diyAudio (
-   Tubes / Valves (
-   -   EL34, KT88, High Voltage?, VTL.. (

kegger 2nd July 2009 12:11 PM

EL34, KT88, High Voltage?, VTL..
I'm working on an amp an area buddy dropped off to me and I can't quite decide which/what
route to take so I thought I'd toss a question out for some possible ideas/thoughts from others.

Here is the scenario,

It's a VTL ST-85 running EL34's in pentode with 600V on both the plate and there SCREENS!
(outputs have no UL taps)

The owner was having an issue with bias and the controls either not getting there required bias
or if they did the pots were turned pretty much all the way with any tubes ever used in this amp,
and a channel not "sounding" quite right. I fixed the bias and the channel had an issue with the
tail resistor in the LTP being 30% off in one channel, fixed that as well.

The amp has gone through tubes quite a bit I'm told an it doesn't surprise me with that voltage.
(and the owner is going to buy a new quad when I get the amp settled, any suggestions?)


So I'm trying to figure out if I want to alter anything in the amp based on the voltage, the tubes
and the fact I'm not a giant fan of push pull pentode, unless say a nice regulated screen volts.

Options I have tossed around are, (if I would be making any changes)

(sticking with EL34's)
Wiring outputs in triode as some pdf's say 600v triode is max (I doubt any new tubes like it)
(would it be better then running pentode at that voltage, humm?)

Taping from the RC stage that feeds the phase splitters has 400v then use it for the screens.
(to stick with pentode operation but still drop the screen voltage down some)

Changing cathode resistors of 10ohms for sense to say 500-700 ohms and do a combo bias,
(which would drop my plate to cathode voltage a bit)

(going to KT88's)
Again wire the outputs in triode, up the cathode resistors to make up the bias voltage needed.
(which again would drop my plate to cathode voltage a bit)

Taping from the RC stage that feeds the phase splitters has 400v then use it for the screens.
(to stick with pentode operation but still drop the screen voltage down some)

(Or combos using any of the above together)


I can modify the bias circuit to get upto about 70v, the original ran 600v B+ and 30ma per tube.
(or mod it to get anything lower as well) (I may try calling VTL, see what power tranny handles)

This thing was rated for 85wpc running a pair of EL34's, do not need anywhere near that power.

(.4a total extra on heater circuit running 4 KT88 verse the 4 EL34)
I'm assuming a .4a difference in heater from EL34 to KT88 shouldn't be a problem, also the amp
runs your typical VTL circuit, Paralell 12AT7 voltage amp cap coupled to a schmidt type LTP an
using a 12AT7 on out to the usual cap coupled outputs with the fixed bias mentioned.
(I have no schematic for the amp, just traced it out and the usual VTL stuff)

It uses 270k grid resistors for the EL34 circuit, if I go KT88 I'll lower those, also G3, pin 1 is tied
to pin 8 so it accepts a 7AC type pinout. It has 1K screen grid resistors and uses a PCB so I will
have to lift one end of the screen grid resistor to either go triode or feed another voltage source.
(nothing here difficult to do)

There isn't really any extra room inside this thing to try and say add another filter stage and drop
some B+ voltage, so I've all but given up altering the B+ (unless someone has a good idea there)
(there is plenty to feed the rest of the amp so dropping voltage wouldn't hurt anywhere though)


Anywho am I off in thinking this original arrangement/voltage is just to stressful on the tubes here?

Any thoughts and or suggestions are welcome, (or any more info needed ask, I tried to give all) :)

Thanks in advance, (hope the post wasn't to long winded)

Allen Wright 2nd July 2009 12:52 PM

1/ 600V on the screens of an EL34 is basically criminal. I went head to head with Luke Manley about this at a hifi show once, and he said "You have to use our specially selected tubes in our amps".

"And where in Europe can I get your special tubes?" I asked.

"From us in California" was the glib answer.

I say BS, design ratings are design ratings and I've also had to service many blown up VTL amps because of exactly this.

One solution I had was to change the amps to cathode bias, and triode operation. That made them bias stable and sounding better. I don't know if an EL34 is universally safely usable with 600V on the plates & screens in triode, but as you say, far safere than in pentode! And these did run for sevearl years at least, until I lost contact with their owner.

The best solution would be to convert the PS to choke input, dropping the B+ down to circa 400V, but there isn't enough space on a VTL to add a decent sized choke.

I feel for you, and can't believe how a company such a VTL can make such badly design stuff and still keep their elevated reputation.

