Direct Coupled tube amp

[Crazy Bill the Eel Killer]

Hello all,

I'm back after a LONG hiatus. I'll explain where I've been in other threads. So let's get going.

My main amp, ( now in pieces ) is an IT coupled design w/ a triode strapped EL84 on the end. I've played around w/ a lot of EL84 designs ( IT and RC coupled ), but would now like to try a DC amp. I know the EL84 is not a good candidate for DC coupling, so I would like to know what output tubes you guys would suggest for a DC amp w/ the following, because I have them in hand.

My OPTs are Transcendar originally designed for for the EL84. 5K primary, 4/8 ohm secondaries, w/ a 35 % UL tap. They're rated for up to 60 ma, so I could use a 45,PX25, or 2A3 w/o issue. But all of those tubes are expensive, and with a grandson in the picture now spare cash is at a premium ( though I do want to try DHT's in the future ) . I have 3 matched pairs of black plate 12b4A in house, with which I will build an EF86/12b4 DC amp, but I really want something with some more guts to try.

So I know you guys have a lot of experience, so what are some output tubes you guys might suggest that would work w/ my present OPTs in a DC config . Feel free to suggest driver stage tubes also. My PT will do approx 290-300 volts DC ( a little less with a strong driver stage, at up to 150ma ( I've got enough iron to do a CLCLC Power supply, so it should be quiet. I've got some CCS's laying around, but no chokes ( but would be willing to source some ). I'm open here, and hoping there's something else out there besides the DHT candidates to lower the costs.

However, that said, if the DHT avenue is truly the only way to go, so be it and let me know. My speakers are single driver units based on a highly modified Fostex FE206E, and that is definitely not changing ( except for a possible upgrade to a Lowther in the future ). So 2 watts or so is all I need ( unless you guys convince me I need more ).

Please remember that I'm a newb, w/ no experience at all w/ DHT's. Complicated stuff need not apply, I'll leave that for you guys that truly know what you're doing, cause I don't.

As always, thanks for all of your help.

Cheers, Crazy Bill

[Mike H]

Hello,

Silly question time ~ why is EL84 not good for DC coupling ..


 

[Paul Barker]

i am not aware of problems with el84.

But 6em7 lends itself.

But the cheap DHT valve with which you could build one of the many DC 2a3 designs is the 6b4g. It sounds just as good as the 2a3 and is identical but with an octal base (which is easier anyway) and requires 6.3v filament supply instead of 2.5 volt. Another cheap alternative the 6a3 has the jumbo 4 pin base but needs also 6.3v heater.

I can sell you a pair of 6b4g's but no pressure.

If you don't mind the lower power the ST shaped 45 as still a cheap valve and there are plentiful dc designs , can't remember if the Morrison micro is dc, but it's simple and worthy.

The 45 and the 6em7 both have the advantage of requiring lower HT voltages, but are lower powered.

I can sell you 6em7's 6b4g's or 45's no problem but not trying to do a hard sell just want to help.

[Crazy Bill the Eel Killer]

Paul,

thanks so much for the offers. I've only had very limited time to look at those tubes, but initially they look promising. I'll try to take a good look this weekend, and if I decide to actually use them in a design, I will most definitely procure them from you.

Paul and Mike,

As I stated previously, my knowledge is limited, so the following statements are not made with authority in mind, especially as I could no longer find the original threads from several years ago on a different forum where the issues of a DC amp were explained to me in great detail. So I'm relying from memory, which in my case is usually not a good thing, but here goes. If I get it wrong, feel free to correct or disagree.

As I understand it, the grid voltage on the output tube is set by the plate voltage on the driver stage. So, let's say that plate voltage drifts 2 volts for whatever reason. While that's certainly insignificant with respect to the driver tube, in the case of the EL84 ( typically biased at -9 v ), that equates to a bias shift of approx 20%, which again, as I was led to understand it, is not good thing. I was led to believe that tubes which allowed greater voltage swings on their grids were better candidates for direct coupling, as the bias drift would have lesser consequences.

