The AM Forum
March 28, 2024, 11:32:04 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: Pi-net component choices  (Read 3208 times)
0 Members and 1 Guest are viewing this topic.
KI4YAN
Member

Offline Offline

Posts: 161


« on: July 23, 2020, 03:54:28 PM »

Working through the RF parts for the 4-65A transmitter.

I've got one of these:





It's a big chunky bandswitch out of some old military gear. The coil that was on the back of it was crushed, so I desoldered and cleaned it up.

What worries me is that it doesn't appear to be a "shorting" switch. each tap would get selected but the unused taps aren't shorted together.

I am using this RCA Ham Tips article to get my pi-network parts values.

http://n4trb.com/AmateurRadio/RCA_Ham_Tips/issues/rcahamtips1402.pdf

I've got a 114pF air variable for C1, the Airdux 2008TL coil stock would be 3.5" long for 80M L, and I have a wide assortment of receiving-type air variables for C2. I'd like to get down to 160M, and I have 7" of the coil stock. I'd need to switch in some 100pF capacitance for C1, though.

For Cb, I have some Murata ceramic doorknob caps leftover from my laser building days-these are 1100pF, 20kV rated capacitors for high peak current work, with low-inductance terminations. (big flat pads, instead of screw studs)

The RFC I have laying around is 6mH pi-wound, dunno where I got it. It had a nick in the insulation on one of the outside turns from being in a hamfest box, so that darker section is clear fingernail lacquer over.



I'm thinking that RFC is not right for this job? Too much inductance and probably resonant in or near 160/80/40/20/10? I have a nanoVNA that I could use to check it.
Logged
w8khk
Member

Online Online

Posts: 1200


This ham got his ticket the old fashioned way.


WWW
« Reply #1 on: July 23, 2020, 04:52:06 PM »

I used a rotary switch identical to the one in your photograph in my dual Eimac 4-400 legal limit linear back in 1968 for the pi net band-switch, and it worked fine for many years.

I removed the switch from a BC-375 transmitter tuning unit like those used on all the Boeing B-17 bombers.  It is a solid, low-loss switch perfect for the application.

Just connect the bottom (common) terminal to the low-band end of the pi net where it connects to the loading capacitor.  Then use the top terminals for taps for each of your bands.  As a result, all the unused inductor turns will be shorted as desired.
Logged

Rick / W8KHK  ex WB2HKX, WB4GNR
"Both politicians and diapers need to be changed often and for the same reason.”   Ronald Reagan

My smart?phone voicetext screws up homophones, but they are crystal clear from my 75 meter plate-modulated AM transmitter
KI4YAN
Member

Offline Offline

Posts: 161


« Reply #2 on: July 30, 2020, 02:39:31 AM »

Got the bandswitch assembled tonight. Should be able to start mocking up the bolt-on bits and measuring to start the front panel layout. The bandswitch moved up to mount directly to the front panel, instead of a folded bracket. This allowed the loading capacitor to move under the chassis, directly below the bandswitch for the inductor.
Logged
Opcom
Patrick J. / KD5OEI
Contributing
Member
*
Offline Offline

Posts: 8308



WWW
« Reply #3 on: July 30, 2020, 10:45:36 AM »

I'd like to know how well that kind of RF choke works for HF. Seems I have seen them in a lot of older gear, but in amps I see the long single layer ones more. I have the impression that because it's in sections it's more suitable that it looks. Did you get to measure it with the VNA?
Logged

Radio Candelstein - Flagship Station of the NRK Radio Network.
w7fox
Member

Offline Offline

Posts: 102


« Reply #4 on: July 30, 2020, 11:01:45 AM »

Thanks for sharing your progress, we all benefit.

A couple of comments:  The rf choke looks like the ones I've seen in the ART-13 transmitter.  It was used for low frequency around 500 kilocycles.  Probably not suited for HF.  A single layer solenoid can be home brewed for pennies.  Be sure to test it for series resonances with a grid dip meter.

The reference for Pi net design is from the fifties, and has been superseded by more accurate formulas.  I would check some handbooks from the eighties or nineties. 

Best regards,
Fox
Logged
KI4YAN
Member

Offline Offline

Posts: 161


« Reply #5 on: July 30, 2020, 06:53:54 PM »

I am also suspect of the choke. I've got plenty of wire here, just need to find a ceramic form to wind it on.

If this choke turns out to be unsuitable, it's only a half-inch diameter but I could use it as the form. I also have a big bag of radio ceramic around here somewhere that I could use, just have to find it.

So far, even my ARRL handbooks still have the same charts in them. I adjusted the inductor value to bring up the Q a bit, instead of 12 I aimed for 15-18, and if I add a bit of fixed capacitance to the loading capacitor I'm still in range for 80M, but I may have to switch it out on 10M. I can adjust taps on the inductor when I get it tuned up.
Logged
WU2D
Contributing
Member
*
Offline Offline

Posts: 1800


CW is just a narrower version of AM


« Reply #6 on: July 30, 2020, 08:46:06 PM »

Single Layer Chokes usually can be designed to have only 1 series and one parallel resonance spot. And they work above resonance - that is until you hit the next one. The trick is to have the first series resonance somewhere between the bands! Usually between 15 and 10 Meters. Now with the 12M band, this becomes tricky.

If you are only interested in 75 M - 20M it gets fairly easy to make a single layer choke.   

Those fancy chokes with the pi windings and multiple pies may outperform the single layer jobs in some bands. They also have resonances, so watch out.

Another trick is to instead of using 1 big choke - to use multiple parallel chokes, one feeding each tube so you can use smaller, lower current chokes. The penalty is more caps.
Logged

These are the good old days of AM
KD6VXI
Contributing
Member
*
Offline Offline

Posts: 2648


Making AM GREAT Again!


