Knock sensors for elite
 

What is the best knock sensor to work with a haltech?

Are they usefull if you already have wideband O2 and EGT?

i'd start with an Rx8 sensor, as its modern, and uses an easy to get connector.

actually identifying knock can be a bit tricky, but its totally worth it.

check out https://www.rx7club.com/general-rota...ensor-1026792/

thanks. good read in that link.

what about the knock sensors on this site http://www.lms-efi.com/sunshop/index...t_detail&p=208

they are broadband and cheaper than rx8 sensors.

I have used the sensors we sell with good results. The reason I don't like using the OE FC, FD, or Renesis sensors is because they have bandpass filtering built into them. We know the frequency at which rotary knock occurs but we don't know, at least I don't know, the frequency that Mazda chose to.monitor with their sensors. Many times a 2nd or 3rd order frequency will be monitored instead of the primary frequency. So if you set your ECU up to monitor the primary frequency but the sensor has built in filtering centered around the 2nd or 3rd order frequency, what do you get?

BTW, the sensor shown on our site has an 8mm bore. The bolt hole on the rotor housings and center iron for the OE sensor is 10mm. We had been using a helicoil threaded into the housing or iron to secure the sensors and that works great. What we've found, through Andre at High Performance Academy, is that the sensors can be bored to 10mm. We're offering that as a service, though it hasn't been added to the website yet. It will be a $10 per sensor option. The result is a direct fit sensor.

good point, Mazda has the luxury of setting the whole thing up on the dyno, so they know they can filter some stuff out.

we're not usually so lucky in private life, and then its better to collect more data. if you have too much signal it can be filtered, but if you don't have enough, there isn't much you can do at the software end

With the Elite, there is a lot you can do. Filtering is adjustable and the spectrograph allows you to see what frequencies are actually in play. For instance, you can look at 2nd and 3rd order frequencies to see if you have a stronger signal to noise ratio at those levels.

thanks Chris. Good info.

Ordered

As far as fasteners, how about a step sized stud like this

https://www.germansupply.com/home/cu...roductid=18135

which those are M10x1.5 and I figure the knock sensor holes are M10x1.25.

somebody's gotta have them though.

just cut the head off of two bolts and tig them together. viola
also, im eyeballing an e1280s, which elite would be comparable to that, mr ludwig?

Much easier to just buy a sensor from us that just fits. ;)

The Elite 2500 would be comparable in functionality to the 1280. However, in terms of ease of use and the end user actually being able to extract the most from what is given to them in the box full of circuits, the Haltech is in a different league compared to the 1280. I'd be leery of any product when the distributor comes out in public and says, "The E1280S I believe will only be sold on a case by case basis. Basically sold to someone who understands how to use the ECU, as it is overwhelming to most people." Since the Elite was introduced we've converted quite a number of Adaptronic users and the feedback has been exceptional.

Combine that with the fact that we're literally down the road from from you, have an outstanding reputation for supporting what we sell, and have won multiple national championships with the products we offer and it seems like a no brainier.

:icon_tup::nod:


Also agree on the haltech. I'm a bit of a fan boy but I have played with aem, megasquirt and a few others on friends piston cars and they don't compare to the haltech in function and ease of use. I have installed and tuned a e8 and now a elite 2500 my self and I have no tuning background, I'm also not a mechanic I'm just a smart guy with a wrench.

I using a 2500 on a REW using two knock sensors. I see knock signals for both sensors but only a knock level for the sensor 1. Any suggestions? Is the knock level is the processed output of the knock signal? If so that may be the issue as one sensor is a stock REW sensor and the other is one from the J&S system I was using prior to getting the Haltech.

I did eventually gets both the J&S sensor and the stock REW sensor to provide knock readings. But the the difference between knock signal values was often 2.3X different and so I concluded the outputs between the sensors must have been on different scales. I did not find a way to setup different scales for two knock sensors and as I thought about it I figured it would be an unwise thing for an ECU manufacturer to support. I tried using two REW sensors and while this worked reasonably well I was concerned about using 30 year old knock sensors and trusting their ability to create a knock signal that I could map properly to the Haltech 2500. So I purchased / installed some new Bosch 0261231006 knock sensors as these were the type many Haltech users seem to have had success and they were reasonably priced. These have been working well for nearly a year now.

Is there a way to differentiate false knock vs real?

Car was on the dyno yesterday. Once we started going up to 6500rpm we started getting knock. Below that the car was making great power. Not sure what to check. Going to swap out the bosch sensor for a new one and do a data log.

This podcast from HPA, https://www.hpacademy.com/blog/046-g...oGRSMUGVmj_B_g, between the 10 and 20 minute mark brings up some good points on knock.

My views are similar and built off these layered ideas, running engines make noise including noises related to igniting of fuel, when looking for noise related to knock one must consider how to ignore mechanical noise and possibly some noise caused by igniting fuel, and not all noise to a given level is necessarily knock. But some some noise level must be chosen to be considered knock to create a knock detection strategy.

When a simple strategy is acceptable then when any noise level declared to be knock is detected then some type of counter-measures are invoked. In more complex strategies the definition of the knock level may have a 'grey area', such as 'almost knock', 'minimal knock', 'severe knock', as well situational application of this grey area based on things such as rpm, map, emap, temps, etc. In these complex strategies the counter-measure strategies would also likely vary due to the situation.

In the rotary world we a challenged with the concept that "no real knock can be tolerated". Even those of us that can gain access to tools such as knock ears find it difficult to define a good real world definition of the bottom edge of the grey area of knock, as intentional knock causing situations are to be avoided. Thus tuning a knock detection technique is often very conservatively approached.

so I wasn't sure if I was picking up false knock or not. So we took the car home and I went over everything. Probably should have done that before I even started driving the car.

I found that the previous shop only had 1 bolt holding the power steering pump / ac bracket to the keg. added the missing nuts and bolts to bracket. I also went ahead and replaced the knock sensor with a new one. Re-torqued it to 20nM.

Took a test drive and 3rd gear pull had no issues. Will be sending a data log to the tuner and schedule another dyno session for later this month. Hopefully that bracket was casuing the noise. Or maybe I just had a defective sensor.

Mike.