I have a box full of these, but haven't tried them yet. That finned heat sink IS the plate connection, so be careful. Sockets are hard to find and usually designed to be mounted in an RF cavity. Forced air cooling is required.
Ham radio operators are using these for moonbounce transmitters at 1296 MHz, they are worth something to them. I am trying to work a trade with mine. I have a friend who operates these tubes in a bath of distilled water to double (or more) the plate dissipation. His amplifier puts out 1KW at 1296 MHz. It uses 4 tubes, and operates on 2200 volts.
Ham radio operators are using these for moonbounce transmitters at 1296 MHz, they are worth something to them. I am trying to work a trade with mine. I have a friend who operates these tubes in a bath of distilled water to double (or more) the plate dissipation. His amplifier puts out 1KW at 1296 MHz. It uses 4 tubes, and operates on 2200 volts.
RF transmitting tubes can be fun to play with for audio projects as long as a few factors are taken into consideration. There is no reason for these tubes to “sing like a tree full of birds" (a great metaphor) if reasonable design precautions are taken. These tubes achieve high bandwidth by way of several techniques: series inductance is radically reduced by the use of planar or ring terminals instead pins and wires. Also a tube like this was designed for short “transit time”. The transit time of electrons between elements can create a low and resistive input impedance at RF frequencies. However, this factor would have no direct relevance at audio frequencies. Interelement capacitances, while designed to be low, are not vanishingly low, due to the constraints of the other designs parameters. Note that for this tube Cgp=2pF and Cgk=6.5pF which is not bad for a 100W triode, but these values could not be ignored at microwave frequencies. These capacitances would have to be incorporated into the tuned input and output circuits at microwave frequencies. Sometimes a technique called neutralization is used to cancel out a part of Cgp. An audio driver would have to contend with the Miller capacitance of up to 150pF to 200pF, depending on plate loading and resultant gain.
So how do you keep these tubes from singing? Basically you use good RF design practices such as are described in the ham manuals like the ARRL Handbook. Keep leads short and grounds close by, well defined and planar if possible. All PS decoupling should be by short leaded, low inductance capacitors. Use a grid stopper (maybe even a ferrite bead) close to the grid. You don’t have to go crazy with these things; just being fairly reasonable is usually enough to prevent oscillation. These same techniques are good for any tube, especially for small RF tubes like the 6DJ8 and 417 that like to sing.
Other unique factors that must be considered: Often RF transmitting tubes are relatively high impedance devices, needing high plate voltages and having relatively high plate resistances. The 2C39 is no exception; its plate resistance will probably be somewhere above 4K or 5K, depending on bias current. This is moderately high for a 100W audio triode. The output transformer will be need a high turns ratio, as well as a high voltage insulation rating. These tubes are often designed to draw grid current, but for audio we usually (but not always) avoid grid current to prevent distortion from the driving circuit. This usually constrains us to higher plate voltages and lower plate currents if we want to approach the full plate dissipation rating.
Still, the 2C39 might be a worthwhile project if suitable forced-air sockets can be found or made and if you are willing to accommodate its other peculiar needs.
So how do you keep these tubes from singing? Basically you use good RF design practices such as are described in the ham manuals like the ARRL Handbook. Keep leads short and grounds close by, well defined and planar if possible. All PS decoupling should be by short leaded, low inductance capacitors. Use a grid stopper (maybe even a ferrite bead) close to the grid. You don’t have to go crazy with these things; just being fairly reasonable is usually enough to prevent oscillation. These same techniques are good for any tube, especially for small RF tubes like the 6DJ8 and 417 that like to sing.
Other unique factors that must be considered: Often RF transmitting tubes are relatively high impedance devices, needing high plate voltages and having relatively high plate resistances. The 2C39 is no exception; its plate resistance will probably be somewhere above 4K or 5K, depending on bias current. This is moderately high for a 100W audio triode. The output transformer will be need a high turns ratio, as well as a high voltage insulation rating. These tubes are often designed to draw grid current, but for audio we usually (but not always) avoid grid current to prevent distortion from the driving circuit. This usually constrains us to higher plate voltages and lower plate currents if we want to approach the full plate dissipation rating.
Still, the 2C39 might be a worthwhile project if suitable forced-air sockets can be found or made and if you are willing to accommodate its other peculiar needs.
Seems to be an awfull lot of trouble to get less than 1Watt from the 2C39.
Class AB2 seems more reasonable from an output-power point of view.
If driven from a cathode-followers low Z-out with a CCS in the tail quality won't suffer imo.
I did just that with 2C34 which is NO cousin of the 2C39 in spite of the typeno.
About forced-air cooling that won't be so practical with audio unless you put the speakes in one room and the amp in another
Here's what I'll do to deal with that particular problem:
http://www.diyaudio.com/forums/tubes-valves/212604-russian-triodes-15-arriving-few-days.html
rgds,
/tri-comp
Class AB2 seems more reasonable from an output-power point of view.
If driven from a cathode-followers low Z-out with a CCS in the tail quality won't suffer imo.
I did just that with 2C34 which is NO cousin of the 2C39 in spite of the typeno.
About forced-air cooling that won't be so practical with audio unless you put the speakes in one room and the amp in another
Here's what I'll do to deal with that particular problem:
http://www.diyaudio.com/forums/tubes-valves/212604-russian-triodes-15-arriving-few-days.html
rgds,
/tri-comp
I think it makes more sense to sell them to someone who can use them for their intended purpose. Even more than most valves, making an audio amp with 2C39 requires microwave-level techniques. If you knew how to do this you would not need to ask. There could be serious safety issues - not just high voltage but high power RF if it starts singing. At the very least, RF burns can be painful and take months to heal.
I have built lots of 811A amps using 600 volts 100 to 120 watts based on Record cutting amps, the P-P load is 1.6k to 2.2 k; I have not used the 2C39 but it has similar mu to the to the DA 42 but max voltage is about 500 volts as you need to bias them at 70 Ma. I find 12DB feedback is plenty; A friend i built an 811a amp for using Russian 811A's he has used for 20 years 811's test as new, your lucky to get 2 years out of 6550's at 560 volts. I use 6BL7's as cathode followers also at 600 volts for the 811a hope to try a mosfet. Max voltage for the 811 is 500volts. Load for 811A at 750 volts is 5k, but can not use 70 ma best to use 572b's. I don't think many Hams use 2C39's any more use to FETS In have built amps using the ceramic TUBE 3CX300A class A 500 volts a computer fan keeps them cool and no Micro -phonics.
Could be worth to try them as triodes, cathodefollower even, they have low g1/g2 gain and decent gm, should work with up to 500V or so.
There are tubes from the same family that do not need high veclocity aircooling, the anodes are directly clamped to a heatsink for further cooling by other means.
If aircooled, and at reduced Pa the temp could be even above 100deg and the heatsink would not really have to be enormous.
Could work nicely with a silent blower.
There are tubes from the same family that do not need high veclocity aircooling, the anodes are directly clamped to a heatsink for further cooling by other means.
If aircooled, and at reduced Pa the temp could be even above 100deg and the heatsink would not really have to be enormous.
Could work nicely with a silent blower.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.