Multi-temperature gauge question...

[kenamond]

I'd like to add some temperature sensors at a few locations in my intake and then have them wired to a switch in either the interior or under the hood and have the single switched output go to a single temperature gauge in the my instrument cluster. Then I can look at compressor inlet and outlet temps and intercooler outlet temps by switching between the sensors.

Has anyone ever done something like this?

 

[spoolup]

Thing about switching thremocoupler leads is that you have to keep in mind that ANY juntion made with a foreign material will also be sensed and will alter the tempature reading. Now if you send the thermocouplers into something similar to the Greddy contrl box that converts the data to voltage for the stepper motor then it may be possilbe but then you must use identical sensors, your post are all air temps which will quailify, but you can not mix a coolant temp with a type K thermocouper since the converter will reference the 2 different sensors the same way. now using software, IE DSMlink. after making a reference table for the sensor it would be possible. It would be quite a experimant,.

The mini cooper guys have done quite a thermo analysis on the stock top mount IC. But the thermocouplers were only installed during their test, I doubt many leave them in all the time. No real use for daily monitroing for them.

Do you have any background in eletronics or SW programming?

 

[DGajre777]

If you really want it, your best bet is to install these.

You must be logged in to view this image or video.

http://www.egauges.com/vdo_ind.asp?Type=Intake_Temp&Series=Phantom&PN=ATM-5772&Cart=1

On another note, I don't see why you would need to monitor the compressor inlet/outlet temps or the intercooler outlet temp everyday. If you use nitrous oxide on the intercooler and need to know the temperature change, that is another thing.

 

[Sivad87]

well how does the sensor send the signal to the guage? is it basically a resistor that changes resitance to different temperatures and the guage senses the voltage and reads it accordingly?

if you you can just make a kind of control box that with some relays changes the sensor the guage is reading with no ill effects. think of it as a tv selector box. when you want cable you swith it to position 1, satalite, position 2, so on and so on.

this all depends on how the guage reads the sensors.

 

[spoolup]

They also make a 2 channel that DGajre linked: http://www.autometer.com/cat_gaugedetail.aspx?gid=2648&sid=7
BUT they cost a pretty penny. http://prostreetonline.com/buy/autometer_phantom_intake_temp_gauges/5773/default.asp?

Oh and here is how a thermocoupler works http://www.omega.com/thermocouples.html it has nothing to do with resistors, there are no passive compnents in the probe. By the 2 joined metals and the small lenght of wire sense the temp change it will develope a voltage (VERY SMALL I may add) and the rcvr then converst that data. But you can not just point ANY metals, they must be matched, where the typical coax switch in your cable box could really care less.

Another good link http://www.temperatures.com/tcs.html

 

[kenamond]

Do you have any background in eletronics or SW programming?

I have quite a bit of SW programming experience, but I don't know if it's in the area you're talking about. I do computational physics for a living (write code that simulates physical phenomena like car crashes, etc.).

If it's a matter of installing 3 identical thermocouples and then plugging one into the control box, doing a test, plugging another into the control box, doing another test, etc., would that work?

My electrical background is limited, but I've made PCBs, rewired my house, made guitar effects boxes, put numerous car audio systems together, things like that, but I'd have no idea how to design a circuit from scratch.

Edit - wow, lots of responses all the sudden while I wrote this.

 

[Sivad87]

well heres a really quick schematic of what i was thinking. but from what was said about the thrermocouplers i highly doubt this would work this way. the relays in the circut would obstruck the voltage (slightly) more than likely and creating an inacurate reading. ill read up on the thermocouplers and see if i cant come up with something that may work.

You must be logged in to view this image or video.

 

[kenamond]

Thanks guys!

Since I'm a science geek, I'd geek out on the data and be able to determine alot about flowrate, IC efficiency, compressor efficiency, water sprayer effectiveness, etc.

Staring at and making sense of the data would probably teach me alot about my car.

I'd like to do the same with boost/vaccuum in similar locations to get a good idea of pressure drops various places in the intake and exhaust, but that's a much easier task (just tap in and add nipples at various locations - and add a baffle in the exhaust vac line to damp out the put-puts and then hook up the boost gauge to the different nipples).

If you know temperature, pressure, and mass flowrate, you can determine VEs of each part of your intake and exhaust, mass flowrate, etc.

 

[kenamond]

On another note, I don't see why you would need to monitor the compressor inlet/outlet temps or the intercooler outlet temp everyday. If you use nitrous oxide on the intercooler and need to know the temperature change, that is another thing.

Agreed. It would be for a handful of experiments up front and then others after I made changes or had sudden problems.

 

[Sivad87]

Thanks guys!

Since I'm a science geek, I'd geek out on the data and be able to determine alot about flowrate, IC efficiency, compressor efficiency, water sprayer effectiveness, etc.

Staring at and making sense of the data would probably teach me alot about my car.

I'd like to do the same with boost/vaccuum in similar locations to get a good idea of pressure drops various places in the intake and exhaust, but that's a much easier task (just tap in and add nipples at various locations - and add a baffle in the exhaust vac line to damp out the put-puts and then hook up the boost gauge to the different nipples).

If you know temperature, pressure, and mass flowrate, you can determine VEs of each part of your intake and exhaust, mass flowrate, etc.

ok since your a super sience geek as you put it. do you know if a 12dc spst relay will add anyresitence or change the voltage of a line? im to lazy to go rip apart the garage to find one of the spare ones and get my volt meter. if it doesnt change the signal line's voltage at all then my diagram will work accuratly on paper.

but we all know that it usually isnt the case.

