities. But I need decent 10k trafos and this project was just to be a junk box amp; cheap cheerful and showing that a transmitter valve amp was easy and safe to build.

I have started to consider how to build a vesion of this amp in a neat way thatis acceptable, safe quite a bit smaller than it is already (about a third teh size of NIck's 211 breadboard) Double decking has to be the way.

[Stuart Dean]

Steve
I recently found a 1250-0-1250 xformer for someone and even more recently a low voltage high current jobbie for someone else.
Would you like me to see if I can find something similar for you.
Can't make any promises but I do occasionally come across iron like this. Last year I had some 3ph 415v control gear carrying a load of around 25kw blow up about 30cm from my face when it tracked across a faulty
isolator.Couldn't see or hear properly for nearly two hours afterwards but the molten copper missed me.
I vowed then that I would never have more than 500v in my living room
STU

[pre65]

Hi-the hammond 10K transformers that Phil (Bluebell) sells were developed with his help by Hammond with the 211 in mind and are £75 each (+VAT from tomorrow).

Philip

[andrew ivimey]

Yikes! Stuart, I appreciate the offer but if it involves such risks, I am happy without. Actually the HT isn't teh problem. I would like some excellent 10k output trafos. I have audionotes but they are very much tied up in Ongaku.

I have the hammonds that give 2.5K, 5k and 10k but their bass response isn't that hot.

As Phil suggests, the hammond big brothers seem a steal at £75. Now I wonder why Bluebell's phone is very busy!

[steve s]

Andrew, dependant on the plate resistance of the valve the grid volts can vary dramaticly.. even on new valves there is variations on auto bias..

i would expect the gm 70 to have quite a high impedance, in line with what i am using.. so the grid resistor need to be quite high..

1 was using 1050 ohms with no by pass cap and getting 17 gridvolts at 300 volts.. i have upped the voltage to 440 and with the same resistor getting 23 grid volts that is causing slight distortion at flat out ( i am under driving the valve by quite a margin)

I'm going to up the grid volts tonight, just enough, then a bit more.. to get rid of the distortion.. trouble is you then draw less current and suffer quite a loss in output.. hence more voltage is easy.. except that the power supply becomes more of a pain.. and i cannot use the old u52's and the like, which sound so nice...

I had quite a bit of fun with the grid resistor, changing that has made quite a variation in the sound.. the lower it is there's more bass but a rolled off top end.. higher sound's much brighter..
it set at 470k at the moment.. seems balanced



I'm not one for maths, I too and prefer to suck it and see..

Stu, thanks for the offer I 'm like you really.. less than 500 volts is ok for me
I don't really need the volume at home..
now a nice 425-0-425 with a couple of 6 .3 heaters and the 5 vt rec heater
would be ideal... but I doubt of you have one like that..

cheers steve

[andrew ivimey]

Okay, just had a long chat with Philip@Bluebell.

He's sending me a pair of his 10k trafos - designed by him and hammond to work happily at 1500volts (for his 211 kits - which are coming on), so they should do me fine; £75 each, not bad!

right next, interstage trafos courtesy of AE ...

[Martin213]

"To hell with those chinese jobbies. They are crap - start your own thread!
More anon, but let's have no more of those chinese junks (on this thread)."

Sour grapes Andrew :p

Just 'coz your amp' went tits up :rolleyes:

Hi-Fi snobbery i say :rolleyes:

For all you know i cudda breadboarded the Yarland by now......not that Paul would approve if i had :D

Kindest Regards

Martin

[Martin213]

(tongue firmly in cheek ;) )

[andrew ivimey]

whaddaya mean, '... your amp went ....' N.B. it's not my amp.

Nevertheless it probably could be improved.

That's not really snobbish, on my part, is it.

GM70s deserve their own thread though; they are the valve equivalent of
Denoon 103 cartridges, if one were to liken GEvt4c to Kontrapunkt.

So, AEs did not go down well. My theory is that using interstage I should go for fixed bias, but can't be bothered so the audionote PIOs are back in circuit. Why on earth are they so heavy?

Carmina Banana on Radio Three - coming across nicely but I can hear the percussion is sadly compressed, pity.

[James D]

Andrew,

What voltage do you need to drive the GM70? I've got a crazy idea :-)

James

[andrew ivimey]

Hi James, crazy can be good!

