more from the innovate forum NTK Vs LS4U

[norbs]

More interesting stuff i thought the buick guys should read


Hi,

The differences between the sensors are:

1. Bosch has high heat sensitivity. It needs to be actively temperature controlled while the NTK has typically a passive temp. control by holding its heater voltage constant (very simple to do).

2. Bosch is sensitive to housing temp. NTK is not

3. Bosch has about 5-10 times faster response speed than NTK due to its planar design (measured delay time between pump cell current change and sensor cell output: NTK ~50msec, Bosch < 5 msec).

4. Bosch has less than 1/2 the NTK's sensitivity to back pressure. This means the readings are richer than actual on the rich side, leaner than actual on the lean side when on the lean side when back pressure is present.

5. Bosch sensor heats up more than twice as fast than NTK. This means it's also more susceptible to heat shock from condensation water because the NTK's slower warmup allows the exhaust system to warm up before the sensor can be damaged. The Bosch's fast warmup is required for smog reasons (it's main application). On the other hand the NTK is a more open design that is more susceptible to damage when hit by debris or water when hot.

6. With respect to survivability with leaded fuel they are about equal.

For most tuning applications, specially where fast response is required (not steady state load) the Bosch is the better sensor to use. The slower response of the NTK causes 'smearing' of the values. This leads to prettier graphs. A case of pretty vs. true and real (as it's often with girls).


Regards,
Klaus

 

[TurboTR]

Well for some flipside info, as far as real world tuning use, my experience is that I'm yet to have to replace an NTK sensor in a FAST unit application for example in normal use. I've had several customers want them just in case, but personally have yet to see one "needing" to be replaced under normal use. My own car still has the original sensor in it, going on 7+ years now... And I run C16 almost exclusively. We have several of the LM1 units that we use for F*rd chip tuning, among other things. They are a very good product, very successful and very useful of course, but sometimes we see results that are well off vs what the engine itself is indicating and what the stock O2(s) are saying. Particularly at low flows (around idle). We have another NTK-based standalone unit to compare against if need be (along with more than one bung to access on the OBD cars) and the results have been... interesting. Have had to replace at least one Bosch sensor already that had less than a year on it, with no leaded fuel either. Good thing that they are a bit cheaper I guess Not knocking the products, love them overall, just some of our experience so far.

Pure specs can be a very misleading and confusing tool, as anyone who sees daily news knows For example, if the response time is/was really up to an order of magnitude different for the two sensors, well that can indeed make very effective marketing schpil, and might matter greatly in say the typical OEM ramp/jumpback fuel control algorithm for example. But as far as what most of us use a standalone wb for- checking the composite a/f ratio of multiple cylinders running at the same time- who cares. It sounds scary on paper, "faster response", oooooh, but in the real world of use it matters ~ nada for the typical user in this case. To be useful in human eyeball terms the data gets low pass filtered anyway Umm, what is the response time of that slick LED display again mr engineer? Oh, > half a second, got it... And that nice data logging output- a whopping 10 Hz, one pt every 100ms? Got it... So then what diff does it make in this case if the sensor itself responds at 5ms or 10ms? Not much really... How about comparing the response time of a healthy switcher vs either wb sensor? It is likely faster still; so does that make it "better" then? Hmmm...

It's interesting reading, but IMO if we really want a useful, unbiased comparo, all sides of the story should be examined IMO, not just the output form one mfgr who probably also has somewhat of a keen sales interest

TurboTR

 

[norbs]

I really don't know what to believe. I can't not see the response time of the ls4u being 10 times faster either. Thanks for the detailed opinion, some things are not true in real life applications as you have found out..........

 

[TurboTR]

Yeah well is just me 1 penny opinion, lol Is great reading- keep it coming And one feature of the LM-1 I believe is the ability to mimic a switching O2 sensor output, so response time is for sure a factor in that case...

Excessivley long response time would throw off the transport delay settings in a ramp/jumpback algorithm, which would then probably not work quite right anymore

TurboTR

 

[bruce]

Originally posted by TurboTR
Excessivley long response time would throw off the transport delay settings in a ramp/jumpback algorithm, which would then probably not work quite right anymore

Not really. I moved the O2 sensor to past the cat converter, and things ran just fine (in several different appls., thou they were heated).
With the way the code just toggles across Stoich., I don't see where it matters much.

 

[TurboTR]

Ok.

TurboTR