"MS4" wishlist thread
Moderators: jsmcortina, muythaibxr
-
- Super MS/Extra'er
- Posts: 884
- Joined: Sat Apr 30, 2011 12:34 pm
- Location: Sandefjord, Norway
Re: "MS4" wishlist thread
One of the most unreliable components you can have in an electronic system is a connector, so I disagree that going to a module based system is a good idea.
OTOH, I think it should be possible to make universal coil/hi-Z/lo-Z drivers if the chosen MOSFET or IGBT is fast enough and correctly dimensioned. Again, it will require programmable hardware (FPGA or SoC) to be feasible, since I believe the PWM current limiter will have to sit upstream of the routing matrix to keep code complexity manageable.
Signal flow:
Spark signal => routing matrix => generic output
Injector signal => opt. PWM => routing matrix => generic output
Aux signal => routing matrix => generic output
The same concept could also be used for inputs, with a routing matrix in hardware all inputs will arrive at the same processor pins in every configuration, reducing the number of branches necessary to perform any function.
OTOH, I think it should be possible to make universal coil/hi-Z/lo-Z drivers if the chosen MOSFET or IGBT is fast enough and correctly dimensioned. Again, it will require programmable hardware (FPGA or SoC) to be feasible, since I believe the PWM current limiter will have to sit upstream of the routing matrix to keep code complexity manageable.
Signal flow:
Spark signal => routing matrix => generic output
Injector signal => opt. PWM => routing matrix => generic output
Aux signal => routing matrix => generic output
The same concept could also be used for inputs, with a routing matrix in hardware all inputs will arrive at the same processor pins in every configuration, reducing the number of branches necessary to perform any function.
Joachim
1974 Jensen-Healey
1990 VW Caravelle Syncro - running MS3+X
2014 Ford Fiesta EcoBoost
1974 Jensen-Healey
1990 VW Caravelle Syncro - running MS3+X
2014 Ford Fiesta EcoBoost
-
- Super MS/Extra'er
- Posts: 9128
- Joined: Sun May 02, 2004 6:51 am
- Location: Quebec, Canada
- Contact:
Re: "MS4" wishlist thread
Injector drivers use 70V and less clamping voltage while coil drivers use 300V and more. You will either seriously compromise the coil driver so that the spark is weak or you will burn injectors. And that's just one aspect.
Connectors might be bad, I agree, but compromised performance and burned components is definitely worse.
Jean
Connectors might be bad, I agree, but compromised performance and burned components is definitely worse.
Jean
Re: "MS4" wishlist thread
What is the typical flyback current of a Hi or Lo-Z injector? I'm thinking a 'universal' component set would feature an ignition-class device, with a 70v Zener and a small-but-tough mosfetin parallel. The secondary mosfet (400v Vdss)prevents ignition flyback from going through the Zener, but allows injector coil (or other device) flyback to be clamped at a much lower level.
FB = Flyback Bypass; drive it high for a non-ignition device.
FB = Flyback Bypass; drive it high for a non-ignition device.
QuadraMAP Sensor Module -- PWM-to-Stepper Controller -- Dual Coil Driver
Coming soon: OctoMAP Sensor Module
TTR Ignition Systems
Coming soon: OctoMAP Sensor Module
TTR Ignition Systems
-
- Experienced MS/Extra'er
- Posts: 153
- Joined: Tue Oct 06, 2009 4:58 pm
- Location: NSW, Australia
- Contact:
Re: "MS4" wishlist thread
The Microsquirt and MS3Pro modules don't seem to have any issues being plugged into the board via a connector...subwoofer wrote:One of the most unreliable components you can have in an electronic system is a connector, so I disagree that going to a module based system is a good idea
'89 Toyota MR2 AW11 - 1MZFE 3L V6 - Need to finish car before ecu
'89 Toyota MR2 AW11 - 16V 4AGZE - DIYPNP 1.5v
'90 Toyota MR2 SW20 - 1MZFE 3L V6 - MS3X
'89 Toyota MR2 AW11 - 16V 4AGZE - DIYPNP 1.5v
'90 Toyota MR2 SW20 - 1MZFE 3L V6 - MS3X
-
- Super MS/Extra'er
- Posts: 884
- Joined: Sat Apr 30, 2011 12:34 pm
- Location: Sandefjord, Norway
Re: "MS4" wishlist thread
Remember how we used to fix computers that started behaving erradically in the old days? Pry every socketed chip out and press it back in, pull off every cable and reseat it - that would usually fix it. Connectors work fine, and can do so for quite some time but it is usually the first problem to occur.
