Page 17 of 20 FirstFirst ... 71516171819 ... LastLast
Results 241 to 255 of 293

Thread: Ecotrons Engine Management

  1. #241
    Join Date
    7th June 2009 - 13:29
    Bike
    Norton Manx
    Location
    Over the Rainbow
    Posts
    379
    Quote Originally Posted by Nath88 View Post
    Quote Originally Posted by TZ350 View Post
    Unlike some other EFI systems the Ecotrons software does not allow for a binary input that switches between two maps. But there is a place for an analog input from the MAP sensor to influence the Ecotrons software to adjust the predicted fueling requirements from the Alpha-N map at any TPS/RPM.
    I was just about to recommend using the map switch I tried it out this weekend, its not dialled yet, but it's the best so far, I can roll along at 8500rpm and open and close the throttle as fast or slow as I like and it always responds. I will tune the main map by lambda feedback (it's pretty much done using values from before), and the misfire map by holding the engine at a load point, then holding the kill switch for a second or two, then release and judge the response when it comes back on as rich or lean, it should be instant and clean like a carb.
    The map switches when the peak pressure between 85 crank degrees and 125 crank degrees, measured at the PV vent (with an adjustable bleed to atmosphere) reaches 102kpa, atmosphere is 101kpa here.
    Means I'm confined to riding at sea level until I sort out a better implementation. Hoping the piezo pulse sensor will be the ticket, since it doesn't care about absolute values, If it works I'll make a bunch of them.

    I think the piezo on the cylinder head idea could work, if the signal is not too noisy.
    Between TZ and Nath88 we may yet see a practical DIY method for fuel injecting two strokes.
    Factual Facts are based on real Fact and Universal Truths. Alternative Facts by definition are not based on Truth.

  2. #242
    Join Date
    18th May 2007 - 20:23
    Bike
    RG50 and 76 Suzuki GP125 Buckets
    Location
    Auckland
    Posts
    10,473
    Quote Originally Posted by Sandokan View Post
    would it be possible to somehow "record" an exhaust harmonic frequency in conjunction with a tps an wideband lambda on a carburated twostroke?
    Good idea Sandokan, I didn't record the exhaust but I guess its possible. I did use the wideband lambda and carb to get an idea of what the fuel map should look like.

    Quote Originally Posted by Niels Abildgaard View Post
    Put a Piezo thing ring under one of the headstud nuts or between head and sparkplug and a very rude instant cylinder pressure signal can be had.
    When piston is halfway down You will know if it fired or not. And then: Who tried it already and why does it not work?
    Another good idea, thanks Niels. It could well work, unfortunately my Ecotrons software does not have a suitable input available, it is short on alternative inputs. But it does expect a MAP sensor which is why I am trying to see what can be done with that.

    Unlike some other EFI systems the Ecotrons software does not allow for a binary input that switches between two maps. But there is a place for an analog input from the MAP sensor to influence the Ecotrons software to adjust the predicted fueling requirements from the Alpha-N map at any TPS/RPM.

    The Ecotrons software adjusts the injector pulse width according to the Alpha-N map and input from the MAP sensor. It is how to make use of that relationship for fuel injecting a two stroke which is the trick I am trying to learn.

  3. #243
    Join Date
    18th May 2007 - 20:23
    Bike
    RG50 and 76 Suzuki GP125 Buckets
    Location
    Auckland
    Posts
    10,473
    Attachment 332794

    Smoothed the MAP sensor reading by adding some volume by replacing the MAP sensor hose with a small chamber.

    Attachment 332793

    MAP sensor Yellow line. Interesting that the line on the left as the engine is still winding down and zero throttle is higher than the MAP pressure line at WOT and as the RPM starts to rise again.

  4. #244
    Join Date
    18th May 2007 - 20:23
    Bike
    RG50 and 76 Suzuki GP125 Buckets
    Location
    Auckland
    Posts
    10,473
    Quote Originally Posted by Nath88 View Post
    Same rpm, same throttle position, different airflow. The pipe effect on a misfire is very weak if it exists at all...

    Attachment 332725

    My graph doesn't show this well, but above 7k RPM the on-pipe and off-pipe air flow is too different, the on-pipe fuelling (what you've tuned for) will not support combustion off-pipe. Way too rich, true lambda probably around 0.5 or worse. As the engine speed falls below 7k rpm, the on-pipe and off-pipe air flow is not so different, true lambda is approaching the rich limit until it supports combustion. Once combustion begins the small increase of airflow from the pipe effect will bring the lambda above the rich limit, 'clearing' the cylinder. You could install any old MAF sensor to prove this.