Things start to go wibbly at around 450volts; its all to do with this linear thingie. Above 750 volts (on the last outing) and they begin to sing. 960+ and they are happy. And then over to NickG. I went to 1100 volts once but only to see what was there.

So 450 volts .... how crazy do you want to be?!

[NickG]

http://www.lurcher.org/nick/audio/GM70.pdf

And A2 curves (just in case)

http://oldradio.qrz.ru/tubes/russian/detail/gm-70_2.shtml

[James D]

My crazy idea was to use the 125FSE as an interstage to drive the GM70. If you use it in 2.5K:16R mode it is a 12.5:1 step down. Using an EL34 in triode mode and 600V Va yields about +/-400V swing on the primary and +/-32V into the grid. But it doesn't work - cos it only generates 2W out from the GM70 with the 10K load... ANd that is less than the EL34 is generating to drive the thingy :)

AHHhhhhh Lateral thinking! If we drive the cathode of the GM70 then the power adds and we get more twice the output! I wonder what it will sound like? We could go A2 but the 10K OPT only takes 90mA and we would need more like 350mA...

Sorry it was too crazy...

J

[NickG]

I have had similar crazy thought, the A2 curves look as it you should be able to do something. I thought about 5k load with light bulbs, and then parafeed, but then you need about 2.5k B+, so that becomes a problem, I could maybe get to 2k, but I haven't got much in the way of caps that would live at that voltage. I could get further but that would need to pinch bits from the 211.

Just thought, that marine power supply would do the job (no stoppit, go and listen to some music)....

But then you may as well do it for real with 833's. BTW James, I think these here HF heater supplies I am using for the gm70 would make the 833 easy to heat, if that helps at all.

Maybe at eggfest we can see what the gm70 sounds like with its plates glowing, meant to sound good with about 200w diss. I have just ordered another quad of the graphite ones, just in case we wear a couple out :-)

[andrew ivimey]

where angels fear to tread!

Oh, I like the ideas and am happy to stand back and watch.

I hadn't been thinking along those lines James but interesting nonetheless.

How about running output transformers in parallel; still need several though - don't know what I am talking about really.

The only time I had plates glowing frighteningly it didn't make the amplifier sound any better. I don't think anodes should glow ;)

I am spending a couple of days just listening to the music. My brain can go on thinking at a deeper level, and work begs my attention.

[Paul Barker]

Actually I agree, don't normally hear sound improvement from raised voltage, but there are exceptions. The 813 sounds better and better as you get up to 750v triode mode, you can't go much higher than that triode mode.

Nick and I just remember a conversation on GM70 forum about running them with good results up there. Col needs high power so gm70 thrashed seems a worthy experiment.

I run my 572b's well over rating as shunt regs with no detrimental sound effects, valve may not last long but it's cheapish and I don't have my equipment on very often. I found parallel shunt valves a failure, one valve hogs the current and goes pop.

[andrew ivimey]

Problems, problems, problems - well they had to start!

I have done a bigger base board because the new Hammond10Ks are too bg to fit on. The Hammond 1638SEAs are nearly as big and reassuringly heavy as my Audionote trafos.

They sound good, authority and a nice overall sound. Again, this is reassuring. They are quieter than either of the other trafos I was using at 5K (understandable) or at '10K' Hammond on which you can vary the primary winding.

I think for £150 the pair they are well worth it. See what yo'all think at Eggmeet.

Anyway, as I'd taken the plunge by actually spending some money I thought I may as well start to rack up the B+. But I am having problems.

Silicon rectifiers of choice are 1N5408, rated at 3amps. In classic configuration e.g 700-0-700, windings of 0.3amps into classic CLC the
1N5408s keep blowing. So I have rigged a 5R4GYB (what size cap should I use? I don't have a spec sheet for this rectifier???) and of course two 5R4GYBs woud be better than one.

This gives me only 700volts and all is fine. I think the amplifier is a tad louder - not at all sure yet.

Why do my diodes keep blowing and should I then double them up by paralleling them and ditto if I decided to go for broke and use a bridge!?


