I know just enough to be dangerous with tube gear so I apologize if these are stupid questions.
I have an Inspire KT88 Firebottle amp by Dennis Had. It is a single ended pentode amp with the screens of the output tube regulated at 260 volts. The voltage on the plate depends on which rectifier you use. This configuration allows you to use a variety of output tubes ranging from 6f6 through kt88 by using rectifiers like 5y3, 5u4, 5ar4, etc. It uses automatic bias and is not user changeable.
Naturally I have rolled a bunch of different tubes through it. Saw the El38, heard it was similar to a KT77 which is compatible with this amp, put it in an adapter and fired it up. Sounds incredible, amazing detail, bass is super deep and controlled, deep soundstage etc. A week later I read that the el38 needs half the bias voltage of an el34 and I’m bound to be killing the el38 with cathode bias set for a reasonable setting for an el34. That does explain how bright they were glowing and the amount of blue they showed…
My question is why they sound so much better than other tubes I have put in my amp. Presumably if I replace the cathode resister to a value better suited to the el38 the sound will change as well. Is there a way to figure out the secret sauce of the el38 and output transformer so it can be replicated (as closely as possible) with another tube? Thanks.
I have an Inspire KT88 Firebottle amp by Dennis Had. It is a single ended pentode amp with the screens of the output tube regulated at 260 volts. The voltage on the plate depends on which rectifier you use. This configuration allows you to use a variety of output tubes ranging from 6f6 through kt88 by using rectifiers like 5y3, 5u4, 5ar4, etc. It uses automatic bias and is not user changeable.
Naturally I have rolled a bunch of different tubes through it. Saw the El38, heard it was similar to a KT77 which is compatible with this amp, put it in an adapter and fired it up. Sounds incredible, amazing detail, bass is super deep and controlled, deep soundstage etc. A week later I read that the el38 needs half the bias voltage of an el34 and I’m bound to be killing the el38 with cathode bias set for a reasonable setting for an el34. That does explain how bright they were glowing and the amount of blue they showed…
My question is why they sound so much better than other tubes I have put in my amp. Presumably if I replace the cathode resister to a value better suited to the el38 the sound will change as well. Is there a way to figure out the secret sauce of the el38 and output transformer so it can be replicated (as closely as possible) with another tube? Thanks.
I don't know anything. In my limited experiments with EL34, it seemed much happier (well behaved) the more current I put through it, I lost my bottle a about 40 watts anode dissipation, heading towards 15 watts output SE iirc.
EL38 is derived from EL37 the predecessor of the EL34, but with a top cap.
A pentode cannot like EL38 cannot sound the same as a beam tetrode (KT88 or KT77), because the IMD levels are different, and screen grid dissipation is higher.
It doesn't have kinks like the KT series and likely works down to really low voltages like the UBL21 which in many ways is very similar.
The EL38 was designed as a TV line driver valve which makes it extremely rugged.
The gain is considerably higher than the the EL34 14ma/v against 11ma/v, and the screen grid much tougher, so you need less drive for full output, but nothing like half bias voltage people claim.
Probably you found a cheap TV valve which works extremely well at close to the limit, and won't burn out easily despite any amount of massive abuse.
I'm a bit puzzled as to how it is possible to run 6F6's in this amplifier but not EL38's. Are you sure about the 6F6's? Does is say so in the manual for your amplifier?
The EL38, as you have discovered, is simply a very good sounding tube. I built a few amps with them last year and compared them to European darling tubes like the EL12. I slightly preferred the EL38.
Because of the higher gain it's possible to use a DHT driver like 26 or best of all 10Y. Or indeed 4P1L, 2P29L etc. This gives you what I call the "Inverted DHT Amp" where the DHT is in the input rather than the output as in 2a3, 300b etc. The 10Y>EL38 amp I ended up with was nearly as good as my 300b amp, with better bass but less lush midrange, so I went back to the 300b. But it was close.
If you can live with an anode top cap, which is a danger for uneducated handlers and should be put in a cage, then you will get some great sounds out of the EL38. I compared it to an EL34 and it was audibly on a higher level soundwise. One of the best output tubes around.
Because of the higher gain it's possible to use a DHT driver like 26 or best of all 10Y. Or indeed 4P1L, 2P29L etc. This gives you what I call the "Inverted DHT Amp" where the DHT is in the input rather than the output as in 2a3, 300b etc. The 10Y>EL38 amp I ended up with was nearly as good as my 300b amp, with better bass but less lush midrange, so I went back to the 300b. But it was close.
