Doppelgänger

iTrader: (12)

Alrighty, new sensor is in. A quick run through the gears with some data logging shows AFRS dropping smoothly under full boost / 100%tps..which is good. The odd is my AFRs only go as high as 16 when I let off the throttle...so I might need to retry the free air calibration. Also, AFRs at WOT were low 10s...which could also be calibration. But the important thing is that the range of AFR seems more normal and linear to TPS/Boost.

Funny thing. When I took the 13246 out of the box, there were no markings on it, but if you pull back the loom/sleeve that covers the very top of the sensor, there is NTK markings on it. I mean, I knew it was a re-boxed NTK by my research, but it's funny to see how Bosch tries to cover it up

y8s

iTrader: (8)

I'm stilll curious about that super cheap one I linked to. it had the right honda P/N...

Doppelgänger

iTrader: (12)

yeah... I was thinking about it, but I can return anything I buy in person much easier if it doesn't work. The only one I would have ordered online is the one from FM.

I'm getting good AFRs under boost, but my off throttle decel is only going as high as 16...so I need to figure out what is up with that.

Doppelgänger

iTrader: (12)

OK, update on the WBO2 issues i've been having.

I re-calibrated yet again last night. I didn't really change much on the Zero Cal and Grade Cal (120 and 115). I was also looking through some of my older maps at various settings and saw that the autotune RPM range was 1500-7500 and the map i've been tinkering with was like 1500-3500, so I changed that back. I also noticed the setting for cutting fuel upon lifting off the throttle..and activated that setting. So I got a little data logging done on my way to work and was pleasantly surprised. On boost AFRs properly dropped to the low 11's and held steady at WOT. But the difference when I lifted off the throttle... AFRs jumped up to 20.5 within ~1-2 second, which tells me the WBO2 is now reading correctly...and that i'm saving a little bit more fuel while driving around

Unfortunatly I was stil l getting a bit of breakup when rolling into WOT...but it smoothed out at high RPMs and one I shifted, the entire next run through the RPMs was smooth. This leads me to believe it's something with the spark plugs. Probably get some new ones, gap them to .032 and go from there.

So for public record.. YES, the Bosch/NTK 13246 DOES wirk with Hydra 2.15

kujitsu

does it work with 2.5 and 2.6 too? think splicing to a gauge is possible? or would it change voltages or sensitivity. I've got 2.1 but waiting for 2.6

schmoo

So does anyone know if 15401 works with 2.6, or only with 2.5?

schmoo

I've been looking at the material referenced in this thread and did some cross-referencing. My findings:

oxygensensor.net says that the L2H2 is NTK 24300 and lists it at $189
wbo2.com says L2H2 is Bosch 13246
y8s experience is that Bosch 15401 will work with 2.5

If I go to NAPA USA and search with part numbers above, I get:

Search Term / Result / Price
24300 / NGK 24300 / $404
13246 / BSH 13246 / $429
15401 / BSH 15401 / $149

If I go to NAPA Canada and repeat the above, I get (prices aren't listed on-line):

Search Term / Result
24300 / nothing valid
13246 / NGK 24300 / $355CDN
15401 / NGK 24302 / $218CDN

NGK 24302 is listed at oxygensensor.net at $119

At Autozone:

Search Term / Result
24300 / nothing valid
13246 / BSH 13246 / $318
15401 / BSH 15401 / $149

schmoo

Added NAPA Canada pricing.

vtjballeng

iTrader: (1)

There are many number crosses but there are basically 2 sensor variations here. One operates on a 4mA pump cell and the other is a 3mA pump cell. The Hydra appears to have been designed around the former and they remove the calibration resistor when they put their connector on. The old sensors they sold come with a Delphi Weatherpack connector and the new ones come with a Delphi Metripack connector. My company deals frequently with these connectors and sensors but as FM sells them I am only going to clarify which sensors will work.

24300 / L1H1 / L2H2 / 13246 & several other numbers from Honda, Bosch, NTK & third party suppliers like SMP, etc is the 4mA pump cell sensor. Any minor variation in sleeving, harness length or connector type will result in a different part number from the same supplier. These are all made by NTK and NTK only, all other suppliers just put this in their own box.