    I'm still working on the Ion sensing, but I'm getting impatient so I'm going back to measuring the pressure at the PV vent since that worked the best so far (I now have an adjustable bleed to atmosphere so I can adjust it's sensitivity), and using that input to try a table switching approach. Have a fuel table for no-pipe effect, and a fuel table that incorporates pipe effect. Theoretically I could put a box-style muffler on and tune 1 table, then put the expansion chamber on and tune the other. Pressure pulses will determine whether the pipe effect is active and ECU switches between the tables accordingly. Away from the pipe's effective rpm the tables should be more and more the same.

    I'm thinking a box-muffler would provide no scavenging effect, and no return wave effect, can someone confirm that?

    I'm also designing a 'pressure pulse' sensor. It's basically a piezo electric microphone hooked up to exhaust pressure, it will more accurately and quickly detect the exhaust pulse from the cylinder than the pressure sensor. I figure I can hear when the engine is firing clear as day, why not use a microphone...
    Quote Originally Posted by Nath88 View Post
    Same rpm, same throttle position, different airflow.

    Attachment 332725
    I understand Nath88's explanation for the change in the LamW02 trace Yellow line. .... good one.

    Attachment 332825

    Pressure in the exhaust header pipe as measured by a MAP sensor, Yellow line, RPM Blue line, LamWo2 Brown line.

    Through a one way valve arrangement the MAP sensor only measures the high positive pressures of the returning pulse in the header and does not see the negative suction pressure.

    Interesting that the MAP pressure line on the left as the engine is still winding down and at zero throttle is higher than ambient and the MAP pressure line on the right at WOT which starts at ambient before obviously rising again as the engine gets on the pipe. There is quite a sudden drop in MAP pressure (Yellow line) when the throttle (Red line) goes from fully closed to fully open.

    But why does the header pipe positive pressure behave like this? ...... .. high and above ambient when there is no throttle and the engine is winding down then goes lower when it's on WOT and starting to pick up again.

    Quote Originally Posted by Nath88 View Post
    I saw something similar with mine.
    Since we're measuring the peak, it implies the average pressure in the pipe must be lower than atmospheric by a fair bit?
    Has anyone witnessed a vacuum in the pipe at any time?

    Once the engine starts making power the exhaust flow creates pressure in the pipe due to the stinger restriction.

    A differential pressure sensor with the average pipe pressure as the reference might be the go. Or my new sensor, should be testing the prototype next week.
    Attachment 332827

    Can you set the crank angle for sampling the MAP sensor with the ecotrons?
    Quote Originally Posted by Grumph View Post
    Can I suggest that the whole system - from inlet via crankcase to muffler - has one end closed when the throttle is shut and the chamber tailpipe is then doing what it's supposed to and acting as a pressure bleed resistor. Open the other end to atmosphere - open the throttle - and the system pressure should drop to atm and possibly lower momentarily.
    That is possible I guess, but on overrun with throttle closed and the motor not firing what could be creating the pressure in the pipe?

    You would think any residual pressure from a full power run would have bleed away quite quickly but on other graphs it actually looks like it builds up a bit.

    Bit of a puzzle

    Quote Originally Posted by Nath88 View Post
    Can you set the crank angle for sampling the MAP sensor with the Ecotrons?
    No. But I am very interested in the sensor you are developing.

    Quote Originally Posted by Nath88 View Post
    That's a shame, timing will be key for this thing.

    I thought that it was showing atmospheric pressure off the throttle, going into vacuum as you opened the throttle. What pressure is atmo?
    But if its producing positive pressure with a closed throttle I'm out of ideas... Definitely qualifies as a puzzle.
    Quote Originally Posted by Frits Overmars View Post
    Pipe pressure varies not only with throttle and rpm, it can also vary locally because of wave superposition.

    It might be instructive to experiment with several pressure take-off points along the pipe.

    Attachment 332834
    Yes I can see how wave superposition can create a localized high pressure spot although I would expect the node to move with frequency as the RPM dropped. I will look for that next time.

    Quote Originally Posted by teriks View Post
    Assuming that the one way valve is working properly, and you still have that bleed between the one way valve and sensor, the lowest pressure you should ever see at the sensor is atmospheric, no matter how low the pressure in the pipe is.
    Yes, that is how I read it too.

    Quote Originally Posted by teriks View Post
    ... the drop in pressure when opening the throttle:
    Closed throttle => ~0 airflow thru the header
    Open throttle (even without firing) => >0 airflow the the header.

    Now, depending on the geometry around your pressure pickup position its not at all unlikely that the pressure drops in that position simply due to flow. -Think venturi..
    The closed throttle no flow pressure was higher than atmospheric and dropped to atmospheric when the throttle was opened again. Your venturie idea is interesting.

    Pressure node and venturie effect could explain it.......

