ad: Schulman-1

Simulating tube amplifiers

Discussion in 'Amateur Radio Amplifiers' started by W8JI, Dec 29, 2011.

Thread Status:
Not open for further replies.
ad: L-HROutlet
ad: l-rl
ad: Left-3
ad: Left-2
ad: L-MFJ
ad: MessiPaoloni-1
ad: K5AB-Elect-1
ad: Subscribe
  1. W1QJ

    W1QJ Ham Member QRZ Page

    I never bought a Kenwood TL-922 because I could never rationalize the cost verses a Drake L4B or Sb-220. However, when I did get to look at one when i was asked to repair it I immediately was shocked to see the mechanical layout as it would pertain to RF. I noticed ALL the drawbacks as pointed out by Tom. ALthough it DOES work, some simple adherence to sound practices could have made it a better "rip off" design. The better strength of the chassis is a big improvement over the SB-220. Try to lift one comapred to the SB-220. It's solid and strong. Thats where they made improvements, but kept the downfalls of all the rest. Although Denny Hadd (had) his faults, he knew enough to directly ground the grids. You won't see any of his amps with floating grids. His faults were of a different breed. Now for the FL 2100 et al. Now this is a nightmare, this is about the only amp I ever saw that could be unstable just sitting on the bench. Fortunately I noticed this problem on someone else's dime. Hence, I have never bought one of these. It makes me laugh how this person tried to stablize this amp. Gives me a pick up when I am down in the dumps. When Eimac first came out with the 3-400/3-500 tubes they offered a simple amplifier design. Their recommendation is clearly a GROUNDED GRID design. All you have to do is look around the internet and you can see this schematic offered by Eimac themselves on how to use the tube ( they designed )in GROUNDED GRID design. Also, as Tom pointed out old things die hard. Many things carry over from old design. Back in the 70's when I was working on car repairs it was a big joke to look at the next generation of engine design to deal with emissons as set by the EPA. Remember the "smog pump", "anti back fire value" etc.? This is how they handled unburned gas emissions. Later we found out the answer to all this was to have a more complete combustion burn. When this problem was solved all these "band aid" add ons disappeared. Of yore, any 300 plus horsepower engine was a gas guzzler, today we have 400 plus HP engines getting 25+ MPG. One fellow in tulip land still buys into the single point ground, this is evident on his website with the TL-922/SB-200 mods. How a single thin wire connection could compare with a full chassis ground plane is beyond me. Want the perfect amp? Build monobanders with all parameters specific for that band. You'll be surprised.
     
  2. G0HZU

    G0HZU QRZ Member

    Getting back to modern simulation and modelling methods I'll ask a slightly more interesting question to Tom


    Tom, if I gave you a sealed box with a single RF input connection and I didn't tell you what passive network was inside it and just told you it was a passive load (actually a really complex passive circuit with hideously poor wiring practice and lots of capacitors and inductors and resistors) and you weren't allowed to open it could you still accurately model the load on a computer up to VHF?

    To keep the task in perspective you would only be allowed (a very generous) 5 minutes from being handed the mystery box to having an accurate model on the computer.
     
    Last edited: Dec 30, 2011
  3. AF6LJ

    AF6LJ Ham Member QRZ Page

    How anybody could conclude that single point grounding could ever be superior in any way over a chassis doesn't make any sense to me at all. Not in audio not in RF not anywhere.
    On the subject of band switching; Has anybody come up with any credible alternatives to rotary switches?
    The contacts and the geometry are the weak link. Roller inductors have their own set of problems not the least of which is all that "L" hanging out in space when you are on fifteen or Ten meters.
     
  4. W1BR

    W1BR Ham Member QRZ Page

    off topic, deleted
     
    Last edited: Dec 30, 2011
  5. W8JI

    W8JI Ham Member QRZ Page

    A single port should be easy to simulate.

    Simulating a complex system of transmission lines and reactances, all of which change with band and control settings, and determining what the anode, cathode, and grid of the tube looks like to the electrons on all frequencies, and accounting for layouts, is not workable as a general principle.

    It is far faster just to run tests.

    I haven't even found a good model for common tubes outside of low frequencies.

    73 Tom
     
  6. W8JI

    W8JI Ham Member QRZ Page

    A roller inductor large enough to cover 160 meters is interesting to look at as the inductance is cranked down towards ten meters. :)

    As the lower frequency resonance moves up, a higher frequency resonance moves down. If they ever cross, it's trouble.
     
  7. AF6LJ

    AF6LJ Ham Member QRZ Page

    As I visualize it; where they cross, you end up with a lot of RF voltage with nowhere for it to go in the unused portion of the roller inductor.
     
  8. AG6K

    AG6K Ham Member QRZ Page


     The entire grid-filament structure in a 3-500Z that sustained a filament-grid short during a big bang:
    3-500_bent-fil.jpg
    No arc mark is seen on the grid. This and the fact that the bang was loud tells me that the arc was outside tells me that there was probably an external arc. Also, before I removed the envelope I performed a leakage check at 8000v and the vacuum was good. I do not see how this short could have involved a gas leak Susan.
    • Rich, ag6k
     
  9. AG6K

    AG6K Ham Member QRZ Page


     The entire grid-filament structure in a 3-500Z that sustained a filament-grid short during a big bang:
    View attachment 82158
    No arc mark is seen on the grid. This and the fact that the bang was loud tells me that the arc was outside tells me that there was probably an external arc. Also, before I removed the envelope I performed a leakage check at 8000v and the vacuum was good. I do not see how this short could have involved a gas leak Susan.
    • Rich, ag6k
     
  10. G0HZU

    G0HZU QRZ Member

    How about a mystery passive box with two RF connectors for network input and output.
    Inside the box could be your anode choke, DC blocker and wires plus band switch and tuning caps and lots of wires and coax all set to an unknown setting.
    It can still be modelled up to UHF with pretty good accuracy. Both for insertion loss, return loss and port impedance right up to UHF. It will measure out very similar to real measurements on test equuipment way beyond the frequencies we are concerned with.

    It only takes minutes to get the mystery model ready on a simulator if you use modern techniques.
     
Thread Status:
Not open for further replies.

Share This Page