Inlet Air Temperature Measurements

Post by **mat_fenwick** » Mon Jun 20, 2011 9:50 am

electrokid wrote:The thermometer I have that could be used to measure inlet air temp on the BX has a lead that is just over 1 metre in length so probably just enough - I expect I have a longer version somewhere (it's a K type thermocouple)

Very OT, but my ears pricked up at the mention of thermocouples! First of all I'll say that K types are pretty poor with their sensitivity, a T type (if your meter will take it) would be much better in this temperature range. Secondly, and probably more usefully, if you need a thermocouple making up to length I can sort one out. The cost is pence for a PTFE sheathed one.

electrokid · Post by **electrokid** » Mon Jun 20, 2011 9:32 pm

Very OT, but my ears pricked up at the mention of thermocouples!

Yes - I realised I was wandering into your territory

Thanks for that Mat - the electronic thermometer I have is K type only - it seems to be ok at ambient and low temps. I have a design on the drawing board using an LM335 but that's not going to happen soon and it's primarily for a computer interface anyway and designed to be good at monitoring small changes in temp rather than have any particular accuracy.

I just figure it would be good to know what temp the air intake should be - that will have an effect on the performance of the carb and its setup. Faults with the temp controlled flap could easily look like a carb problem so finding what its characteristics should be and finding the best way to check it out should be useful.

Post by **mat_fenwick** » Tue Jun 21, 2011 9:48 am

electrokid wrote:designed to be good at monitoring small changes in temp rather than have any particular accuracy.

That is what I actually meant by sensitivity - for any given change in temperature, a T type will have a greater change in voltage output. That said, when I actually checked the figures, there's not that much difference anyway!

Have you considered a thermistor based circuit if/when you go to the trouble of building one? They tend to have a huge change in resistance for a given temperature change, and avoids any need to build in cold junction compensation.

it seems to be ok at ambient and low temps.

You could plug any thermocouple into a meter at ambient temperatures, or even a loop of wire, and it would read exactly the same...

electrokid · Post by **electrokid** » Tue Jun 21, 2011 11:06 am

The choice of meter is determined by... the one I happen to have

which is a Vulcan Lokring GTH 1150 (it's designed for HVAC folks and I bought it via ebay for not a great deal because I didn't have anything that could cope with 1000+ °C at the time). It's max error is 1% + or - 1 digit and I'm confident that it's in spec - recently checked BX temp sensors with it and soon to check a couple of rad fan switches as well.

http://www.directindustry.com/prod/vulk ... 92886.html

The beauty of the LM335 is that it's linear by design. Looking at the schematic page 1 of http://www.national.com/ds/LM/LM135.pdf it's Q1 which is the 'final' control over the voltage between + and - and it's all the other gubbins that controls Q1.

How it works isn't obvious - even to a half decent analogue engineer like wot I is - Q10 and Q11 form a current mirror and R1 is laser cut calibrated during manufacture. The current in R1 is mirrored in Q9 which is a kinky transistor with 2 collectors providing outputs via Q7 and Q8 to what is effectively a 'bandgap multiplier' formed by Q" through Q6. I've ignored the adjusment pin because describing that is hard going !

I think the LM335 is attractive because the math is really easy and precisely defined by the physics of silicon junctions to be 10mV per degree Kelvin - I tend to like anything that makes the maths easier

In this instance though, I'm mainly interested in what the T controlled flap does and how best to test it - I think it's one of those things we normally assume is either working or not that important if it's not working. I think it may be critical to carb operation in some areas.

It would be good to know of a test for the flap controls other than stick a ciggie lighter under it to see if it moves - apart from anything else I've given up smoking now so I've no idea where my ciggie lighter is.

Sadly, life is not so technically challenging at the moment - primarily consisting of sticking some of these on top of each other...

http://www.thakeham.co.uk/index.asp?pag ... g&type=tth

mnde · Post by **mnde** » Tue Jun 21, 2011 1:31 pm

electrokid wrote: It would be good to know of a test for the flap controls other than stick a ciggie lighter under it to see if it moves - apart from anything else I've given up smoking now so I've no idea where my ciggie lighter is.

Hehe, I recognise this from our conversation on Sunday

(for everyone else's benefit: this is what I did to test the thermostatic carb pre-heat flap on the GSA many years ago - using the ciggie lighter from the car that is, not a disposable one!)

Mark.

Post by **mat_fenwick** » Mon Jun 27, 2011 2:49 pm

electrokid wrote:Lots of interesting stuff that is too advanced for my little brain

Ahhh, I had (wrongly) assumed that your LM335 circuit was to be using a K type thermocouple - now I see that it's all pretty much self contained.

If you had been building a measurement circuit from scratch I would have suggested avoiding a thermocouple based system, but seeing as you're cheating (

) by buying in a circuit that doesn't matter...

electrokid · Post by **electrokid** » Tue Jun 28, 2011 12:26 am

The most significant factor is probably SINAD..

Signal In Noise And Distortion.. Thermocouples are very low impedance and if that's factored in correctly when designing the amplifier stages it'll produce similar results to devices that are inherently more sensitive but have a higher impedance.

Actually I went off at a tangent there - yes I want to measure the inlet air temp - firstly so we know that that bit of Mark's car is working properly and secondly so I end up knowing how that bit works (and how that relates to Pentane - the lowest fraction currently included in petrol - which has IIRC a BP of around 35°C).

But since no-one is replying about that I'll prattle on about the temp sensor of the tangent variety

Yes - cheating / buying in - true

but it's a short cut to something that's already way too complex.

It's 'field equipment' in that it (or 'they') will be fitted to an experimental rig where temperatures and temperature changes need to be monitored - connected by a twisted pair of some length - say up to 100m - to an interface board in a computer. The LM335 is ideal for that application (along with other goodies such as a 555 to produce temp to pulse width and op-amps to control charge linearity etc etc). If I do need to monitor higher temperatures where anything other than thermocouples combust then the LM335 can be used (see the application notes as in the link) for cold junction correction.

Vanny · Post by **Vanny** » Tue Jun 28, 2011 4:46 pm

I hate thermocouples.

50+ on the bench, 250+ on the car, its a damn jungle in there.