Parasitics and tube damage.

Discussion in 'Amateur Radio Amplifiers' started by W1QJ, Dec 25, 2011.

Thread Status:
Not open for further replies.
ad: L-HROutlet
ad: l-rl
ad: Subscribe
ad: Left-3
ad: Left-2
ad: K5AB-Elect-1
ad: L-MFJ
ad: MessiPaoloni-1
  1. AF6LJ

    AF6LJ Ham Member QRZ Page

    This reminds me of two situations at Loral Corp.

    Back a long time ago Communications Transistor Corporation (CTC) made some of the best high UHF and low microwave transistors around. Like Eimac they were a Varion (sorry spelling) owned company.
    They had a reliable and consistent product you could teach high school students to tune and build S-band 25W emitter driven power amplifiers. (I'm serious those parts and the design were that good)
    Well unfortunately TRW bought them out and lost the recipe.........

    There were a few months of hard times for us because good parts were few and far between, and one of your contracts calls for fifty-five complete high power (40W) S-band video transmitters every month.....

    The term for TRW was Trashy Radio Work for a time then until they got their act together.

    Well you know Motorola thought they were GOD's Gift to RF techonology and bought TRW's semiconductor division.
    Motorola's experience past 1.2GHZ was lacking for a time also. They couldn't make a good part to save their pitiful life for a time, but then finally got it right.

    Now I know we put TRW under a lot of pressure to preform after they bought CTC....
    Every one of those high power video transmitters used five of one type of transistor and two of another in the stages following the VCO which used a Microwave Associates transistor.

    Just because another company starts to build a part doesn't mean it's going to be as good as the old part.
    It usually won't until or if they get their act together.
     
  2. AG6K

    AG6K Ham Member QRZ Page

     Low VHF-Q parasitic suppressors can be made from copper wire if a Low-L, high enough dissipation @ 28MHZ R-supp is available, or if one is not, a similarly low Q suppressor can be made with an available R-supp of lower dissipative capability if a share of the 28MHz energy is shifted to L-supp by using resistance-wire instead of highly-conductive copper-wire with a suitable at 28MHz. - note - at DC resistance wire is c. 50x more resistive than copper wire. At 50MHz resistance wire has about 5.5 x the R of same-gauge copper wire.

    “The combination of both resistance and inductance is very effective in limiting parasitic oscillations to a negligible value of current.”
    - - F. E. Handy, W1BDI 1926 Ed. The Radio Amateur's Handbook, p. 72
    •• quote dug up by David Newkirk.

    • Rich, ag6k
     
  3. AG6K

    AG6K Ham Member QRZ Page

     the issue is not bad welds, the issue is filaments that could not have been bent when they were welded to the fil. headers or they would not have spanned the distance between headers.

     indeed Pete,indeed.
    • Rich, ag6k.

    "It ain't what you don't know that gets you into trouble. It's what you know for sure that just ain't so."
    — Mark Twain
     
  4. W1BR

    W1BR Ham Member QRZ Page

    Hi Rich

    I don't recall where it was claimed that the filaments were "bent." Given the extremely tight spacing between the filament
    and grid structure, it is feasible that the filament, or even the grid structure, may not have been perfectly aligned during welding.

    Pete
     
  5. W8JI

    W8JI Ham Member QRZ Page

    ....or may not have been stress relieved, or the filament properly processed. They are dozens of reasons why a filament or cathode might contact a grid.
     
  6. AG6K

    AG6K Ham Member QRZ Page

     The explanation that makes sense is that the tungsten filament had to have been bent when it was hot or it would have broken, and it had to have been bent after it was welded to the headers or it would not have bridged the span.
    • Rich, ag6k
     
  7. AG6K

    AG6K Ham Member QRZ Page

     
    Last edited: Dec 29, 2011
  8. VK4TUX

    VK4TUX Ham Member QRZ Page

     
  9. W8JI

    W8JI Ham Member QRZ Page

    I think I'm finally understanding this now.......see if this is right:

    The only thing, or nearly the only thing, that ever makes a tube or anything else in an amplifier fail is the lack of a nichrome suppressor. The only source for nichrome suppressors is also the only source proving that every failure is caused by lack of nichrome, but that is because that source knows more about tubes than the people who make tubes or use tubes by the thousands, or is the only source honest enough to admit what the problem really is.

    Even the manufacturer of the tubes, when they are having problems with tubes, refuses to admit the real problem, that almost everything is cause by a lack of special suppressors. They are in denial so deeply they blame the problem on production difficulties, when really all problems could be corrected by adding nichrome. To save embarrassment, not admit a mistake, or because of secret payoffs or secret pressure, they just continue to claim a manufacturing problem with the tubes, until the wave of parasites that hits only one tube batch, goes away.

    Do I have that right?

    73 Tom
     
  10. G0HZU

    G0HZU QRZ Member

    This thread is going downhill rapidly again...

    What is missing from the bigger picture is a modern technical analysis of why the valve oscillates at the frequency it does and how/why the anode suppressor improves stability and also what other factors internal and external to the valve can aggravate instability.

    The best way to do this is to model it and simulate it. Even if it means simulating just parts of the system first. People can scorn simulations and say 'why bother with simulations?" but if you can't create a (reasonable) working model of the system on a computer then you perhaps don't understand the system in the first place.

    Often it takes several iterations of analyse/measure/revise models/resimulate to get close to the truth but what I find a bit puzzling is that the fundamentals for this mode of oscillation are quite simple to grasp?

    All the clues are there via the manuf datasheet and by looking at typical components used. The parts hardest to model (apart from the valve) could be the big tuning caps and also the input circuits to the cathode.
    But it should be possible to get a good grasp of the major issues via basic initial models for the major parts of the circuit?

    Has anyone ever done this?
     
Thread Status:
Not open for further replies.

Share This Page