Bias feed resistors

[Zouto]

Latest issue here:
http://www.premierguitar.com/Magazin...ifier_Mod.aspx

[Steve Conner]

BackwardsBob wins a cookie!

This is what I love about tube amps. Every single component has at least 5 different functions, so when you're tweaking, like Draper Kaufmann said, "You can never do just one thing."

http://www.salvarsan.org/content/Kau...s%20Rules.html

When it comes to thermal runaway, I just peek in the back of the amp every so often to make sure everything's OK in the "engine room".

EL34s weren't necessarily a poorer vacuum tube. They just have a bigger cathode with the control grid very close to it, which is what gives them their higher gain than 6L6s. This means the control grid is larger and runs hotter than a 6L6, due to radiated heat from the cathode, and that's why they need a lower grid leak resistance.

[R.G.]

EL34s weren't necessarily a poorer vacuum tube. They just have a bigger cathode with the control grid very close to it, which is what gives them their higher gain than 6L6s. This means the control grid is larger and runs hotter than a 6L6, due to radiated heat from the cathode, and that's why they need a lower grid leak resistance.

I agree with the emphasis on "necessarily". However, my experience with EL34s is that they act like the vacuum is softer than 6L6s.

But extra-hot, extra-close grids won't help much.

[martin manning]

If the tube is driven to saturation, grid current increases substantially, charging the coupling cap. The value of the grid leak resistor working with the value of the coupling cap determines the recovery time. Right?

I remember reading this somewhere recently... maybe here in one of Steve C's posts on tube power amp distortion.

So, what effect would this have on tone? I can see where the (charged) coupling cap and resistor would delay the return to the non-saturated state if the signal is driving the tube over and back (with time constant RC), but what does that sound like?

MPM

[dai h.]

EL34s weren't necessarily a poorer vacuum tube. They just have a bigger cathode with the control grid very close to it, which is what gives them their higher gain than 6L6s. This means the control grid is larger and runs hotter than a 6L6, due to radiated heat from the cathode, and that's why they need a lower grid leak resistance.

I remembered reading that EL34s could be used with higher values of grid leak and this lead to a bit of searching, looking over links with the following findings:

this site gives a specific value:

http://www.op316.com/tubes/tips/b260.htm

"...using 300k grid leak Rs (with EL34s) under fixed bias is rock solid stable..."

leading to looking over old Matsushita datasheets:

"On output pentode 6BQ5 and 6CA7"

http://www.ne.jp/asahi/myamada/tube/shicopy/6ca7.gif

some stuff about 6BQ5 having a "barrel-shaped" cathode "helping to give a larger effective cathode surface area contributing to high gm...". Scratching my head about what "barrel-shaped" meant, I think what they mean is that it actually does look like a barrel if you look at the cathode shape from the top end (end with the getters and grid fins). (Looks more clear in the diagram in the next link showing drawing of 6BQ5 and 6AQ5 from top. For 6AQ5 "K" ( = cathode), the cathode is a perfectly round shape compared to the 6BQ5 "K".

http://www.ne.jp/asahi/myamada/tube/shicopy/6ca7_2.gif

And also on that page, apparently to counter the potential grid emission problem, the grid posts are made of a special copper alloy meant to enhance heat conduction and additionally the grid is gold plated, "...therefore, a relatively high value resistance may be used for the grid leak". Then, I took a magnifying glass to a Tesla E34L under a light to scope out the construction and see if I could tell what exactly was being referred to, and what it looked like was three rods on each side, with two copper rods holding g1 with those cooling fins attached to the ends (so big cooling fins look like an additional measure to help with stability), then outside of that silver color rods apparently holding g2, then another pair of rods outside of that for g3 and the getters on the ends. The cathode also appeared to have the "barrel" shape.

(this one is a pic from the same site above of a top of a Matsushita EL34 with glass removed: )

http://www.ne.jp/asahi/myamada/tube/...7/6CA7_M_P.JPG

anyway, I've never tried testing the grid leak R limits (beyond typical Marshall 220k) or anything but that was interesting to learn.

[Steve Conner]

The sonic effect of that self-biasing is a cool kind of pumping as if you had a compressor in there somewhere. If you think about it, the output stage is actually acting as a ghetto vari-mu compressor, or a bias-wiggle tremolo powered by the signal.

Cathode biased amps do it even more prominently.

As for the barrel thing, I think back in the old days most tubes had round cathodes, except beam tetrodes had rectangular box-shaped ones. (the 6550 and KT88 still do.) The electron beams come out of the two long sides, while the two short sides are blocked off by the grid posts and do nothing. I guess they chose the box shape because the geometry of the beam tetrode is basically all rectangular and flat.

The barrel shape is just the box shape with the two long sides bulged out a little. It probably suits a true pentode better, because they don't have neat electron beams, the electrons just splurge out any way they like, except for the shadows cast by the grid posts. The EL37 actually had a neat oval cathode, but you probably have to be a hardcore tube geek to find that interesting.

Yes, the thick copper support posts and cooling fins on the EL34's grid are to help with stability. (Or to allow a bigger grid leak resistor to be used, it amounts to the same thing.)

BTW, I actually have a pair of those Matsushita 6CA7s that that guy dismantled. They look like an exact copy of the Mullard EL34.