Regards, Allen (Vacuum State)

bigwill 2nd July 2009 01:11 PM

How about modifying it slightly for a lower screen voltage - say 300V. You could add a regulator, and pentodes are more linear at lower screen voltages anyway

ColinF 2nd July 2009 03:02 PM

VTL mods
I have modified VTLs over the years. One of them was VTL 300W monoblocks that run on over 600v on eight 6550s per channel at 30ma idle bias current. The owner wanted to change them to el34s because they sound better in other amps that can use both types. He also wanted to run them in ultralinear connection.
I too am shocked that VTL routinely run over 200v in excess of the 6550 and el34 max screen voltage, but the concern here is how much power the screen can dissipate. It will operate at 600v but the average screen current needs to be limited at this high voltage for best reliability.
I came up with a solution that uses high voltage TO220 mosfets to lower the screen voltage to 320v, with the gates of the mosfets operating on 40% of the anode to B+ audio voltage, to simulate ultralinear operation.
The amp put out about 200w with this mod and sounded sweeter and more harmonically resolved. As far as I know the amp is still running fine ever since, over 6 years.
Also, check the input stage power supply capacitors, they go open circuit over time. When the amp is switched on, before the tubes warm up, the voltage goes way over the cap's voltage rating until the tubes warm up.
Hope this helps :)

kegger 2nd July 2009 03:04 PM


That would be an option, there really isn't much room in this thing and
being a PCB adding in other circuits really isn't that easy, but I might..

Thanks for the suggestion..


Allen I am seriously leaning the cathode bias triode deal, glad I'm not
only one thinking this is pretty insane on these tubes, I'd like to try an
keep say a small (like 5-10 volts) adjustable bias still to balance tubes.

And I think going to KT88's is another step in the right direction as well.

Thank you for weighing in with your experience..

kegger 2nd July 2009 03:09 PM


Thank you as well, and I to was concerned for those caps, with being a
solid state rectified amp yah that voltage ramps up there for a while till
things start conducting.

Would you happen to have a schematic/layout showing the type solution
you mention, it does peak my interest, I have to say..:)

Allen Wright 2nd July 2009 03:09 PM

Going to KT88's will solve the overvoltage problem in one hit. And if you do go cathode bias, then allowing for fine tuning of grid V will make the thing sound better than if not. EI OPTs do NOT like out of balance standing current!

Regards, Allen (Vacuum State)

kegger 2nd July 2009 03:17 PM

Well then I may have a plan of attack, I was lost for a while thinking am
I not looking at this right and should I be modding this thing, I'm glad I'd
posted now, I wasn't sure it would lead me to an answer..

Again Thank you for the reassurance..:)

Bandersnatch 2nd July 2009 06:40 PM

Allen has it well thought out. The ground-cathode voltage to idle a KT88 at 35-40 mA with B+ around 600V is going to be 60-65V I'd bet...which puts you down around 540 plate to cathode...a useful step.

You can burn even more B+ with the Blumlein Garter bias/balance circuit in the cathodes...even more useful I think. Balance, and effectively dropping B+.

He's also right about the criminality of the whole exercise...somebody needs their doo-dads put in a vise for that sort of practice, let alone for thinking it is a fine idear.

bozole 2nd July 2009 09:30 PM


Sorry, I just saw this post, maybe a little late ?

What I would do with this amp is decrease the screens voltage to 450 or 500V max, preferably 450V
This way, the primary reflected impedance by the OT will stay good with the circuit. If you use a triode connection, it will be not very optimal ...

Secondly I'd put out the 10 ohms cathode resistor, and put the cathode at ground potential !! If you actually have the g3 at ground, which I suppose is the case, it's a very bad thing, because even if the positive voltage is very small on the cathodes, it's positive relative to g3, and it's really not good for the tube, especially at high voltages ... This mod will make biasing a little longer, but personnaly I bias with a scope and my ears, so it doesn't change anything

And third I'd put out the g3 from ground, which is the case I suppose ? and connect the g3 to the bias supply (before the bias resistors)

And finally rebuild the bias supply

This way is I think the safest one to use EL34 on high voltages, and you will be surprised how robust the amp will become, the tubes too !!

Best regards

All times are GMT. The time now is 02:09 AM.

Search Engine Optimisation provided by DragonByte SEO (Pro) - vBulletin Mods & Addons Copyright © 2019 DragonByte Technologies Ltd.
Resources saved on this page: MySQL 17.65%
vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2019 DragonByte Technologies Ltd.
Copyright ©1999-2019 diyAudio