(A step too far, though, perhaps, so I could go back to my 450-0-450 trafo for that)

[NickG]

I am using 725-0-725 into UF5408 and its fine, but I am using 6 of them in two sets of three in series. Remember with your TX, a single diode will have 1400v RMS across it, which is 1980v peak. So you may get away with two in each leg, but its close, and all you need is for them to fail short circuit and you have a mess on your hands.

Remember to check your cap rating and make sure they have balancing resistors.

If you feed that TX into a bridge, you will get the same peak output 1980v, which will be fine as long as the rest of the supply is expecting 2kv...

[James D]

As Nick says, you have the full secondary winding volts across each diode so 700+700 Vrms =1400Vrms multiply by 1.4 to get the peak voltage and you have Nicks 1960V add 10% for trafo regulation and you have 2156V peak at startup - pop goes the diode as it sees twice its PIV rating. As nicks says you need three in series for this one.

ciao

James

[Paul Barker]

If you decide to stick with the 5r4gy, it is a high voltage low current low value first cap (10uF from memory) rectifier. No problems with your voltage, and if you use two one per half of the winding both plates paralleled you double the current capacity (think it's 250mA per valve at the maximum voltage of 1,000 volts).

The B suffix I think was a little more rugged and can take a slightly larger first cap.

However many people push the first cap boundaries a little with no ill effect. I don't like to, in any case I mostly use choke input.

[andrew ivimey]

Yikes! its that bad - three diodes in series? not parallel, okay. So its the voltage peak, not the current thats a problem.

Using ss rectifiers gives me more options so although I can do a pair of
5R4GYs, I am lazy.

I'll try this tomorrow.

[andrew ivimey]

Okay, a pair of 5R4WGAs; those Chatham navy rectifiers. This gives me 720volts. I can't find my spare 5R4GYs.

I found using three 1N5408s worked but caused the transformer to hum. Too much of a load????? That was a surprise but undoubtedly there. Are these rectifier diodes more of a resistance, less of a resistance???

The valves cause no mechanical hum from the transformer. So I suppose I will stick with them.

And another thing, ... these output trafos are tapped at 40%. What kind of a pentode goes for 10K (KT88??) but if I used the tap for anode would that give say 6K primary? Or is that naughty?

You see, the amp was louder at 5K and B+ is still only 720volts - just a thought.

p.s. looking at the schottky KEL84 thread casue me to look a little more closely. I was using
UF5408s (and getting through a few!)

[pre65]

Andrew-looking at the Bluebell web site it says that the 1638sea does NOT have the primary screen tap !:confused:

Philip

[andrew ivimey]

Hm .... it does doesn't it. There must be an explanation even though the trafos are clearly marked as what they are and their spec.

I wonder if the tap is any use.

[James D]

Hello Andrew,

You can use the tap and it will give you 6K impedance - just be careful to insulate the free ends very well as they may jump up to around 1KV or more...

The much much lower dynamic resistance of the ss diodes will cause the charging current to peak higher and therefore be of shorter durations thereby greatly increasing the magnetostriction (http://en.wikipedia.org/wiki/Magnetostriction) or Joule effect and making the transformer humm... as it pulses every 10 milliseconds.

Fullwave rectifier and cap input will give you 950V or so :)

ciao

James

[pre65]

Hi Andrew-i am watching this thread with great interest.
i joined the GM70 web site some time ago to get some ideas for a circuit,not being able to design one myself but found nothing suitable.
what you are making sounds like it will be a fair performer and affordable.

i like the idea of 700 ish volts (1000 + might frighten me):eek: ,£75 output transformers and valve rectifiers.

PLEASE publish the circuit when you have a finished product.

what i would like is a fairly good basic amp (transformers etc)with relatively inexpensive components to start with (like chinese 845 ?) knowing that when finances permit it can be upgraded with "boutique" components.


Philip

[James D]

Just did some quick calculations.

At B+= 760V, Va=700V Vg=-60V Ia=90mA, staying in A1 you get 8Wrms output with a 10K load. Going hard into A2 up to +40V drive gets you 18Watts output.

So I would stay in A1 and use a 12GN7 in pentode mode very low impedance drive RC coupled to the GM70 and laugh all the way to the bank with 8 good watts of DHT power:)

You can always upgrade with a parafeed IT later and go for broke with A2 drive as well. The 12GN7 is more than up to it.

ciao

James

[andrew ivimey]

Philip, I wondered if that was you. I only joined last week and saw a couple of other peeps you might expect on the boatanchors.

