• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

JJ EL509s in Triode - specs ?

Saw that JJ sell an Octal version they label EL 509s. Never looked at this tube before. I know there are the original top-plate versions out there too.

I've seen it referred to as a Sweep Tube and these are often very nice when wired up in Triode mode. But JJ provides nothing about triode wired operation.

Anybody using this tube in triode mode ? (it's not a low cost tube, not sure who is buying it in enough volume to justify JJ to keep it in production ?)

I always think of the EAR 859 when someone post about the EL509 tube, but that's the original with the top cap. They called that "enhanced triode" which was actually screen drive IIRC.

Looks like Transcendent Sound made a few OTL amps using the JJ version. I don't see any tubes in stock anywhere thou. Eurotubes are out of stock for some time, says check back June 2022.


Generalization about Pentodes and Beam Power tubes:

The maximum screen voltage is Lowest in Pentode/Beam Power modes.
The maximum Quiescent screen voltage is Medium in Ultra Linear Mode.
The maximum Quiescent screen voltage is Highest in Triode Wired mode.

That is not only due to the quiescent state, it is due to what the screen voltage is when the control grid nears 0 Volts, and what the plate voltage is at the same time.

Beam Power example:
When the plate is at 100Volts, and the control grid is at 0V, the screen at 300V is working very hard to act just like a plate, it draws Lots of current (and that current x 300V is the instantaneous screen dissipation). Do not drive the tube into hard clipping, or the integrated screen dissipation will be 1/2 of the peak screen dissipation.

807 beam power mode, screen maximum 300V (factory rating)
807 Triode Wired mode, screen Quiescent maximum 400V (factory rating)
And my estimate: 807 Ultra Linear mode (50% UL tap), screen Quiescent maximum 350V
The integral of dissipation over 1 cycle of a test tone, or with music, the integral of dissipation over time, should not exceed the screen maximum dissipation of 3.5 Watts.

An EL509 or equivalent: The filament and cathode construction is made to work with cathode current, and current, and current.
Last edited:
Web links have a nasty habit of going away so I’m posting those curves here (with acknowledgement of the source added).

The triode curves at 160V plate voltage and Vg -30V gives approx 240mA of current,
The JJ data sheet Transfer Characteristics at 160V on plate and g2 with Vg -30V gives approx 200mA (plate) + 40mA (g2) = 240mA


  • IMG_6600.jpeg
    232 KB · Views: 94
Last edited:
  • Like
Reactions: 2 users
Those curves are from the original TV sweep tube EL509 / PL509.
The octal JJ - EL509 might be quite different.
Its story dates back some 15 years and has been told on radiomuseum.org by a person from T+A who says he had been involved.
Translation of the letter - dated Oct 2011 - follows:

When I asked member Lothar Wiemann about the EL509-3 I received the following interesting answer:

Dear Mr Erb,

I would like to tell you what I know about the EL509 octal sockets without top connections:

Due to its high maximum anode current of over 1A, the original EL509 has enjoyed a certain popularity for OTL (Output Transformerless), i.e. ironless hi-fi power amplifiers (see "Futterman amplifiers") for a long time.

Despite the qualities of this tube, many audiophiles took the position that "a television tube cannot make sound". For this reason, some hi-fi manufacturers requested an "Audio-EL509" - i.e. with an octal socket instead of a Magnoval+top connection.

After Svetlana went into American ownership, this idea was taken up and the so-called EL509-II was created. Electrically a normal EL509 just converted to octal.

This Svetlana version was used (e.g. in the T+A integrated amplifier V10 - I was involved in the development of this .

However, the Svetlana proved to be unstable. In particular, glass breakage was common due to thermal stresses in the very thick glass used by Svetlana.
This problem should then probably be countered by the thicker, bulbous glass bulb a' la 6550. This further development was then named EL509-3.
I don't know how many of the Svetlana EL509-II and -3 were produced in total. We at T+A have installed around 4000 EL509-II units (and returned around the same number due to quality defects). The 509-3 to my knowledge was only built in very small quantities.

Since Svetlana could not eliminate the quality defects, we (T+A) then asked JJ whether they could produce an EL 509 with an octal base. The result was the JJ EL509.
The JJ is actually not a real EL509, but an enlarged EL34 in terms of structure, which has an anode power dissipation of around 40 W and a high-current cathode that can deliver around 1.4 A.
The JJ version wasn't the solution to the quality problems either. Instead of glass breaks and grid shorts, the JJ suffers from thermal drift and has proved problematic in fixed-bias circuits.
Ultimately, we realized that high-quality hi-fi amplifiers are hardly possible with the modern EL509 replicas. We therefore returned to the original NOS EL509 - with the unpopular top connection - about 3 years ago.

I hope that what I was able to say about the "modern" Oktal EL509 will help you a little.

Best regards,
Lothar Wiemann
Last edited:
  • Like
Reactions: 3 users
When using parallel SE, or parallel push pull, I prefer to use individual self bias resistors and individaul bypass caps for each tube.

I would not use fixed bias or fixed adjustable bias. Even if there is individual bias for each tube.
Have you ever tried to go back and forth on multiple bias pots, only to see when you adjusted tube # 1, adjusting tube # 2 changes tube #1 bias.
Oh, you can fix that with regulated B+, but a series regulator requires more B+.
I thought one main reason a lot of people do not like self bias is that you need more B+ versus with fixed bias (Oh . . . fixed bias requires regulated B+). De Ja Vu all over again!
  • Like
Reactions: 1 user