Joachim
1974 Jensen-Healey
1990 VW Caravelle Syncro - running MS3+X
2014 Ford Fiesta EcoBoost
1974 Jensen-Healey
1990 VW Caravelle Syncro - running MS3+X
2014 Ford Fiesta EcoBoost
-
- Experienced MS/Extra'er
- Posts: 153
- Joined: Tue Oct 06, 2009 4:58 pm
- Location: NSW, Australia
- Contact:
Re: "MS4" wishlist thread
While that is indeed true, you will find that it was only on connectors that don't have a locking mechanism. You might have to reseat a PCI card that can wiggle, or a stick of ram or even a IDE cable, but when did you ever have to reseat a processor or molex connector? Considering there have been no issues with the MS1/2/3, Microsquirt or MS3PRO modules being all connectio/socket based, I think a simple low mass injector/ignition driver with a retaining screw isn't going to failure point. I would actually expect the rats nest of wiring you see inside most MegaSquirts to be a higher risk for failure due to vibration then a socket and retainer.subwoofer wrote:Remember how we used to fix computers that started behaving erradically in the old days? Pry every socketed chip out and press it back in, pull off every cable and reseat it - that would usually fix it. Connectors work fine, and can do so for quite some time but it is usually the first problem to occur.
'89 Toyota MR2 AW11 - 1MZFE 3L V6 - Need to finish car before ecu
'89 Toyota MR2 AW11 - 16V 4AGZE - DIYPNP 1.5v
'90 Toyota MR2 SW20 - 1MZFE 3L V6 - MS3X
'89 Toyota MR2 AW11 - 16V 4AGZE - DIYPNP 1.5v
'90 Toyota MR2 SW20 - 1MZFE 3L V6 - MS3X
Re:
Yes, that is what i'm talking about: measuring the peak pressure at the cilinder and making it match 16º ATDC (I know is not that simple, and yes, I've read John B. Heywood also) would avoid lot's of dyno tuning sessions trying to measure power at a chassis dyno with drive train interference trying to distinguish forged pistons and valve train from knocking...Peter Florance wrote: That seems like recipe for disaster: Encouraging tuning without power measurement.
Disasters can also happen at dyno shops... Ion Sense is the way to go.
Last edited by athal on Tue Jun 03, 2014 12:33 pm, edited 1 time in total.
Re: "MS4" wishlist thread
Yes, it seems hard to achieve. And I wish strongly that Megasquirt would be first aftermarket standalone ECU to implement Ion Sensing. I'm very patient, and Megasquirt gurus like James and Matt and lots of others in this forum are always surprising us...racingmini_mtl wrote:Then you won't be buying another MS (or any other ECU for that matter). There are major issues with having a workable ion sensing peak detection for the whole operating range. Why do you think no one has it?athal wrote:I won´t buy another MS until they provide ION SENSING capabilities.
Jean
I agree: As far as i know no aftermarket ECU still has it. But is becoming an industry standard. Even Harley Davidson is using it...
-
- Super MS/Extra'er
- Posts: 9128
- Joined: Sun May 02, 2004 6:51 am
- Location: Quebec, Canada
- Contact:
Re: "MS4" wishlist thread
Not for peak pressure position. It is used for phase, misfiring and knock detection.athal wrote:But is becoming an industry standard. Even Harley Davidson is using it...
Jean
Re: "MS4" wishlist thread
I was trying to find that on my MS2 but was surprised I never found it. Of course it MUST be software defined the firing order. Not messing with the cables. Generic wiring and configurable firing order.
Miata NBFL 1.6 (EU), GT2560R, MS2Extra, Zeitronix.
Fiesta 1.25 Zetec-SE, N/A mods, MS3, AEM.
Fiesta 1.25 Zetec-SE, N/A mods, MS3, AEM.
