Pages: [1] 2   Go Down

Author Topic: Homebrew amp parasitic frequency  (Read 350 times)

K0IZ

  • Member
  • Posts: 1176
Homebrew amp parasitic frequency
« on: July 11, 2021, 07:56:31 PM »

I'm building a two 813 GG amp for 20 meters.  Using my trusty Millen grid dip meter I find a parasitic resonance at about 126 Mhz.  I have a suppressor to each 813 consisting of 4 turns #14 over paralled 82 ohm 2w carbon resistors.  The parasitic frequency doesn't vary hardly at all with settings of tune and load variables.  Now that I found this resonance, should I do anything to make the suppressors more "tuned" to that frequency?

Logged

K6AER

  • Member
  • Posts: 7159
Re: Homebrew amp parasitic frequency
« Reply #1 on: July 11, 2021, 08:51:15 PM »

When you made these measurements were the tubes in conduction?
Logged

K0IZ

  • Member
  • Posts: 1176
Re: Homebrew amp parasitic frequency
« Reply #2 on: July 12, 2021, 08:06:46 AM »

No.  Tubes installed, cathode pi-net tuned, plate pi-net tuned, no power on.
Logged

KM1H

  • Member
  • Posts: 11155
Re: Homebrew amp parasitic frequency
« Reply #3 on: July 12, 2021, 08:58:25 AM »

Then that is not a parasitic frequency but is a real resonance due to layout. Try it again with tubes removed.

Carl
Logged

N8CBX

  • Member
  • Posts: 641
Re: Homebrew amp parasitic frequency
« Reply #4 on: July 12, 2021, 09:10:04 AM »

If one removes the tubes, the tubes output capacitance needs to be emulated & included somehow. 
Jan N8CBX
Logged
Dayton Ohio - The Birthplace of Aviation

KM1H

  • Member
  • Posts: 11155
Re: Homebrew amp parasitic frequency
« Reply #5 on: July 12, 2021, 10:40:00 AM »

Not if you want to find if the enclosure minus tubes has a resonance. This was part of taming the SB-220 as far back as the 80's when the natural resonance AND the VHF choke in the HV lead both dipped around 150 MHz which just happened to be close to the parasitic frequency with tube fired up. The fix was easy, replace RFC-2 with a nichrome wire wound glitch resistor.

The first thing I do when oscillations pop up is to get out my trusty Measurements 59 and find the frequency while sniffing in Diode Mode. Then with amp power off and in GDO mode look for strange resonances.

Ive tamed many old school amps that way and the OP's amp sure sounds Old School.

Carl
Logged

K0IZ

  • Member
  • Posts: 1176
Re: Homebrew amp parasitic frequency
« Reply #6 on: July 12, 2021, 10:50:13 AM »

OK, did some more checking.  Pulled tubes, no change in freq of dip, about 125 Mhz.  Disconnected plate choke (and parasitic suppressors), no change in freq.  Disconnected 2.5mh safety choke on output, no change in freq.  So now have only the plate coil (14t 1 3/8dia #8, 3.17uh) and tuning variable (23pf max) and loading cap (200pf).  No change in freq!  Still about 125 Mhz. 

I do get a nice sharp dip at the expected 20mhz without the tubes.  Tubes measured 23.2pf for the two. 
Logged

W1RKW

  • Member
  • Posts: 219
Re: Homebrew amp parasitic frequency
« Reply #7 on: July 12, 2021, 12:50:02 PM »

I think you need to revisit your plate choke configuration.
Logged

K0IZ

  • Member
  • Posts: 1176
Re: Homebrew amp parasitic frequency
« Reply #8 on: July 12, 2021, 01:50:54 PM »

I suspected a choke resonance, so disconnected from tuned circuit, still same 125 Mhz resonance.
Logged

K0IZ

  • Member
  • Posts: 1176
Re: Homebrew amp parasitic frequency
« Reply #9 on: July 12, 2021, 06:11:59 PM »

I took the plate coil out of the circuit.  Using my Millen GDO, coil all by itself has a self resonance (ends open) of 140 Mhz.  So now my question is, does that matter for a 20M Pi-net?
Logged

G3RZP

  • Member
  • Posts: 2254
Re: Homebrew amp parasitic frequency
« Reply #10 on: July 18, 2021, 10:27:19 AM »

Carl
Quote
The fix was easy, replace RFC-2 with a nichrome wire wound glitch resistor.

Is that where Rich Measures got his ideas from?

Logged

KM1H

  • Member
  • Posts: 11155
Re: Homebrew amp parasitic frequency
« Reply #11 on: July 18, 2021, 11:59:31 AM »

Carl
Quote
The fix was easy, replace RFC-2 with a nichrome wire wound glitch resistor.

Is that where Rich Measures got his ideas from?

I never asked him Peter. I got the idea from the National NCL-2000 when I started there in 1963, no VHF choke and just the resistor. I later saw that RCA published that idea in the 8072/8121/8122 spec sheet so it was either their idea or they got it from someone else.

Carl
Logged

W1RKW

  • Member
  • Posts: 219
Re: Homebrew amp parasitic frequency
« Reply #12 on: July 18, 2021, 02:23:31 PM »

I took the plate coil out of the circuit.  Using my Millen GDO, coil all by itself has a self resonance (ends open) of 140 Mhz.  So now my question is, does that matter for a 20M Pi-net?

not much experience here on this matter but it seems to me the inductance of the plate choke is to small. Why not create a conventional plate choke that covers the ham bands and if you want to expand the bandwidth later you won't have to "reinvent the wheel"?
Logged

K0IZ

  • Member
  • Posts: 1176
Re: Homebrew amp parasitic frequency
« Reply #13 on: July 18, 2021, 04:43:42 PM »

For clarification, what I found out is the resonance is in the plate coil.  Sitting on my work bench, nothing connected, it has a 140mhz self resonance, and 125mhz in the plate circuit.  So doesn't have anything to do with parasitic suppressors, plate choke, tubes, etc.  I'm sure all plate coils have self resonances, but probably no one (but me) bothers to check.  So I'm pretty sure there's nothing wrong with my setup and have put the plate coil back in the circuit.
Logged

VE7RF

  • Member
  • Posts: 1609
Re: Homebrew amp parasitic frequency
« Reply #14 on: July 18, 2021, 05:55:08 PM »

I think you need to revisit your plate choke configuration.

He never mentioned the value of his plate choke.  50 uh is ample ( what's used in the SB-220)  and will series resonate on 40 mhz.   Those  2.5 mh  ....'safety choke'  will typ blow up if the plate block cap  fails.   Measure the DC resistance of your 2.5 mh  safety choke, and u will see why it will fail.  All  u require is a 50 uh choke...wired between   hot side of load cap..and chassis....and wound with 18-22 ga wire.   What else that works is  12 ga magnet wire wound on a small 1.57" OD  toroid.   The safety choke is at a 50 ohm point.

813's  don't require a  parasitic suppressor...even with 3 kv no load on em.   They barely have gain on 10m.
Logged
Pages: [1] 2   Go Up