High Gid Current

Discussion in 'Amateur Radio Amplifiers' started by NE7X, Jun 5, 2011.

Thread Status:
Not open for further replies.
ad: L-HROutlet
ad: l-rl
ad: Left-2
ad: Left-3
ad: L-MFJ
ad: MessiPaoloni-1
ad: Subscribe
  1. W8JI

    W8JI Ham Member QRZ Page

    Here is the right schematic with mods. This will work as drawn, except I am unsure of meter movement values for plate current meter.

    Henry1.jpg
     
  2. KD0CAC

    KD0CAC Ham Member QRZ Page

    Great thread !
    Now I'l have to keep an eye out for a project of my own :)
     
  3. NE7X

    NE7X Ham Member QRZ Page

    I have the 2K Classic wiring 100% to W8JI diagram (minus the “Fault Limit” resistor). I tipple checked and validated the wiring. I also have the 100v zener D105 in circuit and does not blow/short.

    2K Classic
    Keyed, no drive, USB
    Ig meter reads backwards
    Ip meter read 500ma (half amp)

    Keyed, speak into microphone USB ~100 watt drive
    Ig meter reads ~300ma (500ma full scale)
    Ip meter pegs, full scale (10A full scale)

    Reference my 2K-4 and 3K-A amplifiers
    Keyed, no drive, Ig = 0ma (500ma full scale) / Ip = 1A (10A full scale)
    Keyed, speak into microphone, Ig = 250-300ma) / Ip = 4-5A

    Meter movement values per the parts manual show 1ma movement. Both meters have the same part number stamped on them, FS1MADC. The Ig meter is also used for the HV and it shows 2300/3600 volts, so that part of the circuit appears to be working OK. Maybe I need different Ip shunt values and/or different R117/R118 values? If so, how do I identify what values they need to be?

    Both meters have a .01 across them. This is not shown in the schematic, however both my Henry 2K-4 and 3K-A have them, so I am assuming this is ok.

    I have seen in other amplifier schematics where they have two diodes, reversed connected across the meters. What is this for? Is this something maybe to try? The original 2K Classis schematic had a single diode in series with the Ig meter which has been removed. 2K-4 and 3K-A does not have.

    73s, Thomas NE7X…
    http://ne7x.com
     
  4. W8JI

    W8JI Ham Member QRZ Page

    You either do not have it wired the way I posted, or you have a bad component. Backward grid meter is an alert. :)

    If the wiring is as shown, the only component or problem than can cause the Ig meter to read backwards is a short or breakdown from the tube filament to chassis. Maybe you have an intermittent bad tube, or maybe some other problem in the RF deck.

    This all ASSUMES you actually have it wired exactly as I drew it.

    The Ip meter cannot be 10A full scale. Isn't it 1A? Ig current is normal.

    Are the resistors in the meter circuit the values shown?

    If I knew what the meter really was, I could tell you the shunt and multiplier values.

    If you remove the wires from the plate meter and measure the resistance of the plate meter, what is that resistance?

    What your other amplifiers do is unimportant, as are the capacitors and diodes across the meters. That will just divert attention from finding your real problem.

    Measure the plate meter resistance and tell me what it is. For 1 amp full scale if the meter is 1 mA, the 0.2 ohm shunt would require a TOTAL meter resistance plus multiplier resistance of 200 ohms.

    If you wire the system as I drew it, and use R115 of 150 ohms with a meter of 50 ohms 1 mA, then the plate meter will read 0-1 ampere FS.

    The puzzling thing is the grid meter reading backward. That indicates a wiring error, or a bad component in the RF deck area. But I though we checked all that.
    Don't worry about the diodes. That is not your problem, and D106 really eliminates the need for diodes across the meters anyway.

    What happens when you remove D106 temporarily? When you mounted D105, you did not accidentally short it to the chassis did you? If you open R118, you should read about 80 ohms from M101 leads to the chassis. Is that right? If it is not around 80 ohms with R118 disconnected, what is it? Try it with the ohmmeter both ways.

    By the way, C33 on the RF deck is a common problem. It can only be checked on a capacitor tester. It fails from shorted tubes. If C33 is breaking down at 20-30 volts, it will cause the grid meter to read backward when on standby.


    73 Tom
     
  5. NE7X

    NE7X Ham Member QRZ Page

    W8JI

    I checked again, yes the circuit is wired per your recommendations. I don’t want to look like a fool, so I checked again on more time.

    D105 100V zener is not shorted or shorted to chassis, checked and rechecked. It mechanically mounts on Bakelite board. Its working now, not blowing out like it did before.

    Removed D106, powered up, same, no difference if D106 is in or out of circuit. (Could this be a clue?) Ig moves backwards idle, then ~300ma on voice peaks. Ip idle 50ma, pegs meter on voice peaks. I tested D106 again, not shorted or open. Infinity one direction, ~5 ohms other direction.

    When I say the Ig meter moves backwards, it’s off the zero point reverse by two needle widths. It’s just enough backward movement to be noticed. if you are now watching the meter when you key, no drive, you might not even notice it. There is a 12V zener installed, schematic calls for 10V. Could this be causing an issue?

    I lifted the ground end of R118. From the positive terminal on M101, measured resistance one direction is 70 ohms and 100 ohms the other direction.

    I have an “Almost All Digital Electronics” model L/C Meter-IIB capacity tester. I disconnected all wires and C33 measures 1700pf from center to chassis ground. The manual shows C33 as 2000pf. I tested all three feed-throws, C17, C18 & C33, and they all measure the same, 1700pf. Since all three test 1700pf, its most likely my meter that is off and the caps are good.

    I swapped the tubes between my other two amplifiers three times. All three sets of tubes work perfect in the other two amplifiers. I currently have my new set of RF Parts tubes in the 2K Classis.

