Author Topic: nice features to have "request"  (Read 180758 times)

Offline gunni

  • Hero Member
  • *****
  • Posts: 1492
  • BHP: 37
Re: nice features to have "request"
« Reply #75 on: July 20, 2013, 08:26:36 pm »
In actuality there are very few ecu´s that have 0-5v outputs.

It´s very normal to have a seperate sensors for your gauges.

You don´t need to use the MAP sensor as main load function , you can use the TPS as load.

Go to Base Setup - Alpha-N / speed density blending and set
VE table control strategy to Alpha-N
Ign+Lambda table Y axis to TPS
and TPS lookup used below this threshold to 110kpa.

This will make the ecu run solely on the TPS.

If you want and have wide resolution using your throttle you can chance the Y axis in the fuel, ignition and lambda tables to 0-100.

Offline Mads Lund

  • Jr. Member
  • **
  • Posts: 61
  • BHP: 4
    • VEMS member
Re: nice features to have "request"
« Reply #76 on: July 20, 2013, 11:11:18 pm »
Hi

I do use Map for load in the ECU, just a note on that I don't need a boost gauge.

And yes, many aftermarket ecu's don't provide analogue output's, but that don't mean that analogue out's will be a bad idea. Eq for dyno sessions, could it be relevant if you could transfer values form the ecu to the dyno for logging (eg. lambda and spark advance)

And having to add an LC1 or a VEMS round to get an analogue out for the lambda, just seems as money waste and system complication, knowing that the ecu to have plenty of unused analogue output's and it do know all the values.

:-) if the firmware was open source like megasquirt, would it not take a lot to make it :-)

Br Mads

Offline mattias

  • Global Moderator
  • Hero Member
  • *****
  • Posts: 1075
  • BHP: 41
    • Sävar Turbo Site
Re: nice features to have "request"
« Reply #77 on: July 20, 2013, 11:53:40 pm »
When needed, use the bluetooth serial port adapter and an Android tablet/phone with the Vems Display app, you can see everything the ECU sees.

 There is no real use to watch the lambda gauge at all times, it's more of a fault finding tool when something is wrong with EFI components. You can trust the wideband EGO correction to have control of your fuel mixture and make sure it hits lambda target in the load range you require.

If you want warning functions, use any of the misc outputs or warning functions to light up a LED or small light when threshold values are met, plenty of outputs for those. This is usually all you have time to look at in any vehicle  when your eyes have to be on the road.

Offline Mads Lund

  • Jr. Member
  • **
  • Posts: 61
  • BHP: 4
    • VEMS member
Re: nice features to have "request"
« Reply #78 on: July 22, 2013, 08:15:54 pm »
Hi Mattias

Sure, all the dial's is not needed in the cabin. My E60 only have fuel, speedo,tacho and economy.

But I'm building 60' and 70' british sportcars, and in that time was monotoring dials a must, so to keep these cars in original apperance do we need som signals from the ecu. Else will it be a double sensor setup all around. So flexible analogue out's will be highly apriciated in old cars.

In my setup do I have : oil temp, water temp, oil pressure, fuel level and lambda (I got a lambda instead of the original voltmeter), further warning lights from ecu, oil pressure, voltage.

So it will be very nice for older cars if that was possible, and further will it allso be very nice in order to make signals for the dyno work (again, a analogue 0-5v signal for lambda can be connected directly to the input/logging in a dyno)

:-) and it do have a show-of value as well :-)

digital Speedo output will allso be very usefull for older cars (non odb stuff)


Br Mads

Offline Sprocket

  • Hero Member
  • *****
  • Posts: 867
  • BHP: 29
Re: nice features to have "request"
« Reply #79 on: July 22, 2013, 10:27:01 pm »
In the Mini scene, those that fit EFi into what is in the main, old carb cars, the temp gauge uses the original coolant temp sender and a seperate sensor is used for the CLT on the ECU.

In my case I had a later model that had fuel injection from the factory which only had the ECU CLT sensor and the ECU then modulated the temp gauge.

Installing the VEMS, I just installed a seperate temp sender for the gauge.

It is something that is still common practice even on the most modern engine systems.

Granted the digital 'wideband' gauge looks a bit offensive in a classic car with analog gauges, but then so does all the extra wiring and fuel injectors under the bonnet :D

Offline Mads Lund

  • Jr. Member
  • **
  • Posts: 61
  • BHP: 4
    • VEMS member
Re: nice features to have "request"
« Reply #80 on: August 21, 2013, 08:18:48 pm »
I found a solution to the missing analogue output problem, that even solves another request i had (missing digital speedo putput)

My solution is to program a netduino to recieve on the AIM protocol and from that drive my instruments. When I'm done with the coding will I post a link to the solution if anybody else is interested.

