Some unusual tube observations

[jmm22]

I have a 50 watt Marshall hybrid clone, JCM800 preamp section with a JTM45 power section. I've owned it for about a year and absolutely love it. It came with EL34's, which I quickly switched out for Reflector Gold Lion KT77's, which I now want to switch out for vintage KT66's, of which I have eight tubes, six are GEC and Tungsram, and two are new & matched TAD (Shugwang I presume). I tried the TAD first, and preferred the sound to the Gold Lion KT77's I had in there. I biased them at 43/44mV. Then I tried the smoked bottle Tungsram's.

I liked them even more than the TAD. They sounded excellent, but here's the strange thing. They were biased at 50/18mV. Now my amp tech balked at a set of KT66's of mine because they mismatched by 11mV, and here I have a 32mV mismatch that sounds terrific.

My question is, why do they sound good when opinion seems to be that terribly mismatched tubes are not to be used? And is there any compelling reason why I shouldn't use this pair? I plan on testing the others, but felt I could use some insight before moving forward with my testing.

 

[Pete Farrington]

50mA seems high, what’s the HT voltage?
As you’ve got 6 vintage KT66, surely you can find a better matched pair?
It’s a natural human thing to pair the Tungsrams, but I think electrons are indifferent to branding, characteristics are what matters.
Are you loading them with 3k4 ish as per a typical EL34 OT, doubling the load for 6k8, as per typical early JTM45, or what?

 

[jmm22]

Admittedly, I'm not nearly as amp tech savvy as I would like to be, but I'm trying to learn. I don't know what the HT voltage is, but I could measure if I can find a youtube video on it. And as per your last question, I'm similarly in the dark. I asked my amp tech (who's very good, but short on detail when describing things) about switching the KT77's out for KT66's. He said as long as I'm not driving the amp hard it should be fine. I can't get the master past 1 in my small studio, which is where the amp is staying for the time being. He said if I did start to gig with it, he'd be inclined two switch the 1k screen resistors for 2k 5 watt if I wanted to run the KT66's.

I understand that there's a strong consensus to want a much more closely matched pair, but what I'm really trying to understand is why. What is the primary reason for wanting closely matched tubes if greatly mismatched tubes sound just as good? I'm trying to reconcile what I'm hearing with what I'm supposed to be hearing.

 

[Matthews Guitars]

A perfectly matched set of output tubes does a great job of cancelling out odd order harmonics, resulting in lower distortion. A mismatched set of power tubes will not cancel out those odd order harmonics as well, giving a richer tone with more odd order harmonics.

We're talking about good, desirable sounding distortion here. This isn't a high fidelity amplifier even though the power stage is typically treated as if it is one. But mismatched power tubes can add distortion harmonics in the power stage that you like to hear.

 

[jmm22]

I've tried all my tubes. Now just sorting them into pairs matched as best as I can. Although I really liked the sound of mismatched tubes, I haven't put in any real time yet listening to well matched pairs. The KT66's smoke.

 

[Pete Farrington]

I have a 50 watt Marshall hybrid clone, JCM800 preamp section with a JTM45 power section

switch the 1k screen resistors for 2k 5 watt if I wanted to run the KT66's

I wonder what you’ve got there, as JTM45 generally use a pair of 470ohm screen grid resistors (plus a shared/common 1k).
As it came with EL34 and 1k screen grid resistors, might it be more of a JTM50 type output section?
Any further detail / info on the amp / its design & builder?

I can't get the master past 1 in my small studio

In that case, other than a bit more HT ripple from the unbalanced currents in the OT primary, at very low power outputs you may not be hearing any consequences from the mismatch.
The hotter idling valve will be doing most of the work, never being pushed beyond its class A area of operation, especially if the OT primary impedance is up at the typical JTM45 6k6 - 8k, rather than the 3k4 typically used for EL34.

Does the amp have a valve rectifier, eg GZ34 / 5AR4, eg next the the output valves?

To get a rough idea of the HT voltage, power the amp up with an output valve removed, speaker must be connected, master vol and presence turned down.
Set the meter to measure 500V DC, touch the black probe to connect to the chassis metalwork, and insert the red probe into each of the 8 socket holes in turn and note the reading.
Don’t put the probe too far into the socket holes, 1/4” should be sufficient.
Start at the notch, doesn’t really matter whether you go clockwise or counter clockwise.
If, after a minute for the valves to warm up, you measure more than 1V at the holes either side of the notch, then abort.
There should be three holes at around 400-500V, one hole around -50V.

 

[jmm22]

The builder is Steve Moratto in Ajax Ontario. Rectifier is 5U4GB. At least one transformer is a Drake. Voltages on pins clockwise after key:

1. 0
2. 0
3. 476V
4. -52.4V
5. 476
6. 478
7. 0
8. 0

I'll admit I'm sure I haven't played as many Marshall's as many of the fine & savvy folks at Marshall Forum, but this is the best amp I've ever heard.

 

[Pete Farrington]

With a HT voltage that high, 50mA idle cathode current really does seem too high.

FYI from those voltages, the pin numbering is 8,7,6… rather than 1,2,3…

 

[william vogel]

One thing that has to be remembered is that idle bias current is only that and during dynamic playing situations the mismatch at idle may subside and in modulation the balance may become closer together or even flip. The problem with idle mismatch is hum. I have a pair of RFT EL34’s that are spread about 20mA at idle (42-23) but once fed signal they are perfect balance at full clean rms according to the waveform and they sound glorious. I was reluctant to even give them a chance until I played it and I’m glad I did. The amp really isn’t humming either. Moral of the story is if you like the sound, run it.
I will never ship out an amp to a customer with a set of power tubes that are far apart but for me, I let it go.

 

[jmm22]

I've been testing pairs at ~30mV, and haven't detected any depreciation in the tone.

Do you think I would hear a difference if I switched out the 1k screen resistors for 2k?

 

[Pete Farrington]

Do you think I would hear a difference if I switched out the 1k screen resistors for 2k?

Not unless the master volume is turned much higher and the output valves really start trying to pull some HT current.

 

[jmm22]

...and the 2K resistors would mitigate some tube heating issues?

 

[Pete Farrington]

True, but if the master volume is only at 1, the heating issues that might mitigate wont be encountered.
Indeed, but it may be worth considering, if <1W power output is required, is it really worth eating up rare, expensive output valves in a high power amp?

 

[jmm22]

Thank you for your good counsel. Based on that I'll run the amp as is (1k), with vintage tubes, and simply switch in the TAD's (which I'm much less concerned about) if I'm going to play it hard.

 

[Ninjaking67]

Jmm22,
Have you contacted Steve Moratto to ask him? He definitely knows his stuff and would be more than willing to advise you regarding the resistor change.

 

[jmm22]

I have, and Steve obliges as best as he can, but I also understand he can't explain every subtlety of amplifiers and tubes to me, otherwise we'd be talking on the phone all day, hence my questions here in the Marshall forum.

 

[mrrstrat]

Beautiful build!

The builder is Steve Moratto in Ajax Ontario. Rectifier is 5U4GB. At least one transformer is a Drake. Voltages on pins clockwise after key:

1. 0
2. 0
3. 476V
4. -52.4V
5. 476
6. 478
7. 0
8. 0

I'll admit I'm sure I haven't played as many Marshall's as many of the fine & savvy folks at Marshall Forum, but this is the best amp I've ever heard.