Re: "MS4" wishlist thread
OK Jean.racingmini_mtl wrote:Not for peak pressure position. It is used for phase, misfiring and knock detection.athal wrote:But is becoming an industry standard. Even Harley Davidson is using it...
Jean
There have to be a starting point: Ville's software picked raw data from ion sensing mirroring and interpolated some kind of Gaussian curve.
If Ion Sensing can "see" a curve with the knocking shape, or with the misfire shape it may be also able to see a curve with normal/optimal combustion shape.
Comparing the top of the curve with crankshaft position would be the next step. It sounds easy but I'm sure is not.
I was just asking for MS developers not to give up with Ion Sensing. It doesn't have to be perfect, but matching peak pressure curve with let's say 16º ATDC would be a nice starting point to generate an intial spark advance map. Better rather than this:
http://www.useasydocs.com/theory/spktable.htm
More or less in the same way Tuner Studio "Tune Analize Live" is not perfect, but is good starting point for an aftermarket ECU.
Last edited by athal on Wed Jun 04, 2014 10:59 am, edited 2 times in total.
Re: "MS4" wishlist thread
For what it's worth... Saab pioneered the system, they designed the H-D system, and GM uses it because they got access to Saab's intellectual property when they bought the company. Is any other OEM doing ion sensing?athal wrote:But is becoming an industry standard. Even Harley Davidson is using it...
I'm not saying it's a bad idea... I own a Saab with that system, have done some tuning on it, and have great respect for it. But I also know it's very finicky, and requires a LOT of processor power. The 16 mHz 32-bit MC68332 processor they initially used could barely keep up - later they switched to a 20 mHz version.
Eric Law
1990 Audi 80 quattro with AAN turbo engine: happily running on MS3+MS3X
2012 Audi A4 quattro, desperately in need of tweaking
Be alert! America needs more lerts.
1990 Audi 80 quattro with AAN turbo engine: happily running on MS3+MS3X
2012 Audi A4 quattro, desperately in need of tweaking
Be alert! America needs more lerts.
Re: "MS4" wishlist thread
BMW also uses Ion Sensing.elaw wrote:For what it's worth... Saab pioneered the system, they designed the H-D system, and GM uses it because they got access to Saab's intellectual property when they bought the company. Is any other OEM doing ion sensing?athal wrote:But is becoming an industry standard. Even Harley Davidson is using it...
I'm not saying it's a bad idea... I own a Saab with that system, have done some tuning on it, and have great respect for it. But I also know it's very finicky, and requires a LOT of processor power. The 16 mHz 32-bit MC68332 processor they initially used could barely keep up - later they switched to a 20 mHz version.
Yes. Processing speed can be an issue for closed loop cylinder to cylinder advance correction.
But as I said it hasn't to be perfect at the beguining for an aftermarket ECU. An ignition map can be built in similar way Tune Analize live does for VE table...
Lucky you!!! happy owner of a SAAB my engine is a GM stright six 250
-
- Super MS/Extra'er
- Posts: 9128
- Joined: Sun May 02, 2004 6:51 am
- Location: Quebec, Canada
- Contact:
Re: "MS4" wishlist thread
You don't understand. What Ville did was for a very limited RPM and load range. And that's all you can get. Outside that narrow range, you have more noise than signal so fitting would give a random result which would worse than useless. So you would not get an advance map but only a few points at best.athal wrote:OK Jean.racingmini_mtl wrote:Not for peak pressure position. It is used for phase, misfiring and knock detection.athal wrote:But is becoming an industry standard. Even Harley Davidson is using it...
Jean
There have to be a starting point: Ville's software picked raw data from ion sensing mirroring and interpolated some kind of Gaussian curve.
If Ion Sensing can "see" a curve with the knocking shape, or with the misfire shape it may be also able to see a curve with normal/optimal combustion shape.
Comparing the top of the curve with crankshaft position would be the next step. It sounds easy but I'm sure is not.
I was just asking for MS developers not to give up with Ion Sensing. It doesn't have to be perfect, but matching peak pressure curve with let's say 16º ATDC would be a nice starting point to generate an intial spark advance map. Better rather than this:
http://www.useasydocs.com/theory/spktable.htm
More or less in the same way Tuner Studio "Tune Analize Live" is not perfect, but is good starting point for an aftermarket ECU.