Phase 1 will include:
- a pwm (0-5V) scaleable output of lambda (my main request)
- a switch selectable output of airtemp, oiltemp,water temp  on a pwm so all 3 temperatures can be shown on the original water temp gauge
- Driving light output (on when RPM>400) and light switch is on main
- a pwm control of the main beam indicator (my retrofit of led's is way too bright)
- digital output of speed (request from my frind as he did retrofit his mechanical speedo with a drive from a tacho)

Later plan
- enable bluetooth communication, both for VES/AIM and to control the netduino
- make pwm control of rear lights (combined rear and brake light (current rear lights are quite dim, so using pwm on a 21W bulp)
- PWM control for 1 speed dc fan (another classic car problem)



Offline UrSobsessed

  • Jr. Member
  • **
  • Posts: 30
  • BHP: 0
Re: nice features to have "request"
« Reply #81 on: April 25, 2014, 02:37:13 am »
Better varying altitude support for fuel AND boost(7000ft.<).  I live at 9000ft. and travel to 5000ft. and lower a lot.  Having variable high altitude support, instead of just sampling at start up would help the driveability and efficiency of the motor.  I've never been able to get the current barometric correction to work correctly.

Offline VEMS

  • Administrator
  • Sr. Member
  • *****
  • Posts: 487
  • BHP: 22
Re: nice features to have "request"
« Reply #82 on: April 25, 2014, 08:38:32 am »
Hello UrSobssed,

Dynamic barometric compensation is already available even fully configurable when using an anytrim channel for this (of course it requires hooking up a permanent barometric map sensor):

http://vems.hu/vt/help/v3/v3_engine_setup.html

A different Anytrim channel can be used to tweak boost target for barometric compensation using the same sensor.

http://vems.hu/vt/help/v3/v3_anytrim_control.html

Best regards, Dave



 

Offline UrSobsessed

  • Jr. Member
  • **
  • Posts: 30
  • BHP: 0
Re: nice features to have "request"
« Reply #83 on: April 26, 2014, 04:42:09 am »
Thanks for the info Dave.  Do you know if the Audi F96 barometric sensor can work with this?  I'd need to find the range of the sensor, but it seems that it should work as an atmospheric pressure sensor.

I'll have to look into the anytrim controls, as I don't have any experience using that feature in VEMS yet.  If you know of any example configs that use it in this way that would help a lot!

Offline VEMS

  • Administrator
  • Sr. Member
  • *****
  • Posts: 487
  • BHP: 22
Re: nice features to have "request"
« Reply #84 on: April 27, 2014, 01:13:15 pm »
Hello UrSobsessed,

Any sensor which has a linear output can be used, i suspect the Audi F96 is of this kind but make sure to lookup the translation values to verify this.

The help page: http://vems.hu/vt/help/v3/v3_engine_setup.html already has an example curve for Anytrim barometric correction (in the section: Barometric Correction with Anytrim control); Anytrim boost target correction should be equally simple to tune and setup.

Best regards, Dave




Offline UrSobsessed

  • Jr. Member
  • **
  • Posts: 30
  • BHP: 0
Re: nice features to have "request"
« Reply #85 on: April 28, 2014, 03:37:00 am »
Thanks Dave!

Offline Mads Lund

  • Jr. Member
  • **
  • Posts: 61
  • BHP: 4
    • VEMS member
Re: nice features to have "request"
« Reply #86 on: July 18, 2014, 08:25:10 pm »
I know I have mentioned it before, but I'm really hoping you'll make a more flexible analogue output, if you can alter the boost gauge output slightly, so it's possible to select any parameter for the "boost output" instead of that's it locked to MAP.

Last situation is a friend I have just make an VEMS installation fore, and would really like to have a "fuel consumtion meter" like on the BMW's. Data is readyly avaliable in the VEMS box, but impossible to get out on his dash.

Another SW issue, can you add a calculation in vems tune / log viewer that shown dspeed/dt, so it's possible to read car acceleration directly ? that will make the logviewer into a simple dyno, so it's possible to monitor the engine tourque (ofcause only realtively to the same car) but this could be quite usefull in order to monitor the effect of various setup changes. Currently I'm using drpm/dt, but after changing my diff ratio, is it not comparable to my old logs.

