Hey everyone.
I'm currently thinking of designing bulding a SE amplifier, mostly with what I've got laying around. I don't have any iron for this at hand, but I have a chassis I want to stuff everything needed into, and a load of tubes. What I've got, and have considered, are 6550 Winged C:s, EL34's, some NOS 807's, and some old Westinghouse 6B4G's. Valve rectification with GZ30 or 5V4GT (have some NOS Sylvanias around doing nothing...) was what I thought...
Here's what I'm looking for:
This morning I got the idea of using the 807. But then I read the bugger does a measly watt with a truckload of distortion in SE designs. Is this true? I have'nt dug myself into the curves yet, but I thought a 25W tube could give more power even in triode SE?
I was thinking of using cathode feedback on the OP stage, also. That does'nt give a whole lot of NFB, I know, but I have an angst towards gNFB, based on experimentation. Local feedback I think is OK. Actually, at first I was thinking a 6550 pentode connected with local cathode feedback, but that does'nt bring the output impedance down enough with the Hammond OPT:s, basically regardless how you connect them.
Basic question: which output tube to go with? I already have a pair of KT88 PP monoblocks, so the sound of an EL34 would of course be welcome to my house. Then again, the 6550 would be more linear in triode. An 807 I've never heard... And I don't want to use tubes I haven't got.
Oh, whatever the output tube, I was kind of planning to feed them with a NOS 6SL7 dc-connected to a 6J5 cathode follower...
I'm currently thinking of designing bulding a SE amplifier, mostly with what I've got laying around. I don't have any iron for this at hand, but I have a chassis I want to stuff everything needed into, and a load of tubes. What I've got, and have considered, are 6550 Winged C:s, EL34's, some NOS 807's, and some old Westinghouse 6B4G's. Valve rectification with GZ30 or 5V4GT (have some NOS Sylvanias around doing nothing...) was what I thought...
Here's what I'm looking for:
- At least five watts power out with less than 2% distortion
- Has to fit into a Hammond 20cm x 30cm x 5cm chassis
- Hammond's 125xSE iron is cheap and easy to get, so that's what I'll be using for OPT:s
This morning I got the idea of using the 807. But then I read the bugger does a measly watt with a truckload of distortion in SE designs. Is this true? I have'nt dug myself into the curves yet, but I thought a 25W tube could give more power even in triode SE?
I was thinking of using cathode feedback on the OP stage, also. That does'nt give a whole lot of NFB, I know, but I have an angst towards gNFB, based on experimentation. Local feedback I think is OK. Actually, at first I was thinking a 6550 pentode connected with local cathode feedback, but that does'nt bring the output impedance down enough with the Hammond OPT:s, basically regardless how you connect them.
Basic question: which output tube to go with? I already have a pair of KT88 PP monoblocks, so the sound of an EL34 would of course be welcome to my house. Then again, the 6550 would be more linear in triode. An 807 I've never heard... And I don't want to use tubes I haven't got.
Oh, whatever the output tube, I was kind of planning to feed them with a NOS 6SL7 dc-connected to a 6J5 cathode follower...
Doesn't solve the problem. Pentode connected with plate to grid feedback wasn't what I was looking for. Albeit an interesting approach. I was thinking triode mode and perhaps cathode feedback if needed. And still, the selection of the output tube lies unanswered...
parallel 6B4G.
Or push pull, why not?
The other tubes you mentioned I would not consider for SE operation in most cases.
Consider a "parafeed" arrangement if output iron is a cost problem...
I'd take a class A PP 6B4G over a SE pentode, even strapped for "triode" any day.
My opinions are just opinions.
_-_-bear
Or push pull, why not?
The other tubes you mentioned I would not consider for SE operation in most cases.
Consider a "parafeed" arrangement if output iron is a cost problem...
I'd take a class A PP 6B4G over a SE pentode, even strapped for "triode" any day.
My opinions are just opinions.
