Need help identifying tube

[profdc9]

I am trying to identify a tube, in particular, if there is now a source of such a tube from China or Russia as these are likely not made anymore in the USA.

The tube is from a 10 kHz - 250 MHz, 2500 W Amplifier Research Model 2500L MII amplifier. It is one of eighteen tubes in the amplifier finals. The tubes are oil cooled as can be seen by the nipples on the top. The amplifier itself dates from the early 1990s.

The tube is marked

"Amplifier Research"
1003787 - 101C

I have also included the schematic. I was not able to find the item in the BOM corresponding to these tubes however.

Are any similar tubes used in amateur or broadcast amplifiers?

73,
Dan
KW4TI
 

[geggi1]

Have a look at this site.
https://www.vintage-radio.net/forum/showthread.php?t=79910
And this one.
https://www.bunniestudios.com/blog/?p=6175
It could be 4cx250 with some kind of liquid cooling device.

[Wallace Gasiewicz]

I agree probably a 250 or even a 350.
Does not look like the "cooler" can be easily removed, for a look at the tube itself

There are several linears that use these tubes but a lot of them have too much amplification to meet the FCC standard for ham amplifiers
I have an old military radio that uses two 250 or 350 tubes in the final (depends on the model)

[mag_therm]

Licensed amateurs can build or modify amplifiers "without concern for the certification rules" that manufacturers and dealers have to follow. Such non certified amplifiers can also be resold by dealers but only to licensed amateurs.
Reference: Extra Class License Manual section 3.6 and exam questions E1f03 and E1f12

However operation still has to be within all of the  license class, output power and spurious  emission limits applicable to the band and mode.


Edit Above is for USA, .Other countries may differ.

[T3sl4co1l]

NOTES
1.0
ALL TUBES ARE 8122

Does seem to be a conduction cooled variant.

Also, holy shit, splitter into matched delay lines into 6x grid lines each, all feeding one common plate line in order?  Amazing; that thing must've been a terrific pain to tune (or design)!

Tim

[CaptDon]

That 11-pin version came in thermal link conduction cooled versions also. There is a water cooled version of the 4CX250 but the pinout is not the 11-pin version like yours. I think the National NCL-2000 linear amplifier used a pair of the 11-pin air cooled versions. The cost of tube replacement far exceeds the value of the amplifier in today's marketplace and I think you would play hell even finding replacements except through A.R. and they will surely charge 10X their original purchase price from Eimac or Amperex. The A.R. stuff was really untuned broadband and often used in military labs and E.M.I. testing labs where 'Cost was no object'. I was never sure why Eimac changed the basing of the 4CX600 series??? It is physically the same base as the smaller brothers (4X150, 4CX250, 4CX350) but with a different pinout requiring a special socket only available through Eimac. Naturally the Collins military gear used the 4CX600JB tubes with their astronomical price tag!!!! It would have been far cheaper to use 4CX1000 or 4CX1500 variants!!! The first economical water cooled variant across the Eimac line was the 4CW20000.

[coppercone2]

the air heat exchanger might be able to be modified for oil since the specific heat of oil kinda sucks anyway

[JSTCanuck]

Hi,

Did you have any success with identifying this tube with the liquid cooling jacket?

Does this metal jacket pull off the anode of the tube?     Wondering if the cooling jacket could be removed and used on a new tube?

Someone else has commented that they might be RCA 8122 series of RF tubes.

I'd be interested in experimenting with liquid cooling.

John 
 

[Roehrenonkel]

Hi,

8122,
probably modified.

[David Hess]

Also, holy shit, splitter into matched delay lines into 6x grid lines each, all feeding one common plate line in order?  Amazing; that thing must've been a terrific pain to tune (or design)!

I wonder why they did not just make it one array instead of three separate arrays.  Maybe a single layout was prone to oscillation, or it was too mechanically difficult?

Tektronix only used distributed amplifiers up to 100 MHz.

[T3sl4co1l]

You mean a single chain of 18?  Probably, grid losses would be too large for that.  Whereas the delay lines can be made to match... with difficulty.

Hmm, insights I didn't have last time... interesting.

I'd have to check datasheets and see what GBW is like, but it seems likely they obtain more total bandwidth and gain this way, than by ganging stages.  The plate-line loss should be quite low, allowing it as a single chain.

The bandwidth and power make sense given how many tubes are required regardless, while a distributed amp gets the bandwidth up, something a single or couple big transmitter tubes can't do.

Tim