The AM Forum
March 28, 2024, 09:12:14 AM *
Welcome, Guest. Please login or register.

Login with username, password and session length
 
   Home   Help Calendar Links Staff List Gallery Login Register  
Pages: [1]   Go Down
  Print  
Author Topic: Plate chokes  (Read 3107 times)
0 Members and 1 Guest are viewing this topic.
K8DI
Contributing
Member
*
Offline Offline

Posts: 423


« on: January 04, 2022, 01:37:09 PM »

I've been trying to learn what I can about plate chokes.  The RCA BTA-1R1 I am converting came with a choke that measures 1.7mH at 10kHz (as high as my BK LCR meter goes). Using my nanoVNA, I see a series resonance at 3.6MHz.....    Reports of people converting the 1R1 to 160m have no mention of plate choke issues. The one report of a conversion to 80m was more the unconversion back to 160m needing the burned up plate choke rewound...go figure.

To start, I severed the winding at the midpoint, unwound a bunch, connected the now-two windings together and put it on the nanoVNA. After multiple tries, I am left with resonances between ham bands, but below my maximum designed frequency. I am trying to cover 160-20m. 

In addition, I see many many schematics of linear amps and AM transmitters with two series plate chokes, with two bypass capacitors.   All of this is driven by the need for sufficient reactance at lower frequencies without having enough stray capacitance to be resonant at higher frequencies.    What I don't see is much about how much reactance is enough reactance, or how to split the choke.

Right now, I am not happy with the measurements of the two partial windings.  I guesstimated a desire for about 300uH as a minimum value of reactance at 1.8MHz to keep RF power dissipated in the choke to a tolerable level. However, I see no way to do that with a single winding on this core without it having resonances between these bands. One note: the stock choke is fairly large in diameter at this amount of inductance -- the length of winding to get 300uH is less than the diameter of the winding.

Whether 300uH is the right target, I am not certain. Whether splitting it in two and adding another bypass capacitor will work, I am not certain.

I'm seeing as options to buy some choke designed for some other ham amp, buy some older choke designed for a big old transmitter, or figure out how to make this choke/form work. There's some desire to do the last one -- the mounting and wiring is already there/it fits.

Any thoughts or ideas from the folks here?
Logged

Ed, K8DI, warming the air with RF, and working on lighting the shack with thoriated tungsten and mercury vapor...
DMOD
AC0OB - A Place where Thermionic Emitters Rule!
Contributing
Member
*
Offline Offline

Posts: 1768


« Reply #1 on: January 04, 2022, 02:58:27 PM »


Right now, I am not happy with the measurements of the two partial windings.  I guesstimated a desire for about 300uH as a minimum value of reactance at 1.8MHz to keep RF power dissipated in the choke to a tolerable level.

...One note: the stock choke is fairly large in diameter at this amount of inductance -- the length of winding to get 300uH is less than the diameter of the winding.

...Whether 300uH is the right target, I am not certain. Whether splitting it in two and adding another bypass capacitor will work, I am not certain.

Any thoughts or ideas from the folks here?

For comparison, the plate choke I have from a BC-1T measures exactly 200uH, with a winding length of 4.5 inches, a diameter of 2.3125 inches, and 20 turns per inch or 90 turns total. (The calculated inductance is 195 uH). This is the only plate choke in a BC-1T. The XL should be ~ = 0.42XRL.

For a dual set of 4-400  finals, the calculated inductance is 125uH at 1.8 MHz.

The filter cap below the choke is a .001uF.

What are the dimensions of your choke and how many windings?

I think your one-turn measurement of inductance is way off.

Phil - AC0OB
 
Logged

Charlie Eppes: Dad would be so happy if we married a doctor.
Don Eppes: Yeah, well, Dad would be happy if I married someone with a pulse.NUMB3RS   Smiley
W1ITT
Member

Offline Offline

Posts: 573


« Reply #2 on: January 04, 2022, 03:18:49 PM »

When looking for series resonances in a plate choke, it's important to have the choke in the environment it will live in, or at least a reasonable simulation of it.  The reason is that capacitances to surrounding objects can move the resonances around.  Obviously, this is more important as frequency goes up, say 10 mhz and above.  Measurements on a bench, especially if the bench has metal supports underneath can be confusing.  For playing around with a "free space" design, I have built cradles, 3 or 4 inches high out of styrofoam or cardboard to lay the choke in.  When things get close I will then stand the choke up on a piece of sheet metal, with a vertical wall or corner..whatever it takes to look similar to the final location.  Later, measure again in situ.
Again, at 150 and 80 meters, life is easier, but I've spent some time chasing my tail before I wised up to the stray environmental capacitance gremlins.  The NanoVNA rules for this game!
73 de Norm W1ITT
Logged
DMOD
AC0OB - A Place where Thermionic Emitters Rule!
Contributing
Member
*
Offline Offline

Posts: 1768


« Reply #3 on: January 04, 2022, 03:43:14 PM »


For comparison, the plate choke I have from a BC-1T measures exactly 200uH, with a winding length of 4.5 inches, a diameter of 2.3125 inches, and 20 turns per inch or 90 turns total. (The calculated inductance is 195 uH). This is the only plate choke in a BC-1T. The XL should be ~ = 0.42XRL.