Offline Kamuto

  • Hero Member
  • *****
  • Posts: 814
  • BHP: 19
Re: nice features to have "request"
« Reply #87 on: July 29, 2014, 01:09:27 am »
it would be nice that when vems is configured with 8 egt channels warning light comes on when any of these channels are reading more than set value, not only channel 3-4 who is used like default egt channels.
Vems installer in Lithuania
valdas.sunokas@gmail.com

Offline gunni

  • Hero Member
  • *****
  • Posts: 1492
  • BHP: 37
Re: nice features to have "request"
« Reply #88 on: August 14, 2014, 11:09:03 pm »
I would enjoy a graphical representation of injection angle vs cam timing (cam lobes and 0-720deg scaling or -360 - 0 - 360 ) and ignition timing (don´t have enough time myself to do something in excel)

 also that injection angle would be load based, so a 3d table.
Also even without cam sync injection output vs ignition output shown, i.e where is IGN01 vs Inj1.

Also more nice to have

tacho output test function (i.e type in rpm and the tacho output does it) and adjustment curve to make sure tacho isn´t wandering as sometimes they are not accurate at high rpm´s,
4 points should be enough.

injector opening time curve is shown for both traditional and simplified options against X axis (vbatt) graphically

The option to connect the wideband resistor to a analog input for auto calibration. Fit and forget.., replace and forget again. It´s done in the gauge and most people don´t use all their analog inputs anyway.

dTPS in accel window is noted as 20mV/100ms, this should be the degrees per 100ms, if it is, it needs editing.
dTPS minimum angle/100ms to activate Accel Enrich without loosing the first row in the enrichment table

Accel Enrichment window shows a 3d representation of the two other tables in % of Req fuel/VE.
Actual documentation of what Fuel film is all about if anything and represented in a curve

Accel enrichment should also include accel enleanment in the same curve, i.e 8 rows instead of 4, 4 negative movement, 4 positive movement. I see alot of rich spikes when closing throttle would be nice to stop wasting that fuel for nothing.

Boost gauge settings, a 8 point curve based on MAP or turn into a universal output that can use "any" channel, has test function to simulate, to setup any gauge.

Antilag, launch control , wot gear shifts have graphical represenation curves or 3d map displayed

Valve control operates beyond when error is higher then "Near PWM-ing range" , last time I checked it turns off the output/s instead of staying at 100%

Analog input calibration shows resultant curves (all sensor inputs can be reviewed)

Engine can run dual out until cam sync is sorted so fires right away, I believe this is in 1.2.27 or something.

When running S50B32 or S54 or similar the Exhaust table is actually the exhaust table and intake table is intake table. Even if its just a Vemstune menu option that does nothing to the ecu functionally, but when connected a 0 or 1 tells other vemstune how to represent the tables.

closed loop lambda rpm based delay understanding, i.e the ecu knows that the current lambda is Xmilliseconds old if rpm is Y , curve needed. Would be ideal if this is done automatically in the log viewer. might be worth trying a kalmann filter on the lambda input into EGO correction so it accurately predicts the error coming up based on previous errors.
http://en.wikipedia.org/wiki/Kalman_filter

I could go on, but thats alot already.

Offline fphil

  • Sr. Member
  • ****
  • Posts: 398
  • BHP: 6
Re: nice features to have "request"
« Reply #89 on: August 15, 2014, 05:00:50 pm »
An advance ignition ramped when returning from fuel cut in order to smoothen out the recovered engine torque. ( especially when there is some play on the driveshaft ;) )

(comments)
Quote
closed loop lambda rpm based delay understanding, i.e the ecu knows that the current lambda is Xmilliseconds old if rpm is Y , curve needed.
may be difficult to automatically set since it is dependent on the wbo2 pid parameters, however one could enter the delay which has been observed on the logs.

Quote
might be worth trying a kalmann filter
A real Kalman filter is quite involved since it has to be optimal w/r to noise (which statistics is hardly known).  Most of the time one uses clones where the parameters are adjusted by trials. Even in this case it is not so easy to implement because there are in real world several modes of run which need to be recognized in order to commute the parameters of the respective model at the changes.

Better to get on first hand the best information possible from sensors, not have any delay in the computing of the closed loop command, have a fast loop and design efficient filters well tuned for the relevant band frequencies.  Vems algorithms may be eventually improved following this strategy.