24302 / LZA08-H6 & several other numbers from Honda, Bosch, NTK & third party suppliers like SMP, etc is the 3mA pump cell sensor. Any minor variation in sleeving, harness length or connector type will result in a different part number from the same supplier. These are all made by NTK and NTK only, all other suppliers just put this in their own box.

The ability to use one sensor or another depends on the current sense resistor on the board design and whether a controller is using the cal resistor and has all cal values pre-programmed in the unit OR does a free air calibration procedure, which periodically needs to be repeated due to sensor aging, which requires the pump cell curves for various sensors to be pre-programmed in the unit. The Hydra chooses the free air calibration method and appears to be designed around the 4mA sensor. If you use the 3mA sensor, the calibration curve is NOT identical to the 4ma calibration curve. Depending on the current sense resistor Hydra chose, the 3mA sensor may or may not be able to reach a free air calibration point. If it does, the result will be misleading because the non-linear pump cell amperage curve will most likely result in an increasing offset towards lower lambda values. These sensors determine Lambda by evaluating the current passing across the pump cell circuit in a relatively elaborate control circuit where the measurement cell is kept at a constant .45v and the heater circuit is modulated to maintain temperature. These sensors are temperature and pressure dependent so different installation locations may have different results.

To make this shorter and simpler, even if a 3mA sensor passes calibration, the values may be wrong. The difference will be most severe towards lower (richer) lambda values. I recommend you only use the 4mA sensor which is typically $250 or higher in all variations. To clarify further, it appears that all Hydra revisions use the 4mA version so it doesn't matter if you are running version 2.x or 2.y, you need the 4mA pump cell sensor type.

Personally, I have the Hydra sensor installed in the post-turbo location on my MSM. I have an NGK Powerdex AFX (about the same price as a Hydra sensor) installed which has a sensor located in an FM pipe just before the CAT. This sensor location is too far back and straight vertical which results in some issues. The Powerdex AFX can use the 3mA pump cell sensor or a Bosch LSU 4.2 sensor, both of which are less expensive. A better install location would be a few inches from the turbo outlet in the downpipe slightly off vertical such that condensate can drain off the sensor rather than collecting on it's tip as happens in a vertical installation.

vtjballeng

iTrader: (1)

You have a variation of the 3mA sensor and this sensor will calibrate on some units and not on others. On all units, it will have an increasing error as your lambda values drop due to variation in the 4mA vs 3mA pump cell curves.

Interesting that GV Autosport and a few other websites with the correct Honda PN (incorrectly applied) stole a description I wrote for the 24302 verbatim before correcting some of the information from ~2008.

You have a variation of the 4mA sensor which the Hydra was designed for and programmed for. Just be sure your crimps are done properly with cable seals in the connector and you have pull tested them as the silver lined teflon wires on these sensors are a little harder to crimp properly:
http://www.bmotorsports.com/2009/12/...dation-charts/

schmoo

Great info, vtjballeng. Thanks very much.

How did you come to know these intimate details?

vtjballeng

iTrader: (1)

There is much more detail than is relevant to this thread, so I just covered a few things. I wanted to clear up some of the questions because the thread was treating the same sensor from different vendors as different parts while others are using a less expensive sensor that can result in significant errors in high power applications.

At my company, I sell and deal with these sensors daily. We also do engineering & design work around these sensor and automotive/industrial/agricultural/military controls work.

concealer404

iTrader: (3)

Sorry to bump this again.... but here we go.

I've got a dead wideband, using Hydra 2.5.

I just want to clarify before i go on my goose chase, that Bosch 15401 IS a correct sensor that will calibrate and read correctly with the Hydra. I guess ergo... that it IS a 4ma sensor.

Thanks in advance!


Also: Anyone have the part number for the Hydra proprietary replacement?

vtjballeng

iTrader: (1)

Bosch 15401 is NOT the correct sensor and is NOT a 4ma sensor.

NTK 24300 or Bosch 13246 is a 4ma sensor per my prior post. There is no NTK/Bosch pn for the Hydra proprietary version as they just take an NTK sensor and put their connector in it. My Hydra came with a 5 way Delphi Metripack connector but the older ones came with Weatherpack connectors.

concealer404

iTrader: (3)

Ah ok... i believe my confusion is cleared now. Much appreciated!

