Hi guys!

I’m planning to upgrade my D-jet system on a 75 2.0 (2056) with a MegaSquirt system.

I’m still debating how to mount the trigger-wheel-sensor and would like to see some of your solutions.

Thank you for your input,
Dragan

My design is available here.

I bought the McMark trigger wheel. It fits perfectly and the sensor mount is easy to position. I probably could have cobbled something together but it would have cost more and not been nearly as well made.

I've got the same setup and the fit/finish is perfect. Very reasonable as well.

Quick side track, how did you guys plug the distributor hole?

-- Steve

I will go with Van's design (2nd pic from the top) for the plugging the dist. hole:

http://www.ephotomotion.com/914engine/page36.html

BTW, he has done excellent job with this webpage.

Mark, can you pls send me a close-up pic of the final sensor mount? I read you other post and was wondering if the mount shown is/was your final design.

Thanks,
Dragan

The only change in the final sensor mount is that it uses round tube, instead of square tube.

Thanks Mark! Do you have them in stock?

My newest shipment of wheels is shipping right now. So they are 'backorder' with an expected ship date of next week.

We took a new 009 distributor and designed a pickup for it. Works perfectly but very expensive as we had to machine parts to make this functional and reliable. the best thing though is that it uses the original distributor hole so there is no need to machine anything nor drop the engine
Mcmarks is a very economical solution.

What is "very expensive"?

Ya, inquiring minds want to know. I'd love to put my MS back to fuel and spark. The point pickup (or electronic) sucked for it during start.

here is a picture:Click to view attachment

we had all kinds of problems modifying both the flywheel and the fan to install the pickups and NONE were reliable. Plus there was an empty distributor hole AND the fact that to install the megasquirt pickup, the home enthusiast would have to drop the engine, or at least the transmission! What a hassle as compared to just installing a set of carbs. That is why I had my brother in law Galen, an ex GA Tech physics nut case and Porsche perfectionist, design and machine this assembly for us. As you can see by the cover it is very custom and just the one off parts a few at a time take a bunch of precise hours to make.
BUT the result is a direct drop in part that works perfectly and plugs right in, plugging the hole while using it for a modern trigger

McMark's also turns at the rate of the crank... Does MS adjust for cam input to run both fuel and spark?

I'm running something simliar to what Mark has done, only with the V6.

I am doing some MS test work now.. I ended up using an electromotive 60-2 toothed wheel and electromotive sensor to do the job.

our system combines sds electronics with cb performance hardware with our distributor

Which MS setup? Did you get your hands on the new MS-III ?

Jake is the man for all of this stuff. I am excited to have him try our distributor and see what he thinks about it. THEN maybe we can have a tete a tete going with two brilliant minds, Jake and Galen!!!!

Some advice on using spark control with a distributor pickup…

When using the point pickup for ignition control with Megasquirt it is VERY important that your "cranking speed" variable be set correctly to have decent starts, ie. set to what your engine is actually cranking at. Technically all your variables need to be set correctly but I find this one to be overlooked and it can cause lots of problems.

Just guessing at this or going off specs in a book is not good enough, you need to disconnect the ignition and crank the engine for a bit to see what Megasquirt is reading the exact RPMs at and use that for your cranking speed value. If it bounces around a bit chose a value in the middle or on the lower end of what you are seeing. You may also want to delay ignition events for a cycle or two at start to allow the engine to get up to this speed.

When using the points pickup Megasquirt does a time calculation based on engine speed to decide when to fire the spark. The distributor pickup does not give enough sensor events (or any for that matter) to calculate engine speed for itself by the time the first ignition event occurs so it needs to have a proper cranking speed variable given to it to get the timing right on startup. Getting this wrong, or just leaving set to whatever default value it had, and you wind up with an engine that has a hard time starting, wont start at all, or maybe even give yourself a flywheel with broken teeth from kickback against the starter.

Think about it, if Megasquirt thinks your engine is spinning at 200rpm when cranking and you are actually only getting 150rpm it is thinking your engine is running 25% FASTER THEN IT ACTUALLY IS! Now think about where a 25% discrepancy in your timing would put your spark.

Correctly set cranking speed makes a world of difference when starting!

Of course if you are willing to put in the extra time and money, crank sensors are the way to go.