« Reply #7 on: July 30, 2020, 10:41:15 PM »

Use a delrin form.  I have them an inch in diameter


Drill a small hole in the bottom and tap for a 1/4 X 20 bolt or screw.  That's your mounting screw.

Now, from the exposed end, drill a hole as deep as you can.  I start my hole in the drill press to get a straight guide, and then graduate to a 9 inch bit.  I use a 10 inch long form.

Fill the long hole with powdered iron. I've even used powdered iron from the floor of a metal working shop.  A bit of either epoxy (NOT JB Weld!, some is conductive) or wax on the top will hold the powdered iron in their.

Huge amounts of L (relatively) for less windings.

My 160 to 10 amps use this method.  3cx3000 and yc155 sized.


--Shane
KD6VXI

Logged
KI4YAN
Member

Offline Offline

Posts: 161


« Reply #8 on: July 31, 2020, 07:45:21 PM »

So the choke was no good. Measured low ohms (38) on the multimeter, but when checked on the VNA it showed extremely ragged and frequency dependent inductance.

I tried a few different setups, couldn't get a good clean reading, so I thought what the heck, I'll hook it up in an oscillator and see. Breadboarded up a 12AU7 with both sections in parallel, 250V on the plate, and drove it with a signal generator capacitively coupled into the grid. Used all the pi-network parts I already had prepped and drove it into a big 50 ohm dummy load. At no point could I get any kind of result out of it-B+ was all over the place, the choke got hot quickly with only 10mA loading, I would seemingly at random get red plating going on the 12AU7, etc. I figured it was just all the mismatched parts-but then I swapped the choke out for a known good 2.5mH 200ma part and it straightened up. The choke didn't get hot, the amplifier was fairly tame, and I was able to load up to 25mA easily, getting around 5W output. I could have used more grid drive but the signal generator wasn't having it.

Now suspecting the choke, I could see where one of the Pi's had gotten hot and discolored the...maybe shellac? that was painted over the thing. Either this thing was damaged when I got it, or I managed to damage it tuning a 12AU7 to 80 and 40 meters...so it wasn't a good choice either way!

Knowing that it was probably kaput anyway, I decided to unwind it and see what was wrong-turns out it was burned inside on two of the sections, one section was just hot and probably shorted, the other actually was charred and burned in two.

So much for that one. I found my bag of ceramic stand offs, unfortunately none of them are the size or shape needed. Cones, spacers, end insulators, stubby threaded spacers, but nothing long and fairly thin. I'll keep looking.
Logged
VE7RF
Member

Offline Offline

Posts: 56


« Reply #9 on: August 07, 2020, 07:25:25 PM »

I used a rotary switch identical to the one in your photograph in my dual Eimac 4-400 legal limit linear back in 1968 for the pi net band-switch, and it worked fine for many years.

I removed the switch from a BC-375 transmitter tuning unit like those used on all the Boeing B-17 bombers.  It is a solid, low-loss switch perfect for the application.

Just connect the bottom (common) terminal to the low-band end of the pi net where it connects to the loading capacitor.  Then use the top terminals for taps for each of your bands.  As a result, all the unused inductor turns will be shorted as desired.

What he is talking about is called a ..'pick up and hold switch;.... or a ..'shorting switch'....  or...' multiple pole rotor'.  They all consists of multiple rotor fingers.  The 1st rotor finger does the actual tap job, the  remaining  rotor fingers   short out  each un-used   tap on the bandswitch.   With just a single rotor finger as what ur switch is..... shorts out all the  un-used inductor turns.   This will  work on any  80-10m amp.   The problems start when  160m is thrown into the mix.   When on  10m, there is  now  double the inductance to short out.  Buddy's hb amp is a 160-15m  affair  (no 10 + 12) ..and has a low plate load Z..and  uses a single rotor.   Not much uh used in total..due to   low  Z.   He added just a bit more uh onto the end of the  160 portion.....and presto, the upper bands (20-17-15m)  were...'sucked out'

Most of use the superb  GM3SEK  free  pi /  pi-L  software  to design  pi / pi-l  networks.  It factors  in all stray C and stray L.  A small resistor, same value as the calculated plate load Z..is  temp installed between anode..and chassis ( all voltages turned OFF).  MFJ-259  installed on output coax / TR relay  s  ON..or bypassed..and  tune and load tweaked for  1:1  swr on mfj.   Do this on each band.

On lowest band, like 160m, once tweaked,  tune and load left alone.... and  MFJ freq is shifted  up / down....  till MFJ  reads the  2:1 swr points.  repeat for other bands.  You will see that on 80m, the  2:1  points are twice as wide.  on 40m, it will double again....and double again for 20m..and  double again for  10m.   In...'suck out  mode'... the  upper bands  will all have extremely narrow  2:1 points.  That's
the giveaway.   It's  also why a lot of roller inductor setups  will   short out ( via a relay etc)   The last  1/3 of the  coil  (  160 end) when using the  roller >  10 mhz.

The ameritron plate choke is  cheap..and works on all  9 x HF bands...plug and play.   used in both ameritron amps..and also  alpha amps.
225 uh.  Rated  for  4 kv @  1.5 amps CCS.   I tested 2 x new ones recently..and they run stone cold with 1.5 A  flowing for  1 hr  (using my  0-60 vdc @  5A  lab supply).  > 1.5 A..and it gets hot..and fast.   Goes into..'thermal runaway'.

Jim  VE7RF

Jim
Logged
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 0.06 seconds with 19 queries.