The AM Forum
March 28, 2024, 06:54:40 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: 872 Solid State Plug-in  (Read 4418 times)
0 Members and 1 Guest are viewing this topic.
w4fms
Member

Offline Offline

Posts: 36



« on: October 05, 2017, 11:52:02 AM »

I'm working on a Johnson Desk KW.  I haven't been able to find a pair of tall black solid state 872A plug-ins and thought I try building my own.

I wonder if anyone has done this?  Maybe string a series of diodes together on perf board with some kind of clip up top for the plate cap.  I could wire the lower end right in the socket to the screw terminals. 

Any advice appreciated...type and number of diodes etc.

Frank, W4FMS
Logged
WD5JKO
Member

Offline Offline

Posts: 1996


WD5JKO


« Reply #1 on: October 05, 2017, 08:24:28 PM »



I did a quick search on Ebay, and I see one 872 SS replacement. For the 866/3B28 there are a bunch of 1N2637's. Then I noticed several types when searching for, "semtech solid state vacuum tubes". Some might work as a 872 replacement. Maybe someone here on the board has more information about those.

As to a home made SS replacement, from what I have heard, many have failed with the attempts. The biggest problem is transients that exceed the PRV of the diode string. This is common with big transformers with lots of leakage inductance, and choke input filters when the supply is switched ON, and OFF, or with HV on, Morse code Key UP. One school of thought is to double the PRV from the AC RMS requirement, and then perhaps double that again. Another school of thought is to add R-C snubbers to suppress the transient spikes.

Diodes back in the day when the SS replacements were being made had to be screened, matched, and be a type that had a controlled avalanche characteristic such that a punch through transient did not cause a catastrophic failure.

If it were me, I'd find a vintage SS replacement set and move on. They are out there.

Jim
Wd5JKO
Logged
WD8KDG
Member

Offline Offline

Posts: 262



« Reply #2 on: October 06, 2017, 09:20:49 AM »

Solid state if you must, but consider a pair of 4B32's.

Regards,
Craig
Logged

Ham radio is now like the surprise in a box of "Cracker-Jacks". There is a new source of RFI every day.
WD8BIL
Contributing
Member
*
Offline Offline

Posts: 4409


« Reply #3 on: October 06, 2017, 11:09:35 AM »

Here's how I did it for the Viking 1 back in '69. The diodes have been changed out to 5408s which should work fine in the DK. They're 3 amp 1000V PIV diodes. You might want 6 - 8 diodes in each string. The resistor is 15ohm 10W. That'll have to be upgraded a bit for the DK to perhaps 10 ohm at 25 or 50 watts. You can do the math. The plate cap connector goes on the 1/2" hex nut.
Buddly



* 20171006_105622.jpg (1502.71 KB, 3264x2448 - viewed 252 times.)

* 866 SS fix.jpg (353.67 KB, 1700x1470 - viewed 267 times.)
Logged
WD5JKO
Member

Offline Offline

Posts: 1996


WD5JKO


« Reply #4 on: October 07, 2017, 03:24:07 PM »


I like the effort Buddly did with that Viking I.

For a Johnson Desk, not sure of the HV transformer specs, but lets say it is 5500v CT when on high tap. That would be 2750v each side of center. The diode PRV requirement (minimum) for FW CT rectification is 2,828 X V rms (each side of CT), or 7777v. Go ahead and round that up to 8000v, and double that. That gives you 16 1KV rated diodes on each side, or 32 diodes total. (1n5408 X 16 perhaps each side)

I would then consider some transient protection:

1.) Varistor across transformer primary, 20 Joule @ 230v, such as V230LA20A.
2.) A series R-C across the transformer HV secondary. Perhaps a .001uf 10KV and a 10K 10W WW resistor as a place to start. Comments?
3,) A series R-C across the big Choke of similar size as across the HV transformer secondary. Comments?

The idea of the R-C's is to absorb a good portion of a transient generated from the decaying magnetic field from either the HV transformer leakage inductance, or the filter choke. When the load is suddenly removed, these transients often want to rise until the diode string avalanches. 

As far as a series resistance goes, consider that each HV winding has some resistance, so there already is resistance there. Also the filter choke has some DC resistance as well. One thing that can be done though is measure the DC resistance of each HV winding of the Plate transformer. Lets say one side is 100 ohms, and the other is 150 ohms. What causes this is that there is more wire on one side than the other for the same number of turns due to the varying radius between the innermost turns and outermost turns. This causes additional ripple voltage since each loaded half cycle will have a differing voltage. One way around this is to add some DC resistance to the lower ohm side such that they match better. Size the resistor wattage according to the anticipated load.

