8417 compatible with 6550 (or all the other octal power tubes)?

[LooseChange]

The 8417 looks to be pin compatible with all the other octal power tubes we all like. Any thing I should be aware of when subbing a 6550?
Thanks!

[Tiny_Daddy]

Looks OK to me.

[PRR]

8417 has MUCH higher gain; which indirectly means it is more fragile than 6550. Overheat a little too much, the grids sag and short-out.

Also the bias is VERY different. Where a 6550 might need -40V bias, 8417 wants -20V bias. Yes, you could drop a 8417 in a 6550 socket and it would just idle cold. Put a 6550 in a 8417 socket and it will red-plate bad until you majorly increase negative bias. (Most 8417 amps didn't have near enough range to handle this.)

There are no new 8417. They were made into the 1970s by GE and Sylvania and that's it. They are highly sought-after.

I used to make a buck with twenty 8417s making upwards of 700 Watts. When well treated AND you can buy good new tubes at the local TV-repair shop, is a good tube. Beaten in a guitar amp, NO replacements readily avaialble, I would leave it to the few who "must" have 8417. (There were a couple stage-amps and many dark-era Hi-Fi amps which only took 8417.)

[LooseChange]

The amp I am working on is a Sano Supernova. It has two 8417's running at 500v on the plates and cathode bias with a 125 ohm cathode resistor. I do have a few pairs of the 8417's to test before I need to go radical with the 6550's.

Thanks for the great info!

[kagliostro]

Take care of those 8417 you have

a NOS 8417 is very expensive - $ 79.95 each at Tubestore

http://search.store.yahoo.net/thetubestore/cgi-bin/nsearch?catalog=thetubestore&query=8417&.autodone=http%3A%2F%2Fthetubestore.com%2F

and also they are out of stock at the moment

K

[PRR]

$160 to retube a 30W-40W amp is absurd.

Change to 350 ohm 10+Watt common cathode resistor and drop 6550s (or KT88s) in.

(If the existing 200 ohm is in good shape, you could add 150r 5W in series.)

The cathode voltage should be about 45V.

45V/350r= 0.129A cathode current.
Estimate plate current as 90%-95% of cathode current: 0.122A.
Plate-cathode voltage is 500V-45V= 455V.
455V*0.122A= 56 watts plate dissipation per pair, 28W each, comfortable for cathode-bias 6550.

You "could" then drop cathode resistor to get 35W or 42W dissipation per plate. Up to a point, power output may rise; past that point, you are just heating the room. Stop before 0.18 Amp total cathode current (42W per plate). 0.15A is perhaps safer. This may be around 270 ohm common cathode resistor.

Oh, hel.... 300 ohms 10W is probably close enough, but measure cathode voltage and estimate plate dissipation before you put it in service.

The Sano's driver has more than enough drive headroom to handle the larger grid-swings needed by 6550.

Any way you do it, this amp is a cooker. I was very pleased with Sovtek 6550s in another hot-bottle 6550 project. Good as 1970s Dyna/TungSol coke-bottle 6550s but less creaking on cool-down.

*Good* EL34 *might* work a while in this amp. Start with 300-350 ohms bias. Calculate to stay just under 30W Pdiss. (EL34 have high screen current, estimate plate current near 85% of cathode current.) While this is really pushing EL34 ratings, modern for-guitar EL34 will probably stand it for over 1,000 hours, and the price sure is lower than 6550.