Mullard 5-20 in triode mode

[icebear]

Hi

What should the value of the resistor between the screen grid (4) and the anode (3) on the EL34 be to run it in triode-mode? Is there any other things I should be aware off? How about the OPT? Would it be suitable for the task? I'm thinking of just make a switch between pentode and triodemode.

Bjorn

 

[fdegrove]

Hi,

What should the value of the resistor between the screen grid (4) and the anode (3) on the EL34 be to run it in triode-mode?

You can use the same 1K (?) resistor. I'd recommend a 3W one.

Is there any other things I should be aware off?

The most important thing to remember IMHO is that the amp was not designed for this kind of service so the gain will be different when in pseudo-triode mode and the OPT will be a mismatch as well as the impedance will not be optimized for the load either.
No damage should occur from this operation however.
It's just not really fair to compare the two modes of operation this way.

This could be of interest to you as general info in the context of the Mullard amp:

EL34

Cheers,😉

 

[icebear]

OPT

I guess my OPT is ca: 3,2 kohm with a ultralinar tap. Can I use this for a lonly EL34 in SE triodeconnection?

Bjorn

 

[SY]

3.2k is a pretty low load for a triode EL34. A better match would be 6.4K, which you could obtain by connecting a (nominal) 8 ohm load to the 4 ohm taps.

 

[icebear]

?

A resistor or my speaker?
b

 

[SY]

The speaker. We like to pretend that an 8 ohm speaker is 8 ohms, but of course this is a fiction we use to sleep better at night. But whatever the impedance curve looks like, by connecting it to the next lower impedance tap on the secondary, we double to anode-anode load. And in this case, that's probably a good thing.

The other thing to watch out for here is stability and bandwidth. When you convert to triode, you impose a much more severe capacitive load on the driver. This newly-lowered pole can compromise stability, and if it's low enough, you can get noticeable high frequency rolloff. IIRC, the 5-20 uses a 12AX7 as a driver- that's a tube without much ability to charge and discharge a Miller capacitance very quickly.

Anyway, I'd try the connection using a resistor dummy load first, see if the bandwidth and stability are acceptable, then go from there.

 

[icebear]

Opt

Checked my opt, and found that it was 6,6k. How should that affect it? I listened to both 4 and 8 ohm, but I think it sounds fuller and more natural on 8 ohm. The "S" sound is not as sharp as on the 4ohm tap.
Here it is: (http://www.demostenes.no/2.0/endr1be.asp?Kode=6009)

I want to connect the EF86 as a triode. Like its done in (http://www.lundahl.se/pdfs/claus_byrith/amplifier_30wpp.pdf)

How will that affect the feedback when the output is connected in triode . Which values should I use?



Bjorn

 

[SY]

6.6K is close to optimum.

If you change the input tube over to triode, you'll drop the open loop gain and reduce the feedback. It turns out, on closer analysis, that the difference is less than one would imagine, though. In "Valve Amplifiers", Morgan Jones does a very detailed analysis of the 5-20 circuit. If you're serious about experimenting here, it's a book you ought to have.

 

[icebear]

It's done

So both the EL34's and the EF86's are in triode. I must say it sings like never before 🙂 I thougth about completing the mods described on the Lundahl pages, but for now I'm satisfied.

Here is a picture of the setup. The OTL in the rack is not mine. Just for loan.

http://www.memphis-highend.de/diy/galerie+eintrag.htm?gbId=12

Thanks for the help (so far)

Cheers
Bjørn

 

[icebear]

Hi,



You can use the same 1K (?) resistor. I'd recommend a 3W one.

Came along this link: http://diyparadise.com/triode_mode.html

Any use for me to try this, or should I leve it with the 1k?

Bjørn

 

[arniel]

I've got a Maplin Millennium 5-20, which is basically a copy of the Mullard circuit. The biggest benefits I found when converting the input EF86 to triode mode (a la Morgan Jones) was that both noise and input sensitivity were much lower.

A.

 

[GeirW]

Gday.

Low noise would is a plus i assume,
but lowering the innput imedance, how low. (in your case)

 

[fragman56]

input sensitivity

is different from input impedance, I think?

Would simply trioding a tube make it more sensitive? I would think that it would make it less sensitive--less gain in triode? For the same signal, less swing out, no?

 

[SY]

fragman, your understanding is the same as mine- going to triode provides a useful lowering of gain/sensitivity.

 

[GeirW]

Gday.

Ahh.. sorry.
missread.

 

[ray_moth]

In that link, the writer cautions that the output power will be much lower in triode mode than in ultralinear mode. He says he has 97dB spkrs and it's no problem for him. I can tell you that it's no problem for me either, and my spkrs are only 89dB.