Jim
Wd5JKO
Logged
kb3ouk
Member

Offline Offline

Posts: 1640

The Voice of Fulton County


« Reply #5 on: October 07, 2017, 04:13:36 PM »

Somebody just posted a pair of them for sale on here this morning.
Logged

Clarke's Second Law: The only way of discovering the limits of the possible is by venturing a little past them into the impossible
KA2DZT
Member

Offline Offline

Posts: 2192


« Reply #6 on: October 07, 2017, 08:57:18 PM »

16 diodes should be more than enough for 3KV RMS x 1.4 to get the peak voltage.  I still use a 470K resistor and .01 1KV ceramic cap across each diode.  Never had a problem.  I use 1N5408 diodes.  Choke input filters, step-start the HV xfmr primary and I switch the HV supply on and off with receive and xmit.  Works perfect every time.  No need to worry about the different resistances between the two halves of the plate xfmr windings.

I have some plate xfmrs with side by side windings and the resistances of the two halves of the HV secondary are exactly the same.

Fred
Logged
wa1knx
Contributing
Member
*
Offline Offline

Posts: 451



« Reply #7 on: October 08, 2017, 03:30:44 AM »

Frank,
        Those plug in 872's the way I went, its a pretty good deal $50 for a pair in the for sale section.  I used them
in my desk kw years back, worked like a charm.
Logged

am forever!
Opcom
Patrick J. / KD5OEI
Contributing
Member
*
Offline Offline

Posts: 8308



WWW
« Reply #8 on: October 08, 2017, 05:58:56 PM »


I like the effort Buddly did with that Viking I.

For a Johnson Desk, not sure of the HV transformer specs, but lets say it is 5500v CT when on high tap. That would be 2750v each side of center. The diode PRV requirement (minimum) for FW CT rectification is 2,828 X V rms (each side of CT), or 7777v. Go ahead and round that up to 8000v, and double that. That gives you 16 1KV rated diodes on each side, or 32 diodes total. (1n5408 X 16 perhaps each side)

I would then consider some transient protection:

1.) Varistor across transformer primary, 20 Joule @ 230v, such as V230LA20A.
2.) A series R-C across the transformer HV secondary. Perhaps a .001uf 10KV and a 10K 10W WW resistor as a place to start. Comments?
3,) A series R-C across the big Choke of similar size as across the HV transformer secondary. Comments?

The idea of the R-C's is to absorb a good portion of a transient generated from the decaying magnetic field from either the HV transformer leakage inductance, or the filter choke. When the load is suddenly removed, these transients often want to rise until the diode string avalanches. 

As far as a series resistance goes, consider that each HV winding has some resistance, so there already is resistance there. Also the filter choke has some DC resistance as well. One thing that can be done though is measure the DC resistance of each HV winding of the Plate transformer. Lets say one side is 100 ohms, and the other is 150 ohms. What causes this is that there is more wire on one side than the other for the same number of turns due to the varying radius between the innermost turns and outermost turns. This causes additional ripple voltage since each loaded half cycle will have a differing voltage. One way around this is to add some DC resistance to the lower ohm side such that they match better. Size the resistor wattage according to the anticipated load.

Jim
Wd5JKO

I'd tend to vote on this varistor protection idea, and add even more diodes as they are cheap and there is a lot of height in an 872. My thought is, why ever have to repair it? 25 silicon diodes would be close to the a 15V drop instead of the published 10V and should make no real HV difference. I agree with the 1N5408 3A diodes (or larger) since the 1N400n series is only rated 1A and the tube is rated 1.25A

The only different thing I have to add is to consider using PVC pipe as an enclosure and paint it black so it looks nice. No need to glue it beyond repair because it can be tapped for 4-40 or 6-32 screws at each end to hold it in the base and hold the end cap on so it could be repaired. A few small holes near the top and bottom would allow air to circulate. If PVC isn't agreeable, ABS and polycarbonate (Lexan) come in tubes but aren't as cheap as plumbers supplies. Avoid acrylic (Plexiglass, Lucite) because it cracks easily. ABS is self-extinguishing.
Logged

Radio Candelstein - Flagship Station of the NRK Radio Network.
WD8BIL
Contributing
Member
*
Offline Offline

Posts: 4409


« Reply #9 on: October 08, 2017, 08:25:51 PM »

Quote
I agree with the 1N5408 3A diodes (or larger) since the 1N400n series is only rated 1A and the tube is rated 1.25A

Hi Patrick........ the only reason I used 1N442s originally is because it was 1969! Smiley
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 0.04 seconds with 18 queries.