A Simple, Valve Curve Plotter

[AlanBeckett]

One word of warning about those High Power Pots: although the whole Track may be rated at say 50W a small Section won't be.

Alan

[SurreyNick]

Ah. I didn't know that. Thanks Alan
Nick

[ppppenguin]

Without doing the sums my instincts are against using a pot like this. It's simple and that's good. However there's going to be a lot of heat which may or may not matter. What's really worrying me is that the output voltage is going to be very dependent on the load from the valve filament. The only safe way to use a pot is to start at zero and wind it up until the measured filament/heater voltage is correct. This is made more difficult by the fact that a cold filament has a much lower resistance and will take more current until it is warm.

Alan Beckett's warning about part of the track is valid. It's simpler to think about the current rating of the track which doesn't change. For example a 20 ohm 50 watt pot will carry about 1.6 amps. If that current is passed through the whole winding the dissipation will indeed be 50W. Pass it through half the winding and 25W will be disspated. As Alan said, what you can't do is dissipate 50W in half the winding.

Incidentally Alan's warning applies equally to variacs. A 2A variac will supply 2A safely. Whether that's at 200V or 2V. At 200V it's supplying 400W, at 2V just 4W.

[AlanBeckett]

Of course with a Variac you hope that the Losses (Heat) are only a small proportion of the delivered Power.
There's a lot to recommend them.

Alan

[SurreyNick]

I think it's going to be one of those 'suck it and see' situations. Most of these early valves are of the 2v or 4v type, drawing a filament current of between 0.1A and 1A. I might have to reconsider the max resistance of the potentiometer as I may need to go up to 50 ohms, but 2W is the max power for any of the most common valves I'm going to encounter.
Nick

[pwdrive]

Done a few calculations for a DC supply of 8 volts:


Series connected pot (rheostat)

Series R based on some typical minimum and maximum (ish) heater currents that might be commonly encountered, there will be some odd balls but I've left them out to keep things simple.

All calculated from a nominal 8 volt DC supply based on 6.3 volts AC in to the rectifier, the 8 volts out is just a guess....adjust as required

R = series resistance, Pd = dissipated power:

1.4 volt heater at 25 ma:
R = 264 ohms, Pd = .165 watts

1.4 volt heater at 250 ma:
R = 26.4 ohms, Pd = 1.65 watts

4 volt heater at 100 ma:
R = 40 ohms, Pd =.4 watts

4 volt heaters at 300 ma:
R = 13.3 ohms, Pd = 1.2 watts

4 volt heater at 500 ma:
R = 8 ohms, Pd = 2 watts

4 volt heater at 1 amp:
R = 4 ohms, Pd = 4 watts

6.3 volt heater at 100 ma:
R = 17 ohms, Pd = .17 watts

6.3 volt heater at 300 ma:
R = 5.66 ohms, Pd = .51 watts

6.3 volt heater at 500 ma:
R = 3.4 ohms, Pd = .85 watts

6.3 volt heater at 1 amp:
R = 1.7 ohms, Pd = 1.7 watts

Someone else can work out for 5 volt heaters.

Any errors then guilty as charged

Lawrence.

[pwdrive]

Whoops...Forgot the 2 volt figures I jotted down:

2 volt heaters at 25 ma:
R = 240 ohms, Pd = .15 watts

2 volt heaters at 100 ma:
R = 60 ohms, Pd = .6 watts

2 volt heaters at 250 ma:
R = 24 ohms, Pd = 1.5 watts.

Lawrence.

[Nick]

I'd keep it really simple, flexible and safe by just using an external DC bench supply. No worries about power rheostats, variacs, custom transformers etc. All far too ccomplicated.  Two 4mm banana sockets and there you are. 

[Mark Hennessy]

That's quite a range of series resistance - 1.7 to 264 ohms. It would be interesting to plot these on a graph - there would be quite a big gap, with nearly all of them clustered at one end.

Personally, if using a pot like this, I wouldn't wire it as a rheostat; I'd use it as a potential divider. This will be better behaved, but it will dissipate a fair amount of power when it's just doing nothing.

With some very basic calculations (too late in the day for anything else!), you could probably get by with a 10 ohm potentiometer. Higher values would work but again you might find that the scale is cramped at one end. Actually, I might be tempted to add a fixed resistance between the lower (anti-clockwise) end and ground, so that the minimum is just below 1.4V rather then 0V, so at least you're using more of the track.

Anyway, on the assumption we have 6.3V AC going in from the transformer - and we'll keep it as AC for simplicity - then we are wasting 4 watts of heat in the pot alone. Perhaps a bit less if a resistor is added. Perhaps a bit more when a valve is being powered. Probably worse if you converted to DC first, depending on how much smoothing you include.

At least here the circuit is presenting a reasonably low source impedance to the heater, so compared to the rheostat connect, the voltage will hopefully vary rather less during warm-up. Hopefully.

Ultimately, I keep returning to the variac. With that, you have the magic of transformer action, meaning losses are minimised. With a resistance, good old I-squared-R is impossible to avoid.

This is the sort of thing: http://www.ebay.co.uk/itm/Regavolt-variable-Transformer-Type-301-0-5-Amps-/252033736171?pt=LH_DefaultDomain_3&hash=item3aae619deb - it would be on the primary side of a 240->6.3V transformer. Shame it's a bit pricey, but no-doubt one would turn up at the right price eventually. The current rating of 0.5A is generous for this application, as the total load is only 6.3VA (plus a bit of safety margin). On the primary side, that's mere milliamps.

I suppose there is another alternative. You could use a light dimmer ahead of a mains transformer. It's crude, and the waveform won't be pretty, but that doesn't actually matter (perhaps the meters might be unhappy, depending on what sort you choose?). Such things are available commercially: http://www.elenco.com/product/productdet...upply=NTc2 - I've attached the schematic so you can see how it works.

[Mark Hennessy]

I'd keep it really simple, flexible and safe by just using an external DC bench supply. No worries about power rheostats, variacs, custom transformers etc. All far too ccomplicated.  Two 4mm banana sockets and there you are. 

Yes, I agree...

I could find a spare PSU if you were stuck. One of these: http://golbornevintageradio.co.uk/forum/...p?tid=3766

To be honest, they are an essential part of any lab, and you can never have enough. They are also a really good project for beginners, as you can start simple and gradually add more sophistication as you learn more.