Are you going to Eggborough? Nick will have his respectable version and I intend to bring this. I think it is shaping up though last night I thought it was rubbish - probably all part of the process.

Using 450-0-450 and diodes, I got a happy 520 volts or so (CLC) and at 5K output trafos it was loud enough for me.

This version is louder (though I still don't need 10K whether it is more linear or not) but I think the qualitative change has been the trafos. Nick has amorphous and that could be better. Paul is bringing some too.

GM70s are difficult to heat, though Nick has an excellent solution. Heath-Robinson comes to mind concerning mine. There must be a better way than all this heat - I have never gotten used to heat but with valves it is endemic but OTOH&bloodyell my Audionote Ongaku mains trafo only gets warm and stays like that for days. That is how it ought to be.

211s are the dead easiest to drive. What could drive 300b will drive 211. Am I going too far? I don't think so. There are many options for front end and if you use input and driver valves, again, it is easy. Anode choked 6S45PEs do a brilliant job and this is the best I have heard them. I am told it is easy to get a better sound though - and why not! 10volts heating produces no hum on AC and AC is to be preferred - I have found it sounds better - I still have a pair of shuaguang 211s. I was so pleased with them at first; they are as cheap as chips (though more expensive than these russian exiles).

And finally (in this posting ;) roughing up a circuit that works, on a breadboard, using such chunky components, ugly and domestically unacceptable for several reasons, this kind of building really is 'electric legoland' and so much easier than say KEL84.

[andrew ivimey]

All sound advice James. I had reckoned on about 8watts too but blimey -what's the point if all the regular folk down there are getting that from their 300bs - answer - it is a very different sound!

You really like 12GN7s, I know, and last year I did too. I've got a few and this tractor started off with 12GN7 and they sounded awful. Rather than think why I rushed straight for 6S45PEs and they sounded fine - must do a spud ... one day.

I think this heap needs a serious tidy up. I have found the RCA 5R4GYBs which on paper give me a little more than the potato mashers. The latter sound fine though so they can stay over the weekend.
I wonderwonderhow&why&what, next.

[pre65]

Andrew-i did manage to get a copy of a 211 circuit from an "Audion" magazine (in Italian due to a mix up when ordering) which uses a C3m as the driver and mixed that with another circuit for the power supply using diodes and valve rectifier for the B+ (920v) but when i priced up all the parts it was more than i had to spend.
to be honest i would really want more than 8 watts because i found the 300b S/E a bit lacking with KLS3 where the Quad 11's with 15 ish watts make them sing.so 15-20 watts S/E should be my ideal.even the Adire HE10 seem "better" with a few more watts.
i think the 300b S/E will have to go now,which is a shame but need the dosh for this next amp project.

Philip

[James D]

Thats the question I ask, why GM70. I don't like the sound of WE300B although some of the 300B variants are nice (Like VV32Bs or AV32B sor KR300BXLS). So its the different sound that does it, and then is it different from the 211? Does it justify the heater current?

The other point is it is much cheaper to buy the GM70s than any of the nice 300Bs...

ciao

James

[andrew ivimey]

Philip, there is a huge difference in sound quality between PP and SE. Some people actually don't like PP ;) gasp!

Now, as I know adires well, I think this GM70 for all its strengths and weaknesss would blow the cones out. The bass in e.g. the soundtrack to Minority Report is, well, frightening at about 86dBA.

It sounds beefier than any 300b I have yet heard. I have four different ways of driving 300b and now have 32bs being driven by pentodes - not bad. Now I have heard better, I feel that 6SL7 Mu or SRRP 6SN7 and even 6S45 driving 300b are weedy; nice but no welly. There are more variables so this should be guarded.

The heater current gor GM70 is a touch demanding and a little irritating, I admit.

And it is a dreadful mess. If I could get a sound like this on a small board - that would be something; double decker perhaps.

