Syvecs Forums

Maybe I've misunderstood something then.

The pin-out data I have is:

L1H1
1 Signal Ground
2 Ion Pump current
3 Nernst cell voltage
4 NC
5 10k
6 10k
7 Heater +12V
8 Heater drive from ECU

The L2H2 does NOT have the "10k" connections, and I always assumed that was the calibration resistor, but perhaps I'm wrong...I usually am lol.

Yeah, maybe so... but none of these sensors, do not contain such a high valued (10K or so) calibration resistors. Bosch is using 30-300 ohms range (from memory), NTK range is 'somewhat unknown', but i have measured values from around 260 to 607 ohms.

Here is one NTK L2H2 type pinout, with Bosch connector.

Pay attention to pins 2 and 4...

Sensor connector
pin 1 Yellow Heater Gnd
pin 3 Blue Heater (3.8 Ohm @ 25C) +12V
pin 5 Black Ip(-), Vs(-) Gnd
pin 2 Grey Vs(+) Nernst Cell Voltage
pin 4 no wire, since only Vs(+) Nernst Cell Voltage thru calibration resistor (inside connector connected Pin2->Resistor->Pin4)
pin 6 White IpI(+) Ion Pump Current

Note!
Nernst cell Voltage at Pin 2, is without calibration resistor, pin 4 voltage is thru calibration resistor.

I'll take a closer look at an LZA-09-E1 and see if there's any evidence of a calibration resistor between pins. This sensor has a Sumitomo connector and is sold as a Honda L1H1 replacement, albeit 5-wire no 7 wire. It may be that the output pin is, indeed, through a calibration resistor and I just didn't know.

It does make me wonder what the 10k on the L1H1 connector is all about though!

Hi All,

The linearisation you posted is certainly very close .... however I did notice some oddities in the values between 10 and 11 AFR (not that we are likely to be running this rich most of the time, but worth being aware of).

Cheers
Richard

5 pcs Sumitomo connector NTK sensor...

LZA08-H5 5 kohm
LZA08-H4 6,1 kohm
Honda 36532-P07-003 ?? 30 kohm
L1H1 6,1 kohm
L1H1 11,9 kohm

A close look at an LZA-09E1 (L2H2) shows there is a part of its connector blanked off...which I imagine is what contains the calibration resistor. I'm not going to open the sealed packets of my stock to measure it, as I couldn't then sell them.

I think the answer is simply that the L1H1 has the calibration resistor completely isolated (and the drawing I have happens to show it as 10k, whereas it must vary sensor to sensor), whereas more modern versions already have one end connected to the Ion Pump wire, but then there *should* be 6 wires connected from the lambda connector back to the ecu, not 5...

...so, if that's right, and the UEGO connector on the Syvecs does not have a calibration resistor connection, do we infer that individual sensors will need to be manually calibrated, each and every time...unless the ecu is doing some form of free-air calibration (which I doubt!).

Time for Pat to comment, I think

I think that each sensor needs to calibrate individually. All sensors has 'somewhat different' linearization curve and/or slope, when resistor value differs among other.

Btw.... by 'such a high resistor' value - at my earlier post, i mean/meant L2H2 type.

The calibration resistor can/should be used to auto-calibrate the circuitry in the ecu...but I suspect that the Syvecs does not make use of it...so, yes, for a Syvecs, individual calibration is required...which is impractical for most owners

Again, suggest we need Pat's input in case (as is often the case) I have it wrong

Gents,

The EUGO circuitry in the S6 doesn't make use of the cal resistor because it wasn't designed to work with the cheap mass market LZA09-E1 type sensors, it was designed to work with the expensive motorsport version of the sensor. The calibration will be correct if the intended sensor is connected but many people are put off by the cost of that unit. It should be noted that when using a pattern part sensor in place of the correct sensor, one should not take the readings as being accurate until proven to be so. This means that if readings from the sensor are to be used then they should be cross referenced against a known good lambda meter at the time of installation. This is easy enough to do at the point of mapping, most dynos have wideband lambda meters of some description. Please do however take a moment to reflect on the fact that a laboratory grade meter costs more than an S8, and even the de-facto standard MoTeC Professional Lambda Meter costs almost as much as an S6. We could always turn the S6 into a lab grade lambda meter and quadruple the price if that's what people want

Hope this helps,

Pat.

What "motorsport grade" sensor is the S6 designed for then?