When you click on links to various merchants on this site and make a purchase, this can result in this site earning a commission. Affiliate programs and affiliations include, but are not limited to, the eBay Partner Network.
i have been using an air temp type K thermocouple for my IAT since 2013 and really like having 4 accurate readings per second. i started looking for solution after noting that my stock sensor read 27 C throughout my entire 2000 to 9000 run on the dyno.
i am just slightly aware that others are using a "fast" thermistor. the 64 $ question is... just how fast is "fast."
an important question because if your sensor lags you are going to think your IAT is just peachy when it might be getting you close to knockville.
i have spent a bunch of time searching including calling my wonderful EGT guys at The Sensor Connection but have come up w nothing as to response data.
consequently, i am going to acquire a "fast" sensor and install it along w my thermocouple in my UIM. we will then do a match race.
BTW, i have no idea as to the outcome, i just want to see them race.
RX7Clubbers, i need your input... what is considered to be the fastest thermosensor to test?
Thanks!
Last edited by Howard Coleman; Jun 26, 2019 at 08:10 PM.
An exposed thermocouple or platinum resistor with very modest ceramic shroud will likely be fastest. Something that relies on couple or metalic resistance properties. Unfortunately the fastest will tend to be more vunerable to mech damage or contamination although the resolution of automotive analog to digital converters may make that irrelevant anyway. National labs use a mix of platinum resistors and various exposed thermocouples for high accuracy measurement in the ranges we are concerned with.
One of the other great things about thermocouples (apart from being accurate and responsive) is that they're cheap. Sure the amplifiers can get expensive. But once have a 4 channel one, you can measure EGT on each exhaust runner, as well air temp before and after the intercooler.
Howard, out of interest which type K thermocouple and amplifier have you been using for air temp? I've found one from one of the recommendations go to thermocouple supplier in Australia, Auber Instruments. They offer one used for automotive and aircraft air temperature readings and it can be used with adapters to suit Mazda OEM mounting. However, the probe is 150mm long which is a bit impractical. They mention that a thermocouple for intake air temperature should have small thermal mass (which makes sense so it can shed heat quickly when intake air temp changes). See: https://www.auberins.com/index.php?m...roducts_id=589
What Howard is preaching to the masses, a "fast-acting" Triumph/GM sensor (while better than the stock version) should not be our standard/go-to sensor for IAT; rather, we need instantaneous readings.
*Interested to see what the data shows during the pending comparison testing against a K-Type thermocouple.
Last edited by Topolino; Jun 28, 2019 at 07:41 AM.
I guess it seems quite fast compared to the stock one. Actually, wouldn't the mass of the intercooler tend to slow down changes in temperature? Maybe they don't need to be that fast. Obviously stock isn't good enough, but a 30 degrees drop in 10 seconds seems like a fair bit. Interested to see the results. Actually, that might be an ideal sensor to compare against seeing so many people use it... and by many people I mean me!
I might try test it out with a hairdryer or something if you guys want?
K-Type thermocouples are rated for multiple readings in less than a second; compared to 5-15 seconds likely for a "fast-acting" or stock sensor. For a stock or mildly modified car, perhaps this is generally okay. But more frequent/quicker & accurate readings are better in either case.
When the comparison data comes in, you're likely to see a minimum of 20x faster readings using the K-Type thermocouple vs the current standard.
IAT dynamics happen REALLY fast. if there is any lag in your sensor you are in the dark on a very important metric.
here's a recent log proving the point.
boost went from 8.9 psi at 2:59.790 to 19.4 at 3.00.374
just over a half second!
IAT went from 77 F to 126 F according to my thermocouple.
here's 8.9 psi:
here's the log at 3 minutes and .374 seconds:
IAT rose from 77 F to 126 F in less than three quarters of a second. note the bottom section of the log... that's my 100% methanol pressure log. AI pressure rose from 23.2 psi to 121.6 psi. i can't imagine where my IAT would be without it. i do plan to add a bit more meth to drop my IAT below ambient.
thanks to the magic of methanol my IAT trends downward to 117 F at 7722 rpm 24.9 psi boost and approx 585 hp.
my IAT data is generated by a specially engineered "air" thermocouple from The Sensor Connection. the sensor is made by them in the USA and is a product of more than 20 years of engineering to get the right balance between fast and durable. it generates four readings a second and is accurate to around one per cent.
for the record i have no idea as to the speed of other thermister type sensors. looking at the graph in post 5 isn't much help. there are 20 seconds between 140,000 and 160,000 Ms. it appears that the initial rise is about 20% of the 20 seconds so about 4 seconds... the reason we can draw no absolute speed conclusions from the data is that the engine may have taken four seconds to do it's thing. all we can conclude is that it is faster than the glacial speed OE sensor.
i will be adding a thermistor to my UIM shortly and we will have comparison data.
i have a suspicion that if we knew our actual IATs there would be a lot more re-thinking of system design.
^Holler if you need a Triumph fast-acting sensor to benchtest. I previously bought a couple of them off the forum prior to going the k-type thermocouple route myself (via Speedhut air intake temperature gauge).