For a dual set of 4-400  finals, the calculated inductance is 125uH at 1.8 MHz.

The filter cap below the choke is a .001uF.

What are the dimensions of your choke and how many windings?

I think your one-turn measurement of inductance is way off.

Phil - AC0OB
 

Your RL is 3350 ohms, so 0.42X3350 =1407 ohms.

L = 1407/11.3^6 = 125uH@1.8MHz.

Phil - AC0OB
Logged

Charlie Eppes: Dad would be so happy if we married a doctor.
Don Eppes: Yeah, well, Dad would be happy if I married someone with a pulse.NUMB3RS   Smiley
K8DI
Contributing
Member
*
Offline Offline

Posts: 423


« Reply #4 on: January 04, 2022, 03:54:57 PM »

For comparison, the plate choke I have from a BC-1T measures exactly 200uH, with a winding length of 4.5 inches, a diameter of 2.3125 inches, and 20 turns per inch or 90 turns total. (The calculated inductance is 195 uH). This is the only plate choke in a BC-1T. The XL should be ~ = 0.42XRL.

The filter cap below the choke is a .001uF.

What are the dimensions of your choke and how many windings?

I think your one-turn measurement of inductance is way off.

Phil - AC0OB
 

The form is 2" diameter, smooth ceramic. The winding is 5.610" long. My count is either 29 or 30 turns in 0.500", so say 59 per inch times the 5.61 inches -- 331 turns.  Online coil calculators say that is 1.683mH, which is consistent with my BK measured value.

I didn't mean to say a one-turn inductance, I meant I took off turns from the middle, leaving two sections. Attaching a picture. Left section is 1.305", right 1.784" (as shown on caliper), total inductance measures 713uH.  This config has series resonances at around 5, 8, 13, and 17MHz, as indicated by loops on the Smith chart graph on the nanoVNA.

What one turn inductance is off/not sure what you mean?

Ed


* IMG_6215.JPG (674.99 KB, 1512x2016 - viewed 208 times.)
Logged

Ed, K8DI, warming the air with RF, and working on lighting the shack with thoriated tungsten and mercury vapor...
K8DI
Contributing
Member
*
Offline Offline

Posts: 423


« Reply #5 on: January 04, 2022, 04:00:22 PM »


For comparison, the plate choke I have from a BC-1T measures exactly 200uH, with a winding length of 4.5 inches, a diameter of 2.3125 inches, and 20 turns per inch or 90 turns total. (The calculated inductance is 195 uH). This is the only plate choke in a BC-1T. The XL should be ~ = 0.42XRL.

For a dual set of 4-400  finals, the calculated inductance is 125uH at 1.8 MHz.

The filter cap below the choke is a .001uF.

What are the dimensions of your choke and how many windings?

I think your one-turn measurement of inductance is way off.

Phil - AC0OB
 

Your RL is 3350 ohms, so 0.42X3350 =1407 ohms.

L = 1407/11.3^6 = 125uH@1.8MHz.

Phil - AC0OB

So you're saying I don't need anything like 300uH...  half is more than enough. If it were you, would you unwind down to that point, or do something different?

Ed
Logged

Ed, K8DI, warming the air with RF, and working on lighting the shack with thoriated tungsten and mercury vapor...
DMOD
AC0OB - A Place where Thermionic Emitters Rule!
Contributing
Member
*
Offline Offline

Posts: 1768


« Reply #6 on: January 04, 2022, 04:31:42 PM »



Your RL is 3350 ohms, so 0.42X3350 =1407 ohms.

L = 1407/11.3^6 = 125uH@1.8MHz.

Phil - AC0OB

So you're saying I don't need anything like 300uH...  half is more than enough. If it were you, would you unwind down to that point, or do something different?

Ed

[/quote]

Remove windings down to 150uH and test it in circuit.

Is this coil IL303?
Logged

Charlie Eppes: Dad would be so happy if we married a doctor.
Don Eppes: Yeah, well, Dad would be happy if I married someone with a pulse.NUMB3RS   Smiley
K8DI
Contributing
Member
*
Offline Offline

Posts: 423


« Reply #7 on: January 04, 2022, 04:48:27 PM »



Remove windings down to 150uH and test it in circuit.

Is this coil IL303?

Yes; manual attached for others following along.  Will unwind a bunch and install it and then run the nanoVNA on it. I won't be able to actually test for a short time; the output tank/pi network isn't reassembled quite yet..  IL303 sits on that mark between the two black filament transformers with the dangling strap and HV lead...

Ed

* bta1r1.pdf (2380.63 KB - downloaded 108 times.)

* IMG_6208.JPG (796.23 KB, 2016x1512 - viewed 220 times.)
Logged

Ed, K8DI, warming the air with RF, and working on lighting the shack with thoriated tungsten and mercury vapor...
K8DI
Contributing
Member
*
Offline Offline

Posts: 423


« Reply #8 on: January 05, 2022, 07:45:18 AM »

I unwound all but an inch of the bottom part of the winding, ran it up to the top terminal, along with the free end of the top part (leaving that section intact but shorted)  I chose an inch because it was easy.  Low frequency measurement was 177uH, VNA showed nothing but inductance through 15MHz. There was a hiccup in the graphs around 17MHz, not a loop, just a bend.  All this on the bench.  I’m going to re mount it and test it in place today. I get to play around with the thing this afternoon while waiting on a freight delivery instead of working… I’m still holding out hope I’ll get it on air by Friday, even if only on one band. Thanks for the advice, gentlemen!

Ed
Logged

Ed, K8DI, warming the air with RF, and working on lighting the shack with thoriated tungsten and mercury vapor...
WD5JKO
Member

Offline Offline

Posts: 1996


WD5JKO


« Reply #9 on: January 05, 2022, 10:03:50 AM »

Following this thread.....but I have a question...

A plate choke might have the nominal inductance for the frequency, and void of any nearby resonances, but could it still fail?

I'm thinking that the volts per turn might have an upper limit. Is removing side by side turns to lower the inductance the right approach, or is having less turns better when the turns are spaced apart a better approach? I can see winding a choke with two side by side wires, and then when soldering one wire at each end (same strand), then unwind the other leaving a gap between turns.
Test for inductance, and VNA for resonances, and repeat as necessary...

Jim
Wd5JKO
Logged
K8DI
Contributing
Member
*
Offline Offline

Posts: 423


« Reply #10 on: January 05, 2022, 09:33:10 PM »

A plate choke might have the nominal inductance for the frequency, and void of any nearby resonances, but could it still fail?

I'm thinking that the volts per turn might have an upper limit. Is removing side by side turns to lower the inductance the right approach, or is having less turns better when the turns are spaced apart a better approach? I can see winding a choke with two side by side wires, and then when soldering one wire at each end (same strand), then unwind the other leaving a gap between turns.
Test for inductance, and VNA for resonances, and repeat as necessary...

Jim
Wd5JKO

Certainly, a plate choke can still fail....  But I don't think the volts per turn is the next issue.  With 3000v B+, and 60 turns per inch in what's left of the coil, that's 50v per turn, a very reasonable value for insulation.  And, with space-wound coils, you need more turns to get the same inductance, thus you have more stray capacitance and a lower self-resonant frequency.

Barring resonance, your issues will more likely have to do with thermal/power dissipation limits.  The DC current will to some amount heat the wire (and insulation). To a great degree more, the RF current will do the same.  Think about it this way:  the capacitor at the power supply end of the choke is supposed to be basically an RF short to ground at that end, so any RF that does make it through the choke is shunted away to ground. Some RF current WILL flow through the choke. It provides reactance, that is AC resistance, to that flow. The RF current that does flow down through the choke and to ground through the capacitor will create heat.  At some point, the insulation will degrade, coils will short, inductance will drop and at the same time the shorted turns will absorb more energy magnetically, and the whole process will run away.

In my case, this form is 2" diameter, 1/4" wall thickness ceramic. It can take a lot of heat.  The stock choke needed to deal with frequencies as low as 530kHz, hence the high inductance, but also never went any higher than 1620kHz, so self resonance did not become an issue.  I'm looking forward to seeing how it works out when I tune it up on the bands ... soon!

Ed

Logged

Ed, K8DI, warming the air with RF, and working on lighting the shack with thoriated tungsten and mercury vapor...
Pages: [1]   Go Up
  Print  
 
Jump to:  

AMfone - Dedicated to Amplitude Modulation on the Amateur Radio Bands
 AMfone © 2001-2015
Powered by SMF 1.1.21 | SMF © 2015, Simple Machines
Page created in 0.059 seconds with 19 queries.