Replacement of RCA 8122 Tubes

Discussion in 'Amateur Radio Amplifiers' started by N2OTG, Jul 12, 2015.

ad: L-HROutlet
ad: l-rl
ad: Subscribe
ad: Left-3
ad: QSOToday-1
ad: Left-2
ad: L-MFJ
ad: abrind-2
  1. N2OTG

    N2OTG Ham Member QRZ Page


    As many readers will know, 8122s were (and are?) manufactured by several different makers besides RCA, including Burle, Eimac, and Penta (in China?). Each of these latter brands is currently, as of July '15, available at RF Parts and elsewhere. I'm thinking of replacing a couple of RCA 8122s and would consider substituting brands besides RCA, as 8122s are, in general, not getting any cheaper or easier to find.

    I understand that the brands besides RCA have varying characteristics and replacement of RCA 8122s with one of these other brands may require modification of equipment. I'm trying to avoid that. I think I've read that the Burle tubes came in different versions that were indicated by different-color ceramic envelopes and that were designed to emulate different phases of the original RCA production. So, maybe even the RCAs varied a lot. At any rate, the situation seems pretty complicated.

    Which brands of 8122 can serve as direct replacements for the RCA 8122? The equipment in question is a National NCL-2000. Does anyone have any experience with the Penta 8122s? Are they still in production? The Pentas seem to be among the cheapest brands available.

    Thanks for your attention. Any info would be greatly appreciated.

    Bob 73 DE N2OTG
  2. WB2WIK

    WB2WIK Platinum Subscriber Platinum Subscriber QRZ Page

    Carl KM1H might be an expert on this as he has NCL-2000s he keeps running.

    But I had not heard about Burle having different varieties of 8122s; I thought they were the original 'RCA formula.'

    AFAIK Penta doesn't manufacture any tubes. They used to a long time ago, but I think today they're just a reseller who has tubes private labeled for them. They're very local to me, in Chatsworth, and I've been there a few times. Didn't see any tube factory. Did see tube testing, though.
  3. N2OTG

    N2OTG Ham Member QRZ Page

    @WB2WIK Thanks. RF parts lists the Eimac and Burle as NOS but not the Penta. This makes me think that they're still in OEM production by somebody or other but I dunno.
  4. WB2WIK

    WB2WIK Platinum Subscriber Platinum Subscriber QRZ Page

    The Penta tubes are likely Chinese, but I really don't know.

    Their headquarters is about two miles from me and they're not manufacturing any tubes there.

    You could ask Merit at RF Parts -- call him on the phone, he loves to talk tubes.
  5. KM1H

    KM1H Ham Member QRZ Page

    All 8122's work in the NCL-2000 but they still need to be closely matched, you dont just plop any 2 tubes in and expect a clean signal and long tube life.

    Eimac had the 8122 and 8122W and some required a higher bias than the -45V available, its a quick mod.

    Burles come in 3 flavors signified by the dot color on the cooler; 2 out of 3 are fine without a mod.

    The Penta version is a RCA 8072 with a new cooler, you can usually still see the old date code in the same place as the 8122. Ive had two matched pairs since the 90's when they were done in the US, I doubt if it is the current Penta.

    Im currently up to 6 NCL-2000's:

    Prototype that now covers 160-10M including WARC, had since 1965

    Prototype I converted to 6M, in use since 64. It was going to be a product along with a transverter.

    A pair of fairly recent (maybe 10-15 years) near mint originals, one follows the SX-115/HT32B and the other a highly modified 75A4 and CE-100V both Ive owned since 65.

    8122 test jig used in production, got it at the 92 final auction

    The last is a very clean one with a shorted transformer that was a freebie, I had serviced it for over 35 years.

    National Radio 1963-69
    Service Tech, Service Manager, Sr Engineering Aide
    Member of NCL-2000 and NCX-1000 Design Teams
    HRO-500 restorations a speciality
    K9FLY and KU3X like this.
  6. K9STH

    K9STH Ham Member QRZ Page

    8072 tubes are still easy to find, and at reasonable prices, since they were used in a lot of Motorola and General Electric 450 MHz FM base stations and, with Motorola, even in one UHF Motrac mobile version. The "trick" is getting a proper heat sink to "convert" them to an 8122.

    I still have a number of brand new 8072 tubes that I obtained when I acquired the parts, and most of the test equipment, from a self-maintained two-way radio system that went to an Ericsson 896 MHz trunked radio system. I have "toyed" with trying to convert them to the 8122 but haven't found a source of heat sinks, especially in small quantities.

    Glen, K9STH
    N2EY likes this.
  7. KM1H

    KM1H Ham Member QRZ Page

    An Eimac 3-500Z type metal anode will fit the 8072, I use a piece of brass sheet and hardware to hold in place as it doesnt cover the 8072 anode completely, its also where the HV etc is attached. A horizontal fan does the cooling and in a NCL-2000 lash up it worked fine at 800W with no color and I didnt have the fan in a good place. Some day I'll do the conversion in a MLA 2500 as the sockets are the same and the cooling is in place; Id rather build a screen and bias supply than do a hack conversion to some Russian IMD generator.

  8. N2OTG

    N2OTG Ham Member QRZ Page

    @KM1H Thanks Carl. Not sure how I'm going to ensure good matching, since 8122s don't seem to be offered that way very often. I understand why a pair of tubes operating with only one bias control needs matching to avoid things like excessive IMD or one tube working a lot harder than the other. Theoretically, if I were to get hold of some 8122s (or 8072s) and a transconductance meter, about how closely would I have to match a pair?
  9. KM1H

    KM1H Ham Member QRZ Page

    National matched them to within a 1.5V Delta; Ive run to 2.5V before seeing a trace more IMD on the spectrum analyzer.
    You can also add a second bias pot and test point on the back and wire as two separate supplies. While you cant go crazy with different bias voltages it works well enough to separate into 3 categories similar to what Burle does with their color coding. The range Ive experienced with RCA, Eimac, and Burle production covers -25 to -46V which groups them in three 7V segments; of course you can overlap with any 6-7V variable. My 6M amp had been running a 29 and 36 1967 date codes until 2012 until they just flat wore out at 600W. While the idle currents were matched the transfer curves were different but not enough to bother the IMD. I also added the G3SEK circuitry in the mid 90's which improved IMD 5-6dB; others on the band say the signal is very clean using a phase noise free TS-830 as the exciter and an ancient modified Hallicrafters HA-6 transverter.

    NL7W likes this.
  10. N2OTG

    N2OTG Ham Member QRZ Page

    @KM1H Thanks. Suppose I were to rig up two pots, one for each tube. Would getting the idle current the same in each tube provide a decent match? There's only one meter for measuring idle current. If I were to take a tube out, say, and bias the remaining tube to 1/2 the recommended idle current with one pot, then repeat with the other tube and pot, would that ensure a close-enough match? I'm assuming the tubes are from the same manufacturer.

Share This Page