MicroMap / old: SynchroMAP circuits

A forum for discussing the MegaSquirt related (but non-B&G) board development, assembly, installation, and testing.

Moderators: jsmcortina, muythaibxr

MicroMap / old: SynchroMAP circuits

Postby Dennis_Zx7r » Mon Nov 28, 2016 12:34 pm

EDIT:
See below (Whttp://www.msextra.com/forums/viewtopi ... 79#p493713):
No longer trying to replicate SynchroMap circuits, instead developing a microprocessor based solution.

Original first post:
-----------------------------------

I'm looking into building a SynchroMap myself (http://www.synchromap.com/) to experiment with my ITB setup. From old threads, it seems that the circuit diagrams once were public and freely shared in this forum. Unfortunately, all links I could find seem to have expired.
I actually managed to develop a simple diode based solution which seems to also do the job, but I'm not completely satisfied with the results. It seems the original SynchroMap makes use of a little more complex components and I have little doubt that it would work better.

Does anyone still have the circuit diagrams, or the BOM? Alternatively, some pictures of the board would also work.
Last edited by Dennis_Zx7r on Fri Apr 14, 2017 11:11 pm, edited 2 times in total.
My build (german, page 5+): Link
User avatar
Dennis_Zx7r
Experienced MS/Extra'er
 
Posts: 212
Joined: Thu May 26, 2016 1:25 pm
Location: Germany

Re: SynchroMAP circuits

Postby WestfieldMX5 » Tue Nov 29, 2016 6:19 am

Nothing in google cache or in archive.org?
2016 Westfield SEiW MX5 1.8 - 6sp - Torsen - MSM turbo - MS3 Pro - 3kg/hp
User avatar
WestfieldMX5
Experienced MS/Extra'er
 
Posts: 287
Joined: Tue Dec 30, 2008 9:46 am
Location: Belgium

Re: SynchroMAP circuits

Postby lutorm » Tue Nov 29, 2016 2:33 pm

Maybe ask dontz125 if he's willing to share the basic design of his QuadraMap (http://efi.ttrignition.com/quadramap.html) ?

If you want to experiment, maybe the easiest is to use the ADC on an Arduino or something?
lutorm
Helpful MS/Extra'er
 
Posts: 128
Joined: Sat Apr 11, 2015 2:53 pm
Location: The island of Hawaii

Re: SynchroMAP circuits

Postby Dennis_Zx7r » Mon Dec 05, 2016 2:43 pm

Nothing in cache, as the links were mostly of forums.

I don't expect someone selling these things to undermine his position in the market, although I of course would be grateful to build on something that has proven it works in practice. But that's why I won't ask him directly. I actually ordered one of his circuits a while ago, but it seems he's quite busy right now. Just to clarify, I don't aim to commercialise this or something like that.

Anyway, over the weekend, I developed my idea further. Unfortunately I don't have too much time right now, but I'm quite satisfied with what I got so far.

Basically, the idea is that the more or less stable 12V from the Zener diode is feeding a common wire. 12V seems to work better than 5V. A 12V regulator seems overkill as changes in the raw voltage seem to have a negligible effect on the combined signal.
Voltage drops to the level of the lowest sensor output. Diodes prevent current from the other sensors to feed the sensor with the lowest signal.
D5 is there to counter the voltage drop caused by the diode of the lowest sensor. Don't know if I really need R3-R6, but from pictures it seems the original SM had some resistors in line with the sensors.
A 5V regulator delivers supply voltage to the sensors, as I don't want to feed the cluster by loading the regulator in the MS even more and need a 12V supply anyway.
I also want to add a Schmitt trigger to generate a cam-sensor like signal from the detection of the intake phase on cylinder 1 to use sequential.

If anyone sees an obvious problem with this that I missed, I'd be more than happy if you told me. I'm a mechatronics guy, not a pure electrics engineer, so my knowledge in circuit design is unfortunately somewhat limited.
Also, the resistor values and the components have not been selected yet and are just for playing around with the simulation - they just were already available in the simulation tool, and the resistor values seem somewhat reasonable. There are also no capacitors or any hardware filters yet. It's just what I currently think might be an easy and very cost-effective way to realise this.

