How do you match output tubes?

[Bobby Dipole]

What are you trying to match when picking a pair of tubes for a PP stage?

(I'm guessing it's the quiescent plate current.)

I'm restoring an old Pilot AA-309B mono amplifier. It uses EL84s in the output. I have 10 NOS EL84s in my "inherited" tube collection. These are of various brands, and the quiescent plate currents range from 33mA to 70 mA. B+ in this amp is 300V, so some of the tubes exceed the 12W plate dissipation limit! (I already complained about "cherry plate" in another thread.) If you take the fixed "un-adjustable" bias voltage of -10.5V and the plate voltage of 300V, the curves in the tube data sheet say Pilot designed this amp to put the plate dissipation right at the limit! So, it seems to me that one must not only sort tubes for matching, but sort tubes to avoid melt-down! Is this common practice?

Bobby Dipole, Tube Newbie

[Wavebourn]

I bake NOS tubes first. Then write numbers on them using permanent market. Then plug in each one and write down voltage measured on the cathode resistor. Then go through the table and select tubes I want.

[Bobby Dipole]

Originally Posted by Wavebourn

I bake NOS tubes first. Then write numbers on them using permanent market. Then plug in each one and write down voltage measured on the cathode resistor. Then go through the table and select tubes I want.

The Pilot amplifier has no cathode resistors. There's nothing at all to stabilize the bias in this circuit. You would never see this in a transistor circuit! Anyway, I set up a little breadboard with a 9 pin socket, and wired it up to a Heathkit HV power supply for testing. I put a one ohm resistor in series with the plate to measure plate current. I wish my tube tester was versatile enough to do this.

Bobby Dipole

[Wavebourn]

I usually put in a measurement setup cathode resistor needed for an optimal bias.

[taj]

Originally Posted by Wavebourn

I bake NOS tubes first.

Hi Wave ,

How do you bake them? (time and temperature?)

My wife already thinks this a wacko-nut-ball hobby. Wait until she sees this!

..Todd

[Wavebourn]

Originally Posted by taj

Hi Wave ,

How do you bake them? (time and temperature?)

Depends on taste. I prefer brown & crusty.

I buy puff pastry in Trader's Joe, crab meat in Costco, rice and rice cooker in Chinese store, onion and eggs elsewhere... Cook rice, mix with crab meat, chopped onion, add one egg, mix thoroughly, wrap in puff pastry. Preheat oven to 375 degrees F, put a pie there on 22 minutes (until brown and crusty), take it off the oven, eat. Yummy!

Glass tubes need to be baked overnight on the same temperature: they are rated usually for 200 degrees C max. I did not try to bake tubes with Bakelite sockets, but I am pretty sure all 6P3S tubes I have need to be baked.

[vinylkid58]

Originally Posted by taj

My wife already thinks this a wacko-nut-ball hobby.

Hey Todd, who's wife doesn't?

Originally Posted by Wavebourn

Cook rice, mix with crab meat, chopped onion, add one egg, mix thoroughly, wrap in puff pastry. Preheat oven to 375 degrees F, put a pie there on 22 minutes (until brown and crusty), take it off the oven, eat. Yummy!

That does sound yummy.

jeff

[Doz]

Originally Posted by vinylkid58

Hey Todd, who's wife doesn't?



That does sound yummy.

jeff

Sneak the tubes in with the pie ... then only get in trouble for "leaving the oven on"???

I once baked a cylinder head in the oven (to fit new valve guides (ironic!) once)... I only got in trouble, because I made sure the oven was spotless when I'd finished... she noticed it was clean ... There's no pleasing some people...

It does sound tasty btw...

[k9swc]

I've used a fairly simple, if unconventional, technique to match the dozens of 807's I've accumulated to use in my 25L15 replica guitar amps. Bobby's supposition about the quiescent plate current is right, but only half of what I try to match. The other is the tube's amplification factor or transconductance - whatever my TV-2 tube tester measures when I test them. I use Wavebourn's technique of marking each with a Sharpie, noting the Ip and Gm values. Then sort all the tubes by both and select pairs with Ip's +- 1 ma or less and Gm's +-5 (arbitrary values; whatever the test meter reads).

Using these tube pairs has resulted in minimal adjustment of the bias balance pot on the amp.

I have encountered many 807's that exhibit the "runaway" current problem, though. And Unfortunately, all but one of the nearly 100 807's I've bought have bakelite bases, so the baking option won't work. They test OK; just don't work in the Standel PP circuit...

[Soonerorlater]

I once baked a cylinder head in the oven

Everything was going well when I tried this .........until my Mum found the guides in with the frozen peas

Wavebourn - How does baking valves differ from the roasting they get in normal service.


Brgds Bill