MS3X Knock Module
Turbo Mini Forums

Tempted by the 10% discount on offer at DIY Autotune this weekend, I clicked the buy button on the MS3X Knock Module to go with some RTCs and other bits and pieces.

I'm planning on running high boost and a 9.5:1 compression ratio with water injection so I thought this would be a good idea.

Just reading up on what it can do has got me thinking. It has two sensor inputs, you can assign each sensor to a particular cylinder. You set the start and duration of the "knock window" for each cylinder.

As the sensors are not too expensive and the 5 Port is effectively 2 different 2-cylinder engines in terms of VE and cylinder pressures, it would make sense to use the capabilities.

I have read that you guys have tried using the alternator bracket and the end steady mount holes with varying degrees of success, with a preference for the end steady mount location.

So, what has been the particular issues with the alternator bracket location ???

Saul Bellow - "A great deal of intelligence can be invested in ignorance when the need for illusion is deep."
Stephen Hawking - "The greatest enemy of knowledge is not ignorance, it is the illusion of knowledge."

Just managed to procure a pair of Bosch sensor off eBay for £4.99 :)

Seriously thinking of drilling and tapping the block close to the central core plug hole to pick up knock from cyls 2 &3.

Edited by Paul S on 1st Dec, 2013.

Saul Bellow - "A great deal of intelligence can be invested in ignorance when the need for illusion is deep."
Stephen Hawking - "The greatest enemy of knowledge is not ignorance, it is the illusion of knowledge."

Co-incidentally I was reading up on this last week (probably because of the recent group buy thread), mostly just searching for information on the MS-Extra forums.
Although I was looking for a DIY solution (there are several designs on the MS-Extra forums) I obviously came across the DIYAutotune item - a good price but unfortunately (for me) MS3 specific. Also I couldn't really find any info on how it can assign the two sensors to different cylinders - it's probably obvious to anyone who has an MS3 and hence seen it's configuration in TS, but I expected a bit more info from their website.

In one of the very long threads on the forum, there was a German member who said he was getting really good results using the alternator bracket (on an A-series obviously but I don't think Turbo) but if you want to treat it as a pair of eneven two cylinder engines, shouldn't you be aiming to put one as close as possible to an outer cylinder and the other as close as possible to an inner (subject to all the normal criteria of keeping them away from mechanical noise) ?
Even if the MS3 assigns the results per cylinder by engine timing it would seem sensible to have them as equidistant from their assigned cylinder if only to generate the same amplitude signals.

Schrödinger's cat - so which one am I ???

Saul Bellow - "A great deal of intelligence can be invested in ignorance when the need for illusion is deep."
Stephen Hawking - "The greatest enemy of knowledge is not ignorance, it is the illusion of knowledge."

The first bit is what I expected, I just couldn't find it written up anywhere. That explains that module being MS3 only.

"The sound travels through the casting not the water ???? "
Well technically sound travels through both quite well but at very different speeds. I would assume all these systems use the metal path though, faster and probably with less attenuation.
Cast iron is an interesting question - in my field of engineering (ISI/NDT) I've been involved a lot with ultrasonic testing and as a general rule, castings are worse than forgings/extrusions for attenuating ultrasound, especially stainless steeel castings (cast stainless has really large grain sizes and the grain boundaries scatter the sound). But no experience with cast iron

If you're going for what theoretically should be the best possible response, I would consider drilling/tapping the front of the block on the centrelines of 3 and 4 just where those casting numbers are in your photo as high up as possible. That should be drilling into the deck so missing the waterways and give the most direct sound path to the combustion chamber other than the head deck itself. Mill a spotface before drilling then space them out a couple of mm on thick washers so they don't foul the edge of the head.

Just my thoughts.

Schrödinger's cat - so which one am I ???

sound will travel through everything. maybe with a differend speed.

you can even find some older cars with the knock sensor at the intake manifold

OK, I've been taken too literally.

What I meant was that the frequency and amplitude of the noise and vibration from detonation would be severely dampened by the water.

A spotface on the block at the head flange area would be ideal. However, the engine is all bolted up and ready to go into the car. The above picture was just for illustration.

However, the head is not bolted down yet :)

Saul Bellow - "A great deal of intelligence can be invested in ignorance when the need for illusion is deep."
Stephen Hawking - "The greatest enemy of knowledge is not ignorance, it is the illusion of knowledge."

i personally don't think water will dampen enough to make any difference.
the speed of the sound may change, as is i don't know if the speed of sound in water is that much differend than that in cast iron. and even then i don't think it make a big difference as it isn't a very thick water jacket.

Schrödinger's cat - so which one am I ???

My simple minds just thinks: Hit metal with a hammer and it rings.... Hit water with a hammer and you'll probably lose the hammer :)

Right, as the sensor is assigned to a cylinder, then there is no real need to group the inner and outer cylinders with a specific sensor. The software will deal with all that.

Given the potential for valve train noise from cyl. 2 to interfere with knock noise from cyl. 3, then it's probably better not to do it that way.

So, I'm now thinking a sensor between 3 & 4 and a sensor between 1 & 2. On the head at the base of the stud boss. I can drill there without fear of hitting the water jacket and there is a more-or-less direct metal path to the combustion chamber.

