Board logo

Megasquirt repair
clairetoo - 21/10/12 at 12:42 PM

Any recommendations on who to send my faulty megasquirt unit to , and to have it down-graded from Ms2 to Ms1 ?
Rather than muck about trying to map two identical engine setups , I think it would pay me to have both ECU's the same spec - I know Ms2 is better , but where is it better ? I cant find a need for any of the extra features (what ever they are......)


snapper - 21/10/12 at 01:16 PM

ExtraEFI, he wrote the extra code and can repair and advise on spec


coyoteboy - 21/10/12 at 03:18 PM

It's faster with higher resolution on spark and fuel tables IIRC, meaning more accurate calculations and better fueling control IF your car can respond to that (bike engines etc are generally deal with very fast transient fueling situations so tend to like the faster response time). If you have a fairly slow large car engine you won't notice it. If you're not using many special features you won't see much benefit from the MS2 IMO, I run a fairly highly strung 2 litre turbo from an MS1 using the high-res extra code and it's perfectly OK feature/response wise. I've been intending to run my audi v8 from an MS1 too, though Ive recently been considreing the MS2 for futureproofing, but I do like tinkering with electronics

ExtraEFI chappy is probably the best known and most knowledgeable chappy in the UK I think, if anyone can sort you out...

[Edited on 21/10/12 by coyoteboy]


BaileyPerformance - 23/10/12 at 06:20 PM

quote:
Originally posted by clairetoo
Any recommendations on who to send my faulty megasquirt unit to , and to have it down-graded from Ms2 to Ms1 ?
Rather than muck about trying to map two identical engine setups , I think it would pay me to have both ECU's the same spec - I know Ms2 is better , but where is it better ? I cant find a need for any of the extra features (what ever they are......)



I'll do it for free as long as i can keep your MS2 chip!!!

Cheers Dale.


coyoteboy - 24/10/12 at 07:30 PM

From the previous thread on it I think there was some question over the IC itself, rightly or wrongly.


BaileyPerformance - 24/10/12 at 07:55 PM

quote:
Originally posted by coyoteboy
From the previous thread on it I think there was some question over the IC itself, rightly or wrongly.



Maybe, but i would be surprised as we have fitted 10's of megasquirts and converted several more from MS1 to MS2.

If you get a bad chip (MS1 or MS2) normally its dead - no comms - or if its behaving weird re-blowing the code can sometimes sort it.

The MS2 chip is £80, i have at least 5 spare MS1 chips so that's why i offered to downgrade Claire's ECU to MS1!

In Claire's case i would be surprised if its an ECU problem, more like wiring, but you never know!!


clairetoo - 24/10/12 at 08:05 PM

It on its way to Phil at Extra EFI - check/repair and downgrade .
I'm convinced its not a wiring problem - I swapped the ECU for another , and the signal noise was gone.......
I still dont have it running right - but thats one problem less .


BaileyPerformance - 24/10/12 at 08:26 PM

quote:
Originally posted by clairetoo
It on its way to Phil at Extra EFI - check/repair and downgrade .
I'm convinced its not a wiring problem - I swapped the ECU for another , and the signal noise was gone.......
I still dont have it running right - but thats one problem less .


MS2 is far more sensitive to TPS and VR noise than MS1 as the processor is alot faster, this is normal and is cured with screened cable and better grounding.

MS2 is more susceptible to poor grounding and noisy power that MS1.

We always use screened cable for the TPS (as well as the VR) and run all sensor ground back to the ECU 36w plug NOT the block, then the ground wires from the ECU to the block. The later wiring diagrams on the net show the sensors GND connected to the ECU.

MS2 is far better than MS1, the biggest advantage been the 100times better fuel resolution.

