Small heaters

Hi all, I'm mostly just flailing around looking for ideas. I want to make a small heater. Say 1-3 Watts, with resistance in the 10-100 ohm range. (At the moment I'm not quite sure how much power I'll need, it may be less than a watt.) This will be for a pulsed heat capacity measurement. (heat pulse -> measure temperature rise, rinse and repeat.) So it's important for the temp sensor and heater to have a small heat capacity themselves. My first thought was a TO-220 pack resistor.

1.91 grams (outta the box) Trim tab... 1.32 g Trim leads 1.21 g That's still much too big. Now minco has heaters on kapton tape that might work. HK5186R25.0L12 is a one watt heater with a 1/2" diameter And a hole in the center. The 1/2 inch is a bit more than I wanted. I'm looking for other ideas. (?) (Zener diode?, surface mount resistor?...) Well also a transistor as heater, but I'm going to start a second thread for that topic.

thanks, George H.

Reply to
George Herold
Loading thread data ...

How long of a pulse? If the resistor never reaches full temperature then you can put more -- sometimes far more -- than the rated power into it for brief periods.

Some resistor manufacturers (Panasonic, IIRC, is good) rate their resistors for pulsed-current operation.

Ditto if the resistor is in a liquid, or is guaranteed to be below normal ambient.

If none of those float your boat, go looking for some old-style vitreous enamel resistors -- those could get a lot hotter, and so had higher power ratings for their size than "regular" resistors.

--
Tim Wescott 
Control system and signal processing consulting 
 Click to see the full signature
Reply to
Tim Wescott

Strain gauges? Variety of sizes, low thermal mass, often 120 ohms.

PT100 sensor - maybe use to sense as well.

Cheers

--
Syd
Reply to
Syd Rumpo

How long a pulse.. yeah that depends. I'm guessing a few Joules of energy. So let's guess at a few seconds for a pulse. But I also hope to use the same heater in CW mode too... to take the sample to a new temperature. So i'd really like something that can take abuse.

Oh, I need a good way to mount it too! I need to get the heat into a piece of metal.. you can think of a small pcb with exposed copper ground plane. (But it might just be a solid piece of copper.)

No temperatures from 77K to ~400K. And it could be in vacuum! The issue I have with most resistors is that they all have some substrate materail.. and that adds to the heat capacity. If I could get something like nichrome thin films on a pcb type thing that would be nice.

George H.

Reply to
George Herold

Hmm, OK I get four of them at 120 ohms... so I can configure for different resistances... that might be nice. I'll look at those thanks

George h.

Reply to
George Herold

when I was at school I build a thermonstat controleld soldering iron kit that kinda worked that way

The power to the heater was rectified AC from the transformer, when the voltage was close to zero the resistance of the heating element was measured

-Lasse

Reply to
Lasse Langwadt Christensen

How about an RTD? It has about the right resistance, and I hear they make decent sensors, too. :-)

Cheers, James Arthur

Reply to
dagmargoodboat

On a sunny day (Thu, 13 Feb 2014 07:23:58 -0800 (PST)) it happened George Herold wrote in :

I use logic level MOSFET IRLZ34N on a snmall heatsink as controlled heater (by a PIC) in the tritium decay experiment:

formatting link
I use some old BJT be diode as temp sensor. The MOSFET is easy to drive, no power required, and simply screwed on the small heatsink.

Reply to
Jan Panteltje

fine nichrome wire? not very original, but low mass.

Reply to
haiticare2011

Do they make RTD's that can dissipate a Watt? Oh did I mention that this needs to work from 77 to 400K. I'd rather not have the heater resistance change by that much.

George H.

Reply to
George Herold

How did you control the Mosfet? Oh, did you just wrap a control loop around it? I really need a controlled amount of heat and *not* a controlled temperature. (I think Fets are better at low temps than BJT's, is that right?)

George H.

Reply to
George Herold

0-100 ohm range

No.. that's worth thinking about (again). (I might use Phosphur bronze wir e instead of nichrome) I'd dismissed making my own heater out of resistanc e wire as too fussy. The probe will get lots of thermal cycling 77 to 400K and one problem I've always observed when you wrap your own heaters is tha t whatever "gunk" you use to hold the wire in place tends to flake off and come loose. Differential thermal expansion and all that.