Image
My build (german, page 5+): Link
User avatar
Dennis_Zx7r
Experienced MS/Extra'er
 
Posts: 212
Joined: Thu May 26, 2016 1:25 pm
Location: Germany

Re: SynchroMAP circuits

Postby Paul_VR6 » Thu Dec 08, 2016 10:25 am

I have one somewhere that I used for testing. Once we got the ability to choose the map sampling window the need (for me) went away.
-Paul
1995 GTI VR6 MS3/V3
kptuned.com - Megasquirting the World! Megasquirt Sales, Service, Tuning and More!
Image
User avatar
Paul_VR6
Super MS/Extra'er
 
Posts: 876
Joined: Wed Apr 15, 2009 11:31 am
Location: Oxford, PA

Re: SynchroMAP circuits

Postby Dennis_Zx7r » Thu Dec 08, 2016 12:52 pm

But MAP sampling means the value is only taken once per revolution of the crankshaft, or every 720°.
If I understand it correctly, it therefore should have disadvantages in transients at lower RPM due to the lower frequency compared with averaging, where MAP is taken multiple times every 180° (oversampling).
Correct me if I'm wrong, but to me this means that oversampling with a SynroMap should provide a more accurate signal.

If you can find it, could you take some detailed pictures so I can try to reverse engineer it, please?
My build (german, page 5+): Link
User avatar
Dennis_Zx7r
Experienced MS/Extra'er
 
Posts: 212
Joined: Thu May 26, 2016 1:25 pm
Location: Germany

Re: SynchroMAP circuits

Postby Paul_VR6 » Fri Dec 09, 2016 10:09 am

I will look, but check here for more info:

http://www.synchromap.com/
-Paul
1995 GTI VR6 MS3/V3
kptuned.com - Megasquirting the World! Megasquirt Sales, Service, Tuning and More!
Image
User avatar
Paul_VR6
Super MS/Extra'er
 
Posts: 876
Joined: Wed Apr 15, 2009 11:31 am
Location: Oxford, PA

Re: SynchroMAP circuits

Postby lutorm » Sat Dec 10, 2016 1:54 pm

Dennis_Zx7r wrote:But MAP sampling means the value is only taken once per revolution of the crankshaft, or every 720°.


It was my understanding that the MAP sampling happens for each ignition event. The sampling time is relative to the cylinder being sampled.

Also, in your circuit, won't the voltage drop across the diodes mess up your readings?
lutorm
Helpful MS/Extra'er
 
Posts: 128
Joined: Sat Apr 11, 2015 2:53 pm
Location: The island of Hawaii

Re: SynchroMAP circuits

Postby Dennis_Zx7r » Sat Dec 10, 2016 2:23 pm

lutorm wrote:Also, in your circuit, won't the voltage drop across the diodes mess up your readings?

D5 compensates the drop.

lutorm wrote:
Dennis_Zx7r wrote:But MAP sampling means the value is only taken once per revolution of the crankshaft, or every 720°.

It was my understanding that the MAP sampling happens for each ignition event. The sampling time is relative to the cylinder being sampled.

Ok, I just recognized what I said was something different than I wanted to say. In my thoughts, I was referring to my Plan B, which was a single Map sensor on a single inlet port. In this case, there would be only 1 usable event per cycle.

What I want to experiment with, or rather what I hope comes out, would be something like this:
Image
From my current understanding, I expect or rather hope to get something like the yellow line with a SynchroMap and MapAveraging, or the red line for MapSampling, also with a Synchromap (original picture taken from synchromap.com). I plan on taking datalogs and comparing the results to see which "map measuring" method correlates better with PW(minus deadtime), all other things remaining the same.
Unfortunately only the MS3 seems to have a curve to select different MapSampling timings for different RPM, so I'll be limited to compare idle and high rpm for now with my MS2.
My build (german, page 5+): Link
User avatar
Dennis_Zx7r
Experienced MS/Extra'er
 
Posts: 212
Joined: Thu May 26, 2016 1:25 pm
Location: Germany

Re: SynchroMAP circuits / microMap experiment

Postby Dennis_Zx7r » Sun Jan 29, 2017 2:12 pm

Update, for anyone who's interested. I'll update this again when there's actually something to show.
As a background info, this is a high performance ITB engine revving up to 13000rpm, with only 25% TPS as a switchpoint to AlphaN at maximum rev. ITB mode works nice, this is not an issue and what I'm doing here is above all for the fun of it and gaining knowledge. If this works out the way I hope it might, I'll try retuning this engine on SD or Baro%.

I abandoned the circuitry. Instead, I'll use a microcontroller (and call this the microMap :lol:) to run different experimental algorithms, either blending the individual map sensors to a single signal, or outputting a signal based on a single map sensor. This also includes a digital output for use of full sequential. Based on some calculations, I expect to be able to measure at least 50 data points per cycle on maximum Rpm (10Bit ADC), for each individual sensor/cylinder. I expect this to be more than sufficient for my application. Map output will be a 12bit analog output for simplicitys sake in the beginning, but I'm already looking into putting out a Can-bus signal to avoid 2 unnecessary signal transformations. Not that I think this might be much of a problem, but it's just wrong.
Hardware (microcontroller, voltage regulator, Map sensors) is all here and tested, code's done for the most part and tested on the bench. However, this has a low priority, and I don't expect to get the bike running again before march (some major upgrades), and it'll be at least some weeks after that until I have collected some data.