[pre65]

Andrew-my 300b S/E has a 6CG7 driver valve but i just didnt think the sound was to my liking.
it was built in germany for a chap in Hythe,but he preferred his old Quad 11 s with his klipsch speakers.if it were my only amp i would be reasonably happy with it via the Adires but i really feel the KLS3 are a "better" speaker in my surroundings.
i would love to come to the "eggfest" to meet you all and listen to your super equipment but my wife is quite ill and i am looking after her 24/7.
if she responds to her treatment as anticipated i might still make it.

Philip

[andrew ivimey]

6CG7 is I think the same as 6SN7gt, just not in an octal format. Perhaps then, not as much welly as you'd get with a pentode. The thing about PPs, IMT,humble,O is that they give a 'wall of sound' er, sound and I perefer the 'three dimensional quality' and a tad more subtlety that you can get from SE - its a bit more than not being as loud!

Anyway, Philip it would be good to meet. Eggborough is a thoroughly enjoyable day out and you would have a warm welcome, even though some would say, we (well some of us) are a bit odd!

[NickG]

I wouldn't expect too much Andrew, my gm70 is going to be much the same as yours, different driver stage, but other than that (and the HF heating) same sort of thing. I have tried to build it so we can pull it apart and use Pauls TX's and power supply, and the cathode resistors are in a string so we can use clipleads to get the value we want for different B+'s

[andrew ivimey]

This has occurred to me too Nick, hm, that's because we're both onto a band of gold, here! And you have the coppers.

I have messed with cathode resisitors, even none and one ohm. less power, more distortion so settled on 560ohm with 47mfd as a bi-pass. Maybe this is right.

I am using 22ohm resistors as splitters on the heaters. Although they are 25watt ally ones, they get very hot indeed. All this heat is a bit weird but may be related??? I also tried 47ohm; very hot indeed, may get around to 100ohms. Surely then less current would go through them and the heaters would happily draw the 3amps they like, the resistors would be cooler and the toroid would be cooler - or is this wishful thinking?

[NickG]

Well two 22R resistors, would pass 20/44 amp between the two ends of the fillaments, so that a extra half amp the heater supply is having to supply, If you add the current the valve is passing (say 100ma) that means one is passing 600ma and the other 400ma, so the center will be about a volt away from the actual center, and one is dissapating 8W and the other 3.5W, so not that much, but quiet a bit. I am not sure anything is gained with DC heaters from forming a center, I would just connect the cathode resistor to the negative fillament and add 10v to the grid voltage when drawing load lines.

On the subject of lines, I would measure the voltage ont he cathode resistor, and calculate the current and try and see where the 10k load line would take you, I suspect that will help finding a point with best output and lowest distortion.

You may find if you have run with no cathode resistor that the current has been limited by the rectifier you have been using.

[NickG]

Been playing with the gm70 a bit more today, 1kv anode, cathode resistor 640R, giving 115ma and just over 100w diss on the plates. With 10k TX's I am getting a tad over 24w before I go to far into A2 for the driver to cope. I think with an extra 200v B+ it would work well. Doesn't sound bad either. Not a delicate as the 211, but with much more authority, not far from a small version of the 212 at eggfest (not that good, but with similar clout), I think with 1200v B+ and 7K transformers it should have enough to make Colins statics take notice.

[andrew ivimey]

Blimey! So I started to get methodical about it and measure a few useful things like plate current - it is all so depressing because none of the measurements are in the right ball park.

Which means although it 'sounds pretty good, actually' it ain't right.

The hammond transformers give it a solidity on which to build but right at the other end the XXX'd power supply is all wrong. If I go back to safe and solid 540volts or so, I will not get that magic; mayaswell use 300b.

I know roughly what I ought to do but I think I have to start again (and stop guessing). This is the problem with starting from an idea born out of (only) previous experience and having a pinch of luck. For all the mess it just isn't worth it.

time to eat and have a pleasant bottle of wine - tomorrow is another day.

[andrew ivimey]

B+ = 750volts (back to SS rectifiers)
GM70 cathode resistor = 1k, 80mfd bi-pass
no divider resistors on heaters
10k primary
plate current; 35 ma (dead weedy)
grid voltage; -10volts uh!

it sounds respectable, loud, in this small room and there is bass, middle and top but does it make much sense?

anyone for fixed grid bias?