George H.

Reply to
George Herold

Why not a bunch of squiggly lines of thin copper on a PCB? I bet you could use that for both heating and sensing, and possibly with less thermal mass than your transistor idea.

You'd probably have to calibrate each one, but if the shoe fits...

--
Tim Wescott 
Wescott Design Services 
 Click to see the full signature
Reply to
Tim Wescott

I thought about this, but then forgot to put any numbers in. Do you think I can get someting above/near 10 ohms? What is tight on a pcb? Can I do 10 mil wires and 8 mil spacing... on 1/2 oz copper?

Well that's not a problem. The Tempco of a copper heater is not the greatest either, but both can be dealt with.

OK if I use 10 mil traces with 10 mil spacing I get like 5 ohms in a square inch (1/2 oz copper)... I can't really afford a square inch of pcb.. and I need more than 5 ohms. grumble. The little minco heater will be better.

But thanks for the thoughts!

george h.

Reply to
George Herold

A dual pack diode. One diode for heating, another one for reading the temperature (calibrated)

Cheers

Klaus

Reply to
Klaus Kragelund

6 mil traces with 6 mil spacing is pretty common. If you call around you should be able to find someone who'll promise to do better, and who knows? Maybe they actually will!
--
Tim Wescott 
Wescott Design Services 
 Click to see the full signature
Reply to
Tim Wescott

d for that topic.

perature (calibrated)

Hmm, how much current can I push through a dual pack diode? (maybe a quad pack and put three in series). I had this other crazy thought of using a Z ener diode, reverse bias for heating, forward bais as temp sensor. Do they make dual zeners? My one worry with these 'unknown' diodes is doing the t emperature calibration. With diode connected transistors I can do a calibr ation run on a single device and then a single point calibration for all th e other devices. I don't know if that works for 'normal' diodes.

George H.

Reply to
George Herold

Well, if you put a coat of flat black paint on your object, and focus onto it a lamp, the heat transfer can be controlled without actually adding the heat capacity of the lamp. This is like the way fusers in laser printers and copy machines get the paper up to ink-sticks-to-paper temperatures.

As long as it's a brainstorming exercise, you might also cement a thermocouple to a few matchheads. Pulsing the thermocouple will ignite the matchheads, and you get a convenient few seconds of high heat. Then you have the thermocouple in place for the temperature-measure part.

It's not uncommon to run heater current through a thermocouple wire, there are schemes for using AC to heat, and sample the thermocouple during the zero-crossings. Its more a 1 ohm heater than a 100 ohm heater, though. You put a couple of back-to-back diodes in series with the heater, and during the zero crossings there's a millisecond or so that you can measure temperature while both diodes are nonconducting. It takes a low-voltage transformer to drive this safely.

Electrical resistors are intended to dissipate heat, and that means thermally-conductive material next to (for instance) resistance wire. Most are deliberately bulky.

Reply to
whit3rd

Minco or Omega will make any pattern you want, in nickel or nichrome, for a reasonable NRE.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs 
Principal Consultant 
 Click to see the full signature
Reply to
Phil Hobbs

Built-in RTD! Perfect! :^)

Probably wouldn't be unfair to use a Kelvin-connected strip of similar zigzagginess but smaller area for ratio calibration. They should both have similar fab errors and thus give similar resistances. Delivering desired power and checking temperature can then be done as usual.

Could also do tricky things, like weaving the sense trace inbetween squiggles of the power trace, to minimize temp difference between them. Then you could do an unpowered sense as well (but why bother, powered sense would already be easy).

An IPC-2221 trace width calculator

formatting link
says 10 ohms from 6 mil width, 59.5 inch length. Also claims it'll handle

0.37A, which sounds silly to me. Somehow I don't think there is a way to tell it you intend to pile it all up in one spot. :)

6 mil width + 6 mil gap is 12 mil occupied width, or 0.714 in^2 on just one layer (top or bottom), or a square region 845 mils on a side. Or using top and bottom layers, a square region 597 mils on a side. Using four layers gets the region under a half inch square, which isn't too bad. A nice thin 4-layer board (20-30 mils?) would be quite light, maybe 0.1 gram?

Tim

--
Seven Transistor Labs 
Electrical Engineering Consultation 
 Click to see the full signature
Reply to
Tim Williams

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.