Last edited by Topolino; Jun 28, 2019 at 07:13 PM.
I just did a crude test using the Triumph sensor, a multimeter and a hairdryer and it took roughly 4 seconds to go from 14deg to 35deg. I didn't expect that IAT would change so quickly, thanks for that data Howard. Do you have a link to the thermocouple you're using?
Obviously there is typically a lot more air flow over the sensor than a hair dryer can provide, so I just tested using water and saw the temperature go from 15-60 degrees Celcius in 1.5 seconds. Sorry, the car is in bits so I can't test any better than that. Looking forward to seeing your results.
^For reference, TSC's AIT probe is the same as their EGT-AP series probe. With the smaller 1/8" probe tip, curious to know how durable this probe has been for you.
The extreme duty IP series probe is rated the fastest of the bunch (2 add'l readings each second), and the most durable; presumably, due to a 1/4" probe diameter vs. 1/8". Any consideration given to this probe? IP series (extreme duty probe)
Last edited by Topolino; Jun 28, 2019 at 08:03 PM.
^Closed tip thermocouples provide more longevity (& reliability) over the other k-types mentioned, but at the expense of super-fast response time; I.e., just 1 reading per second vs multiple readings per. Still faster than the fastest thermistor probe, though.
For reliability purposes, I suppose this is why the SpeedHut IAT gauges come equipped w closed tip thermocouple probes. Perhaps a happy medium for most enthusiasts (longevity/accuracy vs response time), and the ultimate route I went. But I can definitely see where a high HP engine would prefer the open tip version for instantaneous response time.
I thought my sensor was fast acting till I read this thread ...now Im not so sure. This is the common GM IAT sensor. Log is 0-15psi by 4000rpm then 15 constant to 7500 then back to zero. My thought was that the lag was only about 0.5s judging by the time it takes to peak after throttle is let off. Optimistic ?
Temp is blue line in deg. C
Rpm is red line https://www.diyautotune.com/product/...-with-pigtail/
I managed to dig up the specs in older catalogues and it has a "thermal time constant" of less than or equal to 5 seconds. For this particular figure it was the measured time constant for a temperature jump of –80 °C to +20 °C at an air-flow rate of ≥ 6 m · s–1. While this sounds slow, the other fast acting Bosch Motorsport NTCs suitable for air temp measure 44 seconds. It is not the same as response time. See page 50 here: Wayback Machine
Compare this to the popular fast-response GM Delphi IAT sensor, which has a time constant of less than 15 seconds, 3 times slower than the above product, see: http://pe-ltd.com/assets/air_temp.pdf
I have struggled to find published thermal time constant specs for popular K type thermocouples used in automotive. This article is the best I could find just now: https://www.omega.com/en-us/resource...-response-time
It suggests that for a grounded probe thermocouple with 0.04" diameter probe, the time constant is over 2 seconds in air. This is measured at air flow rate of 65 feet per second (~20ms-1 ie much higher air flow than Bosch test). A 3mm (0.12") diameter grounded probe suitable for use as an automotive air temp sensor is therefore likely much much higher than that figure (response slows with increased thermocouple size). An exposed tip would be much much faster.
Based on this data, it suggests the Bosch Motorsports NTC sensor is likely as fast as or faster than a grounded non-exposed tip TC, but probably not an exposed tip TC.
That's the best data I could find, although I'm sure there is more out there if I had time to dive deeper.
interesting that a post popped up in this thread as i was just going to post here too.
i have been waiting, a couple of months for a new set of studs prior to re-installing my motor. they arrived a few days ago and i expect to have my motor back in a few days and with it do some comparative tests.
i am going to use the IAT sensor that came w my ViPEC (that i never used). it is an open air thermister.
as i understand there is another faster thermister sold by wannaspeed. if i test just my thermister it will solve little as people will say why didn't you test the faster one. i am not about to spend $95 to test it and then have no further use for it. if someone has the sensor lying around please consider sending it to me. i will return it in good shape in a couple of weeks. i believe the two thermisters have differing threads so before i finish putting my UIM on i will need to know if the faster one is heading my way as i am not going to go through the process of re-removing my UIM at a later date to add a different bung. if i don't get it i will test the one i have.
just an update... i did receive the faster IAT sensors from Topolino, thanks, and will return them when finished. it appears they have 10 1.25 M threads and i have looked around for an aluminum bung but haven't been successful. any ideas as to availability? i can always fab one. should be up and running shortly with results. BTW, i have no idea what i will find, just looking for comparative data.
just an update... i did receive the faster IAT sensors from Topolino, thanks, and will return them when finished. it appears they have 10 1.25 M threads and i have looked around for an aluminum bung but haven't been successful. any ideas as to availability? i can always fab one. should be up and running shortly with results. BTW, i have no idea what i will find, just looking for comparative data.
Howard,
I had the same problem. No weldable M10 bungs. Buy a male -6AN weldable aluminum bung. $5 on Ebay. Hold the male AN portion in a vise so as not to ruin the finished usable portion of the bung. Tap the M10 threads through the bung and then cut off the male AN end and sand the end. It works perfectly and looks factory made.
Mike
