I have the LM-1 with the Bosch sensor. I'm using the tailpipe probe that I bought with it. Works great!

 

I'm glad that it's working for somebody with an RX7! You must have a midpipe. Unfortunatley, I have a HF Cat and I'd like to keep it.

 

Yeah, I have a resonated MP. I was worried after reading your post about your problems, I hurried to test out my LM-1 the day I got it. :-)

 

GREAT !

That makes me happy cause I too ordered the tail pipe probe (I always like testing this way, even on cat cars) just to save the sensor and I find it easier to fit on the cars I have and the ones I test.

Fantastic news, I was scared for a minute that it might throw an error with the tail pipe sensor on a rotary. Glad to see that is not the case in practice

 

http://home.comcast.net/~mstrobl2/co...eatshield2.jpg

Now here is an idea that should work.

I think its an issue with the body of the sensor running to hot, shielding worked for this bloke.

I remember when I was at Tech Edge and Peter was heating a 075 LSU with a bare flame of a lighter and it did not throw an error ? Just an idea, worth a try for sure as it fixed the problem in one example.

 

can the sensor be mounted in the cat back section and still function properly(with mid-pipe)??......

 

Hi there,

I am the designer of the LM-1, so I think I can make some qualified comments about the Bosch sensor and the problems some people see on rotary engines.

The issue with the Bosch Sensor is that it requires a relatively low body temperature (max. 500C) at the sensor hexagon where it meets the bung.

Heat conduction from the exhaust pipe through the bung to the sensor body causes overheating and indicated AFR errors, in which case the LM-1 throws an error rather than giving erronious results.
Bosch specs say that if the sensor gets too hot, use a longer bung, which minimizes conduction.

The heat-shield mentioned in a previous post by RICE RACING does not as much work as a shield, but as a heat-sink. It cools the sensor housing by dissipating some of the heat conducted through the bung. A copper sheet of 4" x 4" x 0.060" or thicker, shaped like a V with a flat bottom the width of the sensor and the sides bent up 45 deg. should work even better. Specially if mounted such that it has good airflow over the 'vanes'. Mount the sheet between the sensor hexagon and the sealing washer to further minimize heat conduction from the bung to the sensor and to maximize conduction between the sensor's hexagon and the copper sheet.

Don't worry that the sensor is too much out of the exhaust stream. As long as the very tip of the sensor is in moving gas, it is fine. The sensor tip shape is designed so that the gas-flow will create a slight vacuum at the little hole at the tip so that exhaust gas will be sucked through the sensor.

In tests we heated the sensor front (the part exposed to exhaust gas) to cherry red (and accidentally to yellow hot) for quite a while without getting any errors. The heater circuit regulated all the way down though. Heating was done with a propane torch. Heating the sensor body, even while being careful not to heat the front, causes errors very quickly.

The Bosch sensor temperature is typically controlled via the resistance of the sense-cell, which gets lower with cell temperature. Bosch sensors also have a parasitic resistance in the common current path in the connection of pump-cell and sense-cell. This resistance increases with the sensor HOUSING temperature. All a typical WB controller can see is the sum of both resistances. At some point in the temperature situation the parasitic resistance is so large that it becomes dominant and the sensor can no longer be accurately temperature controlled. If that situation is not accounted for, the sensor may give erronious readings without the controller knowing it. The LM-1 measures and compensates for that effect up to a point, after which it throws an error.

The worst cases of overheating we have seen are combinations of supercharger, large stainless pipes/headers, ceramic coating and mounting the sensor in the stock NBO2 location. These are all things that maximize exhaust heat (and back pressure).
In one case I asked the customer just to take the sensor out and run the engine. Flames where shooting out of the bung hole. The sensor was expected to run bathed in flames?
Make no mistake, overheaters account only for a very small fraction of LM-1 units sold and LM-1's are doing fine in alcohol-burning race cars and dragsters with great success.


I hope this info helps those that experience overheating problems with the Bosch sensor.




Regards,
Klaus

 

Bloody great responce Klaus !

I will be using my mech eng thermodynamic skills and do some tests when I get my unit, heat sink and radiant heat deflection are the two big factors that will eliminate any potential error issues caused by "excess body temperatures" while still retaining hte designed accuracy.

Just want to say I am very impressed with the level of support on this product. Very Very Impressed in this day and age of fast sales and poor post sales support

I will fabricate up something up in Aluminium and post some pictures after I test it, I know copper would be better but I have heaps of Al and a TIG so use what you got

 

Thanks for replying Klaus! I'd also like to say that the customer/technical support that I have experienced first-hand from the Innovate Motor Sports team has been excellent!

I can't wait to get mine up and running. I need some tuning bad...

 

Whoa, whoa, whoa. I just caught this response, and all I can say is that you misunderstood me! I'm a huge fan of the TechEdge wideband! I only used it as the example because it's sort of the "default" low-cost wideband setup. We've used the TE many times with excellent results.