Another issue you could be having with starting is a weak tach signal due to how the tach input circuit on your MS was set up (usually when people mod them to filter noise), but I won’t get into that here.

I had an idea for something similar to this but unfortinatly i lack the manufacturing resources at this point in time to accomplish it.

I wanted to put it where the d-jet injector triggers go though in order to pull the tach signal though a stock wiring harness.

I think i have seen an aftermarket wasted spark setup for VW's that uses a similar sensor on top of the distributor setup to the one you have come up with. I will have to see if i can locate it.

Jake working with megasquirt? Never thought i would see the day! I am glad you were able to put your past encounters with it behind you as I think the system holds a lot of potential in the right hands. Hopefully it does not bite you in the ass again.

I think you will find that Megasquirt is like engines in the sense that the quality of the final product has a lot to do with the person who built the system and how much time/effort/research they put in to doing so. Some of the setups i have seen people patch together are downright scary!

I had initially thought about asking to see if you would be intrested in testing/feedback on the megasquirt based D-jet ECU replacement project i have but didnt bother asking due to knowning your previous feelings on Megasquirt as a whole. If you are intrested though I would love to work with you on it.

Understand clearly that my use of MS is basically not by choice.. I am basically having to interface with the system to gain a first hand set of experiences with the system because it is becoming popular enough and working well enough to demand understanding.. I plan on running two separate FI systems in my 912E when I get it reassembled next month. The car has already ran 3 different systems in the past.

Trust me, I really don't want to have to screw withMS, because I still have scars from it. If I didn't interface with it I'd be just as closed minded and backward as the people who I complain about continually.

That said, I chose MSII rather than MSIII because my MAIN GRIPE with MS is that it is too equipped with bells and whistles that pull the tuner/ user away from the task at hand with the engine...

A system is no good at all if it is too complicated to interface with.. A TIV doesn't need a bunch of crazy features to be properly tuned. Keep it simple.

I have the opportunity to test another system alongside what I already have planned, I am open to working with you on it.. We have already offered a D/L jet retrofit kit in the past and could easily compare some differences.

I have to agree with Jake, I been trying to get MS1 extra too run on my engine.
If you don't use stock parts, injectors, intake manifolds, plenum chamber,etc.
You will run into problems!!! The only help I got was reading through all the manuals MS1,MSII and MSIII, the answers were there but were hard to find, if
you want a example change from MAP to TPS and see how many tables you have to change.

I agree with you completely on the bells and whistles being overkill. I am currently still using the MS1 chip for that very reason. I think the MS1 loaded with the MSExtra code already has way more features then 99% of 914/VW drivers are ever going to use and while the MS2 does increase the resolution I have yet to do any tests to see if that increased resolution makes any difference on a street car. The MS1 has been working for me so far and if it aint broke, don’t fix it. As you said, keep it simple.

The MS3 is just a scary mess.

It does have some cool features and if I had the testing facilities that you do at my disposal I would love to experiment with one and see if sequential injection and per cylinder fuel adjustments make much of a difference, but I don’t think I would ever mess with this stuff on a street car.

I will get some pictures/info together on what I have been working with and send them your way. It might be a little while though as I have been working 7 days a week lately. I am still a little ways off from having a product I want to send out to anyone as I am looking at having another PCB produced that incorporates all my changes into the board itself.

MS1 extra will let you change the 12x12 AFR1 tables to 24x12 ARF1 + AFR2 just by adding the second table and switching to them to switch at a set point.

I knew i had seen something similar

I saw this compufire DIS-IX system on a 914 once
http://www.compufire.com/vw-products-main.html

and found another similar system online
http://www.compu-tronix.com/DIS4009.htm


I am assuming these only have 4 trigger points though.

George,
How many trigger points are you running and did you gear it to run at crank speed?

We haven't seen any benefits of sequential on a street engine, even at high RPM, batch fire also generally makes more torque. Single cylinder fuel trim may not be able to increase performance overall, but they can be used to eliminate EGT delta between cylinders. These generally exist due to induction and exhaust system packaging compromises and they can kill valve jobs.

There is NOTHING worse than a system that is impossible to interface with.

I like where this discussion is heading!

Just a few comments and observation after researching the EFI market for about a year:

There are many options and “comfort” level EFI on the market. IMHO, I believe they are all very comparable with each other.
The difference is mainly the component compatibility, technical support and last but not least the price.

