Ballast replacement help

[n5ama]

I have a piece of equipment that uses two Amperite 3H-11 ballast tubes in a VFO supply circuit. I'm having trouble finding replacements in the event the current ones goes belly up. I could measure the resistance on them but I'm sure that wouldn't do me any good just to replace them with the cold resistance value. Is there a replacement circuit like maybe using zeners that I could substitute?
 
Any help greatly appreciated.
Tom  N5AMA 

[KD1SH]

   My Clegg Zeus uses two of those as well, and yes, they've pretty much become unicornium. Zeners wouldn't work, since we're talking about AC, although you could rectify that filament supply and then regulate it in the conventional manner.
   I suspect, though, that the solution might be a bit simpler, although it would require some experimentation, measurements, and math. Those ballast tubes are essentially PTC devices: resistors with exaggerated positive temperature coefficients. Some people have simply substituted another vacuum tube for the ballast tube, letting the new tube's filament stand in for the ballast tube's PTC resistive element. This works because all "hot wire" elements, whether in incandescent light bulbs or vacuum tube filaments/heaters, have huge positive thermal coefficients.
   You'd need to find the current requirement for your VFO tube's filament, easily found in the tube specs, and then measure the voltage drop across your 3H-11 ballast. The voltage drop and the known current will give you the resistance at that current. Experimentation with a filament transformer, a variac, and some assorted tubes should get you close to what you need. With experimentation, a light bulb, like those used in automotive brake lights, might even work.
   Or, failing that, you could simply replace the ballast tube with an appropriate resistor, following the math again, and be done with it, since the effect of normal variation in line voltage on your VFO's frequency stability is probably very small.
   Good luck; I'll be interested to see what you come up with.

I have a piece of equipment that uses two Amperite 3H-11 ballast tubes in a VFO supply circuit. I'm having trouble finding replacements in the event the current ones goes belly up. I could measure the resistance on them but I'm sure that wouldn't do me any good just to replace them with the cold resistance value. Is there a replacement circuit like maybe using zeners that I could substitute?
 
Any help greatly appreciated.
Tom  N5AMA  

[W1RKW]

on page 2, 2nd graph of the Amperite ballast datasheet (link below), it shows the I/V curve for a typical ballast tube. Also included are some curves of other types of materials. None of those come close to the ballast tube. I thought maybe a small light bulb could be a substitute but even that may not be close enough. I think Bill may be onto something with using a vacuum tube.  maybe a LM317  voltage regulator with temperature compensation??


https://www.technicalaudio.com/pdf/Electronics_Catalog_Extracts/Amperite_ballast_tube_regulator_1960_REM_24.pdf

[Pete, WA2CWA]

National used a 4H4C ballast tube in a number of their receivers back in the day. They recommended a 6V6 tube replacement:
http://emmittsfixitshop.com/files/Download/Projects_HRO_4H4C_3.jpg
Or build a solid-state replacement:
http://emmittsfixitshop.com/files/Projects_HRO_4H4C_4.jpg

Probably something similar could be applied to the 3H-11 ballast tube. Time to get creative.
Some info: https://threeneurons.wordpress.com/wp-content/uploads/2015/05/amperite_ballast.pdf

[W1RKW]

Here's a link for a bigger image of the thumbnail in Pete's post.

Or build a solid-state replacement:
http://emmittsfixitshop.com/files/Download/Projects_HRO_4H4C_4.pdf

[n5ama]

Wow, lots of possible fixes!!! I?ll keep my fingers crossed that the ballast tubes hold , at least until I can find a fix.

I may find something that looks reasonably close

The VFO uses a 6BK7B, so I?ll look it up and play with the drop across the parallel ballasts. It shouldn?t hurt to plug n play different things and see what effect it has on the VFO.

Thanks,
Tom N5AMA

[n5ama]

Is there a reference that would give the specs on ballast tubes that might be used to get close or maybe another maker?s ballast that would work?

Tom

[W1RKW]

A 3-11 or 4-11 would be close electrical alternatives. Not sure about physical size and pinout.  they're probably unobtanium as well.

[Pete, WA2CWA]

Where to get ballast tubes:

https://vacuumtubesinc.com/index.php/ballast-tubes.html

https://vacuumtubes.net/Prices%20Ballasts.htm

https://tubeworldexpress.com/collections/ballast-tubes/ballast?srsltid=AfmBOoo6JLXmFV0mpz-4QP0Vhv5TMc0iOlvC0fZliYpzTxhcZE35fBt_

https://www.tubedepot.com
Search for ballast

https://www.ebay.com/b/radio-ballast-tube/bn_7024792139

[n5ama]

I measured the resistance on the 3H-11 and it was 12.5 ohms cold. What about finding an octal tube with that filament resistance and using it?

