575A Solid State Replacements

[Mike/W8BAC]

I have a pair of large solid state replacements with the same plug/socket and plate cap as a 575A but I can't find any information about them. They have an identification number on them that doesn't seem to help, that number is G953. Anybody have a cross reference? Thanks for any help.

                 Solitron Devices
                        G953
                       

[K6IC]

Hi Mike,

Well Googleing around,  also found nothing for it.  Perhaps someone's in-house part #,  not uncommon for larger mfgs to us special #s,  so customers  would need to buy the real deal.

With a 24-ish volt PS,  a 100 Ohm limiting R,  you could infer the breakdown V from the foreward drop,  assuming 0.7 VDC per diode,  and 1 KV per diode,  but this is not too absolute.

73 GL Vic

[Opcom]

solitron is gone, what we need is a solitiron catalog. Like it or not, so many of the old semiconductor company catalogs are not being kept and scanned. That will be a shame someday. Like right now.

[Pete, WA2CWA]

I have no idea where you guys are searching but Solitron Devices is alive and well.
http://www.solitrondevices.com/

And, under their page "Obsolete Semiconductor Support" http://www.solitrondevices.com/Obsolescence%20Policy.htm , there is this:
Solitron Devices, Inc. is committed to provide viable alternative solutions to current and aging defense and aerospace systems that suffer or may suffer from obsolescence and/or diminishing supplier base. Solitrons product offering is expanded constantly. Thus, all past, current and future products will be supported throughout the life of the Company. The Company will never cease manufacturing its products. It will maintain its capability to design and manufacture "sunset" (diminishing) power components and circuitry which are used by the defense and aerospace industries.

So, if they "maintain capacity to design and manufacture past products" somewhere within the organization there must be some specs for the G953.

Give them a call or send an e-mail.

[Mike/W8BAC]

Thanks Pete,

I don't know why I didn't search the manufacturers name. I assumed it was defunct as well. Thanks for the link. I emailed them for info.

I did a few checks with an ohm meter. One of the SS tube replacements shows continuity (7 K ohms) between one of the filament pins on the base and the plate cap. I'm not sure but I think this is a nonstarter.

Mike

[Mike/W8BAC]

I received a prompt reply from Solitron Devices this morning.

"P/N G953 has been obsolete for over 25 years. Their is no replacement or
data sheet available."

So much for "Thus, all past, current and future products will be supported throughout the life of the Company."

Oh well, they are old and one is probably not in good shape. If anybody has more information please let me know. Thanks

[N3DRB The Derb]

they all lie. service is extremely non profitable for many industries, as such they dont want you to keep using old 'junk'. They want you to buy new product.

[K6IC]

Well, Mike,  That IS support.

Am sure that you can determine the approx V rating of the Rectifiers that you have by measuring the foreward drop at some reasonable If -- say 100 Ma or so foreward.  The reverse V rating would require some megger or better yet,  a current-limited variable HV PS-- HiPot tester.

Rich Measures has a HiPot tester on his site,  but,  at this moment,  it does not come up ... Hope Rich is OK.  Basically his HiPot was variable HV PS,  perhaps up to 20 KVDC,  HV current-limiting resistor,  DC Microammeter and lotsa inculation -- an Acrylic platform for example.  One cranks up the HV on the Device Under Test,  and monitors the REVERSE  leakage I.  One sops increasing the volatge at some chosen current.  This should be non-destructive if one limits the reverse I to some low value ... a few uA.

EDIT:   The Measures site is avail now.  The HiPot tester is here :
http://www.somis.org/BVT.html

  Hope this is not TOO preachy ... you prob know all of this.  73  Vic

[Mike/W8BAC]

Thanks for the info Vic. I haven't learned much over the years about mercury vapor rectifiers. It's much easier to look at a diode or string of diodes to see what is going on than to understand what is happening between the filaments and plate of a 575A. In any case, one of the solid state replacements has a 7K ohm path from plate to filament which I suspect renders it useless. The four pin base and HV plate connections of the SS replacements are identical to the 575A which lead me to think they might be right but at the end of the day it really doesn't matter since it takes two tubes to make this thing fly. Thanks again for your input.

Mike

[w8khk]

In any case, one of the solid state replacements has a 7K ohm path from plate to filament which I suspect renders it useless.