The microprocessor way gives me much more flexibility than a circuit. I personally expect a blended, averaged signal from all sensors to come out as the most practical, and the individual sensors should eliminate the ITB problematic. However, I'll also try some signals based only on the "right" sensor. First, something based on switching on intersection points between the individual graphs. Second, a single minimum like map sampling does (primarily as a comparison/reference) and something based on a averaged minimum, which might result in a signal like the yellow line, drawn in the previous figure.

TLDR: Running some experimental Map-measuring stuff based on a very map-unfriendly engine with multiple sensors and a microcontroller. Expecting first results in about 2 months, will post progress here.
My build (german, page 5+): Link
User avatar
Dennis_Zx7r
Experienced MS/Extra'er
 
Posts: 212
Joined: Thu May 26, 2016 1:25 pm
Location: Germany

Re: SynchroMAP circuits

Postby timiacobucci » Mon Jan 30, 2017 11:40 pm

Thanks allot for the update. I very much look forward to seeing some of your results.
timiacobucci
MS/Extra Newbie
 
Posts: 7
Joined: Mon Jun 23, 2014 3:01 pm

Re: MicroMap / old: SynchroMAP circuits

Postby Dennis_Zx7r » Fri Apr 14, 2017 11:54 pm

A quick update as the 2 months I mentioned have already passed.

Here's the protoype. Microprocessor, DAConverter, a CAN module and of course 4 MAP sensors (NXP 6115) and the power conversion.
The DAC is there as a backup for the output signal and for testing purposes. The breadbox is a more or less standard MS3X with lightened housing, in case you're wondering.

Image

What took the most time by far was/is getting the CANBus to work. I had no prior experience with this, and the fact that the descpription of the CAN module was wrong at one important point didn't help either. Reading from the bus is working as expected for now (I use the baro value in some calculations), but I'm not able to write on the bus yet, so that I can transfer the calculated values to the MS. I expect this to be resolved when I'll continue working on this and did not have the time to look up how the MS handles this yet.
However, there's still no data yet even with the DAC, as the bike's not running due to some major modifications which I'm missing some parts for (and the time to install them :? ). I currently expect completion in about 2 weeks.

------------

Edit:
As of April the 19th, I'm able to transmit the processed MAP value via CAN. This finally eased some concerns I previously had, and even the data rate with preliminary, non-optimized code would be high enough for this to work well in practice.
https://youtu.be/oMh9L1vrWzA

However, I decided not to only transfer some resulting values calculated by the algortihms I want to test by transferring via CAN and using the comparatively slow logging on the MS.
Instead, I will try to record as much realtime data as possible directly on my microcontroller. To achieve this, I will receive the relevant engine data like RPM, PW etc. and log this in conjunction with the recorded raw MAP sensor data, for every individual cylinder. This way data evaluation is going to be much more time-consuming, but this makes it possible to try to tailor some model to the collected data, and not be limited by a preselection.
Ordered some SDCard stuff for this and hope to get this (and the freakin engine) to work in the next week. Well, at least if I blow this self-imposed deadline again I'll already have an excuse :lol:
My build (german, page 5+): Link
User avatar
Dennis_Zx7r
Experienced MS/Extra'er
 
Posts: 212
Joined: Thu May 26, 2016 1:25 pm
Location: Germany

Re: MicroMap / old: SynchroMAP circuits

Postby timiacobucci » Fri Apr 28, 2017 3:35 pm

Thanks for the update, sorry I missed it until just now. I am excited to hear about your progress and look forward to hearing about the results of the data you collect.
timiacobucci
MS/Extra Newbie
 
Posts: 7
Joined: Mon Jun 23, 2014 3:01 pm

MAP data experiment/ old: MicroMap/SynchroMAP circuits

Postby Dennis_Zx7r » Thu Aug 24, 2017 1:59 pm

Well, good news everyone 8)
I'm still alive and fiddling with stuff, and this post is not going to be another excuse why this little side project is delayed further.

Let's just say everything's running fine now, including the Cam sensor emulation and the data collection can begin.
In short as a reminder, what I'm going to try to find is a method of sampling Map on my highly SpeedDensity unfriendly engine to maybe get it to completely run on baro% instead of having to use ITB mode. So the goal is to find a sampling method linking the sampled value and actually needed fuel as good as possible.
At the moment, it's running on Alpha-N without any Map influence on engine load calculation and constant fuel pressure. Next, I will tune the tables to get the engine working reasonably well, which is necessary for the logs to make any sense. I plan on using the VE values as a reference as I don't have to adjust for DT and AFR this way, compared to using PW.
edit: Rewrote last sentence.