That makes sense to me. As I thought about it, with my limited knowledge of what's going on, the drawback that I saw was of course a pretty severe drop in available swing on the EL84 grid, with a corresponding drop in power delivered to the load before severe distortion set in. I don't know anywhere near enough to know if there were other issues involved ( such as instability or oscillation ), and since there seems to be a dearth of DC EL84 designs out there, I decided that the EL84 was a no-go based upon that advice.

That said, I absolutely love the EL84 and would like to try it direct coupled. I know I'm going to need a new PT, but I'll eventually need one for future projects anyway. So it begs this question.

If I build a DC EL84 amp, and the bias slides by a couple of volts, is the only issue the loss of available power, or can I cause damage.

Thanks guys, Crazy Bill

[Paul Barker]

You have posted an interesting objection which I was unaware of.

AS the driver sets the bias for the power stage the simple solution is to regulate the driver supply which is easy top do because of the low currents involved. This might improve all driect coupled amps.

At the output stage the shifting B+ (in my part of the country this wouldn't alter anyway) isn't amplified by the mu of the valve so is not any more an issue with Dc than it is anyway.

SO simple fix VR tubes shunt regulating the HT to the driver, and a worthy thing to do with any amp.

[Nick]

Yep, but hang on, What you say is true if you say it quickly, but just about every direct coupled amp I know of has a cathode resistor under the second valve. So if the voltage on the grid changes by 2v, the voltage across the cathode resistor will rise by almost 2v, so the current will increase a bit, the voltage on the grid will increase at bit, but far less than the 2v you are worrying about, and less than the effect of a 2v rise on the grid if there was no cathode feedback.

[Mike H]

^ Welcome injection of sanity-restored / memory-jogged / back-to-basic-engineering / course-it-is moment. I'll go with that.

I only mention it 'cause I was starting to doubt....


 

[Crazy Bill the Eel Killer]

Thanks for the responses gentlemen, much appreciated.

Paul,

You know, it's kind of funny, but the first thing I thought about after I was told no, was to use shunt gas regs on the driver stage. I scoured the internet, and could find no place where that was tried with the EL84, so I figured there was a reason for that, and didn't pursue it. Now that you think it's viable, I will definitely pursue that fix. I've got Rusky versions of OD3 and OA3 in house, along with a pr K & K audio cascoded FET ccs, so I'll ccs load the gas regs, and ccs load the driver stage to really isolate the PS noise. Now to figure out which tube for a driver stage ( EF86, 6H23N-EV, or 6N1P-EV )
I've played with all three a lot RC coupled, and IT coupled minus the EF86. RC coupled the EF86 won ( 50's vintage Rusky versions, 60's vintage Valvos ). IT coupled was a close call. My 6H23N-EV tubes are a SLN matched cryo pr from the Tubeman, and are unbeliveably good valves. But hit those 6N1P-EV w/ 200v @ 10ma and they really sing. Problem being if I use them I'm at 500vdc for the B+in the DC amp, the other two way less. I'll probably use the 6H23N-EV. 90v @ 6-7ma sounds really nice with those tubes, and they are the quietest of the three.I'll figure it out in the next couple of weeks, as I'm going to have to source a new PT as the ones I have in house will only do 300v with valve rectification, and ss diodes are not an option.

Nick:

I don't have enough knowledge to follow your reasoning on a purely conceptual basis. I'm going to have to print a schematic and actually put some voltages in, change them, and see where it goes. I'll try to get that done in a week or so to see if I follow you.

Again, thanks for your help. I'll be back here after I have some progress to report.

Cheers, Crazy Bill

[Paul Barker]

very bad to ccs load pentodes. the anode load of a pentode stage is the output impedance.

other than that, good.

[Crazy Bill the Eel Killer]

Hello Paul,

My bad, I wasn't clear enough.

