Screen current

[drewl]

Working on a Sunn T50C.....one of the newer Fender designed amps, channel switching, reverb supposedly like the prosonic.
So it has a tube monitoring LED and a 100ma screen grid fuse.
It came in with that fuse blown, replaced it and it seems okay.
I monitored the draw, and it gets to around 80ma or so....so a 100ma SloBlo should be okay, what do you think?
What's the normal screen grid draw for 2 6L6's in a typical Fender type circuit?

[Old Tele man]

...good quality NOS 6L6GC's have a plate-to-screen ratio of about 20-to-one, or 100mA of plate current (Ip) will typically produce about 5mA of screen current (Is), at IDLE:

Is = (Ip*0.05) = (Ip/20)

...it's about double those numbers at MAX power:

Is = (Ip*0.10) = (Ip/10)

...be aware that it's a different "split-ratio" for different tubes, so these values are ONLY for 6L6/5881 and their derivatives (7581, 7027, etc.)

[drewl]

Well that's about right...I'm reading ~4ma at idle.

How about this, was the fuse that blew a slo blo?
It's a mini fuse that reads: T100maL250V
Thanks.

[Old Tele man]

...doesn't the "T" signify "thermal-delay" or "time"?

...what brand and what does their catalog say?

[Enzo]

Yes, T is a slow blow.

[drewl]

Rats....owner syas the amp keeps blowing this 100ma screen fuse.
He had the output tubes replaced and it still blows the fuse.
Must be something intermittant or flaky I need to find.

[Mars Amp Repair]

tube info

...good quality NOS 6L6GC's have a plate-to-screen ratio of about 20-to-one, or 100mA of plate current (Ip) will typically produce about 5mA of screen current (Is), at IDLE:

Is = (Ip*0.05) = (Ip/20)

...it's about double those numbers at MAX power:

Is = (Ip*0.10) = (Ip/10)

...be aware that it's a different "split-ratio" for different tubes, so these values are ONLY for 6L6/5881 and their derivatives (7581, 7027, etc.)

Now. in what currently available books do you find that kind of tube design information in? thanx, glen

[Steve Conner]

Lol, "currently available books" on tube design I can't think of any except the Newnes/Butterworth reprint of RDH.

If you look at the curves on a 6L6GC datasheet, you'll see screen current printed along with the plate curves. Notice that the screen current peaks sharply if the tube is loaded with higher than the recommended load impedance. In particular, an open-circuit load causes massive screen current: pretty much all of the current that should have gone to the plate hits the screen instead. OTM's Ip/10 rule of thumb goes out of the window.

So, I'd be wondering if the owner is using extension speakers with too high an impedance, has the selector set wrong, or has a speaker cable with an intermittent fault. Or is doing some special school stunt like playing it with the speaker unplugged for late night practice

[drewl]

Hmmm, food for thought Steve. I'll check/rework the speaker lead, it looks cheesy.
Possible intermittant impedance switch, or user error. All possibilities.

[Enzo]

Possible intermittant los of bias to the tube as an alternative?


RDH was my first guess as to books.

[drewl]

I've made it a point to clean and retension tube sockets which this one could use.
Thanks again all.

[Old Tele man]

OTM's Ip/10 rule of thumb goes out of the window.

...actually the Ip/10 at "...full output power..." rule-of-thumb falls apart only when Vp is sucked down to the diode line (for whatever reason) at which time, the plate ceases to act as "electron catcher" and the screen becomes the "electron catcher" and it's current skyrockets!

...a "visual approximation" to keep in mind is that the space-charge (Isc) current is essentially a fixed amount of current, and the sum of Ip and Is always equals the fixed space-charge current value...so, what ever the Ip "isn't," the Is "is"...ie: Isc = Ip + Is

...that is, when Ip starts out low, Is is gonna be quite high, but as Ip rapidly climbs along the diode line, conversely Is is quickly declining...and then (approximately) over the Eb-Ib distance beyond the diode line, Ip only slightly increases (ala' constant current source) and Is only slightly decreases...hence it's roughly Ip/20 at idle (very far from diode line) and only about twice as much (Ip/10) back at the diode line...just 'before' its current curves and goes asymptotic.

[drewl]

So in essence, if somehow the plate voltage gets "sucked down" the screen current will increase drastically?

[Old Tele man]

So in essence, if somehow the plate voltage gets "sucked down" the screen current will increase drastically?

...actually, only when the plate voltage (Vp) falls "below" (ie: to the leftside of) the diode line: