Wednesday, Oct 19, 2005 at 23:44
Hi Dozer,
I don't want to bag the opposition, especially when they probably can't respond directly, but a couple of things about the Jaycar kit:
. It's an enthusiasts kit of electronics components, which requires the buyer to have the skills to assemble such a kit, including electronic soldering and the identification of resistors, capacitors, diodes, ICs, etc. I'm not saying that's too hard (I've been doing so for 30-odd years, so it can't be that hard), but it's not for everyone.
. Not sure what was in your kit but the thermocouple shown on the Jaycar website would not be suitable for EGT applications - the lead materials wouldn't last more than a few minutes in engine bay conditions. Did you by any chance use a more robust industrial MIMS thermocouple with fibre-glass insulated, stainless-steel braided leads for your petrol turbo engine application? That would be a much better choice but such sensors do cost more.
. I can't swear to this but I recall that this design (From the Silicon Chip "Performance Electronics for Cars" booklet) provides a linear voltage to temperature conversion using a Seebeck coefficient of 40.6 uV per deg C. This is OK for lower temperatures but is not accurate at higher temperatures (above, say, 400 C). While it's far from the worst, type-K thermocouples do still have significant non-linearity at higher temperatures and an indicator design without linearization will have errors of around 20 degrees at 700 C and worse at higher temps. (If anyone's interested, I can dig out the actual linearization error at various high temps.)
So, while the Jaycar kit is quite good at it's job when correctly and competently constructed, a factory-built and calibrated indicator with appropriate cold-
junction compensation and thermocouple linearization will do a better and simpler job for most users.
Ian
FollowupID:
389283