In other words a functional MS1 installation may not be suitable for an MS2, you need to look at your wiring ;-)

Downgrading from MS2 to MS1 is a massive step backwards, but i can understand your thinking regarding the mapping assuming both your car are EXACTLY the same in every way.


coyoteboy - 24/10/12 at 08:32 PM

100X better fuel resolution makes absolutely no measurable difference at all to the running/tuning of mine If you're finding fueling is THAT sensitive on an MS1 you can run the hires code variant if your hardware doesn't require the additional timer (i.e. if you have high imped. injectors)


BaileyPerformance - 24/10/12 at 08:44 PM

quote:
Originally posted by coyoteboy
100X better fuel resolution makes absolutely no measurable difference at all to the running/tuning of mine If you're finding fueling is THAT sensitive on an MS1 you can run the hires code variant if your hardware doesn't require the additional timer (i.e. if you have high imped. injectors)


It all depends on what injectors you use (or the size of them)we found the high res code to be poor and is no longer supported anyway so therefore should not be used on new installs if it can be avoided.

We are used to mapping high power turbo charged engines with larger injectors - we used 625cc in our 2.0L 350BHP zetecs, with this setup the idle PW is around 1.5ms. So MS2 make a big difference.

You can have a fueling step change on an otherwise rover V8 with MS1, (only on the 3.5/3.9) we find MS2 to be better on the Rover and most other engine we have tuned.

Its also worth noting the closed loop O2 works better on MS2, this is also due to finer steps.


clairetoo - 24/10/12 at 09:10 PM

Thing is - I'm not running a turbo , I'm not running large injectors , and like most people I know nothing about grounding via the Ms to the block........all I want is a running engine , tuning to the Nth degree is just not that important !
What you see as a massive step backwards , if it gets me what I want , is a massive step forwards .

The cars are not exactly identical , but at this point in time I dont care - I just want to drive the damn thing !


BaileyPerformance - 25/10/12 at 04:14 PM

quote:
Originally posted by clairetoo
Thing is - I'm not running a turbo , I'm not running large injectors , and like most people I know nothing about grounding via the Ms to the block........all I want is a running engine , tuning to the Nth degree is just not that important !
What you see as a massive step backwards , if it gets me what I want , is a massive step forwards .

The cars are not exactly identical , but at this point in time I dont care - I just want to drive the damn thing !


You must have some idea of how you grounded the ECU and how its all wired up? assuming you did it yourself?

Anyway, i would be interested to know if the ECU was causing the problem - hope you get it sorted one way or another - as you say its not much use if you cant drive it!


clairetoo - 25/10/12 at 04:33 PM

The earths from the megasquirt are soldered together and earthed to the chassis . The sensor earths are on separate earth studs (I did try adding a jumper and earthing the TPS to the block , but this made no difference)

Latest problem is one non-firing cylinder..........


coyoteboy - 26/10/12 at 12:14 AM

Ideally for any ECU system and associated sensors you want a single ground point taken directly from the battery ground. Distribution blocks are ideal for this. Don't stack connections as they build up contact resistances. When you start grounding to different locations you introduce potential for differences in reference voltages between sections of circuits, often it's not a problem but sometimes it is, and ground faults are a royal PITA to track down.


matt_gsxr - 26/10/12 at 08:30 AM

quote:
Originally posted by coyoteboy
Ideally for any ECU system and associated sensors you want a single ground point taken directly from the battery ground.


MS recommendation on this is to ground everything to the block, not to the battery negative.
I'm not trying to create an argument about this, but if you post a question on the MS forums and have earthed differently to their recommendation then you are likely to get short answers.

http://msextra.com/doc/general/grounding.html


BaileyPerformance - 26/10/12 at 10:41 AM

quote:
Originally posted by matt_gsxr
quote:
Originally posted by coyoteboy
Ideally for any ECU system and associated sensors you want a single ground point taken directly from the battery ground.


MS recommendation on this is to ground everything to the block, not to the battery negative.
I'm not trying to create an argument about this, but if you post a question on the MS forums and have earthed differently to their recommendation then you are likely to get short answers.

http://msextra.com/doc/general/grounding.html


Agreed, that's the new diagram that we should all use. Note the sensors are grounded to the MS 36w connector not the block.
The old diagram showed them to the block.

We use screened cable for the TPS as well as the VR as MS2 is more susceptible to noise (than MS1) on the TPS causing unwanted accel fuel shots.


coyoteboy - 30/10/12 at 12:28 AM

quote:

MS recommendation on this is to ground everything to the block, not to the battery negative. I'm not trying to create an argument about this, but if you post a question on the MS forums and have earthed differently to their recommendation then you are likely to get short answers.