Screenshot of a test log:
Image
My build (german, page 5+): Link
User avatar
Dennis_Zx7r
Experienced MS/Extra'er
 
Posts: 212
Joined: Thu May 26, 2016 1:25 pm
Location: Germany

Re: MicroMap / old: SynchroMAP circuits

Postby Dennis_Zx7r » Tue Aug 29, 2017 1:41 pm

Gathered first real driving data yesterday. VE table needs more reworking than thought before or a first evaluation of the sampling methods can be done, AFRs are still way off when compared to the previous ITB setup.
Already eliminated one sampling method though as noise was very high. Check the characteristics of the curves below. You can actually see quite a difference in the WOT part in the middle of the screenshot when comparing two of the minimum-samplers to the casual map-average.
Image

However, even the synchromap-like sampling (only the lowest sensor is fed to the MS) would not enable this engine to run on SD/baro% at higher or even medium loads. 16% TPS at 10k RPM definately is not WOT... something like Map sampling seems to be needed to improve this experiment's chances of success.
Image

Edit 20170904:
I went all out and sampled a single sensor at nearly 60kHz, recording the minimum pressure encountered over a time period of 200ms (works from idle to max rpm that way) in the log. It's basically something like the MinSample the MS offers, I however doubt it's anywhere near as fast as my experimental setup.
I think the data gathered today may be another nail in the coffin for my little experiment. Have a look at this table. Even though the reading are lower overall, WOT is still indicated way too early. It's also not much of help that the (peak!) minimum pressure decreases with more throttle above the (wrongly) indicated WOT. Guess the increasing velocity of the moving gas might be responsible :?
Image
There's only one idea I have left at the moment, and that's to move the sensor feed a little towards the inlet valve. Currently it's right behind the butterflies of the ITBs, and there's already a hole in there intended to be used for balancing the carburetors originally fitted.
My build (german, page 5+): Link
User avatar
Dennis_Zx7r
Experienced MS/Extra'er
 
Posts: 212
Joined: Thu May 26, 2016 1:25 pm
Location: Germany

MAP data experiment/ old: MicroMap/SynchroMAP circuits

Postby Dennis_Zx7r » Sun Nov 26, 2017 8:40 am

Didn't really have much spare time for this since the last post.

However, here's some of my data gathered since.

Image
As a reference, the "classical" averaging usually found on ITBs. Note this is done with 4 individual sensors and delivers a little lower MAP values in idle than actually physically teeing the vacuum lines together.

Image
Experimental idea, signals of cylinders 1/4 and 2/3 are averaged, the lower signal is forwarded.

Image
Absolute minimum value encountered over 720 degree. Something like MAP sampling in the MS3, but not limited to a window and done with 60kHz.

Image
Synchromap approach with the current minimum value put through.

----
Note that on all methods with the exception of the averaging(!), the MAP value rises to a maximum and then drops again with further opening of the throttle, or with the actual demand for fuel. Note that there's no bijective relation between VE and the MAP values above this point, somewhere around VE60 on the Y-axis.
As of now, the use of AlphaN (or ITB mode) seems to be required no matter the measuring method.

So, what's the best way to get a MAP value out of this ITB motor? I compared the different approaches in resolution (kpa/deltaVE), switching point (%TPS to switch to AlphaN) and linearity.
Image
Here the switchpoint where to switch to AlphaN is depicted. Note I included a curve for the standard 4-1 tee'd vacuum lines and that it's identical to the averaged 4x sensors above around 4000RPM, respecting the measuring inaccuracies.
However, the synchromap approach or the sampling for the minimum value clearly improves the range at lower RPM.


Image
kpa/VE (higher load)
Note that the Synchromap-approach still has the best resolution in the higher loads (in the baro% part of the table) despite having a high switchpoint to alphaN.
Image
kpa/VE (lower load)
Image

----
Summary:
Overall at this time, the Synchromap-approach (forwarding the lowest value) seems to be the best compromise to me when optimizing the MAP signal on ITB motors for the typical street riding. It however only seems beneficial at lower RPM and I therefore doubt if using it on a racing vehicle is advisable regarding the added cost and risk of component failure. It sure makes low-load tuning a lot more fun though.
My build (german, page 5+): Link
User avatar
Dennis_Zx7r
Experienced MS/Extra'er
 
Posts: 212
Joined: Thu May 26, 2016 1:25 pm
Location: Germany


Return to Expansion boards

Who is online

Users browsing this forum: No registered users and 2 guests