Tom N5AMA

[KD1SH]

   The VFO filament winding on the transformer supplies 20 volts, I believe (I don't have my good copy of the schematic in front of me right now; just a blurry PDF). The 6BK7B, according to the RCA Receiving Tube Manual, will draw 0.45 amps at 6.3V. So, you'll need a resistance that will drop 13.7 volts at 0.45 amps. Easy math gives a resistance of 30.44 ohms (actually, 60.88 ohms, since there are two ballasts in parallel), but the tricky part is that you need more than just that static value; you need the resistance/current curve. The cold value doesn't even get you close to that.
   As the Amperite document that Bob provided a link to shows, those ballast tubes have resistance/current curves that, rather than being linear, are highly tailored to provide proper regulation. Using a series element with the wrong curve would likely be worse than simply using a fixed resistor.
   It is a bit of a conundrum. Experimentation might reveal a tube with a filament that comes close to the curve of the ballast tube, though it's very unlikely to duplicate it. The LM317 circuit that Pete linked to would probably provide the best starting point for getting yourself free of the ballast issue altogether. If I ever need to replace my 3H-11's, I'm thinking I'll go with the LM317 circuit. Or, maybe you'll find the 3H-11 somewhere.

I measured the resistance on the 3H-11 and it was 12.5 ohms cold. What about finding an octal tube with that filament resistance and using it?

Tom N5AMA

[n5ama]

Pete, could you provide the component values to your SS regulator? I?ll make up one of those and put it aside in the Clegg spares box for future emergencies.

In the meantime, I?ll look for Zeus power supplies to purchase for spare 3H-11 ballasts.LOL
Tom N5AMA

[Tom WA3KLR]

I think you should just put in an appropriate series resistor value and prove whether that is good enough or not before going fanatical on this.

[n5ama]

That would certainly be done initially. I?m just trying to cover my bases in the event the simpler things don?t work.
Thanks,
Tom

[KD1SH]

   I wish you luck in finding a Zeus power supply/modulator. I've got several examples of the Zeus, but only two supply/modulators, and I've only seen a few at hamfests. The problem is that when a ham goes SK, and his family or non-ham friends dispose of his gear, the impressive-looking stuff like transmitters and receivers, with high-tech looking meters and dials, are given priority, while the big, ugly, and very unattractive power supplies sitting in the dust under the bench often wind up in the scrap metal bin. The Gonset G-76, Clegg Thor-Six, and Multi-Elmac AF-67 external supplies often meet the same fate, which is why you find so many of the rigs at hamfests, orphaned from their power supplies.
   Zeus supply/modulators sometimes turn up on eBay, but shipping costs can be discouraging.

Pete, could you provide the component values to your SS regulator? I?ll make up one of those and put it aside in the Clegg spares box for future emergencies.

In the meantime, I?ll look for Zeus power supplies to purchase for spare 3H-11 ballasts.LOL
Tom N5AMA

[KD1SH]

   I agree with Tom: do the math, solder the correct resistor into an octal plug, and plug it in. I commend Clegg for their diligence in keeping VFO drift to an absolute minimum, but in my experience with my own Zeus, the drift due to the natural aging of the VFO components over their 60+ years of life will surely render any drift due to line-voltage fluctuations insignificant. As an exercise in engineering for its own sake, finding a modern approach to regulating the filament voltage could be an enjoyable rabbit hole to go down, but if you just want to enjoy running the rig, you might want to avoid the rabbit hole.

I think you should just put in an appropriate series resistor value and prove whether that is good enough or not before going fanatical on this.

[Pete, WA2CWA]

Pete, could you provide the component values to your SS regulator? I?ll make up one of those and put it aside in the Clegg spares box for future emergencies.

In the meantime, I?ll look for Zeus power supplies to purchase for spare 3H-11 ballasts.LOL
Tom N5AMA

What SS regulator?

[KD6VXI]

Pete, could you provide the component values to your SS regulator? I?ll make up one of those and put it aside in the Clegg spares box for future emergencies.

In the meantime, I?ll look for Zeus power supplies to purchase for spare 3H-11 ballasts.LOL
Tom N5AMA

What SS regulator?

The one you posted a link to.

Bottom of this screenshot.