    Attachment 332835

    Quote Originally Posted by Nath88 View Post
    I was just about to recommend using the map switch I tried it out this weekend, its not dialled yet, but it's the best so far, I can roll along at 8500rpm and open and close the throttle as fast or slow as I like and it always responds. I will tune the main map by lambda feedback (it's pretty much done using values from before), and the misfire map by holding the engine at a load point, then holding the kill switch for a second or two, then release and judge the response when it comes back on as rich or lean, it should be instant and clean like a carb.
    The map switches when the peak pressure between 85 crank degrees and 125 crank degrees, measured at the PV vent (with an adjustable bleed to atmosphere) reaches 102kpa, atmosphere is 101kpa here.
    Means I'm confined to riding at sea level until I sort out a better implementation. Hoping the piezo pulse sensor will be the ticket, since it doesn't care about absolute values, If it works I'll make a bunch of them.
    Attachment 332861

    Just realised/remembered that the ecotrons EFI system does have a switch for switching between two maps. I had been using it for so long as an input for the wide band O2 sensor that I had forgotten about it.

    But after I have got the on the pipe map sorted it looks like I may be able to use one of your switches to then develop an off pipe or miss fire map.

    The performance switch just switches in another table of values that are used to divide the values in the main map. Ie if you have a load value of 120 in the Alpha-N map and switch in the performance map that has a Lambda value of 0.85 then the Alpha-N map gets divided by 0.85, 120/0.85 = 142 ie the result is a richer mixture for exceleration.

    I guess there is no reason why the performance switch can't be used to lean the Alpha-N map off when the engine has misfired or is off the pipe.

    Quote Originally Posted by Nath88 View Post
    If it's another TPSxRPM table with multipliers it should work great. I'm doing the same thing right now but in a spreadsheet, then dropping the numbers into the tuning software. The multiplier is the pipe effect on airflow so it's a more sensible value to use anyway. You'll need to extrapolate the tuned 'firing' map you have into the skip-fire/misfire regions as best you can. Making the switched input to feed into the ECU will be the hard part, but I want to do that for my setup as well to free up the MAP input, and make the sensor unit a plug and play product for any ECU.

    Had a play with the piezo disc for the exhaust pulse sensor, looks promising, very sensitive, minimal noise, nice.
    Quote Originally Posted by Nath88 View Post
    If it's another TPSxRPM table with multipliers it should work great. I'm doing the same thing right now but in a spreadsheet, then dropping the numbers into the tuning software. The multiplier is the pipe effect on airflow so it's a more sensible value to use anyway.

    Had a play with the piezo disc for the exhaust pulse sensor, looks promising, very sensitive, minimal noise, nice.
    Yes that piezo looks promising as the basis for a switch. I will definitely be a customer for one.

    Attachment 332883

    This is the Ecotrons EFI softwares Basic 16x12 Alpha-N map.

    Attachment 332884

    When the "Performance" switch is activated the corresponding cells in this 16x12 map are used to divide their related values in the original Alpha-N map to give an alternative fueling value.

    This seriously looks like it could be used in the way you have described previously of leaning the fuel off when there is less air flow when there is no pipe action drawing air through the motor.

  5. #245
    Join Date
    18th May 2007 - 20:23
    Bike
    RG50 and 76 Suzuki GP125 Buckets
    Location
    Auckland
    Posts
    10,473
    Quote Originally Posted by wobbly View Post
    Quote Originally Posted by TZ350 View Post
    Another unfulfilling and downright frustrating experience with Ecotrons support Dept.... .
    Did it occur to you TeeZee that just possibly, you know more about the system now, than they do.
    With small capacity high performance 2T EFI that maybe true ... ... and sure, in fairness you can't expect them to be onto everything.

    But what I asked for was a simple change to a basic cal file to turn on the LamO2 sensor. I want my cal file to have both "Injection Angle End" and "LamWO2" options.

    Its easy enough, other EFI software packages come with tick box's for that sort of thing. With Ecotrons you have to enlist their assistance and it is often a slow difficult negative experience dealing with their help dept.

  6. #246
    Join Date
    18th May 2007 - 20:23
    Bike
    RG50 and 76 Suzuki GP125 Buckets
    Location
    Auckland
    Posts
    10,473
    I received a reply from Ecotrons lastnight. The advanced calibration file needs these changes to be able to monitor and record results from the wide band oxygen sensor.

    CV_SSFLAM = 4
    CV_SSWO2 = 1
    VAL_uLamSlope = 0.3
    VAL_uLamOfs = 0.5

    Hopefully I will get to try it out this afternoon.

    I used to think that it was easier for the local 2T Kawasaki 350 and Yam 250 that are successfully running EFI because they did not rev much past 9,000 and so only needed one logical injector and that the bigger capacity helped mask fueling miss steps.

    I started to have doubts about that when I saw that some small 2T drone aero engines 25cc or less were successfully running EFI. But these basically run at a constant speed.