[James D]

Hi Andrew,

Interesting! These voltages don't make sense to me. If we assume Va is 650-700V then Ia=35mA yields a Vgc of -70ish volts yet you seem to have -10Volts. I guess we should ask relative to what is the voltage measured but I can't see how you can get -10V. I 'm sure thats what you measured just not sure what it is telling us...

On the other hand if we pass 35mA through a 1K cathode resistor that put the cathode at 35V and if grid is at ground then that yields 180-200mA Ia!!!

Lets assume that grid is 10V negative relative to ground then we -45V bias and that gets us around 140mA Ia...

I assume the heater is floating? if it isn't then we may have a mix of fixed and cathode bias - actually we may have a funny situation with strange currents flowing and various components overheating that shouldn't be...

I'm really not sure what is going on with this amp. It seems to outside of normal explanations...


Nick, anyone any idea?

[Paul Barker]

Yes something seriously wrong. Since it's listened to and sounding ok we know that it isn't something really stupid like grid not referenced to ground at all (I've known it done).

Valve is massively underperforming, is the filament voltage 10 instead of 20?

Is it by mistake a 10k cathode resistor?

Is it connected inverted?

At 750v if you use one of the 1k resistors shared you should get -60v bias 120mA that would be interesting.

[andrew ivimey]

okay, I've had breakfast. I'll have another look but even in my own befuddled way, what you have said James is the merrygoround I've had, thoughts wise.

And Paul, I know, I know.

Heaters are 20volts d.c. 1K cathode res. (with 80mfd cap; not important at this stage) The heaters are not floating as this produces no sound at all.

Grid voltage measurement is between grid and earth and last night was unchangeingly -10 volts, which is silly.

weird, then.

[NickG]

Andrew: OK, first, you need to get that grid to 0v, what value grid resistor do you have?, I am guessing you are cap coupling from the previous stage.

1k was about what I first used a couple of years ago, should be fine, and give you about 75v on the cathode, and so 75ma current.

Your comment about the fillament not floating worries me, any chance you could sketch out just what you have, there is something odd going on here. If you have grounded part of the fillament, that would make sense, the voltage on the grid could be from grid current with the cathode at 0v.

Paul: I found last night, that the graphite gm70's do start to sound better with more current, going from 115ma to 130ma opened the sound out in a very good way, for a lash up its sounding rather good. Or was last night, may sound naff today :-). Only problem with the higher current is I hit A2 sooner, its more than enough here, but I think for the statics it will need a bit more B+.

James: I think I am going to have to try your cathode drive suggestion. I don't think the GM70 would work as a grounded grid, but with the grid voltage regulated by a shunt supply that is happy to sink 10-20ma of current it should work.

[andrew ivimey]

Hi Nick, so this morning ....

Coming out of the coupling cap from the anode of the 6S45PE, there is a
300k resistor to ground, the signal goes straight onto GM70 grid. Raised eyebrows at 330K? - I have tried 47K, 100k and now 330k - it sounds best.

Now, I found this morning that with the volume control of the passive that feeds CD into 6S45PE the GM70 grid is at 0 volts. By turning up to about 1 o'clock I have -40 volts. With the volume control full on the grid voltage is back to zero.

the phone rings ...

[James D]

If the heater is not floating we have a big problem. With cathode bias the heater needs to be 'up in the air' ontop of the cathode resistor so the heater has to float. If one side of the heater is tied to ground then the depending on which side, the mid point of the cathode is either at +10V or -10V assuming a 20V heater suppy. This is locked by the ground connection. This should cause very high anode current at 700V Va too - it isn't implies a lot of negative volts on the grid.

Now we have this funny volume control causing negative volts on the GM70 grid... don't know how but this seems to be controlling the anode current of the GM70... But where is the negative voltage coming from? And we have a cap in the way here to???

I'm very confused as to what is happening here - intrigued but confused :)

James

[Paul Barker]

First let me correct what I cocked up. Don't share 1k across cathodes, use 500 ohm per cathode, or share 250 ohm.

Second, if you were measuring current across th cathode resistor with the cathode grounded at one end you were not measuring cathode current, that must have been up there in the 250's surprised the gm70's didn't indicate this by going red. No doubt the limiting factor would be the rectifier.

In short it's wired up wrong.

As James says let filament supply float, attach one end of cathode resistor to - of filament supply other to ground, same for cap.

300k grid leak resistor will cause bias alteration too, back off.

