To get potentiometer recommendation

[Stonestreecty]

TL;DR Summary

I need a recommendation for a project to insert a potentiometer inline between the transformer and bulb to vary the intensity of light the bulb generates.

Hi, folks
I need a recommendation for a potentiometer.

I have an old 8mm projector that uses a 8v 50w bulb. The bulb works off a simple on/off switch. I would like to insert a potentiometer inline between the transformer and bulb to vary the intensity of light the bulb generates.

Even if I read theories like How to Choose a Potentiometer? (search recommendation, maybe there are some mistakes in it), my knowledge of potentiometer is not enough to figure out what part to use when there is no old part to use as a reference.

I would sincerely appreciate any suggestions. Thank you in advance!

 

[Bystander]

See "Rheostat."

 

[tech99]

Summary:: I need a recommendation for a project to insert a potentiometer inline between the transformer and bulb to vary the intensity of light the bulb generates.

Hi, folks
I need a recommendation for a potentiometer.

I have an old 8mm projector that uses a 8v 50w bulb. The bulb works off a simple on/off switch. I would like to insert a potentiometer inline between the transformer and bulb to vary the intensity of light the bulb generates.

Even if I read theories like How to Choose a Potentiometer? (search recommendation, maybe there are some mistakes in it), my knowledge of potentiometer is not enough to figure out what part to use when there is no old part to use as a reference.

I would sincerely appreciate any suggestions. Thank you in advance!

The lamp takes a very large current so a big and heavy variable resistance or rheostat is needed, in series with the bulb. Maybe one of the school laboratory rheostats would be suitable, about 8 Ohms maybe. Alternatively, is it possible to use a 100W dimmer switch in the mains supply?

 

[DaveE]

Your rheostat will get hot (≤12.5W, in this case). It will also be kind of expensive. So, I would try a dimmer on the AC source first.

However, the algebra is pretty simple to define what you would get, as shown below.

You can look for parts online at places like this:
https://www.mouser.com/Passive-Components/Potentiometers-Trimmers-Rheostats/_/N-9q0yi?P=1z0wn8sZ1z0wn9mZ1z0wn7mZ1z0x737Z1z0x703Z1z0x6umZ1z0x77jZ1z0x707

 

[berkeman]

is it possible to use a 100W dimmer switch in the mains supply?

So, I would try a dimmer on the AC source first.

Do you guys think that the choppiness of the dimmer on the AC source may cause flicker in the projector lamp? Google says that the 8mm frame rate is 16fps, so it seems like you could get some visible mixing of the 120Hz chopping (with its extra cooling/dimming of the bulb) and the 16fps and its harmonics.

@Stonestreecty -- For the dimming, could you live with something like 4 discrete choices instead of a continuously variable dimming? If so, you could then go with 3 power resistor and a rotary switch...

 

[DaveE]

Do you guys think that the choppiness of the dimmer on the AC source may cause flicker in the projector lamp? Google says that the 8mm frame rate is 16fps, so it seems like you could get some visible mixing of the 120Hz chopping (with its extra cooling/dimming of the bulb) and the 16fps and its harmonics.

I doubt it. It would seem insignificant compared to the frame rate flicker, which can be annoying in home movies. To mix down from 120Hz to, let's say, the 30Hz your eyes can see, you would need something like the 5th harmonic of 16Hz (for first order products). I suspect it's there if you have a good detector, a spectrum analyzer, and you know where to look; but I don't think a real human watching a movie would see it.

 

[DaveE]

I wonder if there could be an issue with the inductive load of the transformer and the dimmer, also, perhaps saturation issues due to unbalanced switching making a DC component. Many years ago I tried to design a phase angle dimmer to operate a cathode transformer at about 200W and had a very difficult time with these issues. I finally gave up because I just didn't trust the reliability in production. We did a linear regulator instead with a big MOSFET. This was a very different application than the OPs though.

We also did a 20KHz drive circuit (to reduce the transformer size, mostly) that was scrapped because the 20KHz modulated the beam. The point with a 50/60Hz phase angle dimmer was that the harmonics wouldn't be noticed because of all of the other unavoidable line frequency noise that was already there. It was probably good that we weren't adding extra though. The result was the sort of design that you would look at and think there's a better way, but when you actually looked into the subtleties the simple version wasn't so bad.

 

[Windadct]

Umm why will a variable DC supply not work?

 

[berkeman]

Umm why will a variable DC supply not work?

If it uses a linear regulator, you still have the waste heat issue. If it uses a switching regulator, you may need to worry about flicker (or maybe not).

 

[DaveE]

DC could work. However, the lifetime of incandescent bulbs is reduced with DC excitation. Some more than others, depending on the design of the filament.

 

[Windadct]

DC could work. However, the lifetime of incandescent bulbs is reduced with DC excitation. Some more than others, depending on the design of the filament.

But the point here is dimming - so running below nominal power, just switch to the direct AC feed for full power?

 

[DaveE]

But the point here is dimming - so running below nominal power, just switch to the direct AC feed for full power?

Yes, good point. It does sound kind of complicated though. Really, the rheostat idea isn't bad. It's simple and 13 watts isn't so much if you design for it.

 

[Stonestreecty]

Thank you all for replies!
I know any device used to lower the power to the lamp would need sufficiently high enough power rating.
Some thoughts
If dc, then it may be possible to dim the lamp with pwm but not easy given your knowledge level.
If ac, then some form of thyristor phase angle control.

