Monitoring Heatsink Temperature... / General Science and Electronics / Forums

Forums

4hv.org :: Forums :: General Science and Electronics

« Previous topic | Next topic »

Monitoring Heatsink Temperature...

first previous 2 3 4 5

Move Thread

LAN_403

Ash Small

Mon Dec 10 2012, 10:38PM

Registered Member #3414 Joined: Sun Nov 14 2010, 05:05PM
Location: UK
Posts: 4245

klugesmith wrote ...

Good work, measuring the component.

1. Does the response agree with what you expect from datasheet?

I don't have a datasheet, or even a manufacturer's part number. These were the only thermistors that Maplin stock. I had a 0C and 100C figure, and I didn't expect a linear response. It seems to be within a % or so of the figures I remember (5 k Ohm @100 C), but I can't guarantee the accuracy of the thermometer I'm using, and there may be some error in my 'experimental method', but, as I pointed out earlier in the thread, I don't need a huge amount of accuracy, I just want to know when the heatsinks are 'hot'. (I'm just plotting the graph from 40 C to 100 C at the moment, it will be posted as soon as it's finished).

klugesmith wrote ...

2. Regarding the "wrong-way" nonlinearity:
Can you show us a chart of predicted meter current with respect to temperature?
It certainly won't be proportional to the thermistor resistance.
Dr. Slack's original recommendation mentioned a reasonably linear voltage variation around the temperature where NTC and fixed resistor divide the reference voltage in half.

I'm working on this at the moment. I don't want the meter to rise much above a third FSD until ~80C, then I want it to rise to ~two thirds FSD @~100C (This is where it enters the 'red'), but I want it to rise progressively faster as the temp. rises,

The basic idea is that if the meter is reading less than 1/3 FSD I don't need to worry, if it reads between 1/3 and 2/3 FSD I need to keep an eye on it, if it gets to 2/3 FSD I need to start thinking about switching the power off.

I'll post the latest graph as soon as it's been plotted.

Ash Small

Mon Dec 10 2012, 11:29PM

Registered Member #3414 Joined: Sun Nov 14 2010, 05:05PM
Location: UK
Posts: 4245

Here's the latest graph. The red line is from the 'jam thermometer', the black line is from the 'cheap room thermometer'. I still have some more readings to take, and I've not checked anything yet, but it does give a 'rough idea' of the 'non-linear curve'.

I'll post again tomorrow, when I have some free time.

lightlinked

Tue Dec 11 2012, 12:40AM

Registered Member #2087 Joined: Tue Apr 21 2009, 08:32AM
Location:
Posts: 115

how do you regulate current for tig welding? do you have to scratch start it?
i still dont get why you aren't using common k probes.

Ash Small

Tue Dec 11 2012, 01:06AM

Registered Member #3414 Joined: Sun Nov 14 2010, 05:05PM
Location: UK
Posts: 4245

lightlinked wrote ...

how do you regulate current for tig welding? do you have to scratch start it?
i still dont get why you aren't using common k probes.

I built an HF TIG starter around a 2N3055 and TV flyback (LOPT)

I was originally going to use T probes (copper/constantan), but my meters have an internal resistance of 750 Ohms, so was persuaded by the 'good folks here' to 'explore the possibilities of thermistors'. If you read the whole thread you'll see how I ended up here.

EDIT: current regulation is done by the 'buzz box'.

...

Tue Dec 11 2012, 01:06AM

Registered Member #56 Joined: Thu Feb 09 2006, 05:02AM
Location: Southern Califorina, USA
Posts: 2445

Judging by your graph you have 100k NTC thermistor, there is an analytic formula to calculate the resistance for a given temperature, as R = R0 * e^[b] where R0 is the resistance at T0 (ex, 100k at 25c for your device) and B is as constant of about 4200 (varies slightly with manufacture).

There is also a nice plot of resistance vs temperature for a similar device at

(calculated for R0 = 100k, T0 = 25c, and B = 4200)

For most purposes you can use this datasheet as the datasheet for your device, the curves seem to match up pretty well

Ash Small

Tue Dec 11 2012, 02:16AM

Registered Member #3414 Joined: Sun Nov 14 2010, 05:05PM
Location: UK
Posts: 4245

[quote]
Judging by your graph you have 100k NTC thermistor, there is an analytic formula to calculate the resistance for a given temperature, as R = R0 * e^[b] where R0 is the resistance at T0 (ex, 100k at 25c for your device) and B is as constant of about 4200 (varies slightly with manufacture).

There is also a nice plot of resistance vs temperature for a similar device at

(calculated for R0 = 100k, T0 = 25c, and B = 4200)

For most purposes you can use this datasheet as the datasheet for your device, the curves seem to match up pretty well

[/quote1355192083]

The NXFpWF104 does look pretty similar. I'll have another look in the morning, as I've had a drink or three. Thanks for posting the link.

EDIT: The first link doesn't appear to work

Ash Small

Tue Dec 11 2012, 06:57PM

Registered Member #3414 Joined: Sun Nov 14 2010, 05:05PM
Location: UK
Posts: 4245

OK, I've now done some maths

and, at the moment, things are looking pretty good. I think I was mistaken in my earlier post (ten posts above) when I said the graph seems to curve 'the wrong way'.

I'm only really interested in the range from 80C to 125C. I want around 1/4-1/3 FSD @80C, ~2/3 FSD @100C (where the meters enter the red), and FSD @~125C.

I'm told (by the vendor at the hamfest where I bought the meters) that they require ~200uA for FSD, and ~100uA where they enter the red. I'm assuming that 50uA will give me ~ 1/4-1/3 FSD.

I've designed the power supply to give ~0.5V to each thermistor/meter. This is adjustable by adjusting a variable resistor connected to the 'adjust' pin on the LM317T voltage regulator (It will nominally be set for 2V, then fed into a string of four resistors, to divide the voltage between the four thermistor/meter circuits. (see earlier post)

@ 80C, thermistor resistance is ~ 10 kOhm. Ohm's law gives us 50uA.

@100C, thermistor resistance is ~5 kOhm. Ohm's law gives us 100uA.

@125C, thermistor resistance is ~ 2.5 kOhm. Ohm's law gives us 200uA. I've not measured this yet, but the Murata datasheet linked to a few posts above gives us this figure for a 100K thermistor)

So, the meter won't move much until ~ 80C, should enter the red @ ~100C, and FSD will be ~125C. Exactly what I wanted in the first place

The above figures are all approximate, but are close enough to give me the accuracy required for this project.

I will have two basic adjustments for 'fine tuning, the adjustment of the LM317T to 'tweak' the voltage, and a variable resistor (5 kOhm) in parallel with each meter for 'fine tuning' of each thermistor/meter circuit.

It looks like Dr. Slack and Klugesmith (and others) were 'spot on' with their advice.

Next job is to check the meters to see if the information that the vendor gave me was accurate.........

(Unless anyone can find any mistakes in my maths)

first previous 2 3 4 5

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob