Syvecs Forums

UPDATE: I've had 10 different 'cheap' NTK's sent to me now and upon calibration they all required different curves, some were out allot others tight within the 0.85L to 1L range, each one so far has been very different (so no generic curve will do).
If using these lower cost/grade sensors I would recommend to send it to someone who can do you up a custom sensor calibration and mark the sensor as such, least then you know it will read what it is mean too across full operating range.

How much would you charge for this service Peter, roughly?

email me (or pm here)

Benkku wrote:

TimH wrote:Only the L1H1 has a calibration resistor - NTK L2H2, and I imagine the Walkers copy, do not have the calibration resistor.

Are you absolutely sure about that?

Asked this directly from NGK Europe subsidiary, and they responded that sensors contains resistor.

"we are not selling a trimmed wide band sensor. All wide band sensors are calibrated by resistor."

No, I was wrong. It was a very long time ago that I wrote that lol. But the Syvecs does not read the trim resistor so it's a bit of a moot point.

CT9A wrote: ↑Fri Jun 10, 2016 11:44 pm Thanks for posting this info, I bought the Autronic model B with a good low used NTK sensor. I will buy a new sensor anyway, then we can calibrate the cheaper NTK for Syvecs use.

How did you get on with this?

Did a stack of new LZA sensors last few days. I convert these over to the DTM plug and when pulling the OE one apart there as some massive spreads in the resistor fitted to each unit and corresponding large differences in the readings they put out across virtually all the output range.

Did a little video showing the process for those who want to try and do this themselves?

It is not that hard but you do need a set up and the time, as described previously its not just done in one test, takes this repeated 3 times or more depending on how it comes up.... I generally aim for better than 0.5% accuracy across the whole calibration curve.

https://www.youtube.com/watch?v=PvArF4Y ... e=youtu.be

You really do need to do this, especially if you are running at the richer end of the scale for a port injected engine (turbocharged), depending on your 'luck' you could have a real turd of a NTK LZA which will be up to 5+ percent incorrect (either way + -) and you just dont know till you check it.

Here it is in picture form, as some hate Lambda or understand a percentage differential
Once completed you can get it very close or as much effort that you want to put into a bespoke individual sensor calibration that is

Did a batch of varying NTK consumer grade sensors and thought would put this up showing the process some more for those interested.

Video > https://www.youtube.com/watch?v=r1r2wl4 ... e=youtu.be

A few hundred iterations later you can get it to this level of precision, these are LZA09 sensors to lab grade NTK reference.

Another RR enhanced 'shit box LZA09' off the production line thinking its a Lab Grade NTK