 

[DaveE]

In my previous analysis I assumed that the lamp resistance is constant. It's not. The resistance of incandescent filaments is a function of temperature, with the resistance increasing with increasing excitation. Unfortunately, this also depends on the details of the filament design and the lamp thermal environment, so there isn't a fixed answer for all applications. I'm going to guess a variation of something like 10% for steady state and ignoring the very low "cold" resistance near 0 excitation. Less resistance when dimmed will increase the power dissipation in the rheostat, but I'd bet the effect is noticeable, but not huge. I'll (probably) add another post with a better model.

 

[jrmichler]

I found an H3 halogen bulb in my junk box, and could not resist measuring the current vs voltage. The plot is below:

The current is proportional to the square root of the voltage. That relationship is within about 2% for over 6 volts. It glows red at 4 volts, and starts making useful light at 6 volts. I would size a rheostat for your bulb as follows:

Supply voltage = 8 volts
Minimum bulb voltage = 8 volts / 3 = 2.7 volts
Current at 8 volts = 50 watts / 8 volts = 6.25 amps
Current at 2.7 volts = (2.7 / 8)^0.5 * 6.25 = 3.6 amps
Maximum voltage drop across rheostat = 8 - 2.7 = 5.3 volts
Maximum rheostat resistance = 5.3 volts / 3.6 amps = 1.5 ohms

I see that Digikey has a large selection of 2 ohm rheostats with power ratings from 12.5 up to 1000 watts. I would expect that Mouser and Newark (and others) would also have these. Digikey's website says that they ship all over the world.

 

[DaveE]

I measured the 12V, 50W halogen lamp in my microscope (not the 8V flavor the OP has). The square law model works well, which makes sense since I'd expect the filament resistance to vary (mostly) linearly with temperature, but the temperature should vary as the square of the voltage. My model included a small series resistance for the leads*, since I couldn't actually measure just the filament. For my setup V = 0.545⋅I² + 0.33⋅I fit very well, <1% error from 4 to 12V.

Then I modeled the rheostat dimmer comparing the constant resistance model (post #4) with the non-linear model and found much more power in the series resistor in the non-linear model; 19W versus 13W (max.).

* edit: 0.33 ohms is way too much for just the lead length in my setup, which was more like 0.02 ohms or so. So this is mostly from the filament deviation from the square law.
Excel said V = 0.4658I² + 0.883I - 0.9104 was the best fit.

 

[hutchphd]

https://www.ebay.com/itm/303960382478?_trkparms=aid%3D1110006%26algo%3DHOMESPLICE.SIM%26ao%3D1%26asc%3D233120%26meid%3Dd8ce9f18460844eb80849790e1f88463%26pid%3D101195%26rk%3D3%26rkt%3D12%26sd%3D253375823473%26itm%3D303960382478%26pmt%3D1%26noa%3D0%26pg%3D2047675%26algv%3DSimplAMLv9PairwiseWithPLXWeb%26brand%3DUnbranded&_trksid=p2047675.c101195.m1851&amdata=cksum%3A303960382478d8ce9f18460844eb80849790e1f88463%7Cenc%3AAQAGAAACADtA4wkuHIt8rrd%252BW0KQRY4Uyw34ksnk66TxC1lHRr30e3cNI0j9EBVuTyzkhn0VSz7LslJHhbuVsulNRF4mLg3bLaal4UJus1n70PhuLFPtcZHy3C42JLZqVtRmyG7IWoqUUQpiW8Bbk3QrJxcR7iq7TUzrC756laXnSq6nCu2ilylC2kLgAvTUzK3twJDhTleJkSGfCv8tX0RRMw8zmZuDu7iXXTriPxgXwnhVdJj5ip88ghB39YpTK%252BCNAcGs3bsqyMr0A%252Bg0ZQcZPjfgTLhdzZeVf71hhGpofRkRs%252F5aCiDzL7N%252F9XGqQR%252B8dR0OsnTR7bDHy%252FRwYEXSnji0b8A5EcnjYfJgRBf0OIx0w3j9b9iEYR35k9VdfK3NG3nUXwiIWFAyOsfcvEiE22lLlnFxUBMyuISnndl9f2QtRuF3KuL5313%252BgyLqm28M8pEXF3q3Bm7T87KwZS%252BDMV68MydWUQzBEdgeOJe8vmvcHx2JnCUDWPsq2OOks%252FVRYcWQlTuoajSs954igRcR%252BFrWvMPmamjBXpWA%252FZfRBHZK344F1ERQsAXaea0aukftU5apXT8iOiQC7gEcm7B2y3NiVup3P5OLHORae%252BViIfjPixjhTaLaRjQEXu5alxFfkWFNwmBDtSfgTfqGQ6S%252FXX0%252FyhMh21nGp9HOoxLpm%252BYDJLkk%7Campid%3APL_CLK%7Cclp%3A2047675

Here is a buck regulator from ebay. Looks like it is current limited. 4 bucks. No I haven't used one but I'll bet it would work fine.

 

[Stonestreecty]

In a word, my thought, rather use a proper electronic dimmer module, any pot used will burn out at the lower resistance settings.

 

[DaveE]

any pot used will burn out at the lower resistance settings.

Not if you buy a big one. They have power dissipation ratings.
However, I'd probably try a dimmer on the AC source, if it doesn't work you won't have wasted much time, effort, or money. Maybe something like this.

 

[sophiecentaur]

Do you guys think that the choppiness of the dimmer on the AC source may cause flicker in the projector lamp?

I'm a bit late to the party dudes and this question made me try to remember. Afair, film projectors used to be so jerky and flickery that, if the AC supply to the lamp had any effect, beats were not noticeable.
TV impairments tend to be far more noticeable in that regard.

The filament on a 50W low voltage lamp would be fairly chunky so the luminosity variation in time would be pretty soggy due to the thermal inertial. It used to be a common trick to use low voltage lamps in machine shops to reduce the risk of strobing on rotating machines.