    The issue for me is drivability, i.e. throttling off for a corner and then on again exiting. After Nath88 pointed out that you can have two different air flows through the engine due to pipe action/non action for the same Throttle/Rpm position on the Alpha-N EFI fueling map and remembering Frits's explanation of how the wave action in a pipe collapses when throttling off and how it takes some time to re establish after throttling on again.

    I now think that one of the big differences is how aggressive the pipe is. On a small capacity high performance race engine the pipe action in drawing air through the engine would be much more variable than the pipe action on a larger capacity MX or Trail bike engine setup more for power range and usability than outright power hit.

    So in getting 2T EFI to work, it turns out the pipe makes a huge difference, who would have thought ......

  7. #247
    Join Date
    20th January 2010 - 14:41
    Bike
    husaberg
    Location
    The Wild Wild West
    Posts
    11,823
    Quote Originally Posted by TZ350 View Post
    I received a reply from Ecotrons lastnight. The advanced calibration file needs these changes to be able to monitor and record results from the wide band oxygen sensor.

    CV_SSFLAM = 4
    CV_SSWO2 = 1
    VAL_uLamSlope = 0.3
    VAL_uLamOfs = 0.5

    Hopefully I will get to try it out this afternoon.

    I used to think that it was easier for the local 2T Kawasaki 350 and Yam 250 that are successfully running EFI because they did not rev much past 9,000 and so only needed one logical injector and that the bigger capacity helped mask fueling miss steps.

    I started to have doubts about that when I saw that some small 2T drone aero engines 25cc or less were successfully running EFI. But these basically run at a constant speed.

    The issue for me is drivability, i.e. throttling off for a corner and then on again exiting. After Nath88 pointed out that you can have two different air flows through the engine due to pipe action/non action for the same Throttle/Rpm position on the Alpha-N EFI fueling map and remembering Frits's explanation of how the wave action in a pipe collapses when throttling off and how it takes some time to re establish after throttling on again.

    I now think that one of the big differences is how aggressive the pipe is. On a small capacity high performance race engine the pipe action in drawing air through the engine would be much more variable than the pipe action on a larger capacity MX or Trail bike engine setup more for power range and usability than outright power hit.

    So in getting 2T EFI to work, it turns out the pipe makes a huge difference, who would have thought ......
    Maybe the bang bang set up.
    https://www.pressreader.com/australi...82050504983923
    Quote Originally Posted by Katman View Post
    I reminder distinctly .




    Kinky is using a feather. Perverted is using the whole chicken

  8. #248
    Join Date
    18th May 2007 - 20:23
    Bike
    RG50 and 76 Suzuki GP125 Buckets
    Location
    Auckland
    Posts
    10,473
    Quote Originally Posted by jbiplane View Post
    Thank Norman, Frits and others. I found here many interesting ideas. We are small company which works like "engine atelier" for small aircrafts, UAV and other applicarion where weight and fuel economy are vital. To the moment numerous codes like ~bimotion ~mota ~Ricardo Wave ~Lotus concept
    gives irreal results on boxer 2-strokes engines I produce. So one of my present objectives is to make parametric CAD model linked with CFD simulation of 2-stroke engine complete from intake to exhaust. I believe will sucseed in ~4+ month by myself and understand what have to be changed.

    Quote Originally Posted by guyhockley View Post
    Quote Originally Posted by FastFred View Post
    There you are TZ, maybe a successful 2T EFI system, looks like an injector port in that throttle body.
    Lot easier to do FI on an aero-engine...
    Quote Originally Posted by jbiplane View Post
    This throttle body we got from Ecotron. We often have problem with components quality and have to do inhouse.
    Now we 5-axis mill bodies and make our proprietary 2T EFI system, very lightweight and poverfull
    We place some of sensors in one ECU box and got by wires cylinder head themperature and throttle angle.
    May be one day we will make 2T ECU as a "box product"
    Quote Originally Posted by Nath88 View Post
    The exhaust pressure sensor I was using died by drowning in exhaust oil, so it motivated me to rig up the new exhaust pulse sensor for a test.

    Attachment 333019

    Green line is the exhaust pulse, rate of change of pressure. Red lines are the sample window (ECU takes the lowest value inside this window). White lines are the trigger wheel teeth as the crank rotates. There's 20 teeth on the wheel, first tooth is at 72°BTDC. Primary exhaust pulse hits the sensor at around tooth 11 (126°). On this trace we see 2 firing primary pulses followed by a misfire pulse, caused by a throttle blip at 2400rpm, however engine speed increases an unknown amount over the 3 rotations. Either way the pipe is way out of sync with the engine at these revs. I later tweaked the sample window from 30 to 45 degrees duration to properly catch the pulses.

