ad: Radclub22-1

Homebrew HF amplifier mis-match

Discussion in 'Homebrew and Kit Projects' started by KM4HGU, Jul 15, 2017.

ad: L-HROutlet
ad: l-rl
ad: L-rfparts
ad: Subscribe
  1. KM4HGU

    KM4HGU Ham Member QRZ Page

    I ran across a schematic for an amp using an 813 tube. I had pretty much all the components needed in a box of junk. So, I decided to give it a go.
    Got it all assembled, and fired it up. No majik smoke. burned in the tubes, and shot the HV to it. Still no majik smoke, and the diodes giving a nice purple glow. HV tested to 1kv. Shut it down.
    So far all seems well. I don't know why, but I decided to hook up my antenna analyzer to the RF input to make sure that there would be a 50 ohm impedance, prior to hooking up the driver. The analyzer shows
    66 ohms, and swr of 1.5:1. Hmmm, sez I. I fire up the filaments, and key it, impedance jumps to about 75 ohms, swr off the scale. I know that the analyzer doesn't have enough power to drive the amp, (something in the mw range) But I decided to give it a go anyway, and shot the HV to the plates. Of course there was no change.
    But my concern is, that with the amp off, and dummy load connected, the analyzer is showing me a 66+/- ohm impedance. And an swr of 1.5:1. Which is what my transmitter will see when hooked to the amp.
    There's something between the RFin, and the RFout, causing this, but I can't find it.
     
  2. SM0AOM

    SM0AOM Ham Member QRZ Page

    You cannot measure the actual input impedance of especially a Grounded-Grid amplifier with an antenna analyser.
    The input impedance is non-linear and highly drive dependent, especially if there is no cathode tuned circuit.

    To interpret your measurements, it is necessary to know more about the circuit, but if you get a "cold" impedance of about 66 ohms, you are seeing the circuit losses.

    73/
    Karl-Arne
    SM0AOM
     
    N2EY likes this.
  3. G0JUR

    G0JUR Ham Member QRZ Page

    If you mean you see 66 ohm and swr 1.5:1 without the amp keyed I would clean the relay contacts.
     
  4. KM4HGU

    KM4HGU Ham Member QRZ Page

    Thanx guys. That's what it is, "cold" @ 66 ohms, and 1.5:1, amp not even plugged in.
    I suspected the relay right off the bat, and even though it is new, I replaced it with another. (I have several)
    I guess I'll just hook it up to the xmitter, and see what happens.
     
    KG7SWP likes this.
  5. KM4HGU

    KM4HGU Ham Member QRZ Page

    Update: After checking various suspect items, and making a few changes, I hooked up the Xmiter. Found it to be a good match "cold" but, when hot,
    on keydown it went to 1.5:1, then creeped up, as the plate voltage went down. I found that I had low voltage on the HV supply, and bad regulation as well. I suspect the amp is fine. I've been working on the HV supply, off and on, for the past weekend.
    I managed to get the proper voltage (2kv+) by hooking up two transformers in series. I'll get better regulation by re-arranging my filter caps.
    I just need two more 872 Diodes, and another filament transformer. Or go solid state with the rectifier. (which opens yet another can of worms) I'll keep y'all posted.

    73
    KM4HGU
     
  6. K2XT

    K2XT Subscriber QRZ Page

    I'm not sure about this, so it's more of a question or something to discuss than a statement of fact.
    But you know, when an amp is keyed, with hv on, etc. it creates noise, broadband noise that can be heard in a receiver.
    An antenna analyzer operates at low signal levels, around +10 dbm, so it's detector would be looking at reflected power levels in the -10 dbm range. If the 813s were generating noise in that range it could confuse the analyzer.
     
    N2EY likes this.
  7. WA9WVX

    WA9WVX Subscriber QRZ Page

    I think you'll need a tuned input circuit on each band you're operating on HF and a multiband ceramic switch so the
    L-C circuits don't interfere with each other and the same holds true for output circuits where you'll need a tapped coil for each band too. I checked my 1977 ARRL Handbook and it suggested 2500 VDC on the Plates of both 813 in parallel should produce 650 to 700 W PEP Output running them in Class AB2 for Linear Operation. You'll probably need HV +3KV Filter Capacitors, a HV Choke & a Full Wave Bridge circuit with Solid State HV Diodes in your Power Supply. JMHO

    Dan
    WA9WVX
     
  8. KM4HGU

    KM4HGU Ham Member QRZ Page

    Here's the amp schematic, The power supply schematic is pretty much a voltage doubler, fed by a 1.1kv (sec.) transformer.
    I don't like doublers, And the highest PIV solid state rec. diodes that I have found are only 1kv. so I gotta run several in series with a resistor of about a half meg. and several watts. I have on hand a couple of 10h filter chokes, and a couple of 872s, Some 470mf 450v caps, and a handfull of 20k, 10watt resistors.
    The max plate voltage for the 813 is 2kv for ICAS, and 1.6kv for CCS, in gg configuration (plate modulated). I'm sure that one could run more for ICAS, but at what cost?
    Anyway, my two transformers in series put out about 2.4kv no load. After the rec./filter I get about 2.2kv.
    I'd rather find a 4kvCT transformer, and be done with it. But apparently they're kinda far, and few between.

    Apparently I can't post the schematic. It has the wrong file extension. (bmp) But it's a standard, Run-of-the-mill grounded grid single 813 amplifier, with LC input, and PI/L tank. I just need to get a better regulated 2kv supply.
    Well, I managed to convert to PDF so here it is:
     

    Attached Files:

    Last edited: Aug 14, 2017
  9. WA9WVX

    WA9WVX Subscriber QRZ Page

    Dave,

    That schematic doesn't appear as Grounded Grid circuit and although it has a L-C Tuned Input some how you must be able to align the Input circuit for each band of operation. Here's a website from AA8V with a schematic diagram:

    http://faculty.frostburg.edu/phys/latta/ee/wing813amp/813schematic.html

    Here's W8JI Grounded Grid RF Amplifier https://www.w8ji.com/grounded_grid_amplifiers.htm

    Hammond (Transformers) Manufacturing http://www.hammondmfg.com/5cpwr.htm

    Surplus Sales High Voltage Transformer http://www.surplussales.com/item/_tp/600086-512-001.html

    Surplus Sales High Voltage Transformer / Power Supply http://www.surplussales.com/item/_ps/ar-150a.html

    Surplus Sales High Voltage, High Current Solid State Diode Assembly

    http://www.surplussales.com/Semiconductors/Diodes-Rectifiers-3.html

    (SDI) 67-6257 IR Double Diode, 20kv, 2 amp, 16"L on Page # 3 I can't believe you could ever blow or short this Diode Assembly.

    There are my thoughts about building a homebrew RF Amplifier for Grounded Grid Operation.

    Dan
    WA9WVX
     
  10. SM0AOM

    SM0AOM Ham Member QRZ Page

    This very much a grounded-grid amplifier, and the circuit comes straight out of the 70's W6SAI Radio Handbooks.
    A cathode circuit of this kind would work well on one or maybe two adjacent bands, but would be impractical over the full tuning range of the amplifier.

    For a workable circuit Q in the parallel-tuned cathode circuit, a good approximation is about 5-20 pF/metre wavelength, so 500 or 800 pF would be OK for 80m with a single 813.

    If you just increase the frequency without altering the value of the coil, the Q will decrease and the "flywheel" properties of the cathode tank vanish.

    It is "good engineering practice" to have bandswitched cathode circuits.

    73/
    Karl-Arne
    SM0AOM
     
    WA9WVX likes this.

Share This Page