Saul Bellow - "A great deal of intelligence can be invested in ignorance when the need for illusion is deep."
Stephen Hawking - "The greatest enemy of knowledge is not ignorance, it is the illusion of knowledge."

my 2pence worth not that it matters much, but do you really need two sensors? I mean, an engine block has a harmonic. That is to say, if you hit the block with a hammer, it will 'ring out'. so why do you need two sensors for a 4 cylinder inline engine? The firmware code will be able to determine which cylinder has fired, and the one sensor will pick up the noise as the other would.

Bolting the sensor to as close to the deck of the block would be a better location since detonation will likely occur around 14 degrees ATDC (right at the top of the block). Its a solid slab of iron across the top.

Obviously there can be unwanted noise from other parts mounted to the block, and I would have thought that the engine steady bracket can become as noisey as the alternator if its not quite tight enough, more so if its spherical jointed.

Test the theory by connecting the knock sensor to the sound car of your PC or scope and monitor the frequency when you hit the block at the opposite end. then hit the block again on the same end as the sensor is mounted. Will there be a significnt difference that one sensor will not pick up both?

No, I don't need 2 sensors, but for the sake of a bit of wire and a connector I can have 2 :)

There a lot about the build that I don't "Need" like the wedged crank, the lightened rods, the coils-near-plus etc etc.

The Aggregation of Marginal Gains - Sir David Brailsford - 2008.

Saul Bellow - "A great deal of intelligence can be invested in ignorance when the need for illusion is deep."
Stephen Hawking - "The greatest enemy of knowledge is not ignorance, it is the illusion of knowledge."

Schrödinger's cat - so which one am I ???

Paul, what size (ie, bore diameter) engine are you considering.

I'm just looking at suitable OpAmp circuits so I can scope this.

The test obviously won't exactly match the actual frequency but I should be aiming for 7.85KHz by the "normal" maths on my 73mm bore.

For the test, rather than a hammer, I'm thinking of my percussion (SP?) centre punch against the block deck - it should produce the same impact energy every time.

Schrödinger's cat - so which one am I ???

The bore of the latest engine is 65.1mm.

Is this the bandpass frequency? I can see settings for this, up to around 20kHz, but I've got no idea what it means. Care to explain?

The other thing that currently concerns me is electrical noise. I now got the sensors and can start milling the head, but the sensor footprint is quite large, about 28mm diameter and with the connector off the side. Best orientation would be with the connector vertical above.

However, I've got the LS coils in the general area and I'm worried about interference. I'm going to use screened mic cable, but will that be enough?

Saul Bellow - "A great deal of intelligence can be invested in ignorance when the need for illusion is deep."
Stephen Hawking - "The greatest enemy of knowledge is not ignorance, it is the illusion of knowledge."

Edited by Rod S on 4th Dec, 2013.

Schrödinger's cat - so which one am I ???

This is the MPi block. The coil pack bracket mounting holes. The top one is almost perfectly inline with the cylinder 4 and couldn't be any closer to the top of the block than it is, we, no actually lol it can as can be seen in the other pic of a Miglia engine LMAO. "How much do you want to skim off the block ?!"......

Attachments:

46.79kb

147.33kb

Rod, setup screens from TS may help:









I see a lot of "%" settings. Percentage of what I ask myself?

Saul Bellow - "A great deal of intelligence can be invested in ignorance when the need for illusion is deep."
Stephen Hawking - "The greatest enemy of knowledge is not ignorance, it is the illusion of knowledge."

Schrödinger's cat - so which one am I ???

I've emailed you a full project folder. Hope that works.

BTW at 20 degrees ATDC (end of window) the piston is only about 2mm down the bore.

Saul Bellow - "A great deal of intelligence can be invested in ignorance when the need for illusion is deep."
Stephen Hawking - "The greatest enemy of knowledge is not ignorance, it is the illusion of knowledge."

Got it and loaded it.

I suspect the end of the window is a lot more than 20 ADTC (esp. on a turbo) but I guess there's not too much difference between block deck and head deck in terms of sound from the combustion chamber, they are both pretty solid (well apart from that picture Colin posted....).

Schrödinger's cat - so which one am I ???

Nearly a year has passed and I'm back on this.

Sensors fitted to head:



I've only milled off the minimum (~2mm at the flange, ~1mm on the stud post) and achieved about 80% contact area with the sensor. Let's see if that is enough before milling off any more.

Now I can get the engine in the car

Edited by Paul S on 30th Oct, 2014.

Saul Bellow - "A great deal of intelligence can be invested in ignorance when the need for illusion is deep."
Stephen Hawking - "The greatest enemy of knowledge is not ignorance, it is the illusion of knowledge."

isnt that the same stuff some 2 temp race bike's use?

you can do anything if you set your mind to it...
i rather blow it up winning than keep it together losing.

finish date set for project 1... march 2018

I'm now glad I've done it this way as the MS developers are working on knock retard per cylinder. So the ECU will only back off the timing on the cylinder suffering knock :)

Saul Bellow - "A great deal of intelligence can be invested in ignorance when the need for illusion is deep."
Stephen Hawking - "The greatest enemy of knowledge is not ignorance, it is the illusion of knowledge."

Do you suspect that there is significant or intrinsic knock sensitivity differences between cylinder? I understand the desire to adjust timing only on affected cylinder but I was wondering if were expecting something. I guess you could use timing trim if you do see some pattern.

http://www.jbperf.com/