Testing rear window defroster grid

Hello All,
I would like to develop a testing method to detect window defroster grid failures. Rear windows has 12 heating wires painted to the glass and cca. 2% are faulty pieces. I tried to measure the resistance, current and voltage drop of good parts and compare results with those where was at least 1 broken wire, but the difference was smaller than the tolerance specified by the supplier. (Other problem is that the resistance is greatly depending on the temperature)

Now we are testing visually with a thermal paper/foil but it takes more time and sometimes the operator mistakes or can forget to correctly check the thermo foil. So I need an automated system which blocks broken pieces from further process. It is also very important to not hurt or scratch parts during the testing process, so best option would be contactless testing if possible...

I'am thinking about to make a tool with 12 sensors, 1 sensor to every each heating wire to sense a flowing current in a wire. But I'am not sure that there is any sensor available in the market for such a purpose. Finally PLC have to check the outputs of every sensors, if all the 12 heating wires are good (current is flowing through them) gives visual signal or in case of failure sound alarm or red light flashing...

Here is the specification of rear window defroster:
Power: 167,37W ±10%
Current: 12,87A ±1A
Resistance: 1,0 Ohm -10%/+20%
Wire thickness: 0,48mm
In the attachment you can find the tool used for window checking and cleaning and technical drawing of the window, I placed red dots where I would place the sensors.

I would like to try to use digital hall effect sensors (0 if no current, 1 if any current) because this seems the fastest, the most accurate method and could be easily automated and evaluated with PLC. Could you recommend any appropriate sensor, small and sensitive enough? (current in each wire ~1A, measuring distance >2mm) I have 4cm gap between the window and the tool if I place window into that. Would it be better to use lower voltage or use AC power instead of DC 12V?

I am little bit sceptic about thermo cameras, there could be many confounding factors, I think. (operators hand, fingerprint etc., not enough contrast for the second measurement if the window is already heated...)

Any ideas and suggestions are welcome.
Best regards,
Ate

Under low pressure, you could find discontinuities with HV (e.g., small Tesla coil), but I think a Hall sensor, either analog so you get an idea of the amount of current of binary/digital as you suggest, would be far better and faster.

John

Why not just pulse current through the heating element and then scan for thermal change across all heating legs? You could probably build a trustworthy thermal test profile quite quickly for each window type that you process

Hi,

What you need is a thermal imaging camera. That takes a picture but 'sees' in heat rather than light. The picture would show a grid with one line missing.
The cheapest ones i know of that will take a pic of the whole thing in one shot is about 200 dollars USD. It connects to Android but they might make one that connects to the regular PC too.
If the resolution happens to be too low, take two snap shots.
Of course the thing has to be energized to do this.

I dont know much more about them, except you need something like Android 4.4 to run them,but check that before you buy any if you decide to go this route.

My rear window defroster makes lines in the dew or frost every morning about 1 minute after I turn it on, and they are gone 2 minutes later when the entire window has heated up. Thermally, isn't there a time after power is applied but before the window comes to thermal equilibrium, when there is a temperature gradient between the heater lines that is observable with a thermal camera?

Electrically it sounds like the heater resembles twelve 1 ohm resistors in parallel, so one line failure (8.3%) can be masked by the 10% overall component tolerance. Are you testing just the heater foil, or has it already been mounted on one sheet of glass?

ak

Hi,

My guess is that there is always a temperature gradient because the glass dissipates while the elements provide the heating power. The gradient wont show up as melted or non melted ice because there is no more ice once it all melts
If the camera is sensitive enough it will pick it up. Most can detect human body heat easily.

Pretty much any b&w camera that is reasonably sensitive to infra red, and most security night vision cameras are, will happily highlight the legs of the heating element within moments of being pulsed with power. The idea is not to actually heat the elements all the way up to working temperature, just to create sufficient delta to ambient for successful detection

What are the test parameters? Instrumentation cost budget? Pieces per hour? Also, as with any test process, which is less desirable, false positives or false negatives?

ak

How about powering it from an AC supply and use a small coil as a probe to detect the AC magnetic field in the track. If there is a break in the track then there will be no current in the track. OR feed it with high frequency with one side grounded. Then follow the tracks with a small plate on the end of an oscilloscope probe. If there was a break in the track then there would be a sudden change from a full signal to no signal. On a good track the signal would change slowly from one end of a track to the other.

Les.

A simple aproach would be to use a linear o/p hall effect sensor and detect the magnetic field around each conductor, assuming they are far enough apart to discriminate one from the others.

The sensing distance is greater than 2 mm. I've used Hall sensors with magnets, but not just laying one up against a single conductor. I'm sure the fine people at Allegro semi can help you determine if one of their parts will work.

ak

Very simple using a cheep Infrared non contact thermometer, took about 2 minutes to find the faulty strip on my vehicle from start to finish. Before all the glass heats up there is a large difference in the temp of the strips. I was using the thermometer on the outside of the glass & touching the glass & just sliding it up or down. Use in the middle of the window.