    History Refresh

    The amplifier was working perfect for years, then one of the 3-500 tubes shorted, taking out the zener. The guy who I purchased the amp from replaced both tubes and D1 zener (wiring it backwards) and that is when the Ig only meter circuit was pegging full scale. He tried to resolve the Ig meter issue for over 6 months, then finally gave up and sold me the amplifier.

    What I first powered up the amp the frame for the band selector switch was loose where it mounts to the chassis and was causing arching for the band selector ground. I also found the roller inductor wheel arching to the shaft. I removed both, cleaned up all metal contact points and now mechanically all the arcing issues have been resolved.

    Found D105 shorted, replaced and it shorted again. Removed and left out.

    Next the amp was working, putting out RF, however the tubes were running supper red HOT! We discovered the zener was in backwards. Reversed the zener connections and now the amp puts out a full KW+ and the tubes are running a nice soft rose pink.

    Current

    From everything I can tell, the amp is working. Full KW+ output, nice rose pink to the tubes, nice peak on all bands, no arcing or popping, HV looks good. The only issue is the meter circuit are not reading correctly. Ip meter reads 50ma on key, no drive. Ig meter reads slight backwards movement, key, no drive. When applying 90 watts drive, Ip meter pegs full scale, Ig meter read 250-300ma. D105 in circuit, not shorting out anymore.

    73s, Thomas NE7X…
    http://ne7x.com
     
  6. NE7X

    NE7X Ham Member QRZ Page

    I have been doing some experimenting. I have been playing with resistor values in series with M101 Ip meter off the B- line where D107 use to be.
    The interesting thing I see, when I add resistance, Ip meter reading goes down and Ig meter reading goes up. Also when I add resistance, the backward movement of the Ig meter M102 is reduced or eliminated, and if I add high resistance, the Ig meter starts to go forward off the zero scale.

    With 90 watts drive

    33 ohms: Ip = 300ma / Ig = meter pegged (Ig meter +30ma reading forward when keyed no drive)
    12 ohms: Ip = 450ma / Ig = 450ma (Ig meter zero reading when keyed no drive) (both Ip & Ig will peg with 100 watts CW drive)
    No ohms: Ip = pegged meter / Ig =300 ma (Ig meter backward of zero when keyed no drive)

    So does this information tell us anything?

    73s, Thomas NE7X…
    http://ne7x.com
     
  7. W8JI

    W8JI Ham Member QRZ Page

    You have a wire in the wrong spot or an accidental short. Changing the resistor in series with the plate meter cannot possibly change grid current reading, unless something is wired wrong (either through an accidental short or an error).

    Let's look at the schematic again....

    Henry1.jpg

    R115, the meter multiplier, is ONLY in series with the plate current meter M101 (lower meter in my drawing). It cannot affect the current in any other mesh unless there is a short or something miswired. It just will not happen unless there is some accidental short or connection. It is electrically impossible.

    Since in your amplifier the value change of R115 affects the grid meter (M102, the upper meter), it tells me somehow the grid meter is connected to the plate meter on the meter's negative terminal. There may be another place for an error, but one source I see is if the negative terminal of the plate meter is tied to pin 5 of the interconnect socket or R118.

    Also, you never did answer this:


    That question is critical.
     
  8. NE7X

    NE7X Ham Member QRZ Page

    Per my Tektronix digital DVM, with all the wire disconnected from the meter, the coil resistance of the plate current meter measures 51.8 ohms.

    I keep looking and looking, checking and checking, I just don’t see any wiring errors. I checked 10 times today to make sure its wired to the diagram you gave me. Its 100% (minus the HV limit resistor) wired to the diagram.

    What I noticed visually, it acts like the meters are not linear. If I speak softly into the microphone, only hit the amp with 30 watts drive, the meters do not peg. If I speak loud or give full 100 watts CW carrier, both meters peg.

    Right now key, no drive, Ip = ~200ma and Ig slight backwards just off the needle.

    73s, Thomas NE7X…
    http://ne7x.com
     
  9. NE7X

    NE7X Ham Member QRZ Page

    Checked again:

    M102 (Ig meter) the negative terminal goes to chassis ground.
    The positive terminal goes to the switch, then to one end of R117 (30 ohm)

    Other end of R117 ties to M101 positive terminal and R116/R118/P5

    Other end of R118 (.2 ohm) goes to chassis ground

    M101 (Ip meter) the negative terminal goes to R115
    The positive terminal goes to R116/R117/R118/P5

    Other ends of R116/R115 tied together and goes to B- cathode of D106

    I ohm checked between connections looking for shorts, nothing found.
    I validated the meters polarity, there is a + on the back. The cases are clear and the + terminal has red wire and the neg terminal has black wire.

    M102 (Ig meter), BIAS voltage is negative? should the positive terminal go to chassis ground and the negative terminal goes to R117?
     
  10. W8JI

    W8JI Ham Member QRZ Page

    I suggested you might want to measure voltage breakdown of C33. I said:

    You measured capacitance, which does not tell you voltage breakdown. You said:
    Also, you never did answer this:

    I'm at a loss. All I can tell you is the circuit will work perfectly the way I drew it. The circuit would not work reliably the way Henry drew it. I'm 100% sure of both.

    Provided you are not doing something like readjusting drive power when you change the multiplier resistor on M101 (R115) , R115 should have NO effect on grid current reading.

    I'm assuming the chassis ground for the grid meter is connected to the chassis ground for the power supply components, and you have not unbolted something that has isolated those grounds.

    The circuit will work as drawn, the few questions above are the last unanswered. No other parts are needed.

    73 Tom
     
Thread Status:
Not open for further replies.

Share This Page