Brandon
BR7 Racing

 

will placing the sensor in the midpipe affect the "quickness" of the readings?in other words,will the display readings lag behind the engine's true ratio.i have the sensor mounted in the midpipe(3")without any problems.

also,i dont have any experience with other units but i'm really happy with this one.showing what the outputs do in the manual lets you build your own cable for datalogging,
etc.without having to buy theirs.also having programmable outputs(2)makes setting up easy.i already hooked up an lcd for readings.awesome.

 

Hi platinum,

I did the calculations for that a while ago for the following assumptions:

a. N/A engine of 2 liter displacement .
b. VE of 1.0.
c. RPM of 2000.
d. Exhaust pipe diameter of 3".
e. Sensor mounted 6 ft from engine exhaust ports.
f. EGT of 800 deg C.

The time it takes for exhaust gas to travel to the sensor is 63 msec. The formula used is simplyfied, assuming the intake air is just heated to EGT and does not take into account the added gas volume provided by the fuel itself. This added volume will shorten the time. It also assumes no back-pressure.

The formula is:

t (in sec) = (5.89934 * d * d * pi * L *IAT) /
(Vd * VE * rpm * EGT)

where

d = Exhaust pipe (inside diameter in inches)
L = length from engine to sensor in feet
IAT = Intake air Temp in deg. Kelvin
Vd = engine displacement in liter
VE = volumetric efficency
EGT = exhaust gas temperature in deg. Kelvin

regards,
Klaus

 

any word on a group purchase??...

 

Anymore updates on people's experiences with this wideband setup?

I'm thinking of buying one soon.

 

PM "Rice Racing",

He may have figured out a way to make the LM-1 Work with the FD. I was not able to get it to work with my car for more that 15 minutes at a time before getting sensor timing errors due to high EGT's

John

 

Did you try to weld the bung in down stream in the exhaust or do the heat sink thing they suggested?
I saw you were going to try the ntk sensor that failed also?

Did you return it or get stuck with it?

 

Yup, tried it all. Nothing worked. Maybe my heat sink/sheild was poorly constructed... IMS refunded my money w/ no questions asked. They gave me incredible customer service. I would buy another uniot later if they figure out a solution that would work on my car.

Tried it out. It may work for you.

 

I have had 3 LM-1 units.

1st one worked about 2min and never came back on.
2nd one the replacement for the 1st one worked for 5 min and then the display went wacked unreadable.
it also say sensor error sensor timming.
3rd one works for about 30 - 45 sec and get sensor timming errors.


I ask the Innovate guys if we can return them they said yes but we would be black balled dealer and would never be able to be a dealer for them if we returned them. so basicly we are out for 2 units or never to be dealer again with them.

their customer support was great. I have no problem with that. but these items are useless to turbo rotary tuners.

and Im not about to spend the time to build some tail pipe probe and heatsink to try to make a work around.
also measuring the a/f at the tail pipe is not where you want to measure it. that is way down stream. to me its like measuring your egt at your muffler.

Im just frustrated on what to do with them eat the cost. bend over and smile. or send them back

 

Where did you have it mounted? How far downstream?

I want to get one, but this is the second post I heard bad about them on turbo rotaries

 

Here's my .02
The Innovate unit was prematurely advertised and sold as a reliable and complete unit without appropriate testing. In the long run, it may be made reliable, but that may be a long way off. If you want to stay on with them as a dealer bend over. If you really want a state of the art and almost daily upgradeable product that benefits from a world wide support and development team of users and programmers, try the TE v 2.0. They have addressed the issues of high temp breakup of the LSU4 and are also working on a firmware falsh that would allow interchangeability between the NTK and LSU sensors without needing two different controllers.
I have had three DIY units from TE (1.0, 1.5 and 2.0) and have been thrilled with each. My 1.0 is permanently installed in my FD and I have over 30K miles on the NTK sensor which was bought used from a Civic VX that had 60K miles on it. The sensor and controller when coupled with my datatlogit still track the AFRs as between 0 and +.2 leaner than a MOTEC. My TE unit NEVER reports a AFR as richer than a MOTEC. I use the 2.0 to test and log other people's cars (rotary and piston) and it and the 1.0 are in perfect sync.

That's my recomendation and experience. Good luck. PM me if you want more information about the TE.

 

Thanks for the great info Jeff48, I will most likely buy a Techedge.

 

Everyone has their aliances I guess ?

Measuring AF at the tail pipe is no problem (see my last logs with TPS recorded). No amount of software will over come an over heated sensor, it will always be inacurate.

I after seeing the results would fabricate a heat shield/sink for ANY WB system (as is done in some road racing cars), obviously done for a reason and after my experience with the LSU4.2 I can see why.

Sure its a hassel to make heat shields/sinks (TP probes are not necessary) but its also proven to give far more accurate Lambda/AFR readings when the sensor is within Boschs specified range of operating temperature. If you think Peter G or Adam from TE have even measured such a thing you are dreaming I saw some of the analytical tests he did on my Autronic meter v's his TE latest unit. A butane cigarette lighter is not the most scientific controled test I have ever seen But if your unit measures to within 0.2 of a PLM then thats all that matters (assuming the PLM is correct) which after my discussions with a particular Motec distributor here in Australia varies greatly on exactly where you place the sensor (read "heat effected"). They recommend you place the sensor underneath the drivers door to get consistent repeatable readings from the PLM. The bloke I spoke to does allot of tuning of race cars and that was his strong recommendation.

