[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [TCML] Taylor TUBES



Ed Phillips wrote:
Comments at end.

jiml

The LANL report "High Power Microwave Tube Transmitters" (or similar..) from William North (available in various places on the web) covers all this stuff in detail. (FWIW, North just recently died, presumably not from HV, although I didn't ask)

Purpose of the note was to say let the buyer beware! The guy who wrote that particular note is in the business of "audio equipment" and tubes therefor [in other words, high power receiving tubes] which most definitely are not expected to arc over or short under any circumstances although equipment using them is usually fused in the input line to protect against catastrophic failures.

Context is everything, I suppose.


  The same can be said of
relatively low power transmitting tubes of the 833 class [< 5 kV plate voltage, < 1 kW plate dissipation]. Such failures would not be expected and equipment designed using them wouldn't have any stored energy constraints. I'm not aware of any higher power tube transmitters which have power supplies deliberately designed to protect against arcing but there may well be some.

Funny that should come up. Sometimes, the protection came as a byproduct of standard designs in days gone by. Consider something like a linear amp for hams. The power supply likely used a thermionic rectifier and used a choke/cap type filter and had terrible regulation (by modern standards) because of the series resistance, but that same terrible regulation has the side effect of limiting the fault current in a flashover (either internal or external: the spider crawling in).

Today, though, with semiconductor HV rectifiers, and cheap capacitors, someone might "improve" the design to stiffen up the regulation and reduce some of the losses in that choke, etc. Maybe even use a FET series pass regulator that's floated (after all the regulator only has to deal with a few tens of volts of ripple/sag). Now when the flashover or fault occurs, you've got a lot more stored energy, a lower impedance path to the energy source, and maybe a series regulator that can't take the full supply voltage across the pass device so it fails shorted.

The guy back in the 60s probably never really thought about all the inherent protection in the circuit... just designing with rules of thumb and following what folks had done before (the same way we design today, really), and doing some typical analyses to get it right.


Several members of the TCA group from which this note came are ex-Eimac and they have often described the great lengths to which Eimac went to pre-condition high voltage tubes to prevent sparking and arcs at voltages well above ratings. Shorts are never expected.

Not in the tube, but in the circuit around it, I think they're something that, except in a sort of experimental or cost sensitive (e.g. hacking) environment, one might casually expect. Perhaps not to the point of formally addressing it in a design, but just "not do something stupid". Even the ARRL handbook (which is sort of the epitome of cost sensitive hacking design: the unspoken assumption is that if you built it, you can probably fix it) talks about the value of a resistor in series with the B+ supply in case of shorts downstream.


  I am very
familiar with the design of high power MICROWAVE transmitters for radar purposes. Cathode or anode pulsed transmitters are inherently protected against arcing. However, with the advent of high power gridded TWT and Klystron transmitters which may operate continuously with applied HV matters become entirely different and such protection is mandatory. A rule of thumb we used at Hughes [in connection with X-Band radar transmitters] was the "aluminum foil test". The supplies were designed to dump [with crobar switch] HV fast enough in the event of a short that if the output was shorted to a sheet of kitchen foil the size of the resulting hole was very tiny.

Indeed. And now, with switching supplies, you can get HV with small stored energy, and it's almost self protecting.The primary side can shut down instantly, and there's just not much stored energy... if the PWM rate is 50kHz, you only need enough to supply the output power for 50-100 microseconds or so. Even at a kilowatt, that's 0.1 Joule. Hardly even exciting if you short it.


With VTTC's, particularly those built by enthusiasts without a lot of tube experience, failures due to circuit design can be expected and the consequences are merely monetary if they do get poor tubes. The general reputation of cheapie Chinese transmitting tubes is very poor with respect to gassing and low emission and I was trying to suggest that guys look for something better if they can find it.

I agree.. and because of the generally terrible and uncertain quality of tubes these days, more sophisticated designs (or at least ones that tolerate failures without catastrophe) are a good idea.

It's the "pole pig to diode string to 10 uF, 10kV filter capacitor" sorts of the designs that are doomed to excitement.



_______________________________________________
Tesla mailing list
Tesla@xxxxxxxxxxxxxx
http://www.pupman.com/mailman/listinfo/tesla