[andrew ivimey]

I'm just confused!

maybe the heater is 'floating' and I am as above. I thought 'floating' meant that the heaters have no connection to ground at all, (like e.g. my ongaku before I put in a circuit to measure current)

the negative side of the heater is tied to ground by the cathode resistor and bipass cap. There is nothing on the + heater side.

All yesterday there was no audible distortioin at any output level. Today there is.

B+ is 830 volts as I am back to using 3 aside uf5408s.

[Paul Barker]

Try 500 ohm per cathode then.

[NickG]

Ok, either 1k or 500R should give numbers that make sense. I wonder if the negative voltage is the result of the 6c45 going off on its own in the mHz region, and you are measuring the rectified result of that.

Even though 300k sounds better, I would use 47k for the moment untill the DC conditions of the valve make sense. It will take very little grid current to get the grid to rise under its own steam. I would be tempted to connect the grid directly to ground untill the DC voltages make sense.

I suspect the power supply is limiting the current. I found that 130ma @ 950v will get the graphite plate a very dull red, only visible in the dark when the fillaments are off, thats just under at the spec limit at 123w but I guess that a lot of the 60w heat the fillament is giving out ends up on the anode as well. (thats with 570R in the cathode)

[andrew ivimey]

As the crowds begin to pack up their picnic hampers, some already slipping away before the excited but satiated folk realise the chaos there will be in the carpark, I suppose you’d like to know what has been happening.

This morning’s hysteria gave way to a few important decisions being made. For some reason the grid resistors needed to go back down to where they had been a while back; 100k. This was good, then, back on solid state rectification it gave a healthy 800 plus volts but the diodes made the transformer protest too much, which it wasn’t doing last night. The 1K cathode resistors had to be wrong too so, what the hell! Back to basics! As Mr Major once said, when your back is against the wall you just have to turn round and march boldly on…’ no more audible distortion.

And then a return to solid state 540 volts with a transformer that promises an HT current of 450ma. No probs with all six diodes and this original transformer. Cathode resistors now 560ohms and it feels right that it should be there. So, the crowd settled into a fine afternoon’s performance.

To be further thrilled, I rewired the old WE91 (6SJ7 at the front and AE C core trafos at the rear) clone so for a little evening’s listening the warm mellow tones of 300bs can be heard in this part of Bedford.

GM70s are louder, even as I have them now, than 300b. I wonder, if as Nick suggested, the 6S45Pes weren’t doing their own supersonickingthing thus causing more probs … I don’t know but things have settled down and the amp isn’t sounding that bad again. I get the message, but Hey! 211s are so very much easier. 300bs are a doddle but I have had times with these too.

Maybe time to tidy up the GM70s and round them off. The circle is nearly completed from where I started. If I had loads of money more I’d try a few more things like a massive HT transformer, but I haven’t, so won’t. Now I can start to measure with a method rather than splashing about…….

[pre65]

Hi-you have probably seen this

http://diyparadise.com/nikcgm70.html

but if not it may be of use.

Philip

[andrew ivimey]

Hi Philip - that was the inspiration or last year's barrow load of ....

Last year the 12GN7 was marvellous, as triode and anode choke.

Last year I was pulling my hair out over how I could get reliable and safe 20volts for the heaters, even did a voltage doubler - it worked but oh so much heat.

The Tractor sounded rough, to say the least.

This year's model has sounded floorboard ripping up amazing and it has sounded poor. ThisY'sM has 6S45PE up to today and I have been pleased but James has been right all along; use 12GN7 but the power of pentode, not tremulous triode, so this weekend I'll give it a go.

Take the back end of Nick Chua's GM70. That is more or less what I am using. I have had probs and hiccups but the 10k txs are lovely and I have settled back down to 540volts. let's see / hear what Nick's high HT does.

Its the boat anchor lot on Yahoo who got me thinking about 813s.

[IslandPink]

The power of pentode, you know it makes sense ..
MJ

[andrew ivimey]

a late night last night - hurry hurry hurry

After a couple of pentode configs with 12GN7 I decided to try Peter Millet's version from his 813 amp.

No more gain than 6S45PE (less distortion), and I found this with other variations. But this one sounds better. There is very audible distortion as I turn the volume up beyond 1 o'clock but imn my room 80dBa is easy to get yto before I can hear it.