_-_-bear
Hi,
You are right about the 807 in triode mode. When you study the very good STC Datasheet you will notice that there is little use not use it in pentode connection.
I used the 807/1625 with local cathode feedback with good results. Here is more
Furthermore you don't need a cathode follower to drive a typical pentode like 807 etc.
And...don't expect too much with those OPTs
You are right about the 807 in triode mode. When you study the very good STC Datasheet you will notice that there is little use not use it in pentode connection.
I used the 807/1625 with local cathode feedback with good results. Here is more
Furthermore you don't need a cathode follower to drive a typical pentode like 807 etc.
And...don't expect too much with those OPTs
It's quite microphonic though , went through my stock of 20 of these and all were unuseable as a linestage tube . If using as an output stage with sensitive speakers (98dB/1w) the same applies , even with damping and viton rings on the tube I can hit the pause button during a loud passge and hear 'cowbells' for a good few seconds . Reducing filament voltage helps but the issue does not really seem to go away completely . I found there is a distinct sweetspot in terms of operating point and I could not get anywhere near the 2.5w you quoted , more like 1.2w with 5k optx to get best sound . I am running a push pull 4P1L amp at work with an old pair of Kef Cadenza and it sounds very nice , albeit at lowish SPLs
2A3 is what I would recommend as an SE output stage . easy to light up , easy drive requirements and very forgiving
Triode connected EL34 may be the best of the bunch on the posters list of available types...316A
Thank you for correcting my mistake.
I wanted to delete my message yesterday, but it was too late. Then I calmed down myself thinking that people who are going to compete against my commercial products don't know how to cook them, to get 2.5W of nice clean power from a tube, with no microphonics. Looks interesting, that design of yours. What kind of output impedance did you get out of that? And what kind of OPT?
Using the CF is only a neurosis of mine; I want a low output impedance driving the output tube if it is wired in triode. I've heard such a design, albeit fully DC-coupled from input to output tube, and it was a very quick sounding amp. And used the notorious 125ESE... Output was an ECC99 with halves paralleled. ECC81 input directly coupled to an ECC82 cathode follower directly coupled to the output tube. Efficiency was - of course - around nothing.
I know the Hammond 125xSE isn't exactly the bees knees, but having used the 125ESE in an E55L spud I think it's better than it's reputation. The only thing is that the gap is too generously sized, which might pose a problem. At 20Hz I started seeing saturation after a couple of watts. The OPT is gapped for 80mA, and I was thinking of using the bigger irons for this project, but they have even larger gaps for more current, which I don't need.
I've started leaning towards the EL34; I have these, they don't need DC heating, fairly easy to drive, but could they give the power I specified in the OP? UL mode should get them there, but then I'd have to resort to feedback... And again, I might be tempted to go all the way to building a fully pentode thing... Having heard a F2a driven by a C3m both in pentode mode, I do feel the urge of trying a pentode 6550 driven by an E83F...
I quickly scribbled together a draft schematic of what I was thinking about. At first I thought of making it fully DC-coupled, but I would've needed too much voltage for the power tube, and would have been dumping around 10 watts of power to a resistor at the EL34's cathode. B+ would be something like 350-400VDC, with around 50mA of quiescent current through the EL34. Cathode biasing makes the thing a plug'n'pray sort of machine, or as the company supplying all the testing equipment to the place I work for would say, "Maintenance free" (meaning that it will fail, and when it fails repair - my job - will be a major PITA).
This actually should, if I looked at the curves correctly, land in the 5-6 watts ballpark. Interesting, since I just finished off a schematic for a self inverting EL84 push-pull amplifier, that should give approximately the same amount of power. That would enable me to make comparisons between similarly powered different constructions.
This actually should, if I looked at the curves correctly, land in the 5-6 watts ballpark. Interesting, since I just finished off a schematic for a self inverting EL84 push-pull amplifier, that should give approximately the same amount of power. That would enable me to make comparisons between similarly powered different constructions.