Mike, I would not be too quick to reject the SS replacement because of a high-resistance reading on an ohmmeter.  I recall working on the F4 Phantom radar system, we used similar HV stacks for the magnetron power supply.  They read high-resistance in the forward direction on the ohmmeter.  The battery supply was insufficient to overcome the forward drop on the stack of series diodes.  A higher voltage is needed.   Try performing the test suggested by K6IC, and see if they perform in that scenario.

[K6IC]

OK Mike,

Too bad,  but there are lots of these devices around.  Some Hams do go back to MVs,  and  this can make some avail.  Also,  some stations carried a few spares and so on.

73  GL  Vic

[Mike/W8BAC]

Hi Rick

I would not be too quick to reject the SS replacement because of a high-resistance reading on an ohmmeter

That is what I'm thinking as well. That is why both SS replacements are still on the bench. What has me doubting is that only one of the SS stacks has the 7K ohm reading from plate to one of the filament pins. The other SS tube is open from plate to filament pins. Both tube replacements are identical and same manufacturer. possibly different lots.

Seems odd that only one filament pin in the suspect SS tube has continuity to plate. I would think both filament pins would be common to the single diode string. What do I know. The filament pins are 2 and 4 of four. In the dual 575A circuit filament pin 4 of V111 and pin 2 of V112 connect to one end of the filament transformer secondary. Pin2 of V111 and pin 4 of V112 connect to the other end of the filament transformer secondary. B+ exits the filament transformer secondary center tap.

As a foot note, I don't see continuity between pins of any combination on the SS base.

[Todd, KA1KAQ]

Seems odd that only one filament pin in the suspect SS tube has continuity to plate. I would think both filament pins would be common to the single diode string.

As a foot note, I don't see continuity between pins of any combination on the SS base.

As it should be, Mike. With the tube there are only two pins on the base used for the cathode/filament. Unlike the filament of a tube which requires both legs to complete the filament circuit though, SS stacks only require one connection, just like a regular diode. Been going through this recently making some temp repairs to the big rig while getting the rect. filament transformer rewound.

Check for a color dot by one of the filament pins. This will be the one connected to the cathode.

[Opcom]

You are a saint Pete! Now I can look up some old items. or not. Teach me to believe a salesman about what's gone and what's not..

[Opcom]

I received a prompt reply from Solitron Devices this morning.

"P/N G953 has been obsolete for over 25 years. Their is no replacement or
data sheet available."

So much for "Thus, all past, current and future products will be supported throughout the life of the Company."

Oh well, they are old and one is probably not in good shape. If anybody has more information please let me know. Thanks

Would it be worth the trouble to have someone in authority ?? or some "august figure" from here, contact them and speak to someone in a 'position' and ask for copies of the old catalogs or for certain types of parts such as SS replacements?

Cite the "customer/industry goodwill" and "visibility to engineers" such as Triad Transformer's online catalog, and the Sylvania catalogs they gave permission to scan and post, while keeping copyright (the ones on Bunker still have the notices of such). 

I can dream. I have a dream..

[w8khk]

In any case, one of the solid state replacements has a 7K ohm path from plate to filament which I suspect renders it useless.

Mike, I would not be too quick to reject the SS replacement because of a high-resistance reading on an ohmmeter.  I recall working on the F4 Phantom radar system, we used similar HV stacks for the magnetron power supply.  They read high-resistance in the forward direction on the ohmmeter.  The battery supply was insufficient to overcome the forward drop on the stack of series diodes.  A higher voltage is needed.   Try performing the test suggested by K6IC, and see if they perform in that scenario.

Mike, your post sparked my curiousity, (no pun intended) so I ran a few tests on my 8008 replacement rectifier stacks.

The diode stacks are Solitron S6698, and the red dot appears to be on pin 2.

With a DVM with an internal 9 volt battery, there was no continuity between the top cap and any pin - totally open circuit.

I connected a 9 volt battery in series with the diode stack and a 470 ohm load resistor as follows: plus on the battery to the top cap, the red dot pin to the resistor, and the other end of the resistor to the battery minus terminal.  I connected the DVM as a DC voltmeter across the resistor.  With one battery, the meter reads zero volts.  Not enough to make the stack conduct.

I added another nine volt battery in series with the first.  The rest of the circuit remains unchanged.  With the 18 volts source, and I measured 6 volts across the resistor, indicating the diode stack is dropping 12 volts.