That's one option, personally I don't like using the block/ancills as a distribution point and run separate grounds for every sensor back to the battery neg. Either is acceptable so long as you stick to one so all of your ground refs are to the same point. The issues start when people ground to the block on a "normal" car, which can then often be grounded to the chassis, then back to battery neg, or just back to battery neg, or some other schemes I've seen. Then you start getting HR connections in one of many ground wires and the whole lot becomes a nightmare. Been there [block grounding], done that - won't do it again unless I'm starting from absolute scratch.

[Edited on 30/10/12 by coyoteboy]


BaileyPerformance - 30/10/12 at 10:10 AM

quote:
Originally posted by coyoteboy
quote:

MS recommendation on this is to ground everything to the block, not to the battery negative. I'm not trying to create an argument about this, but if you post a question on the MS forums and have earthed differently to their recommendation then you are likely to get short answers.



That's one option, personally I don't like using the block/ancills as a distribution point and run separate grounds for every sensor back to the battery neg. Either is acceptable so long as you stick to one so all of your ground refs are to the same point. The issues start when people ground to the block on a "normal" car, which can then often be grounded to the chassis, then back to battery neg, or just back to battery neg, or some other schemes I've seen. Then you start getting HR connections in one of many ground wires and the whole lot becomes a nightmare. Been there [block grounding], done that - won't do it again unless I'm starting from absolute scratch.

[Edited on 30/10/12 by coyoteboy]


The best way to ground the system is as per the link shown on one of the previous posts, it is wrong to ground the sensors to the anything other than the megasquirt plug. You can chose to ground the MS at the battery/block or the body as long as good ground straps are installed from the engine+battery+body.

The reason for this is the potential difference between the MS itself and the ground point, the MS ground wires carry the injector and sometime ignition current (assuming you are using direct drive coils) back to ground, this will be several amps. Even if the ground wires from the MS are kept short, some voltage drop will be present between the MS and the ground point, as the MS 0v reference for all sensors is on board (A to D ref is the 0v rail within the MS) any difference between the sensor and the board WILL cause sensor error or noise. The simple solution to this is to run the sensor ground separately back to the MS, so this ground wires are only carrying the sensor current. (mA)

If you look at Ford EEC engine management this is how its done, also the same for Cosworth. Note - Cosworth used the inlet manifold as the ground point for the ECU but all sensor grounds/returns went back to the ECU.


coyoteboy - 30/10/12 at 03:18 PM

Indeed, I apologise, it was late and I wrote some rubbish, I meant ground the sensors to the squirt then squirt via many wires to the battery neg/distrib block. What I dislike is grounding the ECU to the block, using the same upstream ground for the ecu and sensors as for the biggest power draw, inductive spike filled beast of a motor and the alt isn't in any electronics design strategy I know of. Obviously you can end up stuck having to do that if your sensors are engine grounded, personally I replaced all my engine grounded sensors for that reason.

Are the injector/coil grounds in the ms v3 board separate to the sensor grounds, I don't remember the schems. On the v2.2 they are shared, another woe.


BaileyPerformance - 30/10/12 at 06:11 PM

coyoteboy,

I think we are on the same page ;-)

The way Ford do the grounding is to ground the ECU to the inlet manifold or the block, then run an extra 6mm2 ground wire to the body. (as well as the main 25mm2 ground from the battery to the block and the 10mm2 from the battery to the body)

The trick here is to run more than one engine ground, voltage drop is one issue which can be solved with one heavy cable but high frequency noise is best grounded with several small cables.

In some ways the logical place for the MS grounds is the battery negative (this is the way i used to do it) but, the source of energy (and most of the noise) in a car is the alternator not the battery - so it makes more sense to ground to the block as thats where the alternator is grounded.

The MS V3.0 has 1 common ground, several pins are in parallel to form a heavy low resistance common connection point for the sensor returns and the ground wires to the block. I just solder the sensor grounds to the common pins.

Another interesting thing is MS2 is far more sensitive to electrical noise than MS1, i can only assume this is due to the much faster processor speed on MS2. I always add a 1000mf 25v cap inside the MS case across the incoming 12v, and a 100mf cap across the 5v. Screened cable is a good idea for the TPS too.