 

[kenamond]

I think that the problem is that the thermocouple is based entirely on precise resistance measurements. They weld two dissimilar metals together, and the resistance at that location varies predictably with temperature. So if you change the resistance of the system, you throw off the reading. You could potentially recalibrate the sensor, but I don't know if that would mean ripping into the guts of the control box or what.

I'm not an electrical engineer, so I don't know what options there are for the recalibration part. Perhaps the control box has a calibration knob or there is a variable pot in there that you can tweak. I just know the basics that they taught us in our engineering measurement/experimentation class about thermocouples. In class, we calibrated using icewater (known temperature of 0*C), but we took all of our readings with a 75lb oscillascope (doesn't fit in my gauge pod and not very intuitive to read). I think the teacher made his own thermocouples, too. That was 15 years ago, though, and I'm having to remove masses of cobwebs from my brain to author this post.

 

[Sivad87]

I think that the problem is that the thermocouple is based entirely on precise resistance measurements. They weld two dissimilar metals together, and the resistance at that location varies predictably with temperature. So if you change the resistance of the system, you throw off the reading. You could potentially recalibrate the sensor, but I don't know if that would mean ripping into the guts of the control box or what.

I'm not an electrical engineer, so I don't know what options there are for the recalibration part. Perhaps the control box has a calibration knob or there is a variable pot in there that you can tweak. I just know the basics that they taught us in our engineering measurement/experimentation class about thermocouples. In class, we calibrated using icewater (known temperature of 0*C), but we took all of our readings with a 75lb oscillascope (doesn't fit in my gauge pod and not very intuitive to read). I think the teacher made his own thermocouples, too. That was 15 years ago, though, and I'm having to remove masses of cobwebs from my brain to author this post.

about the resistance to the system thing, thats why i said that if a simple relay doesnt add any resistance then my diagram will work. otherwise you would need to find some other way do dissconnect and reconnect the thermocoupler from the guage without interupting the reading.

im going to try and find out some more info on the effects of a relay. it shouldnt add any resistance because all they are is an electromagnet that moves a connection inside it. so there is nothing but a little bit of metal inside them interupting the signal.

 

[Sivad87]

ok so a dry reed relay has 0.1 and 2 ohms on average. does anyone know of a way to calculate voltage drop from the resistance of something?

 

[spoolup]

sivad, since you are using dis-similar metals, Ie copper, steel or gold plated lugs on the standart bosh style relay, will alter the tempature reading from the thrmocoupler. The Ktype coupler has 1 lead made of chromel the other of alumel. If you were to try and extend the couple using speaker wire. you will have a VERY false reading, but now you use the correct wire BUT swap the leads of the maiting wire. You will read a value of the item you are trying to measure minus the tempature of the air around where the extension piece of material is residing. IE EGT is 1600F, and the splice is made in the car wiht AC on at a nice 71F, the gauge will read @1529F.

As for using your common fluke multimeter to measure the resistance of a piece of metal is, for the lack of better words, meaningless. The correct way to do it is to use a milliohm meter (4 probes, not 2) that can have resolution to the 0.000000 ohm where your typical DVM is only reading 100th of a ohm 0.00 so you seeing a 0.01 could actually be a 0.000235 ohm. plus the uncalibrated leght of the testleads. Connect the 2 leads together and you may not even get 0.00ohms. Flukes have a lead loss calculation that try to guess but it is still a guess.

Kenamond, a quick google search and I found this little guy. http://www.aprsworld.com/thermok/ not a gauge but is pretty nice. with 232 output for a logging, and runs off of 6 AAA's. I am sure you could get this guy to work for you.

sivad, ohms law, e=i*r, i=e/r, r=e/i get 2 find the 3rd. A REALLY good website ofr mobile eletronics with a great forum is http://www.the12volt.com/

Your diagram above would work for autometer water temp sensors ect. since they use standard wire with a crimped eyelet. For thermocoulers however, it is not feasible. Sorry.

 

[Sivad87]

sivad, since you are using dis-similar metals, Ie copper, steel or gold plated lugs on the standart bosh style relay, will alter the tempature reading from the thrmocoupler. The Ktype coupler has 1 lead made of chromel the other of alumel. If you were to try and extend the couple using speaker wire. you will have a VERY false reading, but now you use the correct wire BUT swap the leads of the maiting wire. You will read a value of the item you are trying to measure minus the tempature of the air around where the extension piece of material is residing. IE EGT is 1600F, and the splice is made in the car wiht AC on at a nice 71F, the gauge will read @1529F.

As for using your common fluke multimeter to measure the resistance of a piece of metal is, for the lack of better words, meaningless. The correct way to do it is to use a milliohm meter (4 probes, not 2) that can have resolution to the 0.000000 ohm where your typical DVM is only reading 100th of a ohm 0.00 so you seeing a 0.01 could actually be a 0.000235 ohm. plus the uncalibrated leght of the testleads. Connect the 2 leads together and you may not even get 0.00ohms. Flukes have a lead loss calculation that try to guess but it is still a guess.

Kenamond, a quick google search and I found this little guy. http://www.aprsworld.com/thermok/ not a gauge but is pretty nice. with 232 output for a logging, and runs off of 6 AAA's. I am sure you could get this guy to work for you.

sivad, ohms law, e=i*r, i=e/r, r=e/i get 2 find the 3rd. A REALLY good website ofr mobile eletronics with a great forum is http://www.the12volt.com/

Your diagram above would work for autometer water temp sensors ect. since they use standard wire with a crimped eyelet. For thermocoulers however, it is not feasible. Sorry.

i see now. i didnt know that they used a certain type of metal for the whole run of wire. it would just throw the guage off by who knows how much. i made that diagram from the understanding that they work similar to a water temp type sensor. so with a thermocoupler it would be really hard/expensive to make something like that work.