There is NOTHING wrong with the LM-1, there is plenty wrong with the AFR/Lambda readings if your LSU sensor is heat soaked beyond Bosch's factory specifications (same happens for all sensors). I even through all my frustrations would rather have a unit thats throws an error or has a hunting display rather than be inacurate and inconsistent. And no matter if your talking Motec or M&W or TE you simply will not know if its up to 1.0 AFR out because of an overheated sensor. I will tell you right now that non of these companies have tested their products for accuracy at elevated sensor temperatures. Least the LM-1 will error out rather than GUESS the actual AFR

 

87GTR

I have tuned 2 turbo rotaries (using my sensor mounted in the DP 5" from the turbine exit) with the heat shield and AL heat sink, no errors, no glitches, accurate readings............... from idle to WOT for as long as you like

Again, having to fabricate extra **** is FRUSTRATING. I agree fully !

Knowing that the readings will be accurate because the sensor will remain within Bosch's intended operating parameters (see my posts in the LM-1 thread, giving detail) is reason enough to make these things.

Dont know what else to say ? My intention is not to sell any unit (please dont think that !) I can understand that what you are saying is that the LM-1 is useless the way it arrives in the box. But for very little effort you can make it a gem, and rest assured that it is going to be spot on accurate and repeatable no matter what you throw at it with a bit of re-engineering and close attention to the sensor operating conditions.

 

Rice,
Sorry for the confusion but, actually the TE v2.0 will "error out" when the LSU4 is demonstrating a too high temp (anything above 850C.). The software fix compensates for a too sensitive temp threshold built into the original TE v2.0 which resulted in the 2.0 reporting false errors. Peter and Adam have always recommended placement of the sensor as far as possible from the turbo outlet, yet still in front of the cat. Shields, heat sinks, extra tall bungs, etc etc may help, but there is still nothing as dependable as placing the sensor where recommended.

The rest of my post addressed 87GTR's experiences, only a couple of which can be directly attributable to an overheated sensor. In any case where my LSU4 overheated (I always try it in a stock position 1st), the senor returned to usable after cooling down. (Yes it is still accurate as tested against another LSU4 and the NTK v1.0 setup).

I am not selling anything either. I don't ewven know Peter other than to say, I am impressed by the way he handles customer inquiries and provides one forum (the TE yahoo group) and supports 3rd pary software development. My experience with them and hearing of others problems with Innovate, were the bases for my comments.

I wish Innovate nothing but good luck, we need more, not fewer companies turning out affordable quality diy7 tuning options.

 

Hi Jeff

All the companies should be encouraged, they have done an amazing thing in destroyin ght e hype or "myth" of WB units. What was normal prices years ago is looked upon as a rip off now days !

M&W units sell for as little as $999 Aus in LSU sensor config where as 2 years ago they were double that amount !

PLM's (motec) sell for around $1640 Aus again a fair bit cheaper than a few years back. Some other big brands are still amazing amounts but they will be forced to lower their price especially when they are no where near as fast or acurate or have anywhere near the features of what is offered by TE and Innovate.

Again I repeat, heat and its effect on sensor body (hexagon and up) temperatures is the KEY point here, I will tell you right now not many people have tested these effects in detail (I have been forced to). It is rare indeed for the EGA to exceed 850 deg C in any location especially under steady state conditions at mild loads. What is REAL EASY is for allot of heat transfer to conduct what ever heat is in the sensor tip and exhaust piping up into the sensor body (electrics region "measurement cell etc") When you have NILL AIR FLOW or LOW COOLING AIR FLOW this situation compounds as the radiated heat from surrounding components (turbo !!!) realy procludes any cooling of the sensor body, hence constant errors or innacurate displays of AFR/Lambda. Shure the software can be written in a way that the increased resistance experienced in an over heated body will not throw an error but there is no way it will retain its accuracy. This is why the Motec dealer I made reference too suggested to move the sensor to under the drivers door (for good cooling = accurate repeatable readings).

My whole point here is that you can have your cake and eat it too with a bit of effort. With a good heat shield and a remote heat sink (like pictured in my LM-1 thread) you can place the sensor ANYWHERE and if available the closer to the merged collector of all exhaust pulses. In a turbo right after the turbine exit ( I stuck mine around 5" to 6" away) It works perfectly

With a "re engineered" sensor I have been able to heat the tip up to and well beyond ANY temperature it will ever see in the DP ( a level above bright red, almost yellow white) and it did not throw an error, after the test I put the same sensor back in the test vehicle and ran my std test procedure and the sensor is dead accurate.

The issue is heat rejection, if you can get rid of even small (500deg C) generated at the sensor tip through the exhaust (through a heavily heat soaked engine bay or low cooling air flow over the sensor body) then it will be innacurate all the time or throw errors. If on the other hand you shield the body from radiated heat source and then pull heat away from the body with a heat sink even with NO cooling air flow (or very little) I challenge anybody to get an error with the LM-1 or TE units