So...is anyone using the KT120's yet?

[jmaf]

I feel really tempted by the idea of a 2 tube 150 Watt amplifier.... I think it'd be a real attention grabber.

I heard there's a design issue with the KT120 requiring 50k max grid leak resistors, but what exactly is meant by that when using fixed bias and no grid leak?

Other than that, if I use a Marshall Major output transformer with two KT120's - does that sound like a plausible setup?

I'm looking for an excuse to throw U$ 90 at a pair - Any pointers, previous experiences appreciated.

[es345]

I have built an ultralinear bass poweramp stage evaluation with 2 KT120. It is running with Anode Voltage of ~850V idle and ~750V under full load. The OT has a separate winding for G2 (40%) running the G2 at ~600V. Primary impedance of the OT is 4K.
The resulting Power is about 200W, impressive clear sound. This machine is well suited for bass used in clean mode.
Very heavy overdrive would exceed the current limitation of the cathode, so if you want to go in that direction then change to 5K Primary and get about 150W, thats more safe.
I don't know the values oft the Marshall Major output transformer, so I cannot comment on that one.

[jmaf]

I have built an ultralinear bass poweramp stage evaluation with 2 KT120. It is running with Anode Voltage of ~850V idle and ~750V under full load. The OT has a separate winding for G2 (40%) running the G2 at ~600V. Primary impedance of the OT is 4K.
The resulting Power is about 200W, impressive clear sound. This machine is well suited for bass used in clean mode.
Very heavy overdrive would exceed the current limitation of the cathode, so if you want to go in that direction then change to 5K Primary and get about 150W, thats more safe.
I don't know the values oft the Marshall Major output transformer, so I cannot comment on that one.

Thanks for that, es345. 150W is all I wanted, to be able to overdrive it but I confess I'm a bit surprised at the high impedance of your primary. I was guessing 2k P-P.

What about the "50k grid leak requirement", what do you think that means? Problem is mentioned here The Amp Garage :: View topic - KT120 anyone?, here Kt-120 ? - Page 3 - diyAudio and here New Tung-Sol KT120 output tube [Archive] - AudioKarma.org Home Audio Stereo Discussion Forums

That means we need a low Z negative bias power supply to drive it / higher bias current?

[Wilder Amplification]

At 850V B+ 4K is the perfect load for a 200W amp.

[Bruce / Mission Amps]

Hmmm... would the term used, "grid leak requirement" imply that the grid load resistors and the bias supply needs to be at least 50K?

[octal]

I would assume that the grid leak must be a MAXIMUM of 50K for stable bias. (I doubt they mean grid leak in the sense of 'grid leak bias' but rather in the sense of the maximum resistance between the control grid and the bias supply.) With other output tubes, a max control grid resistance is specced in data sheets and a lot of the classic amp designs ignore this recommendation.

[kg]

basically a low value of Rgk means the tube has a less than stellar vacuum hardness.

the low DCR path is necessary to prevent thermal runaway from ionization.

[Steve Conner]

Or, it has a great big cathode and a grid very close to it, as required for high gm and high current. This means the grid runs almost red-hot from radiation, and risks developing emission as the tube ages.

(KT120s might very well work with larger grid bias resistors when new, but go into thermal runaway as they age, whereas with 50k bias resistors they would have "lasted" longer.)

[jmaf]

So, the entire (bias supply filter chain + eventual trim pot + "bias resistors") must add up to less than 50k to be able to drive this tube safely? For instance, if I had 10k + 10k pot + 10k to ground, the bias resistors could only be 20k instead of the usual 150k to 220k?

Thanks all who replied, appreciate the info.

[black_labb]

no, it refers to the resistors between the grids of the kt120's and the bias.

so for your bias you would have a negative voltage supply into a voltage divider and a resistor to the grid along with a signal coming through a coupling cap.
most amps tend to have 100k-220k. usually the 100k amps have multiple sets of tubes.

a 50k grid leak resistor could make a post PI driver stage a useful thing if using a LTP.

http://www.ozvalveamps.org/holden/sl...rcit200wka.gif

see the 100k resistors between the 12ax7's and 6550's. these are the grid leak resistors. there are also grid leak resistors on the LTP which are the 1m reistors between the grids and the junction with the 820 ohm resistor and the 10k resistor

The circuit also has some post LTP drivers in the form of DC coupled cathode followers. Keeping them dc coupled is important as you dont want the extra caps in the feedback loop otherwise the phasing of the feedback is wrong.

[jmaf]

Thanks for that black_labb. I'm familiar with grid leak resistors in the cathode-biased stages, I just didn't know they were called that in fixed bias as well. Appreciate the clarification.

The holden wasp schem is interesting.

Or maybe I oughta drive the KT120 via a transformer coupled single ended el84 ? I've got some extra trafos lying around....

[es345]

Thanks for that, es345. 150W is all I wanted, to be able to overdrive it but I confess I'm a bit surprised at the high impedance of your primary. I was guessing 2k P-P.

About 150W output with a 2K P-P OT requires parallel push pull config, e.g. 4x EL34 (140W) with 500V Ub. In this example the resulting average current at the 2 cathodes at one side is about 300 mA before clipping. This would kill a single KT120 in less than one hour.
When I start a design I use a small excel to estimate and set the target area for Ub, Power dissipation and OT. The minimum anode voltage is what you need in addition, that is normally somewhere between 70V and 100V and the idle current. With this values you get a draft estimate of the Power dissipation and the anode current. In the attachment you will find the excel. The input data are set to PP EL34 at 800V, 11K (config from data sheet). It fits quite well in this case. Just play around with the inputs.

[kg]

no, it refers to the resistors between the grids of the kt120's and the bias.

that assumes the bias supply is relatively low Z. not all of them are.

[Steve Conner]

Yes- as jmaf suggested, the output impedance (Thevenin impedance if you want to sound scientific) of the bias supply should be added in there.

Normally it is negligible, but I have an old ham radio manual that warns about it. The case they mentioned is that Class AB2 amps can feed a lot of current back into the bias supply and pump the voltage up.

[jmaf]

Gotya. And if I'm not mistaken, that bias supply is also an AC signal path to ground, right? So lowering Z there to 50k will require a heck of a driver too.... I guess that's the kt120's achiles heel?

Maybe es345 can share with us how he drove that kt120 - Something like black_labb's 2 triode dc-coupled driver?