Swamping resistor diode?

Discussion in 'Amateur Radio Amplifiers' started by N2RDQ, Jun 2, 2018.

ad: L-HROutlet
ad: l-rl
ad: abrind-2
ad: L-MFJ
ad: Left-2
ad: Subscribe
ad: Left-3
  1. N2RDQ

    N2RDQ Ham Member QRZ Page

    I'm going through mods that were part of a kit that i have from Rich, AG6K(SK), for a dentron mla-2500

    he shows to replace the 100ohm inductive swamping resistor with 200 ohms of non-inductive resistance and adding a fast-acting diode...
    so ok, i can understand going to non inductive resistance, and he has notes on the resistor values. ... but.. whats the diode for?

    best thought i had was for balancing out some non-symmetrical load?
    Anyone know, or is this one of those thoughts that should not be done???

    [​IMG]
     
  2. KD2ACO

    KD2ACO Platinum Subscriber Platinum Subscriber Life Member QRZ Page

    That looks like a tactical error to me, or an experiment gone awry.
     
    Last edited: Jun 2, 2018
  3. KA0HCP

    KA0HCP XML Subscriber QRZ Page

    Can you give us a link to his plans so we can read his concept?
     
    KD2ACO likes this.
  4. N2RDQ

    N2RDQ Ham Member QRZ Page

    I will see if I can get permission to post,. this was a purchased kit along with associated docs. Most of what he talks about he had published in various places. I don't want to run afoul of someone's copyright rights though.
    there's no documentation on the diode other than a schematic as far as the diode goes and including it in the parts provided.
     
  5. WB2WIK

    WB2WIK Platinum Subscriber Platinum Subscriber QRZ Page

    Rich is SK so I doubt he cares.

    Looking at what you posted, I don't see the purpose of the diode. It would just rectify positive peaks of the RF signal, blocking negative peaks, and I don't see why that would serve any particular purpose.

    Too late to ask him.:(
     
  6. K7TRF

    K7TRF Premium Subscriber QRZ Page

    It's certainly unusual but look at what it does. On positive voltage peaks above the diode's turn on voltage (~0.6V) the input sees the shunt resistor. Those positive voltage peaks just drive the tubes into cutoff in this pseudo grounded grid design, it's the negative going voltage peaks that increase current through the tube and amplify in a typical Class AB design.

    Can't say why he thought it was a good idea to swamp roughly half the input signal and just look into the cathode circuit impedance during the other half cycle.

    I see Steve just said roughly the same thing...

    FWIW, I don't perform mods that I don't understand and probably wouldn't add the diode as it's value is unclear even if it's clear what it does.
     
  7. K7TRF

    K7TRF Premium Subscriber QRZ Page

    P.S. Thinking about this some more, I suspect Rich's idea was to present a known, non-inductive resistor while the tube was in cut-off while presenting the actual cathode circuit impedance while the tube was conducting. Tube input impedance varies with drive level so it makes sense that the input impedance would look quite different when the tube is cut off vs when it is conducting though I've never seen an amp design that worried about that.

    Also FWIW, I've added swamping, sometimes called feedback resistors to the input of grounded grid amps. It tends to flatten out the input SWR and helps with amps that prefer to be driven with roughly 60-70 watts when most rigs these days easily generate 100 watts. But the resistors I've added were always in series with the input circuitry placed between the tuned input networks and the cathode circuitry and usually 10 to 20 ohms non-inductive. That also has the benefit of minimizing changes in tube input impedance on overall impedance looking into the amp but it seems simpler and more conventional than a diode switched shunt resistance.
     
    KD2ACO likes this.
  8. N2RDQ

    N2RDQ Ham Member QRZ Page

    WB2WIK, even though Rich is an SK, I'm sure his widow does care about her husbands intellectual property rights and copyrights...... she informed me he was an SK when I inquired about the kit. She told me that they would stop selling them once they sold off the remaining inventory of kits.

    I do want to be respectful of the family, and have asked permission.

    i'm hoping I can post the complete docs and schematic.
    he has several changes including adding resistors to the cathode as "feedback resistors" to eliminate the need for ALC, an rlc circuit on the cathode for parasitic suppression.

    There are other changes too which are more his fairly well published parasitic suppression changes in the anode side of the circuit.

    K7TRF to walk down the road your going down, I did find a document were Rich did talk about the asymmetrical load at the input of a GG amp specificly an sb-220 http://www.somis.org/sb-220ci.html look int he inproving swr section, it is a good description, at least for me, and he talks about the asymmetrical loading(not his words but mine) depending on if your on the positive or negative going signal ...and how the input tuned circuit deals with it by basicly storing the energy of the unloaded side of the cycle ...... the mla-2500 has no input tuned circuit so maybe this was the bandied for handling that, being that there is no tuned input circuit do deal with it, basically giving load to the peaks that are not loaded by the tubes????. though if so, why not make the load 50 ohms?

    having just learned what load is presented by a GG AB amplifier with no tuned circuit does show what a bandied using a swamping resistor to correct the swr of the input with no tuned cirucit is, sure it will make the swr read better, but the load is not symmetrical. and I wonder what havoc that plays with the output tuned circuit of your transceiver.. … wow I might have to throw all this up on a spectrum analyzer and scope just to see, actually i'm not sure I have taps to handle that without overloading my spectrum analyzer, but, id love to see the effects of all this... this is getting interesting. Adding a swamping resistor with an amp with a tuned circuit to adjust the impedance as you described seems like the right way to go.

    it could explain many things I've read comments on how much imd you get from a solid state transceiver when using an amplifier without a tuned input.

    I've searched the internet, I've yet to see a design like that either.... nothing even remotely close...but, then again, the right way to do it seems to be using a tuned input circuit. or at least let me say the accepted method. Tuned circuit certainly would seem more efficient.

    BTW there is an input tuned circuit addon for the mla-2500 called the cm-2.. it actual looks like a lowpass filter to me... I actually have that in hand. i was supposed to help with swr and requested you re insert the swamping resistor if it was removed before you use the cm-2. removing the swamping resistor was a change made for lower power transmitters from what I understand.

    so if I can actually test it, I could, in theory, try it both ways. with a tuned circuit and with the diode, … this is assuming that's the reason for the diode... which is a bit of a stretch I guess.

    i'm with you K7TRF, I don't add / modify things I don't understand.

    there's always something to learn :)
     
  9. K7TRF

    K7TRF Premium Subscriber QRZ Page

    A tuned input circuit can also help control harmonic content prior to amplification, particularly if it's the common low pass PI network topology. I'd go ahead and add the CM-2 tuned input circuit kit that you have as it seems like a good way to go all around.
     
  10. N2RDQ

    N2RDQ Ham Member QRZ Page

    circuit in the cm2 is your typical CLC lowpass pi filter, I just did a quick test of the cm2 with a rig expert, and got some strange results into 50 ohms, though it appears to be attempting to be a low pass filter, the frequencies are way off on the upper bands.

    i;ll have to check it on the spectrum analyzer to be sure, .. first time using the rig expert on anything so, i've got no confidence in it.
    though its probably deigned to match the impedance of the amp, so, ill need to retest that way . hmm thinking about it, it probably explains the strange results.

    all in all though, yea if it works, the cm2 is probably the right way to go. which would likely mean also not touching the swamping resistor as that would upset the impedance the cm2 is designed for.
     

Share This Page