I don't do pentodes unless they're triode strapped. When I was playing around with the RC coupled EF86/EL84 ( both triode strapped ), my favorite op point was B+ @ 250v, plate resistor 87-91K, cathode resistor of 1.5k w/ 47uf cathode bypass cap.
Gave me about 115v on the plate @ approx 1.5ma. Very,very sweet.
Took out the plate load resistor and cathode bypass cap and threw in the ccs. Kicked it up a notch, sweeter than before w/ greater dynamic swing and top end air. Not much, mind you, but better.

So I'll try the ccs loaded triode strapped EF86, I really liked it a lot RC coupled. We'll see what happens.

thanks so much for your input and feedback.

Cheers, Crazy Bill

BTW, just curious. You wouldn't, by any chance, be the same Paul Barker
with a spot on Dave Slagle's Intact Audio forum ?

[Paul Barker]

sorry I misread you, yes you know your stuff.

yes Dave gave me a spot, but my interest in valve audio ran out when I started geting sounds far worse than ever some 3-4 years ago.

I have begun pinpointing the problem now though.

[Crazy Bill the Eel Killer]

Hello Paul,

sorry to hear of your troubles. I do hope you get it sorted out soon. I can think of nothing more frustrating than not being able to enjoy my music. I've been going through that sort of thing since just before Christmas when my turntable died ( Mitch Cotter based on the Denon DP6000 ). I can't find anyone able to repair it. I think its a cap on a specific board, but I don't have enough knowledge to figure it out ( and visually it's not obvious ). I've got a friend who makes custom turntables, and he was kind enough to offer his platter, bearing and DC motor drive assembly which I will fit to the Cotter base, keeping the Fidelity Research FR66s tonearm. Should have it back operational in under three weeks now. It's forced me to listen to CD's and HD downloads, which ( I hope I don't offend anyone ) frankly suck. The HD downloads are listened to through a friends very high end system, so its not the gear. I just can't get lost in the music w/ digital which I can easily do w/ vinyl. Can't wait.

BTW, nothing to be sorry about as you didn't misread anything. I was not clear about the triode strapping. I should try to post earlier in the evening before I finish my glass of vodka. Tend to be a little clearer minded.

Thanks again for you help, Crazy Bill

[Mike H]

I just went back to the beginning to refresh meself what this is about

DC-coupled wiv DHT o/p, possiblee....

I envisage [a something] as input/driver, anode DC connected to DHT. DHT cathode jacked up on appropriately valued resistor and bypass cap. If input anode is at 100V-ish, not difficult. Probs need a PSU giving 400 - 500V though really.

If o/p is EL84, most probs manage with 350V-ish.

[Crazy Bill the Eel Killer]

Hey Mike,

thanks for posting. it is getting a little sidetracked ( my fault ).

For now, the o/p will be an EL84, ( DHT in the future, though ) I've got very good iron spec'ed and wound for that tube, so it makes sense to use it ( plus I love the way it sounds ).

350 B+ would be a little on the short side for me. I like to juice the EL84 quite a bit ( 290v @ 45 ma ), so I would be looking for more in the 400 volt range for the B+ to use with either the EF86 or the 6H23N-ev. I originally figured I'd need a new PT, but then I got to thinking ( usually dangerous in my case ). I have a pr of Hammond units 300-0-300 @ 150ma , and w/ sand diodes I should be able to get to the 400volt mark to breadboard the amps.

I've always liked to use tube rectification, as that is all I had and I liked the results, and verbal descriptions I read on the web usind sand did not sound appealing to me. But I quess its time to expand my horizons a bit, especially as I won't have to buy a new PT right away to experiment with. So I'll get some ss diodes and have a go at it.

Problem is, w/ no experience, I'm not sure what to procure. Should I use cheap generic diodes and put snubbers on them, or should I use the more esoteric stuff, say high speed soft recovery Schottkys ? Any help would be appreciated.

Cheers, Crazy Bill

[pre65]

Hi Crazy Bill.

You could use 2 sand diodes and a valve rectifier in a hybrid bridge. Quite a few of us on here use them in our projects.