Is there a Hydra part number for their proprietary beast? I don't see anything listed on FM's website for it. :(


Hrmmm.... $205 shipped for NTK 24300. Not too shabby!

John Dollison

Here is everything I've learned while researching the Wideband Oxygen Sensor (WBO2) for the Hydra Nemesis ECU, firmware versions 2.5 - 2.7:

FM sells a replacement oxygen sensor for the Hydra for $250. They sell a cheaper sensor but it's for a different AFR meter and is not compatible with the Hydra. They also sell a sensor with a Hydra "unlock code" for Long Term Trim on the Hydra Nemesis 2.7 firmware for a total cost of $450.

So I called NGK/NTK Tech Support and gave them the following information:

The Flyin' Miata sensor has the following stamped on the body: "LHA:LZA08-H4 NTK JAPAN 6YN6" (the last part is just the manufacturing lot code stamped on the nut). The end of the sensor is a metal can with one hole in the end, and six holes around the circumference. It is a 5-wire sensor with a 6-pin WeatherPack WPT-6 connector which says "2 P.E.D." on the side. The pin-out is:
A = gray (VS reference +)
B = white (IP power +)
C = yellow (heater ground -)
D = black (IP & VS cell ground -)
E = blue (heater power +)
F = no connection
(I looked inside the connector body and there are no jumpers.)


NGK/NTK Tech Support looked it up in their files, and told me:
--------------------------------------------------------------
* This sensor is a NTK stock # 24302 and has a 136ms response time, which is one of the fastest-response units they make.

* The only 'catch' is that in some cases the reseller (Flyin' Miata, in this case) add their own custom connector or jumper, then charge an additional fee for the sensor. This sensor comes from the factory with a connector designed for most Honda / Acura V6 VTEC SOHC engines (3.0 - 3.5L made from 2003-2008).

* This unit is available from most car-parts stores for about $180, but might be special order. (In fact, I called NAPA and they had one in a local warehouse, and got it to me the same day for $150 + tax.)

* 5-wire sensors are designed for maximum accuracy and response times, compared to cheaper 4-wire sensors. They are often labeled "laboratory grade".

* Some 5-wire sensors don't require a free-air calibration because they have some method of auto-calibrating. This sensor is not that fancy; it DOES require a free-air calibration. NTK recommends a minimum of 30-45 minutes of free-air calibration, but longer calibration time is better.

* NTK owns exclusive rights to 5-wire oxygen sensors, but only manufactures about 40% of them, and they contract with Bosch and others to make the rest. (Later I called Bosch tech support and found that their equivalent part number is Bosch 15401.)

* Each sensor model includes a custom ASIC chip to give it proprietary calibration values, so it is not possible to simply swap out a sensor with a sensor from another brand. There are no cross-reference tables like there are for spark plugs.

* Using the wrong sensor can actually fry your ECU. Don't make a very expensive mistake!

* Leaded fuel (some types of racing fuel) will kill the sensor.

* Running rich (like under WOT) can shorten sensor life.

* This particular sensor was designed for GM engines. Although it might last 100,000 miles for that use, two years and 25,000 miles isn't unusual for an aftermarket turbo kit.

* NTK also sells a PowerDex AFX wideband monitoring and datalogging kit for $280, for people who don't have a Hydra ECU (or similar).

* Be cautious about buying sensors online; even if you get exactly the part you expected, warranty service is provided through the reseller, not through NGK/NTK. So you might be out of luck if you have any problems with the sensor.

* Regarding the life of the sensor: NGK/NTK Tech Support said it's hard to say how long these sensors will last; in a OEM application they might last 100,000 miles; but my experience of two years and 25,000 miles seems about right for an aftermarket setup. Several turbo-Miata Hydra owners have told me of similar experiences.

* No NOT use the 24300, it is NOT the same, the internal circuitry is very different.



Also, here are some questions I asked, along with NTK's response:
-----------------------------------------------------------------
Q: What does the "-H4" on the part number mean? I ask because I see that your stock # 24302 crosses to an LZA-08-H6 at this website:
http://www.oxygensensor.net/ntk_o2/24302.php

A: The -H4 not critical; it's a code stamped for OEM applications (in this case Honda and GM) that tell you the harness length and connector style. (So an -H6 or -HE4 or -HA4 suffix should still use the same sensor body.) The exact definition of the -H4 would need to be answered by a OEM O2 engineer back in the office in Michigan. Most people on the aftermarket are not worried about things like this as they are replacing OEM parts with OEM parts.