I added another nine volt battery in series.  With 27 volts source, I measured 15 volts across the resistor.  Still 12 volts across the stack.  In all tests, I verified the battery voltage was not dropping under load, using a separate DVM.

Assuming approximately .7 volts drop per diode, we may assume the diode stack consists of roughly 17 diodes in series.  (Your mileage may vary under load with HV AC applied.)

I did no hi-pot tests, and I did no tests at higher current or voltage, but this basic test should get past the useless results from testing with a typical ohmmeter or DVM.

I built a diode stack for my Viking II this evening to replace the two 5R4 vacuum rectifiers.  I ran the same test on my stack (one half, as the stack is full-wave).  I have 6 diodes in each leg, 1000 piv at 10 A.  The test revealed just over 4 volts, consistent with around .7 volts drop per series diode.

You may want to perform this basic test on your stacks, and if they check out, perform additional hipot and HV load tests.   Hope this helps.....

[Mike/W8BAC]

Todd/KAQ said

Check for a color dot by one of the filament pins. This will be the one connected to the cathode.

Interesting stuff. No dots on the Solitron stacks or the known working stacks in my transmitter but it dose explain the dot next to pin 2 in the schematic for both the 866's and the 575A's. See below for more. Great information Todd. I didn't see the logic of having only one filament pin DC connected at first. I see it now.

Richard/KHK said

You may want to perform this basic test on your stacks, and if they check out, perform additional hipot and HV load tests.   Hope this helps.....

Yes, boiling the test down to my level helped very much. I pulled out 3 nine volt batteries and a 470 ohm resistor, some clip leads,  tube socket and two meters.

I tested the stack with no continuity and found a voltage drop of roughly 14 volts. the stack that showed 7K ohms continuity anode/plate to cathode/filament had only 5 volt differential at each voltage applied. A quick ohms law check tells me the voltage drop is almost equal to the internal (7K ohm) plus the external (470 ohm) resistance. Negative outcome. I placed a dot next to pin 4 on these two stacks.

I than pulled the two 575A stacks in my transmitter. Why didn't I do this earlier? Because I assumed what I was told about the Solitron stacks was true. That is "they are 575A replacements". and I thought they should be identical and a lack of understanding of how the stacks differ in circuit.

The first test took me by surprise. My test jig was wired to pin 4 of the NO. 434 socket and I had an open circuit. I moved the clip lead to pin 2 and BAM, the lights came on. This explained what the dots on the schematic meant and what Todd had said AND it also sort of explains what might have happened to one of the Solitrons. IF the previous owner had made the same assumption I made, that these might be 575A stacks and plugged them in to test, Both might have shorted hard. "OOPS, these don't work, I'll sell em at the next fest!" And that's when I showed up.

Thanks Rich and Todd and everyone else for the lesson. Like I said earlier I never paid much attention to MV rectifier circuits and now I have enough knowledge to be dangerous.  

[Opcom]

that is so depressing:

"OOPS, these don't work, I'll sell em at the next fest!"

But true. How can a fine gentleman's hobby have so many Dick Dastardlies? Come on Opcom, move from the '40's to the present.

Now thinking of your bad rectifiers, and of others, It seems to us..., to swipe a phrase, that one could take the base from the stick or from a deceased 575A, and also the anode cap or a 1/2" metal dowel or piece of copper pipe, and cut a piece of grey PVC to the proper height, and build up such a stick.

If 6A 1KV diodes were used, which cost $1.29 each, the peak and surge ratings would be similar or better than the 575A. The diodes could be free-hanging to each other or supported on a piece of plastic, it would not matter much.

The PVC could be epoxied into the base, and a PVC pipe cap could be put on the top of the assembly to support the plate cap and hold it on by solder. This way, you can make some nice looking sticks and if you assemble them right, they could be repairable.

If heat is a concern, maybe, 3 small 1/4" holes could be drilled radially at each end of the assembly to let air pass by convection. If debris is a worry, place some window screen inside the holes.

Or, just don't worry much about heat and pot them in resin like the originals if that is what is preferred. The sticks do run warm but I have never seen one cracked.

You could even integrate into this, a piece of #30 or so wire to act as a HV fuse. Put it in a circle around the outside, at the top of the PVC tube and fix it with a drop of silicon glue. Possibilities abound for fun in this situation.