    I also took the bike to the dyno, it was running too rich, perhaps due to the dead sensor, maybe some other factors. Made 45hp on the dynojet, sounded rich all the way through but the curve was pretty consistent with carb 2 strokes, maybe a bit stronger coming onto the pipe. I'll scan the chart at some point and post it.
    Quote Originally Posted by Nath88 View Post
    I did some experimentation with the piezo sensor on the road yesterday, saw much less signal amplitude than the previous test and DC voltage seemed to vary a bit over time, but I think the hose was kinked and I know the wiring was crap. So I've fixed those things and will see if it's better/back to normal today.

    Also leaned it out from where it was on the dyno, target AFR is 14:1 before 5300rpm (anti-resonance point) then 13.4:1 AFR to 8500rpm (peak power) then it slowly leans out to 13.8:1 at 9800rpm. It feels pretty happy, probably more to gain here. I think it still blows a little smoke on full throttle before anti-resonance, might be able to go a bit leaner still.
    Totally looking forward to hearing how your piezo sensor experiment goes. I have been trying to re map my CPU for all low throttle fully on pipe running in the hope I will be able to get a piezo sensor switch from you sometime to see if I can use it to switch between low throttle on and off pipe maps.

    Quote Originally Posted by Nath88 View Post
    I did another test just free revving, signal output looks better than before, I'll do another road test this afternoon. The sensor follows the rise and fall of the overall pressure in the pipe, throws the cycle to cycle numbers off a bit so I'll add a filter to get rid of that. Hopefully a switch isn't too far off, putting the pressure on!

    Current test rig:
    Attachment 333094

    A more general question, if the engine is 'clear' but it's blowing some blue smoke at low-rpm full throttle, is it too rich? Is there any harm in going leaner until the smoke clears?

  9. #249
    Join Date
    18th May 2007 - 20:23
    Bike
    RG50 and 76 Suzuki GP125 Buckets
    Location
    Auckland
    Posts
    10,473
    Quote Originally Posted by Nath88 View Post
    Quote Originally Posted by TZ350 View Post
    Attachment 333134Attachment 333135

    I took a look at a Piezo switch and thought that I might be able to use the exhaust pulse instead of a finger press to activate it.

    Attachment 333136
    There is a 100Meg resistor across the piezo sensor which acts as the load, piezos need high impedance on the output to measure low frequencies, so that's fine. However the next 100Meg resistor and 100pF capacitor form a low pass filter of 15hz. This acts to smooth and slow the input, probably so it works decently as a switch instead of flicking on and off with any variation of pressure applied to the piezo. It may work, I would expect a delay between the pressure input and when the mosfet has enough voltage at it's input (gate) to 'switch'. The switch threshold would be set by the gate voltage of the mosfet combined with the filter.

    My pulse sensor gives clear indication at high rpm light throttle, which was the problem area with the pressure sensor used earlier. At 7700rpm the exhaust pulses are the strongest, so that's probably the peak resonance of the pipe. However at around 4600rpm the indication is poor, not a big difference between full throttle and closed throttle. The exhaust pulse from the port is probably being cancelled by the existing wave, why at 4600? It's giving me a headache so I've made the maps the same below 5000rpm, low rpm response was never an issue.
    Attachment 333187

    Got a Piezo switch today, 24V 200mA max.

    Attachment 333188

    Tried it out on my trusty PLC test bed. Piezo switch works great, pulse output non latching. Will check later how it goes with 12V and how square the wave form looks on a scope.

  10. #250
    Join Date
    18th May 2007 - 20:23
    Bike
    RG50 and 76 Suzuki GP125 Buckets
    Location
    Auckland
    Posts
    10,473
    .

    Following on from Nath88’s suggestions about the amount of air flowing through the motor that needs fuelling depends on whether the motor has fired and the pipe action has sucked air through or not.


    I have been looking for ways to see if the engine has fired or not. And I like Clints idea that measuring the in cylinder pressure just before exhaust port opening would be a reliable way to tell if the motor has fired or not. It would not take a very sophisticated pressure sensor to do that. One would not need to know the real pressure, only that there is more pressure and a simple home made sensor using a piezoelectric disk from a hobby shop may be enough to do that.

    It would not have to be that clever, with the help of an Aduino mini all I would need to know is that between two consecutive ignition trigger pulses there was an in cylinder pressure rise of a predetermined magnitude or not.



    Clint Gray - www.tfxengine.com sent me some comments by email that made sense to me and with Clint’s permission I can share them.

    The equipment that we sell is for engine tuning via measuring cylinder pressure, port pressure waves etc. per degree of crank rotation. That might be overkill for what you need.

    There are 2 ways that you could do it in the cylinder.

    The first would be to use a true cylinder pressure sensor which is necessarily exposed to the whole cycle. This would involve drilling and tapping the head.

    The second way to do it would be to install a pressure sensor part way down the cylinder wall. It would involve drilling a tiny hole roughly 0.75mm to 1mm diameter straight into the cylinder wall then mounting a sensor somewhat back from the hole external to the cylinder.