Q: I've read some discussions in the forums about whether Hydra needs a sensor with a 3mA or 4mA pump cell. What's the difference, and will the wrong one damage my Hydra or just not sense correctly?

A: The current rating of the pumping cell is directly correlated to the stamping number on the sensor. There is no need to worry about the pumping cell in the sensor body, if the stamping number directly matches our number there is no reason to worry about the amperage of the pumping cell.


Q: I thought you said that Bosch and NTK sensors were made differently and wouldn't cross-reference each other, but the OxygenSensor.net website above says they do. What's going on?

A: NTK and other competitors patent 5 wire sensors, this one may have close similarities, or may actually be the same as one of our sensors as we used this variant in our wideband logging system and supplied the Bosch early in the sale, and switched to supplying the NTK. I would be curious to know where oxygensensor.net got their information. We get ours from our engineering department and this is done in the engineering process with many OEM suppliers, this is the most accurate data you will ever get. The bottom line here is that if the stamping number directly crosses then it’s the same sensor body and internal construction. At this point we just need to verify the wiring on the connector.


Q: You said that each sensor model includes a custom ASIC chip to give it proprietary calibration values, so it is not possible to simply swap out a sensor with a sensor from another brand. There are no cross-reference tables like there are for spark plugs. Is that correct?

A: Yes, the Asic chip is preprogrammed with the calibration values.

NGK Spark Plugs ( USA) Inc.
Aftermarket Division
46929 Magellan Drive
Wixom, MI 48393


I found this tidbit on the NTK website's FAQ's: "Most [oxygen] sensors slowly degrade in performance and send a false rich signal to the ECU, running the engine too lean." That would sure match my symptoms of light pinging under moderate- to heavy-acceleration between 2200-2900 RPM.


It turns out that if you do a Google search for "ntk 24302 hydra" you'll find several threads about replacing the FM sensor with this one. They talk about putting a new connector on the cable from the Hydra, so swapping out sensors becomes an easy plug-n-play solution.

Some other possibly useful pages are below, although the information provided is possibly old or inaccurate:
http://boschautoparts.com/FAQs/Pages...Oxygen+Sensors
http://www.mazda-speed.com/forum2/in...c,19513.0.html
http://www.techedge.com.au/vehicle/wbo2/wbntk.htm
http://forum.miata.net/vb/showthread.php?t=239491
(Look for klatinn's post near the bottom.)


I've seen some claims that others have successfully used an $80 sensor; that doesn't mean it's accurate or fast enough for all conditions, but maybe it's good enough for most situations. I don't really know.


John Dollison
'99 Mazda Miata
FM-II turbo kit with Hydra Nemesis ECU ver. 2.6.

vtjballeng

iTrader: (1)

John,

Good work researching these issues. There are some issues here which are a little confusing in my view. I am not going to comment on everything but only choose to comment on a few points.

Nearly all modern "wideband" or linear response UEGO/lambda sensors are double nernst cell sensors. One cell measures pump cell oxygen as compared to exhaust stream oxygen. The measurement cell is kept at .45v within the linear response range of a traditional zirconia sensor element while the pump cell is operated in reverse. The amount of current used and some other corrective parameters determines the measured oxygen concentration level. There are only 2 present variations of "wideband" sensor that comes in a 4 wire configuration and neither is currently considered laboratory grade. One is the Bosch LSM11 variation used on many applications in the late 90's which is essentially a highly tolerant, modified narrowband sensor whose non-linear range is somewhat more forgiving. The second configuration is a newer Denso modified design used on many Toyota's. This design will likely remain OEM only and is unlikely to be used in the aftermarket.

The use of free air calibration vs using the in-connector calibration resistor is purely a design choice and not sensor based. The NTK sensors come with a connector with a calibration resistor which has a little over two dozen trim options based on the "custom asic" determination at the factory for trimming. Bosch designs use laser etched trim or calibration resistors. Most high quality systems require that the user calibrate the sensor on a regular basis in free air against some fixed value to account for sensor drift. Some popular systems and most OEMs simply use the provided calibration resistor and attempt to account for sensor drift in other ways. Once a manufacturer chooses to change the connector due to availability or other reasons, the cal resistor option is lost and a free air calibration method must be used.