If you can live with an anode top cap, which is a danger for uneducated handlers and should be put in a cage, then you will get some great sounds out of the EL38. I compared it to an EL34 and it was audibly on a higher level soundwise. One of the best output tubes around.
I think EL38 tubes are good sounding to so many people because they are pretty alike the most common tubes in amplifiers. The more common ones are usually theoretically worse, but as we already know: audio enthusiasts have a tendency to like distortion when it pleases their pickle. Thus, it's simply a matter of what flavour of distortion you like, and that will be what sounds best to you.
While you may there for like the EL38 tubes, someone may think entirely differently about that. It is just about the flavour of distortion you prefer 🙂
While you may there for like the EL38 tubes, someone may think entirely differently about that. It is just about the flavour of distortion you prefer 🙂
EL38 is derived from EL37 but with a reinforced cathode and screen.
They even look internally the same, but the EL37 is silly money now.
UBL21 being a philips design, is a quantum leap better than the EL34, & of which like the EL38,- zillions of them available NOS for next to nothing.
+
With a 55V heater can have the heater powered in series directly off US 110-115V mains.
Use a matched parallel pair of them and they are the SAME output as the EL37/EL60/EL34 with lower distortion, (no top cap) just LOCTAL base.
PL38 is even more interesting - with the 30V heater from rectified 24V AC transfo, you can use DC rectified heaters as an ultra stable -30>-32V negative bias supply for fixed bias op.
Cute trick with the PL38 - I hadn't thought of that.
The UBL21/71 is new to me. Only 11W dissipation, so more like EL11 and EL33, two excellent sounding tubes I've been using. Have you used it in triode, or seen any triode curves?
We've had similar experience with many other tubes. The last experience was with a 212. Red hotting an anode usually gives a tube more even, feel of a stable tone. A hypothesis of mine is the change of vibration behavior of the innards of the tube due to increasing softness of metals due to overheating.
Both Fisher and Bogen did their bias this way decades ago. TBQH, it's a well known thing about running tube idle dissipation above 70-80% max. 6L6 / 807 are most linear when they are on the point of melting. If it's 50p TV tube, who cares what the lifetime is, if it sounds good, it IS good.
The manual says 6f6 tubes are usable if you use a 5y3 rectifier. I am using a 5u4g with the el38.
I was told to check for red spots on the plate. Trouble is I don’t know which of the internal structures are the plate. The entire center section that is hidden behind the main grey structure is lit up brightly but the Grey part is not. Usually only the heater is the only thing that glows but if that is the heater I’ve never seen one so large and easy to see before.
Well that’s the problem, el38 tubes aren’t that cheap anymore. Or if they are could you point me to those prices?
If it turns out I’m not actually damaging the tubes I would be very happy.
Which part of the tube is the anode? Is it possible that using the same bias point as a 807 won’t damage an el38?
The anode, or plate, is usually the bigest piece of conductive material inside the tube. It surrounds all other electrical tube parts, such as the grids, cathode and (heater).
Here's a randomly picked internet schematic for you:
Here's a randomly picked internet schematic for you:
An externally hosted image should be here but it was not working when we last tested it.
eh??
If you look on EBAY (UK) you will find 4 off for about 1-2 UKP each. 🙄
I never imagined it would be desirable, but hey 50p was a few years ago when people used to shoot them with airguns for fun.
The question of bias is pretty simple.
The higher the gain of the valve the less bias you need.
Think of it as a switch, where you have to start from a reasonable "slightly on" base within about 70-80% of its maximum.
A typical 807 would need a bias of about -30>-32V at 600V, because it's an ancient design of valve with low gain. That would make it settle at say 35ma or around 20W idle.
By comparison a 8417 (one of the highest gain devices made) would only need about -15>-17V at 650V. (about half). The idle current would be roughly the same.
The 8417 has roughly 4x the gain of the 6L6/807, so it needs less drive to turn it on and off.
Being as the EL38 has higher gain than the EL34 it's likely the using the same bias voltage would turn it OFF, rather than making it melt.
Last edited:
As above - I'm in the UK and EL38 is not uncommon here. A typical operating point for 18W dissipation for the EL38 in triode would be 300v, 60mA, -15v. An EL34 in triode would be 300v, 60mA, -22v. So yes, running an EL38 as above at 300v and -22v bias would give you about 18mA current or a dissipation of only 5W.
Good to know. What about running it in SE pentode with the same settings as an 807?
I’m going to very happy if I can keep running them like this.