Hoffman Amplifiers Tube Amplifier Forum
Amp Stuff => Tube Amp Building - Tweaks - Repairs => Topic started by: timbraun on March 20, 2014, 01:51:48 pm
-
Hi, Folks!
I just did a little maintenance on a JCM800 4104. It had a busted presence pot and really noisy pots. The EL34's were mis-matched, in that "only one blew, so only one is replaced" kind of way.
I found some Garnet EL34's locally, and plunked them in. Biasing them up to B+ 459V, 38.1 mA, 36.2 mA
Two questions :
1. seems like a reasonable bias level? Warm without shortening tube life too much?
2. matching - I don't have many options other than a completely new pair of tubes. Is 6% considered a good match? I'm thinking 10% is a match.
Nice place you got here... :thumbsup:
-
I found some Garnet EL34's locally, and plunked them in. Biasing them up to B+ 459V, 38.1 mA, 36.2 mA
Two questions :
1. seems like a reasonable bias level? Warm without shortening tube life too much?
Sure, the hottest tube is not quite 70%, so I doubt that will shorten either of their lives. Bottom-line, as long as you don't see redplating under strong signals, you're fine with any bias you choose ("under strong signals" implies you also get no redplating at idle).
2. matching - I don't have many options other than a completely new pair of tubes. Is 6% considered a good match? I'm thinking 10% is a match.
Depending on how you do the math, the present tubes are matched roughly to 4-5%.
I would be fine with that level of matching. Absolute perfect matching only gets you a hair more clean output power. Perfect matching & output stage balance also cancels all even harmonic distortion generated in the output section. Then again, your output transformer and phase inverter may not be perfectly balanced, negating the effect of perfectly matched tubes.
Gross mismatching (like 70mA, 10mA) may rob some power through-put capability of the output transformer, especially at low frequencies. It may also hinder the hum-cancelling of push-pull (or not; I'm not sure at what point a severe imbalance will result in loss of the cancellation of hum that normally happens in the OT).
In Radiotron Designer's Handbook, 4th Edition, it was shown that using 2 dissimilar triode output tubes with characteristics 50% apart only resulted in about 5% THD instead of a much lower THD due to perfect even harmonic cancellation in the output transformer.
Also note that for power output tubes, idle current was at best the secondary method of matching tubes; the primary way was testing for output power given a specified input signal and load impedance (according to RDH4). That told the tester that the tubes operated the same under dynamic conditions. This is not impossible to do now, but I can't think of any vendor that specifically claims to match by this method.
Read into the above comments however you like. I tend to try to buy matched tubes, maybe because of a false sense that perhaps they'll wear evenly or behave the same over time. I also have bias adjust and bias balance controls in my latest amp build, so that 2 non-matched tubes can be biased to the exact same idle current. However, even if I do that I still would not get the maximum possible clean output power, because of differences between the tubes, unless I have dynamically matched tubes. And with all of this said, I still only plug & bias the tubes once, and hardly ever look at biasing again (even though I know tubes can and do drift).
So I'm saying I'm contradictory: I like having the ability to bias tubes to a gnat's-ass, but I really don't use that ability much.
-
I found some Garnet EL34's locally, and plunked them in. Biasing them up to B+ 459V, 38.1 mA, 36.2 mA
If your readings are good , this is a very good bias and well matched tubes .
If the amp is used to play loud , try colder bias if sound /tone stay good at your ears.
-
Thanks, guys! It's nice to talk to knowledgeable folks. I'll keep my eyes and ears open. This amp won't be cranked too often. If I do crank it for a while, I'll check the level of glow.
I've been playing tube amps for ... well I got my Garnet Jammer back in '78 or so. I mod'd a Garnet PA90 into a high gain 6v6 based amp in the early '90s, but I let the amp internals lay fallow for a couple decades. Never did get into biasing until last year. Stuck with Garnet's and Mesa's, which are all fixed bias.
Marshalls and Fenders and Vox's. Oh, my!
-
Garnet , Canadian amp . I have one ; Revolition 1. Great little amp
Gar Giilis the builder publish a very good Book on amp , easy to understand .
Look at garnetamp.com
-
As HotBluePlates stated as long as the hottest biased tube is a 70% your fine and 2mA difference is fine.
-
Ho ho ho, I didn't get back to this thread too quickly, but I have to say the 4104 sounds fantastic with the old Garnet EL-34s in there.
This won't be the first time you will have heard : it sounds best with a bit of volume, it does thin out if it's volume knobs are too quiet.
Thanks HotPlateBlues and Jack1962!
And yes, stratele52, the Garnet amps are well worth checking out. The web site is important for us Garnet fans...