Each sensor has a cal resistor. If you have a unit that does a free air cal (like the Hydra), then you can swap to any sensor of the same model. Some older ECUs were locked to one cal value for one sensor model which was problematic. You do not want to mix sensor types or families such that you have varying heater circuits or basic pump cell designs.

No. The variation in pump cell design between the Bosch and NTK models is insufficient to cause outright damage (short of poor design practice). The variation in heater circuits is enough to overheat say a 4.9 driver trying to drive a 4.2 sensor but that is somewhat unique to Bosch designs. Any good designer should account for the possible variations as some installation locations will cause the heater circuit to be driven full bore over its entire life (rendering poor results) where other locations might see a 15% lifetime duty cycle door to overheating.

Absolutely true. You can kill a sensor in minutes with a crazy rich race fuel. Or with a good tune and unleaded fuels it can last for 50k miles. Most tuned cars will be rich and have tuning problems and as such a standard lifetime for such a sensor tends to be in the 10k-25k range. Most aftermarket systems have poor testing for sensor failure because they cannot control fueling parameters and test for feedback validation. Aftermarket ECUs could choose to do this but almost none do.

The NGK Powerdex AFX has the same performance as some OEM calibration systems. We use this system for any user wanting a solution below $1500 (we sell it for a lot less than $280). I have one hooked up to my Hydra as an external AFR input.

These sensors are not supposed to be warrantied once used because, as mentioned above, you can kill one in 5 minutes with little visible damage in a poor-running environment. Many customers are able to talk their way around this manufacturer dictated policy with some vendors. It largely depends on the vendor's knowledge and/or how much the person at the counter knows.

This is where I will take a stance and largely why I responded. 24300 = 4mA pump cell. 24302 = 3mA pump cell. The service rep is correct, replace what you have with the same sensor because mixing them will cause issues. I take the same view as the service rep for a number of reasons and this is what I have been saying on this forum and recommending. Prior Hydra systems have been coming with the 24300 sensor (not the LZA08). My Hydra purchased earlier this year from FM came with a 24300 sensor with a 5 way Delphi Metripack connector.

So now the question is, where did your LZA08 sensor come from and how did it make it in the mix here? When did you purchase this sensor? Is Hydra now shipping with the 24302 sensor indicating a hardware and/or firmware change? Is FM mixing stock, thinking they can modify the 24302 sensors to work on 24300 applications? Has anyone else received a new Hydra with a sensor stamped with LZA08? Most systems using the 24300 cannot also drive the 24302 where many systems using the 24302 can also drive the 24300.

VanMSM

I have a hydra 2.5 that I bought many years ago and the WBO2 that came with it is the NTK LZA08-H4 which based upon the above is the 24302. Clearly my Hydra was purchased before vtjballeng's (which it is indicated was purchased earlier this year) thus maybe the Hydra 2.6/2.7 are using a different WBO2? Or perhaps Hydra and/or FM actually are mixing up stock?

This for sale thread is also interesting, since it is a Hydra 2.6 WBO2 from FM that doesn't even have the proper connector: https://www.miataturbo.net/miata-parts-sale-trade-5/hydra-2-6-wb02-sensor-brand-new-61919/ That is odd.

VanMSM

This is probably unrelated but this guy had his WBO2 die after upgrading from 2.5 to 2.6: https://www.miataturbo.net/showpost....79&postcount=2

It was claimed to be a defective part in the 2.6 upgrade. But in the interest of creating a conspiracy out of nothing, maybe they did change their WBO2 circuitry and thus the WBO2 that is supported?

vtjballeng

iTrader: (1)

Your LZA08 is a variation of the 24302 sensor family which operates at the lower pump cell current. The sale thread sensor uses a 5 way Delphi Metripack connector which is actually the same connector mine came with. Previously, Hydra used a Weatherpack connector. Neither are original to the NTK sensor which is an 8 way gray connector with cal resistor in the front lock mechanism. I don't know why Hydra changed connectors or why there isn't a significant distinction but the Metripack series is significantly newer. Think of weatherpack as a 70s design, metripack as an 80s design and there are newer families from the 90s like the GT series.