I haven't much time to do anything else as I have several very 'pressing engagements'. I'll tidy this version up but I think this is what is coming to eggborough.

[andrew ivimey]

okay, some might say that the Tractor is 'tidied' up.

swapped the audionotes for jupiter beeswax coupling cap too - anyone want a pair of audionotes!

shorted out the 6 volt heater winding until things started to smell - now sorted.

So now 'Apostrophe is spinning on the JVC / Denon - great googly moogly!

It sounds fine, oh for a pair of coppers and some hard work to make it look better, and BE more safe.

Anyway that's it for now - a low voltage (500volts) transmitter amp.

(hope it travels well)

[IslandPink]

Watch out for the yellow snow .
See you at Eggmeat .
MJ

[Martin213]

GM70s are direct heated triode (DHT) transmitter valves from the old soviet union. The last thing I heard about them was that they were invented as pulse transmitters to send morse signals down the length of the trans siberian railway. As that is quite a long way, er they needed to be powerful. GM70s have one draw back. The heaters take 20volts at 3amps. You can't use ac, which is very unfortunate. B+ can be quite happy at 1500volts. They have enormous potential for power at those hair raising levels that Paul and Nick seem completely at home with.

The ordinary and bog standard graphite anode GM70 costs less than £10 each on eBay. The copper anodes are not quite a rare as hen's teeth and are reputed (by e.g. Paul and Nick) to be a superior valve. I have been surprised to find there is a difference in graphite ones too.

211s as NOS GE VT4Cs are beautiful, silky smooth, less power and happy with about 1100 volts tho' Philip @ Bluebell has a cunning way of running them at 1500volts, too. F@#~in* awesome! They are very easy to drive and only use 10volts for heaters - this can be easily achieved using AC (as I do in my ongaku - which is superb!)

845s look like 211s and are somewhere in between 211 and GM70 in character. But getting NOS (or old good, but used, is much more difficult) I bet they are superb, but have never had the pleasure.

Shuaguang do both 211 and 845s. We have had some pleasure from Shuaguang 845s, haven't we! They are very cheap and some would say, sound very cheap too.

But you really can't beat the best DHT single ended. full stop.

Sure, each DHT valve has its signature, that's why tears come to some peoples' eyes over 300bs. 2A3 and 45s have their place, but at the top, in terms of power and excitement (or is it anxiety) the transmitter triodes have it; and they can drive modern difficult speakers.

(We are hoping for great things from a powerful GM70 setup from Nick to drive the notorious ESL57s of Colin. My speakers at home are such an easy load and 93dB sensitive they make rubbish sound not bad so Am happy to give this Tractor an airing).

I have probably missed quite a bit out, but building transmitter SE amps is like electric lego; dead easy.

Getting it right is something else :rolleyes: :D ;)

So,all i have to do now is decide GM70,211 or 845
Yes,i'm considering building one Andrew,the Yarland did it for me!!!
All i need now are the parts to build one :rolleyes:

[Martin213]

Just found this:

http://www.audionote.co.uk/ankoruwritings.htm

[andrew ivimey]

Martin, these amps are only easy to build when you have built one or two (Catch 212), then had a good think about it and learned lots. Using a big chunk of MDF is good because you can really see what you are doing.

211 seems the easiest way to go but here's a thing.... if 211s are so easy to drive why do all those ebay chinese amps use every transmitter valve but the 211?

The best and safest way to start on this might be to talk to Philip at Bluebell for a lovely 211, but you will have to drill, file, cut and spray as well as the tedious soldering, anxiety,stress, electrocuting and leaving a wife and child to fend for their existence without their beloved pa.

I have found another way of driving GM70 which is simple and I am itching ... but I still will keep the B+ down to around 700volts at the max. And then there's the pentode version, GK71 (about £8 each from Anthony welsh) BUT.

Right now I am having a very interesting listen to what spuds on the one hand and 45s can do, on the other. 300b is a constant ongoing trial too. You could get into fostex speakers and cunning cabinets, so you could enjoy the fascinating world of single drivers and single watts.

You know how Paul and Nick got those anodes glowing at 17,000,000 volts. well you can do this with smaller valves and still listen with baited breath.