    As an example, if you drilled the hole anywhere in the cylinder wall circumference so the piston ring exposed the hole to the pressure from about 70 ATDC through all the crank angles to 70 BTDC, all you would need is a low pressure sensor mounted say an inch behind that hole (i.e. the hole is a 1 inch long passageway feeding the sensor), then take your reading at 80 ATDC and 80 ABDC. So twice per revolution 180 degrees apart. There would be quite a difference in pressure if the engine fired, yet the peak pressure and average temp the sensor would be seeing would not be all that high.

    We also have a sensor that will work in the cylinder wall like described but it is super-fast etc. For what you are doing, if you mounted a sensor as above, the sensor would not have to be super-fast, nor would it see high pressures, and average temps wouldn't be all that high considering its location and the part of the cycle it is seeing. Also the hole is a dead end hole so there is no significant flow in that hole, not like a hole that is open to the atmosphere. You might be able to find/use a much cheaper sensor if you go through the cylinder wall.

    Further to what I wrote above, regarding a small hole in the cylinder wall. As I mentioned you could measure twice per revolution 80 ATDC and 80 ABDC, or even just once per revolution at 80 ATDC and use a threshold pressure value to determine if it fired or not.

    However, it would be simpler to take a reading at 80 ATDC and 80 BTDC then compare the two pressure values per revolution. That way you don't need to have a threshold pressure value, just compare the two directly every revolution, and since the readings are not 180 degrees apart, they are 160 and 200 degrees apart, it would be easy to know which reading is the ATDC reading and which is the BTDC reading.

    The fact that the sensor is slow, say a ms delay shouldn't matter either. It is still going to "see" the pressure at 70 ATDC it's just that it's going to be a little bit of time for the sensor to fully register what it sees.

    You will be able to tell right away whether you need to delay the reading a full ms or much less than that, because when the hole is closed the pressure the sensor is seeing is only going to be crankcase pressure, so not much, and then when the hole is uncovered it will jump up very quickly, to say 150-200 psi. A slow sensor may take a full ms to fully register a pressure to the nth degree but it will probably register most of that pressure rise in much less time than a ms. Lets say 200 psi is applied instantaneously to your ms delay sensor, well it may only take 1/10th ms for the sensor to register 50-100 psi of that 200 psi and that would be plenty to tell you that it fired. If you go this route use a sensor good for 300 psi or a little better so it has a safety factor.

    If you run a small tube that seals (need good contact for heat transfer) to the engine, and it runs between the cylinder wall hole and the sensor, then the actual sensor (which is likely sizeable) can sit outside the engine where it has space. You still want the sensor fairly close to the engine of course.

    FYI we have several customers that drill their cylinder walls and install one of our threaded M5 sensors, so it is not uncommon. They don't do it for the reason you are doing it for, they do it to more accurately measure cylinder pressures during the low pressure portions of the cycle. A low pressure sensor is used instead of a combustion sensor.

    I think that you can still put a hole in at 70 ATDC and get a cheap and "slow" sensor to work. The average temperature from 70 ATDC to 70 BTDC is not going to be very high, and it's a dead end hole so not much flow.

    With your 1/4"BSPT thread try to sleeve the inside hole of the sensor so that there is as little volume inside there as possible, without contacting the diaphragm in the sensor of course.
    You want to have as little transfer of gases as possible moving to and fro, makes for less heat and a quicker fill (reading).

    Clint.

    Quote Originally Posted by Frits Overmars View Post
    Excellent lecture from Clint Gray, combining theoretical knowledge and practical thinking. I especially like the simple asymmetrical signal spacing.
    Here is some more injection lecture for the Christmas holidays:

    EDIT: uploading failed because the file is too big (2793 kB). You might try to find it yourself on the internet.
    Search for WO2016193902A2 : Internal combustion engine having two fuel injectors per cylinder and control method therefor.
    Quote Originally Posted by Nath88 View Post
    I believe this is the same patent.
    http://www.freepatentsonline.com/20170159598.pdf

    Update on my thing:

    -I'm overdriving the piezo transducer, they're rated at 15V AC, I'm seeing an output of 50V. The signal diminishes over time, quickly at first, then it settles. Could market it as 'self calibrating', but the better option would be to use a smaller disk to reduce the output.

    -The signal to noise ratio is still not that great, and I'm hampered by being fixed to 1 value where the fuel table switch occurs. The ideal switch value varies by RPM.

    Attachment 333724

    -I have found that the pipe effect on overall air flow is greater when the throttle is more closed, it's about 2.5x 8000rpm 100% throttle, and about 4x at 0% throttle (by extrapolation). I guess the pipe is providing a bit more intake vacuum, overcoming some of the throttle restriction.

    -Richening the air fuel ratio to approx 14.4:1 from 15:1 has stopped the unwanted ARC/HCCI operation at light throttle.