Again ask yourself why it's not available or used anywhere (in that specific way). If people with huge budgets don't do it there's usually a reason. What you want is the equivalent of a wideband O2 sensor for spark; what you can get is the equivalent of a narrow band for idle only. Also, have you ever seen anyone being able to get some operational setup based on Ville's work?
So unless you find someone who has a reproducible working setup (for the full RPM and load range) that just needs to be integrated into an engine management setup, I don't see why there should be time and money spent on this feature. It is something that could require huge amounts of both. I personally thinks that it is not theoretically possible based on what I've read.
Jean
Re: "MS4" wishlist thread
Jean, is not my intention creating sterile polemics in this forum.
May be no one had good resoults developing Ville's work.
But there are guys that are fare more ahead: The Ed VanDyne team from Adrenaline Research (now Motorola) and it's aftermarket SMARTFIRE IGNITION module with ION SENSING capabilities and PLASMA enhanced discharge and an astronomic price out of reach for me and for most in this forum.
http://www.peaker.com/PDFs/ACC.pdf
http://papers.sae.org/2000-01-1377/
And this was fifteen years ago.
What I mean is this stuff is not out of reach.
Jean[/quote]
Again ask yourself why it's not available or used anywhere (in that specific way). If people with huge budgets don't do it there's usually a reason. What you want is the equivalent of a wideband O2 sensor for spark; what you can get is the equivalent of a narrow band for idle only. Also, have you ever seen anyone being able to get some operational setup based on Ville's work?
So unless you find someone who has a reproducible working setup (for the full RPM and load range) that just needs to be integrated into an engine management setup, I don't see why there should be time and money spent on this feature. It is something that could require huge amounts of both. I personally thinks that it is not theoretically possible based on what I've read.
Jean[/quote]
May be no one had good resoults developing Ville's work.
But there are guys that are fare more ahead: The Ed VanDyne team from Adrenaline Research (now Motorola) and it's aftermarket SMARTFIRE IGNITION module with ION SENSING capabilities and PLASMA enhanced discharge and an astronomic price out of reach for me and for most in this forum.
http://www.peaker.com/PDFs/ACC.pdf
http://papers.sae.org/2000-01-1377/
And this was fifteen years ago.
What I mean is this stuff is not out of reach.
Jean[/quote]
Again ask yourself why it's not available or used anywhere (in that specific way). If people with huge budgets don't do it there's usually a reason. What you want is the equivalent of a wideband O2 sensor for spark; what you can get is the equivalent of a narrow band for idle only. Also, have you ever seen anyone being able to get some operational setup based on Ville's work?
So unless you find someone who has a reproducible working setup (for the full RPM and load range) that just needs to be integrated into an engine management setup, I don't see why there should be time and money spent on this feature. It is something that could require huge amounts of both. I personally thinks that it is not theoretically possible based on what I've read.
Jean[/quote]
-
- Super MS/Extra'er
- Posts: 17499
- Joined: Thu Apr 16, 2009 8:08 pm
Re: "MS4" wishlist thread
There's been some looking at ion sensing before: single cylinder test engine Yes, I know it's an older thread.
But when doing research, by definition, it's hard to predict what you will find. No telling whether the end result will be a viable product, or a conclusion that the technology needed just isn't there yet.
But when doing research, by definition, it's hard to predict what you will find. No telling whether the end result will be a viable product, or a conclusion that the technology needed just isn't there yet.
Last edited by Matt Cramer on Thu Jun 05, 2014 6:42 am, edited 1 time in total.
Matt Cramer -1966 Dodge Dart slant six running on MS3X
Re: "MS4" wishlist thread
Not to be too argumentative, but the sources you cite don't really offer any support for the goal of determining whether timing is correct.
Ignoring for a moment knocking conditions, determining the best ignition timing is a matter of adjusting so the the peak *pressure* point from combustion occurs at the correct point in the engine's cycle.
The Woodward documentation about the SmartFire system does not talk about measuring pressure. They say the system can be used to identify pre-ignition and misfires - both of which require only making a yes/no determination about whether combustion took place (and of course when). The SAE paper only discusses misfire detection, which again only requires making a determination about whether combustion occurred or not. Detecting whether combustion did or did not take place is relatively easy... because when the gas mixture burns, it creates a plasma which is electrically conductive, and that conductivity can be measured at the spark plug gap. Measuring cylinder pressure and determining its peak is a whole different thing.
Just to be clear, I *do* think it would be cool if MS could do preignition and misfire detection via ion sensing... if the CPU has enough horsepower which I'm doubtful about. But I have a hard time seeing how the leap could be made from that to "spark advance autotune".
Which brings me to my "wishlist for ms4" idea: an MC68332 processor! Now *that's* a processor with some horsepower, and is specifically designed for automotive applications. Of course it's also near end-of-life so maybe a ColdFire MCF532x type instead.
Ignoring for a moment knocking conditions, determining the best ignition timing is a matter of adjusting so the the peak *pressure* point from combustion occurs at the correct point in the engine's cycle.
The Woodward documentation about the SmartFire system does not talk about measuring pressure. They say the system can be used to identify pre-ignition and misfires - both of which require only making a yes/no determination about whether combustion took place (and of course when). The SAE paper only discusses misfire detection, which again only requires making a determination about whether combustion occurred or not. Detecting whether combustion did or did not take place is relatively easy... because when the gas mixture burns, it creates a plasma which is electrically conductive, and that conductivity can be measured at the spark plug gap. Measuring cylinder pressure and determining its peak is a whole different thing.
Just to be clear, I *do* think it would be cool if MS could do preignition and misfire detection via ion sensing... if the CPU has enough horsepower which I'm doubtful about. But I have a hard time seeing how the leap could be made from that to "spark advance autotune".
Which brings me to my "wishlist for ms4" idea: an MC68332 processor! Now *that's* a processor with some horsepower, and is specifically designed for automotive applications. Of course it's also near end-of-life so maybe a ColdFire MCF532x type instead.
Eric Law
1990 Audi 80 quattro with AAN turbo engine: happily running on MS3+MS3X
2012 Audi A4 quattro, desperately in need of tweaking
Be alert! America needs more lerts.
1990 Audi 80 quattro with AAN turbo engine: happily running on MS3+MS3X
2012 Audi A4 quattro, desperately in need of tweaking
Be alert! America needs more lerts.
-
- Super MS/Extra'er
- Posts: 3653
- Joined: Fri Apr 02, 2004 8:40 pm
- Location: Virginia Beach, VA
- Contact:
Re: "MS4" wishlist thread
My fear is ion sensing is considered as an alternative to dyno testing.
Peter Florance
PF Tuning
81 BMW Euro 528i ESP Car
60-2 Wheel LS2 Coils, Low Z Inj
Co-Driver 1999 BMW E46 DSP car.
PF Tuning
81 BMW Euro 528i ESP Car
60-2 Wheel LS2 Coils, Low Z Inj
Co-Driver 1999 BMW E46 DSP car.
Re: "MS4" wishlist thread
If it gave the peak pressure info, it would be. As it (apparently) stands, it isn't.
QuadraMAP Sensor Module -- PWM-to-Stepper Controller -- Dual Coil Driver
Coming soon: OctoMAP Sensor Module
TTR Ignition Systems
Coming soon: OctoMAP Sensor Module
TTR Ignition Systems
-
- Super MS/Extra'er
- Posts: 1679
- Joined: Tue Oct 27, 2009 6:24 am
- Location: Van Alstyne, Texas
Re: "MS4" wishlist thread
subwoofer wrote:Ion sense engine phase detection, knock detection and misfire detection should be within reach though. I don't know how much hardware is needed to support those functions for HD and SAAB coils.
:
Perhaps within reach with direct pressure measurement, which should be easy + cheap now that the MEMS sensors are being built into factory diesel glow plugs, but I digress...
Ion sense unfortunately only provides a good signal/noise ratio at medium>WOT loads, at idle>cruise and even light acceleration extracting the data is a PITA, there is nothing to see vs the raw flamefront ionization.
Is useful for knock/det sense but not PPP except at high loads, as the knock signal stands out.
Direct pressure measurement does not seem to have as many issues, and the sensors put out a decent signal.
Always doing things the hard way, MS2 sequential w/ v1.01 mainboard, LS2 coils. 80 mile/day commuter status.