Thanks Eric, sorry, I should have been clearer, I don't have a thermistor symbol in LTSpice so I labeled the lower 10k res in the bridge "TR" to indicate it is meant to be a thermistor.

I was getting Vout to be around the 1V to 4V range so it's looks like the bridge will need re-balancing given the values you have. I guess in practice at least R3 (and may be R4) would be variable to allow for zero setting. I will revise the circuit this evening.

 

hi,
If you join the free Yahoo LTS group, you will access to these thermistor models.
There are two models available Non self heating and Self heating, you want the latter.
https://groups.yahoo.com/neo/groups.../neo/groups/LTspice/files/%20Lib?prop=eupdate

Eric

 

Wow, that's a bit of a learning curve! Anyway, joined the group, downloaded the NTC universal thermistor and sort of got it working. I have swapped the second Op Amp for an instrumentation amp since the feedback resistors were affecting the bridge balance.

The thing now is that the 10k therm changes more than the 100R therm. Any ideas?

<edit> re-uploaded the file because to remove the redundant voltage divider on the U2 ref pin

 

hi Chris,
Downloaded your #23 file, will run it today.

Pleased to see that you have taken the first step with LTS, understanding it will bring you many benefits.

Eric

 

hi Chris,
Your circuit looks OK in simulation.

Ref this image, note the different dissipation in the thermistors.

Have you considered how you are going to mount the 100R thermistor?
Dont have it soldered to a substantial heat shunt as this will slow its response time.

Perhaps a 'wand' at the end of a short lead from the main unit.

With very low air flow rates it would advisable to have the ability to hold the wand at arms length.?

Eric

 

Hi Eric,

many thanks for all the help and time you have put into this. I'm glad it looks OK. I'll have to learn how to plot power, that's very useful.

With regards to mounting, the plan is that the 100R will be a small (0603 package) SMD device mounted on the thinnest possible PCB traces an inch or two long and then some kind of wand as you suggest. The device will be connected to a data logger and left in place for weeks/months to sense air flow across a range of atmospheric conditions.

I may also develop a hand-held version of the device but we tend to use joss sticks or canned smoke for immediate tests.

It may take a few weeks but I'll get one built and let you know how it works out.

 

many thanks for all the help and time you have put into this. I'm glad it looks OK. I'll have to learn how to plot power, that's very useful.

hi Chris,
To plot power.
Run the Sim.
Place the cursor pointer on any component
Press down and hold down the ALT key on your keyboard, the cursor will change to a thermometer.
Left mouse click while you are holding the ALT will run the Sim and plot the Power.

Eric,
BTW:
I am a Yorkshire man and I was a weekend troglodyte in the Peak district when I was younger.

 

Great, one day I'll RTFM for LTS...

I'm in South Cumbria but do most of my caving in the Dales, if you ever fancy going underground again just let me know

 

Just thought that I would say that I have now built one of these and it works great. I am using an INA122 instrumentation amp so the feedback resistor is different (about 10k) but other than that it is a very sensitive draft detector.

 

hi Chris,
Thats good to hear.

Any chance you could post your final circuit with a plot so that others could learn from your project.?

Eric

 

FWIW here is what I implemented.