I was a professional auto repair tech in my previous life and used to test heated window grids with a test light. I would test for voltage at the positive buss, then about half way across each line and any line that showed battery voltage at the half way mark was open circuit. As the test light moved toward the ground buss the voltage dropped if the circuit was complete. I don't know if a non contact voltage detector is sensitive enough for the low voltage but might be worth considering.

Hello All,
Thank you so much for your time and efforts really appreciate!
I uploaded a video to show you how we are testing heating wires currently. Visually it is totally OK, fast and accurate, thermal paper almost instantly changes its color. I have to extend our PLC controlled Poka-yoke system with Go/no-go testing to automatically block broken windows from further processing. With a regulated power supply the difference between 12 and 11 wires is very small. I measured current consumption using a non regulated power supply and the difference was clearer, 5-8% less Amps but this is still much smaller than the tolerance defined by the manufacturer. So I am afraid that 2 weeks later for example from the next delivery all parts could be evaluated as a broken ones. That's why I'm trying to detect wires separately. Than summarize outputs into one digital go/no-go signal.
Thermal camera I think would be very problematic, can catch fingerprints, or glass can be accidentally preheated, in this case the contrast will be insufficient. We are using infrared thermometers, pyrometers and heat switches in other departments in the factory but I thought that is difficult to measure glass surface, and these kind of sensors detects bigger area not just 0,5mm thickness.

lansmnet said:

Hello All,
Thank you so much for your time and efforts really appreciate!
I uploaded a video to show you how we are testing heating wires currently. Visually it is totally OK, fast and accurate, thermal paper almost instantly changes its color. I have to extend our PLC controlled Poka-yoke system with Go/no-go testing to automatically block broken windows from further processing. With a regulated power supply the difference between 12 and 11 wires is very small. I measured current consumption using a non regulated power supply and the difference was clearer, 5-8% less Amps but this is still much smaller than the tolerance defined by the manufacturer. So I am afraid that 2 weeks later for example from the next delivery all parts could be evaluated as a broken ones. That's why I'm trying to detect wires separately. Than summarize outputs into one digital go/no-go signal.
Thermal camera I think would be very problematic, can catch fingerprints, or glass can be accidentally preheated, in this case the contrast will be insufficient. We are using infrared thermometers, pyrometers and heat switches in other departments in the factory but I thought that is difficult to measure glass surface, and these kind of sensors detects bigger area not just 0,5mm thickness.

Click to expand...

Hi there,

QUOTE:
"Thermal camera I think would be very problematic..."

Surely you jest
Thermal imaging has been used in automated process control for decades. The advantage of a thermal image is you can see the whole picture at once, and if that isnt enough, from various angles as it passes below on a conveyor belt for example. For one example BMW has used it for years now for various processes, for one, looking at engine exhaust.

I dont know how much you want to spend though, but to find out more about this do a quick search on "Thermal imaging for process control", and see what millions of applications you can find
You should even be able to tell units that have elements that are marginally conducting, not just totally not conducting at all. In other words, quality control.

Depending on how sophisticated you want to get, you can even have a computer program analyze the photo(s) and determine which ones are bad or marginal. That would be automatic quality control.

One of my bosses worked for pilkingtons I think they made car windows too.

Telemecanique now scheider do a couple of cameras with built in image recognition and a relay o/p, tey used one of these at cussons the soap place to make sure that the machine that screwed a bottle cap on had actually done so, one of these might be handy for automated testing if youwant to go that way.

lansmnet said:

Hello All,
Thank you so much for your time and efforts really appreciate!
I uploaded a video to show you how we are testing heating wires currently. Visually it is totally OK, fast and accurate, thermal paper almost instantly changes its color. I have to extend our PLC controlled Poka-yoke system with Go/no-go testing to automatically block broken windows from further processing. With a regulated power supply the difference between 12 and 11 wires is very small. I measured current consumption using a non regulated power supply and the difference was clearer, 5-8% less Amps but this is still much smaller than the tolerance defined by the manufacturer. So I am afraid that 2 weeks later for example from the next delivery all parts could be evaluated as a broken ones. That's why I'm trying to detect wires separately. Than summarize outputs into one digital go/no-go signal.
Thermal camera I think would be very problematic, can catch fingerprints, or glass can be accidentally preheated, in this case the contrast will be insufficient. We are using infrared thermometers, pyrometers and heat switches in other departments in the factory but I thought that is difficult to measure glass surface, and these kind of sensors detects bigger area not just 0,5mm thickness.

Click to expand...

Dust, fingerprints and even warm glass won't make any real difference to your detection cycle using a thermally sensitive camera. You are looking at the delta temperature between ambient and the heating element. Also, unless you are frequently and randomly taking your glass outside your environment, that ambient temperature is going to be nearly constant, or at least very, very similar from window to window, so once you profile one type of window, it's not going to change much at all for all windows of that type. As for the area of detection being small, as pointed out by others, you are not looking at just a single heating leg but rather the whole window heating element, in one pass, i.e you see the whole picture in one go

If you were to automate the process, you could not only profile good window heaters, but also the poor ones too, like those with hot/ cold spots or other anomalies and set some sort of tolerance values to test against for QA. Even if you only use it for go/ no go testing along with a current measurement, it should be reasonably straightforward to implement using a camera that is sensitive to infra red