    -The next trial will be to use a differential pressure sensor between the intake and the crankcase. I've previously found crankcase vacuum increases with pipe effect, but it's relative to the total pressure in the system, which is influenced by atmospheric pressure and throttle restriction. Accounting for both with differential pressure might give me a usable signal. KTM must be doing something like this, since they are measuring crankcase vacuum.

  11. #251
    Join Date
    18th May 2007 - 20:23
    Bike
    RG50 and 76 Suzuki GP125 Buckets
    Location
    Auckland
    Posts
    10,473
    .
    If a tuned two stroke engine does not fire, less air is drawn through the motor by the wave action of the pipe.

    This change in wave action is more noticeable on small capacity high performance engines and significant on small throttle openings < 25%

    The project is to develop a Arduino routine for switching between Alpha-N maps on a fuel injected 2T engine.
    By determining from residual cylinder pressure if the EFI 2T engine has fired or not and switches to an alternative map if the engine did not fire on the last cycle and needs a leaner map.

    Click image for larger version. 

Name:	Arduino Nano and Relay.jpg 
Views:	26 
Size:	281.2 KB 
ID:	334206

    My Aduino Nano project for switching maps between the Alpha-N map and the driver demand Lambda map on the Ecotrons EFI system.

    Click image for larger version. 

Name:	4-20mA pressure transducer..jpg 
Views:	49 
Size:	361.8 KB 
ID:	334207

    A pressure sensor is used to measure the residual cylinder pressure just before exhaust port opening.

    Click image for larger version. 

Name:	The Scope.jpg 
Views:	36 
Size:	561.8 KB 
ID:	334205

    I am using a signal generator and oscilloscope to mimic the ignition and cylinder pressure cycles so I can develop the project on my bench top.

    The plan is based on determining if there is a cylinder pressure pulse between two ignition pickup pulses.

    But of course there would be no pressure pulse if there was no ignition pickup pulse so we only need to look at the cylinder pressure pulse followed by an ignition pickup pulse.

    Click image for larger version. 

Name:	The Plan.jpg 
Views:	21 
Size:	676.9 KB 
ID:	334204

    The green shows how there is higher air flow through the motor on the following cycle if there has been combustion and a proper cylinder pressure pulse due to the wave action of the pipe but if there is no combustion then the air flow is much reduced during the following cycle.

    Below is my code, it looks to work on the test bench Ok. When the new pressure sensors arrive I will get to test the Arduino relay switch idea on the bike properly.


    // Arduino Nano map switch routine.
    // Routine for Switching between Alpha-N maps on a 2T engine.
    // By determining from residual cylinder pressure if the EFI 2T engine has fired or not.
    // Switches to an alternative map if the engine did not fire on the last cycle.
    // Because if the engine does not fire, less air is drawn through the motor by the wave action of the pipe.
    // This change in wave action is more noticeable on small high performance engines and throttle openings notable < 25%

    const int Relay = 8; // the number of the relay pin
    const byte EngIgn = 2; // the number of the ignition pulser interupt pin
    boolean EngIgnState = false; // Boolan EngIgnState
    boolean CylPressureState = false; // Boolan CylPressureState
    int CylPressure = (17); // the number of the analog cylinder pressure sensor pin

    void setup()
    {
    pinMode(LED_BUILTIN, OUTPUT); // initialize digital pin LED_BUILTIN as an output.
    pinMode(8, OUTPUT); // initialize digital pin D8 as an output.
    pinMode(2, INPUT); // initialize digital interupt pin D2 as an input.
    pinMode(17,INPUT); // initialize analog pin A3 as an input.
    analogReference(DEFAULT);
    attachInterrupt(digitalPinToInterrupt(2), IgnitionPulse, RISING); // ignition pickup pulse.
    digitalWrite(8, HIGH); // start with the relay off
    digitalWrite(LED_BUILTIN, LOW); // start with the LED off
    }

    void IgnitionPulse () {EngIgnState = true;} // on this revolution we have seen a pulse from the ignition pickup.

    void loop()
    {do // now to see if the cylinder fired last time and the pipe sucked extra air through the motor.
    // if the pipe did not suck extra air then we need to switch to a leaner map.