Attachments
Had a busy day today but I didn't forget this thread
The OPTs are exactly these here: Pair 5k Transcendar Audio Output Transformers for SE amp, 8 ohm 10 watt | eBay
Back then, these were only 90$ the pair and the Euro peaked (1€ ~ 1.5$). So this was an absolute bargain even with shipping to germany. They do their job suprisingly well. Much better than any Hammonds I got so far (1627SE and some cheap ones)
Just calculated with the standard voltage-feedback equation and I get around 6Ohms output resistance. So damping-factor is around 1.5
That's not superb but the amp is used with some fullrange speakers. For sure, it's a different story with some heavy bass-reflex enclosures.
I have a similar arrangement using an EF40 to drive a EL38, both in pentode with plenty GNF. I like the 1625 with local feedback much more. Sounds more lively.
Of course, F2A is a superb tube. Like EL156 and EL503 these were the peak of tube manufacturing. I would always choose these over EL34, KTxx, 6L6 etc. but they cost a fortune by now.
low impedance driving is fine. The question is if it's worth to waste a tube for that. George from Tubelab explains how to use a Mosfet here. Using the small ZVN0545 from Zetex is a really good alternative for a tube.
Fully DC coupled designs are tricky. Either you waste a hell of a lot power in cathode resistors or you need multiple power sources and very good calibration of anode voltages. The PDF titled "Direct Coupled Amplifiers" here tells the story.
Yes you will get over 6W without any trouble with 350-400V Ub. The UL circuit brings the output resistance down to an even lower value. A bit of calculating and you're there.
The OPTs are exactly these here: Pair 5k Transcendar Audio Output Transformers for SE amp, 8 ohm 10 watt | eBay
Back then, these were only 90$ the pair and the Euro peaked (1€ ~ 1.5$). So this was an absolute bargain even with shipping to germany. They do their job suprisingly well. Much better than any Hammonds I got so far (1627SE and some cheap ones)
Just calculated with the standard voltage-feedback equation and I get around 6Ohms output resistance. So damping-factor is around 1.5
That's not superb but the amp is used with some fullrange speakers. For sure, it's a different story with some heavy bass-reflex enclosures.
I have a similar arrangement using an EF40 to drive a EL38, both in pentode with plenty GNF. I like the 1625 with local feedback much more. Sounds more lively.
Of course, F2A is a superb tube. Like EL156 and EL503 these were the peak of tube manufacturing. I would always choose these over EL34, KTxx, 6L6 etc. but they cost a fortune by now.
low impedance driving is fine. The question is if it's worth to waste a tube for that. George from Tubelab explains how to use a Mosfet here. Using the small ZVN0545 from Zetex is a really good alternative for a tube.
Fully DC coupled designs are tricky. Either you waste a hell of a lot power in cathode resistors or you need multiple power sources and very good calibration of anode voltages. The PDF titled "Direct Coupled Amplifiers" here tells the story.
Yes you will get over 6W without any trouble with 350-400V Ub. The UL circuit brings the output resistance down to an even lower value. A bit of calculating and you're there.
Had a busy day today but I didn't forget this thread
The OPTs are exactly these here: Pair 5k Transcendar Audio Output Transformers for SE amp, 8 ohm 10 watt | eBay
Back then, these were only 90$ the pair and the Euro peaked (1€ ~ 1.5$). So this was an absolute bargain even with shipping to germany. They do their job suprisingly well. Much better than any Hammonds I got so far (1627SE and some cheap ones)
Just calculated with the standard voltage-feedback equation and I get around 6Ohms output resistance. So damping-factor is around 1.5
That's not superb but the amp is used with some fullrange speakers. For sure, it's a different story with some heavy bass-reflex enclosures.
The problem is, that even though this amp is not likely to be a permanent part of my main system, that still is where it's going to be evaluated. And it's going to have to drive an open baffle speaker with dual Eminence Beta 15A's per side in parallel, and a Beta 10CX. Not a dreadful load, but it still dips to under 4 ohms. I'm driving it currently with a pair of class A KT88 PP monoblocks without feedback, and even with just 20W they are more thunderous I can handle in my livingroom.
EL34, 6550 and the others I stated in the OP are tubes I have around. Of course, I also have more than 70pcs Philips PL519, but they're going to be used for OTL some day. Not really suitable for a normal SE, I think.
I could also use a bipolar, but this time I want to go with a tube. I have some 20 6J5GT tubes NOS around, and they are just begging to be in a circuit like this. I also have a nice stash of NOS 6SL7... Somehow I want this thing to be all-octal. Incidentally, I had this idea of building a discrete, zero global feedback FET preamplifier with only or mostly SMD components... But this amplifier is going to have no sand, 5V4 or GZ30 rectifier...
Doesn't UL bring the output impedance up from triode? Rp is higher..? Calculating the circuit from what I scribbled up is a breeze, actually, but I just made that to demonstrate what I was up to. Still, the 807 is in my thoughts, but if it's hopeless, then I'll take the advice of those wiser.
Thanks for the input! I still have to evaluate the distortion from the EL34 and 6SL7 stages, but now I have at least some basis for the output tube decision. I'll leave the 6B4G's for later, and another chassis. The one I'm trying to get rid of is just too small for the DC heater supplies. It's just so sad to watch an empty chassis on the shelf. It clearly wants to be an amplifier.
Wooow wait a second !
There is an interesting way to use PL509 and even DC couple the whole amp. Look here. There is another Version with an ECL82 driver around. I think it was from Bob Danielak. I have never heard or built such an Amp. But the screen driven PL509 seems to sound good.
That's OK. Just keep in mind that the output resistance of a cathode follower is ~1/gm. So you may prefer high gm tubes. For a simple driver/CF (no grid current) those 6SL7/6SN7 work well.
Yes it does. But my basis was a pentode. So UL has a lower Rout as a pentode but as you said a higher as a triode.
Rout: Pentode > UL > Triode > "SupraTriode" (article)
The 2A3, I had a marvellous, although short experience. Built with F&S Trading custom outputs, mains..., but ages ago, sorry bad memory. I splashed out on Ansar (RATA branded, then.) polyprops, thoughout incl. psu (fun bill!).
But made the mad as me mistake of strapping the 'cathode' (transfo ct) resistors to their bypass caps.
Sounded great thru my AN-J's on FRED stands, etc..... For about 15-20 seconds!!!!!
...Then it went into a pyrotechnic display before the fuses tripped.
I can very much reccommend the 2A3 from that, but easier and cheapper, just use some EL84 (PSE for more power, can't rmember impedances). They have sounded better to me than the EL34 in SE, they could be found, then for UK £1.55 ea from Sovtek. I used a SRPP ECC82 as gain stage and stupidly enough, no nfb. But they worked a treat!
But made the mad as me mistake of strapping the 'cathode' (transfo ct) resistors to their bypass caps.
Sounded great thru my AN-J's on FRED stands, etc..... For about 15-20 seconds!!!!!
...Then it went into a pyrotechnic display before the fuses tripped.
I can very much reccommend the 2A3 from that, but easier and cheapper, just use some EL84 (PSE for more power, can't rmember impedances). They have sounded better to me than the EL34 in SE, they could be found, then for UK £1.55 ea from Sovtek. I used a SRPP ECC82 as gain stage and stupidly enough, no nfb. But they worked a treat!
Oh, yes. I am aware of screen driving these buggers. But this would imply very high output impedances and local feedback might not suffice to get it down. A friend of mine did give a thumbs up when he heard the venerable EAR 869 using a screen driven EL519. Maybe one day, but it isn't what I was going after this time. A basic SE amp is something I've never heard in my own system, so that might be the reason I'm trying to go with something fairly cheap, basic, and something that could be modified later to experiment with different concepts. I don't remember building a thing that would've lasted a year unmodified. Usually if I buy something, it doesn't remain unopened and unmodified for more than 12 hours. Every project is a work in progress.
I once considered using an ECC81 as a CF stage, because of it's high µ. But usually high-µ tubes don't like current. And I do... Slew rate is defined basically by current and capacitance. And triode grids are caps... Again, a neurosis of mine, but I've usually preferred the higher current designs. Whatever the real reason, I don't know if I even care. My preamp (tube) has a slew rate of 160V/µS (measured), theoretically 175V/µS. ECC99 running at a moderate 5mA per side, the latter half of each working as a CF. I did a tube output DAC with 5687's, never measured it though, but sounded nice. Amazing low end grunt compared to an op-amp based internal converter in my CD player. And a sparkling clean top end.
One possibility, of course, would be running the 6SL7 at a higher than average current in a µ-follower setup. Would give good output impedance, low distortion and be simpler than what I suggested in my earlier draft schema. Might have to take this opportunity into consideration as well.
It seems that I've taken on some enthusiasm towards this SE design of mine. I just brought home a bunch of Hammond iron from the store, and am going to make a working piece of equipment out of this. 372FX power tranny and 125FSE output irons on my table...
If this makes up anything worth mentioning, I'll try to document it as well as I can, so that someone able to sling some solder would be able to make himself one of these. I'll also try to make it so that it could use several output tubes, such as EL34, KT66, 6L6 and their variants such as KT77 (a marvellous tube).
I'm still trying to cram all this into a 20 x 30 x 5cm Hammond chassis, and after an evening of playing iron tetris, it seems like this is actually possible. The finished thingy should have 3 inputs, a volume pot and outputs for 4, 8 and 16 ohms.
If this makes up anything worth mentioning, I'll try to document it as well as I can, so that someone able to sling some solder would be able to make himself one of these. I'll also try to make it so that it could use several output tubes, such as EL34, KT66, 6L6 and their variants such as KT77 (a marvellous tube).
I'm still trying to cram all this into a 20 x 30 x 5cm Hammond chassis, and after an evening of playing iron tetris, it seems like this is actually possible. The finished thingy should have 3 inputs, a volume pot and outputs for 4, 8 and 16 ohms.
Pay attention to what Jane said above - GOOD advise.
.
I had some communication with the Hammond Application Engineers about the 125ESE.
They stated:
The 125xSE range was not intended for HiFi Applications, the problem being limited primary inductance, HOWEVER, very good results can be obtained by driving them with a tube with low rp or a tube in triode mode (which gives low rp).
I was particularly asking about using the 125ESE driven by a triode strapped EL34. My question to them was about what would happen if I exceeded the 80mA idle current rating and had 100mA idle instead. The response was that you would loose some of the primary inductance but since I was driving it with a triode strapped EL34 the reduced rp would more than make up for the primary inductance loss.
My experience suported that advise from Hammond.
I see that you are heading for a triode strapped EL34 as well - don't forget to add a 150R resistor in the screen to anode strap.
I have since started a "Super Champ" Guitar Amp build with a 125ESE driven by an 845 (approx 850V for 20W out).
Cheers,
Ian
.
I had some communication with the Hammond Application Engineers about the 125ESE.
They stated:
The 125xSE range was not intended for HiFi Applications, the problem being limited primary inductance, HOWEVER, very good results can be obtained by driving them with a tube with low rp or a tube in triode mode (which gives low rp).
I was particularly asking about using the 125ESE driven by a triode strapped EL34. My question to them was about what would happen if I exceeded the 80mA idle current rating and had 100mA idle instead. The response was that you would loose some of the primary inductance but since I was driving it with a triode strapped EL34 the reduced rp would more than make up for the primary inductance loss.
My experience suported that advise from Hammond.
I see that you are heading for a triode strapped EL34 as well - don't forget to add a 150R resistor in the screen to anode strap.
I have since started a "Super Champ" Guitar Amp build with a 125ESE driven by an 845 (approx 850V for 20W out).
Cheers,
Ian
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.