How to reduce filament voltage

[W8ACR]

What is the preferred method to reduce AC filament voltage for large transmitting tubes? I have a filament transformer that has a nominal secondary rating of 10VCT@10A (Stancor P-6461). The actual secondary voltage is 11.2VAC. The tube is an RCA 8000 running about 1800VDC on the plate. I can think of the following options.

1. a variac on the transformer primary

2. a rheostat on the transformer primary or secondary

3. run the tube at 11.2VAC

I do have some large rotary rheostats that would be suitable. I also have a couple of variacs that would work. Waddyyall think? Can somebody point me to a reference section in one of the standard texts?

Ron

[WQ9E]

I don't have a reference to point you to but I prefer the low resistance rheostat in the secondary.  They are low cost (even if you have to buy one less than $5 from Fair Radio), it provides a bit of a "soft start" for the filament since it will help control cold start inrush current, and it is easy to set for the exact voltage you want.  If you have to drop a lot of voltage then a variac in the primary would probably be better or if you go with the rheostat option use a large low ohm fixed resistor to dissipate most of the power and use a very low value rheostat for fine adjustment.

[WA1GFZ]

Variac will make less heat and you can bring it up slowly. Either will work.

[K3YA]

For that small of a drop I would use a rheostat in series with the primary.  It's smaller, cheaper, doesn't dissipate that much power, and I like the  idea of resistance to limit inrush current.  I have used both rheostats and variacs but would only use the latter if  it was a big voltage drop and a big tube.  

My 4-1000 plate modulated transmitter uses a 6.3 volt transformer and a 3 position switch to buck/boost 115 VAC primaries  of the filament transformers, which is yet another option.

[W4NEQ]

If the overvoltage is present under normal load, I would vote for some R in the primary - the value will likely be more practical than that for the secondary.
 

[KE6DF]

A couple of other thoughts.

I have two autotransformers (one by UTC and one by Kenyon) that are made for solving this problem.

The have a bunch of taps at about 5 volt intervals from 90 to 135 or so.

So, if you have 120 VAC line voltage, attach the line to the 120V tap and the filament transformer to the 115V tap.

Otherwise, I like the idea of putting a 6.3 volt transformer as a bucking device. It will buck the 120Vac down to about 114 or so.

A 10amp filament transformer will buck a 10 amp line circuit.

Probably best to not run the 8000s on 11+ volts. Those babies are getting hard to find and you want them to last.

[W7TFO]

BC transmitters most always use a pot in the primary rather than a Variac. 

Reason being the inrush current will not be limited by a VT as they are a constant-current device and that thump is hard on filaments.  Their tubes cost a lot and they are cheap bastards is another reason for that topology.

Pure R in the primary will give a bit of sag upon start-up.

BTW, the pot doesn't get a lot of adjustment after the initial setting, especially if you have a regulator ahead of it.  Those are usually found with fil trans that have a lot of reactance to get the most out of the 'soft start' requirement.

73DG

[W8ACR]

I'm pretty sure the T-368 has a rheostat in the circuit to adjust the filament voltage to the 4-400 final. Does anybody know the values of this rheostat and where it is in the circuit?

[Rob K2CU]

RCA specs the filament at 4.5A at 10V AC or DC.

Did you measure your filament transformer with a 4.5A load, and with what primary voltage? Stancor specs it at 117VAC primary voltage for 10.0VAC @ 10 A.

The P-6458 @ 5A might be a better choice.

IF you get 11.2VAC @ 4.5A, then you would need .2667 Ohms in the secondary to get 10.0 V.  It will also dissipate 5.4W. The primary current should be on the order of a half Amp, so you could play with dropping the primary. Frank's variac would give you the most control to correct for line voltage, or you could play with a reostat there as well. You would need something like 20-30 Ohms or so at 5W. A small 6.3 or 12.6V filament transformer set up in bucking mode with the primary is a good solution as well.

Then again, the tube is rated for DC on the filament, and you could gin up a DC supply, but make sure it has crowbar over-voltage protection. It could current limit at 5A, provide soft startup by ramping up the voltage, and have a crowbar set to 10.5V.

In either case, you need to know what the transformer does under load, as the transformer is the simplest, yet not ideal, solution.

[W3RSW]

I put an old 50 ohm 100 watt slider variable resistor in the primary of my twin 813 amp 's fil. Xformer.  The fil. Xfor was a Raytheon with twice the current needed.

[KM1H]

It doesnt take much wire to drop voltage when a healthy current is involved.

With Teflon or other hi temp insulation a smaller gauge wire takes less length.

Aim for the bogey value at no load on the AC line and recheck that it stays within the 5% window at maximum drop. Here the drops are less since Im running stiff 120 and 240V circuits. Going over the bogey every 1% reduces filament emission life.

Carl

[WD5JKO]

Interesting issue here.