    {CylPressure = analogRead(17);

    if (CylPressure >= 250) // looking for cylinder pressure.
    {CylPressureState = true;}
    else
    {CylPressureState = false;}

    if (CylPressureState == true) // is there cylinder pressure after an ignition event?
    {digitalWrite(8, HIGH); // turn the relay off
    digitalWrite(LED_BUILTIN, LOW); // turn the LED off
    CylPressureState = false;}
    else
    {digitalWrite(8, LOW); // turn the relay on
    digitalWrite(LED_BUILTIN, HIGH);} // turn the LED on

    EngIgnState = false;
    }while(EngIgnState != true);
    }

  12. #252
    Join Date
    17th May 2018 - 01:50
    Bike
    Moto3
    Location
    Northern Ireland
    Posts
    7

    Looking for Ecotrons help

    I bought a small engine EFI kit and they assured me it was for a two stroke machine
    It Has 38mm TB and not CDI
    ()
    I have two versions of a trigger wheel
    12/2 and 12/1
    ()
    I have checked the ignition position and at zero all is good and I even got this motor to start and idle
    (0
    Problem is that it seems to be stuck in a four stroke cycle no matter what I do as even the end of injection angle has to be 360+200 so 540 to 560 which is the 2nd rotation of a four stroke cycle
    I have confirmed pulses with a scope and no matter what I do the ignition is only firing every 2 rotation

  13. #253
    Join Date
    18th May 2007 - 20:23
    Bike
    RG50 and 76 Suzuki GP125 Buckets
    Location
    Auckland
    Posts
    10,473
    Quote Originally Posted by Ulstermanone View Post
    I bought a small engine EFI kit and they assured me it was for a two stroke machine

    The problem is that it seems to be stuck in a four stroke cycle no matter what I do.
    In the advanced calibration options there is a field that allows you to select 1 or 4 cycle.

    There are two versions of Ecotrons software. ProCal and the latest EcoCal.

    In EcoCal go to "EFI Basic Settings" at the bottom of the page.

    Select "Variables" at the top of the page and then "Add Calibrations".

    You are looking for "Val_nRevPerCycle" which is the one that allows you to select 1 or 4 cycle.

    Have a look at the screen shots below to get an idea of the changes possible.
    Attached Thumbnails Attached Thumbnails Click image for larger version. 

Name:	Screenshot (14).png 
Views:	25 
Size:	219.5 KB 
ID:	336815   Click image for larger version. 

Name:	Screenshot (13).png 
Views:	38 
Size:	220.0 KB 
ID:	336816   Click image for larger version. 

Name:	Screenshot (12).png 
Views:	24 
Size:	217.7 KB 
ID:	336817   Click image for larger version. 

Name:	Screenshot (11).png 
Views:	53 
Size:	219.7 KB 
ID:	336818   Click image for larger version. 

Name:	Screenshot (16).png 
Views:	27 
Size:	217.7 KB 
ID:	336819   Click image for larger version. 

Name:	Screenshot (15).png 
Views:	21 
Size:	220.2 KB 
ID:	336820  

  14. #254
    Join Date
    18th May 2007 - 20:23
    Bike
    RG50 and 76 Suzuki GP125 Buckets
    Location
    Auckland
    Posts
    10,473
    Click image for larger version. 

Name:	MAPin to MAPout.jpg 
Views:	82 
Size:	865.1 KB 
ID:	336823

    My current EFI efforts have been to use an Arduino Nano to read the analogue signal from a MAP sensor in the crankcase and find the Difference between the Highest and Lowest reading and output the difference as an analogue pseudo MAP signal to the EFI (electronic fuel injection)'s CPU (central processing unit).

    The Arduino talks about “analogue” output on their card but they are fibbing a little bit because it is really PWM, pulse width modulation and not analogue at all. So not at all useful at all for use as a true analogue output but PWM is pretty good for controlling electric motors, LED’s etc.

    Click image for larger version. 

Name:	Arduino Analog PWM.png 
Views:	88 
Size:	210.1 KB 
ID:	336821

    Arduino PWM (Analog) Tutorial:- https://www.arduino.cc/en/Tutorial/PWM

    Click image for larger version. 

Name:	Arduno Nano with DAC card.jpg 
Views:	65 
Size:	791.7 KB 
ID:	336822 Click image for larger version. 

Name:	MCP4725 DAC Card.jpg 
Views:	61 
Size:	559.4 KB 
ID:	336824

    To get true analogue output with the Arduino Nano you need a DAC, digital to analogue converter like the AdaFruit MCP4725 and AdaFruits software library addon.

    Click image for larger version. 

Name:	AdaFruit MCP4725.jpg 
Views:	81 
Size:	235.4 KB 
ID:	336826

    AdaFruit MCP4725:- https://www.adafruit.com/product/935

    AdaFruit Tutorial:- https://cdn-learn.adafruit.com/downl...c-tutorial.pdf

    Click image for larger version. 

Name:	AdaFruit Libary.png 
Views:	61 
Size:	470.7 KB 
ID:	336825

    AdaFruit Library:- https://learn.adafruit.com/mcp4725-1...g-with-arduino

  15. #255
    Join Date
    5th April 2004 - 20:04
    Bike
    Exxon Valdez
    Location
    wellington
    Posts
    13,371
    What allowance for the information you wish to send, is your ECU equipped with?

    The easiest way I can imagine is an ECU that has allowance for boost.

Thread Information

Users Browsing this Thread

There are currently 1 users browsing this thread. (0 members and 1 guests)

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •