Differential Temp Meter - Part 2

This article was first published in 2002.

Last week we showed you how to build a very cheap and effective digital differential thermometer - one that reads out the temp difference between two points. To recapitulate, a meter of this sort lets you measure how close to outside air temp a cold-air intake is, what the temp drop is across an intercooler or a radiators, and lots of other uses. That 'base model' was very cheap to build, but you needed to use a normal digital multimeter to read the output.

This week we add to the system a digital display, so that the meter can operate completely independently of other instruments.

If you have not seen the article that we ran last issue, go read it now - otherwise most of what follows won't make any sense at all...

Digital Meter

Because the output of the differential temp meter is 10mV per degree Celsius temp difference, it's easy to read the temp on a voltmeter. Which is handy, because most cheap LCD panel meters are in fact voltmeters! So, just connect up the output to a meter, power it up via the same battery, and things are cool? Not quite.

To keep things as absolutely as simple as possible (see "More Complex?" breakout), we wanted a meter that could be powered direct from a battery - preferably 9 volts since we already have one of those in the box. But pretty well all cheap and readily available LCD panel meters that work off 9V don't like being powered from the same battery that's being used to power the circuit being measured. (Meters that are happy working in this configuration are called 'common ground' meters.)

And while we considered a number of options, in the end we decided it was simplest to just add another 9V battery, so that there are two batteries in the box. One powers the temp measuring circuit, and the other the LCD display.

The meter that we selected was the Dick Smith Q2220 LCD Panel Meter, which costs AUS$24. (Many other 9V-powered digital LCD meters will also be suitable.) As it arrives out of the box, this meter is scaled to read up to 200mV. But with our temp measuring system, 200mV is only 20 degrees C temp difference (ie 20 x 10mV per degree C) and so this isn't high enough.

However, by the addition of two resistors, the meter can be alternatively scaled to read up to 2V, which is fine for this application. (Other meters are similar in the way that their scaling can be changed.) Two resistors - 1 meg-ohm and 9 meg-ohm - need to be soldered onto the back of the meter. The two resistors are shown in this diagram. Be careful when doing this, as the solder pads are very small.

The position of the decimal place on the meter's display can also be set, and in this application it needs to be placed so that the numbers are shown to one decimal place. This is achieved by placing a link across 'P1', as shown here.

Once that is done, you can connect power to the meter, using the solder pads that can be seen here. Note that the meter instruction sheet has an error - go by the polarity shown on the meter board, rather than on the instruction sheet! The correct wiring is shown here.

The final meter connections are to the signal output of the temp sensing system - where last week you connected the multimeter. It doesn't matter which away around these wires go.

Meter Power

The extra 9V battery for the meter can also fit into the box, although it's all a tight squeeze. To switch both the meter and temp sensing power on together, use a DPST switch, so that the two circuits can be operated together.

Calibrating the Meter

Once you have the differential temp meter up and running with the LCD display, you can use the LCD panel meter's adjustable trimpot to fine-tune its readout. Connect a multimeter to the signal leads in parallel with the panel meter, and turn the LCD panel meter knob (it's on the back of the panel meter board) until the two meters agree at all different temps (or as close as you can make the agreement).

Tweaks

There are a couple of possible tweaks. You'll find that the zero adjustment pot is very sensitive. There are two ways of overcoming this. One is to sandwich a thin rubber grommet between the knob and the retaining nut of the pot, so that the pot becomes stiffer and so cannot be easily bumped. Another way is to buy a 10-turn 20K pot from a company such as RS Components (cat no 173-423). Or you can do as we ended up doing - and fit both the 10-turn pot and the rubber grommet.

Finally, there's no battery level meter in this design. When the battery powering the panel meter starts to go flat the display dims, although it keeps measuring accurately until this happens. When the battery that powers the temp measuring circuit goes flat the correction that needs to be applied by the zero adjustment will increase. The current draw of the two circuits is very small, so the batteries should stay fine for intermittent use for a very long time.

Conclusion

Depending on how you do it, you can end up with a measuring system that costs from very little, all the way up to about AUS$65. Which way you go is up to you. But we're sure of one thing - you'll end up with an instrument that will take away a lot of the guesswork about car mods that can potentially cost a helluva lot...

Testing Day 1 One very quick and easy test is to use the meter to measure the difference between outside air temp and the temp of your car's air intake pick-up. We used the meter on a dead standard '98 Lexus LS400, placing one probe on the front bumper, and the other just inside the mouth of the over-radiator air intake. This was to see whether the location that the air was being drawn in from was hotter than outside air. The car was then driven on the highway, through suburban areas, and idled around a carpark. Certainly not a full test, but indicative. And the temp difference? It didn't exceed 1 degree C in all moving conditions, and rose by only 3 degrees C when the car was idled for a period. So the standard intake is picking up air about as "cold" (ie as close to ambient) as it's possible to get. But, just for comparison purposes, we then put the 'air intake' probe in the engine bay, about where an exposed pod filter might go. Here the difference in temps was amazing. The underbonnet location was hotter than outside temp by a minimum of 8 degrees C - even when driving at 120 km/h. When stopped at traffic lights, it rose to being up to 40 degrees hotter than outside air, skyrocketing to this temp difference at a rate of about 3-4 degrees per second. And, when climbing a steep hill at about 60 km/h, it rose to being 30 degrees above the temp of the air in front of the car. There are some interesting points here. The first is that - on this car at least - those people who say that the underbonnet temps soon drop once the car is moving have it only half right. They drop all right - but still stay a lot hotter than the air that the engine should be breathing! Secondly, launching from a set of traffic lights in a car equipped with an exposed ram-pod could be giving the engine very hot inhalations. And the third point? A very high underbonnet temp was measured under heavy, low-speed load - something usually not remarked upon. Incidentally, that last temp difference was measured while climbing a hill where there is almost always an altitude-based temp drop - it's always cooler outside at the top of the hill than the bottom. But of course when using this meter to measure a temp difference, this outside variation is being taken into account the whole time... Day 2 The next day we did some further intake air testing, this time with the 'intake' probe just inside the airbox (as opposed to being at the mouth of the intake duct). As expected, the temperatures here were higher than at the mouth of the duct, but again there were some unexpected characteristics to the temp variations. Compared with the temp of the air in front of the car, at a constant 100 km/h the airbox air temp was typically about 4 degrees C above ambient. With the car stopped in traffic, this could rise to about 12 above ambient. However - and here's the interesting point - the temperature of the intake air depended a lot on the amount of air flowing through the duct. The duct (positioned over the radiator) transfers heat to the air flowing thought it, and the less air that's passing through the duct, the hotter that the air gets. This was very easy to see by driving along at a constant speed, noting the temp of the intake air, then planting your right foot. Even before the car had a chance to pick up any speed, the intake air temp immediately started to drop! Conversely, lifting off at any speed caused an immediate increase in intake air temp. The response speed of the meter allows these subtleties to be seen - something impossible with many temp measuring systems.

More Complex? The electronics gurus amongst you will recognise that there are many other ways of building this system. The extra battery can be deleted by using a LCD panel meter that's happy sharing its ground with the signal being mentioned - and adding the voltage regulator that will probably be required to feed 5V to the meter. The extra resistor used across the second sensor could be turned into a variable resistor - so allowing adjustment of the drift characteristic. Consulting with the LM355H data sheet will give you a circuit showing how an op-amp can be used to add degrees Celsius and degrees Fahrenheit outputs - in addition to some other very interesting circuits. You'll find the data sheet at http://www.national.com/ds/LM/LM135.pdf. However, our number one priority in designing these projects is to make them absolutely as simple as possible for a novice to construct - after all, AutoSpeed is a car publication, not an electronics one!