The link to a much larger schematic is:  http://emmittsfixitshop.com/files/Download/Projects_HRO_4H4C_4.pdf


--Shane
WP2ASS / ex KD6VXI

[KD1SH]

  The values are all in the schematic that Pete posted the link to. I didn't get the impression that he was claiming it was his design.
   If you do decide to use that approach, I would suggest ensuring that both of those regulators, the LM317 and LM337, are well heat-sunk, and you might want to put a resistor...20 ohms or thereabouts and rated for at least 5 watts...in series with the input, to lessen the total input/output differential, since your input will be 20vac and your output only 6.3 vac. They will handle that differential, but it's always better to let a dumb resistor be the beast-of-burden, rather than your solid-state parts.
   I also noted that the schematic says that the output is 6.3vac RMS, but then states that the output "approximates a square wave." RMS values for sine waves are calculated differently from those for square waves, but the end result should be within the tolerance of the tube. If you're concerned about that, you could make the circuit adjustable; the app-notes for the LM317 and 337 will show you how. Just make sure your digital voltmeter is capable of displaying the RMS value of waveforms other than sine.

Pete, could you provide the component values to your SS regulator?

[n5ama]

Pete, could you provide the component values to your SS regulator? I?ll make up one of those and put it aside in the Clegg spares box for future emergencies.

In the meantime, I?ll look for Zeus power supplies to purchase for spare 3H-11 ballasts.LOL
Tom N5AMA

What SS regulator?

I?m sorry, SS= solid state versus my generation of vacuum tube technology.

[n5ama]

 The values are all in the schematic that Pete posted the link to. I didn't get the impression that he was claiming it was his design.
   If you do decide to use that approach, I would suggest ensuring that both of those regulators, the LM317 and LM337, are well heat-sunk, and you might want to put a resistor...20 ohms or thereabouts and rated for at least 5 watts...in series with the input, to lessen the total input/output differential, since your input will be 20vac and your output only 6.3 vac. They will handle that differential, but it's always better to let a dumb resistor be the beast-of-burden, rather than your solid-state parts.
   I also noted that the schematic says that the output is 6.3vac RMS, but then states that the output "approximates a square wave." RMS values for sine waves are calculated differently from those for square waves, but the end result should be within the tolerance of the tube. If you're concerned about that, you could make the circuit adjustable; the app-notes for the LM317 and 337 will show you how. Just make sure your digital voltmeter is capable of displaying the RMS value of waveforms other than sine.

Pete, could you provide the component values to your SS regulator?

My copy of Pete?s schematic printed almost as bad as my copy of the Zeus manual. LOL

I could tell what component was there but not a clue of values or component part number.
Sorry

[KD1SH]

One post down from Pete's, Bob, W1RKW posted a link to a larger version that will probably print better.

 The values are all in the schematic that Pete posted the link to. I didn't get the impression that he was claiming it was his design.
   If you do decide to use that approach, I would suggest ensuring that both of those regulators, the LM317 and LM337, are well heat-sunk, and you might want to put a resistor...20 ohms or thereabouts and rated for at least 5 watts...in series with the input, to lessen the total input/output differential, since your input will be 20vac and your output only 6.3 vac. They will handle that differential, but it's always better to let a dumb resistor be the beast-of-burden, rather than your solid-state parts.
   I also noted that the schematic says that the output is 6.3vac RMS, but then states that the output "approximates a square wave." RMS values for sine waves are calculated differently from those for square waves, but the end result should be within the tolerance of the tube. If you're concerned about that, you could make the circuit adjustable; the app-notes for the LM317 and 337 will show you how. Just make sure your digital voltmeter is capable of displaying the RMS value of waveforms other than sine.

Pete, could you provide the component values to your SS regulator?

My copy of Pete?s schematic printed almost as bad as my copy of the Zeus manual. LOL

I could tell what component was there but not a clue of values or component part number.
Sorry

[Pete, WA2CWA]

Duh! Who's On First?

[KD1SH]

Slowly I turned...step by step...inch by inch...

Duh! Who's On First?

[Opcom]

I really like the electronic one idea but I wonder if the RMS current is 1A as it would be with a clean sine wave and a vintage ballast, because I think I see the current waveform being clipped and no longer able to be measured the same way by a  ammeter. or is the math and waveform wrong in my head?

When I first saw the topic, envisioned a bridge rectifier with a regulator across it but also a large filter cap there, to convert the AC voltage to smooth DC and then regulate the current to the exact value the tube would want. Theoretically it would therefore be 6.3V. I hope. after midnight thinking I guess.