If the measured voltage is 11.2 when 10v is desired, then it only makes sense to me if the filament transformer is designed for 110v primary when it is getting 124v instead. So the ratios of 11.2/10 is close to the ratio of 124/110. Both are about 1.12, or 12% high.

If so, then the turns ratio is 110/10 = 11.1:1

If the secondary current is 4.5A, then the primary current would be 4.5/11.1 = .405A (not counting XFMR losses).

So then if the primary voltage is 124V and we want 110v, then we need to drop 14V @ .405A. So R=14/.405 = 34.6 ohms.

The wattage would be .405^2 * 34.6 = 5.7 watts. Best to pick a 10W resistor.

OR a loss-less way is to use a series AC capacitor. So a 76 mfd AC motor starting cap has around 34 ohms reactance at 60 hz. Not very practical IMHO. But if there is 240VAC in the same transmitter, then a ~ 8 MFD
AC capacitor should work. 

Jim
WD5JKO

[KC2FXE]

 I use a 2.5 amp variacs with my homebrew works great

[k4kyv]

With a variac or other autoformer, I would set it for over-voltage and then use a fixed resistor in the primary of the filament transformer to bring the voltage back down to normal and limit inrush current.  Excessive inrush current will eventually kill a transmitting tube.

[KM1H]

So then if the primary voltage is 124V and we want 110v, then we need to drop 14V @ .405A. So R=14/.405 = 34.6 ohms

In that particular example a simple 12-14V 1A bucking transformer would also do the trick with minimal heat. Fine tune with a low value resistor or use an air wound resistance wire or use an open power resistor for a form.

[W7TFO]

If you make the value fixed, you'll never be able to compensate for falling emission as the tube ages.

Pots are cheap and work best, Variacs are expensive, thump filaments pretty hard on start-up, and make hum fields.

73DG

[KE6DF]

With a variac or other autoformer, I would set it for over-voltage and then use a fixed resistor in the primary of the filament transformer to bring the voltage back down to normal and limit inrush current.  Excessive inrush current will eventually kill a transmitting tube.

An NTC thermistor in the primary of the filament transformer (or in the secondary depending on value) would also help with this problem.

[The Slab Bacon]

Just FWIW, when I built my 4X1 transmitter 12 years ago I didn't have a 7.5V
@21A filament tranny laying around. I did however have a 6.3V @ 20A tranny with a 105v primary tap laying on the shelf.

A little head scratching and looking also turned up a 2A variac as well.
So........I just used the 105V primary tap and the variac and away I went. Since our line voltage here is all over the place, especially winter / summer.
Usually right around 114-115v in the summer and 125-126 in the winter, it seemed like the way to go.

I mounted the variac right on the side panel with a screwdriver slot shaft end and installed a pair of meter probe jacks to read the actual filament voltage.
I usually run it between 7.0 and 7.25v and that provides ample emission for the power levels I am running. It is happy as a clam and has had the same setup and same 4X1 final tube in it for 12 years now...........

The bottom line is just to keep it as simple (and inexpensive) as you can and don't over complicate something that doesn't have to be.

[KM1H]

My philosphy is to make do with what is on hand even if it takes a little extra work.

When starting with a new tube and adjusting the filament V for dead nuts on the bogey as an absolute max and maybe 1% or so on the low side with no other load on the AC line the emission will outlast most of us on here.

[Bill, KD0HG]

Could a turn of wire be removed from the secondary?

Bill

[Opcom]

I'm surprised no one has suggested cannibalizing an old space heater and winding some of the nichrome wire around a ceramic stand-off for a primary resistor. It would not be so much as to get very warm and it would have a slight current regulating effect but I suppose that is a lot of work to fix a simple issue.

[KM1H]

The space heater or clothes dryer suggestion has been around for so many decades as to be part of ham lore....I thought everyone knew that method

It works for meter shunts also.

[Opcom]

Well yes but I guess I hopelessly like the old ways. Not that I'd wear a coat and tie to hamcom in June.

[WD5JKO]

I recall someone posting a salt water 50 ohm RF dummy load article a while back. Could use the same concept here for the resistance in series with the filament transformer primary. If the filament voltage is a little low, add some salt and stir. If a little high, add a little water and stir..

http://www.qsl.net/k5lxp/projects/SaltLoad/SaltLoad.html

Jim
WD5JKO

[W7TFO]

I recall someone posting a salt water 50 ohm RF dummy load article a while back. Could use the same concept here for the resistance in series with the filament transformer primary. If the filament voltage is a little low, add some salt and stir. If a little high, add a little water and stir..

http://www.qsl.net/k5lxp/projects/SaltLoad/SaltLoad.html

Jim
WD5JKO

I think you're all wet!

............OR..................

Add a little garlic and olive oil and wait for Slab to show up.

73DG