Most of the bolt on solutions are clearly focused for mainstream applications like domestic new and old type V8 and imports (Asian and European).

My personnel choice would be Electromotive, MegaSquirt and SDS.

Nevertheless, as of today my choice is MS.

Electromotive has a great product with their TEC 3 units and great support. I called the main number and got a real technical person that was patient enough to talk to me and answer my questions. The only con is the price – at last for me. I had the same great 30 min. experience with SDS on the phone! However, I don’t like the interface w/o a PC - as a EE I like to play with computers.

Hence, I will go with MS. I like the new MS/III options (i.e. data logging on SD card, UBS connectivity and LC touch screen display, just to mention a few) , but like with all new products there are still a few bugs with the SW. Hence, as of today, I would go with MSII. But since I’m in no rush, I will wait a few more months how MSIII develops and make my choice accordantly.

Jake, I appreciate your comments about MS and your public outing, in particular since I happened to read some of your older post about MS!

Beside of Mark’s picture (I know George you posted as well some pictures but not from a trigger wheel as the title of the post implies ) I’m still waiting to see what the member of the great club have fabricated for their trigger wheels!!! Bring it on!!!

PS: I have my rebuild engine (2056 and running stock intake) on my work bench, hence accessibility is not a issue for me.
I’m not looking for top performance but rather a good reliable daily driver. I guess this is what everybody said .

From what i have read on the theory behind it the only advantage to sequential injection would be at low RPMs and most likely only have a slight effect on emmissions and possibly efficency. The bennifit supposedly being that with sequential injection you are shooting the fuel into an open valve where as in batch injection at least half of your injectors are shooting on closed valves. This is all fine and good when engine speed is low, injector pulses are short, and the time the valve stays open is long. As the engine speeds up under load the time the valves are open gets shorter and the injector pulses get longer. Before long you are firing your injectors on valves that have not yet opened or already closed.

At what point in the engines range this occurs depends on the engines fuel needs and and the injectors used but you can do some quick math to get a rough idea.

3000 RPM is 50 RPS or one revolution every 2ms and that is for the entire rotation, depending on the cam used the time the valve stays open is going to be some fraction of that. You can look at Megasquirt to see how long your injection pulses are but in most cases an injection time of 2ms is passed well before 3000 rpm, in fact the megasquirt manual states that you want your pulse width at IDLE to be no less then 2ms, meaning even with sequential injection, the majority of the time you are still injecting on a closed valve.

I actually prefer the handheld programmer of SDS as my users don't need to have to learn software and tables before tuning their engine.. The more interface and adapataion thats necessary the more phone calls we'll get and the more mis-tuning thats possible.

I first used Electromotive in 1994, Haltech shortly thereafter and in some way I have tested or applied most every system available except the newer CB system, that even includes Motec.

When people "play" with my engines they are playing with my reputation and IF they screw it up you can bet your ass they aren't going to be accountable, it'll be my fault and that takes my time to square away.

What it all boils down to is what makes sense, what people can understand and whats simple to comprehend. I tend to like simple with most everything, right down to the optics on my rifles. Doing more with less is what I prefer.

The tables and software for both Autronic and Electromotive both make a lot of sense to me.. Thats what matters most. Electromotive is setting me up with a new, simplified ECU that doesn't have all the bells and whistles and is simpler to install as well as tune. I will be using it in parallel with MS on the 912E, or I may possibly install it on my Wife's Vert which now has Dells on the 2270 roller cam, A/C equipped engine.

I am using SDS on my double cab, that engine makes 161HP at the rear wheels and is a true beast.. All the systems have pros and cons.

About the time MS was really starting to impress me a really well built system failed in the middle of a trip acorss Africa in the hands of one of my customers that has an older engine kit. This person already drove that Bus from Amsterdam to Beijing in 2008 with stock FI without issues. This year they made a trip across africa, 18,000 miles in total and MS didn't make the grade, it failed and had to be replaced with stock FI for the remainder of the trip...

BTW- The theory you have read concerning sequential arrangements has not proven to apply to TIV engine in my direct experience back to back. The TIV has exceptional port velocity, mixture motion and chamber filling compared to many other engines. We spend more time gaining power in the lower RPM range than anywhere, because thats where the street engines really need to shine, even those that make 200 net HP..

The main point i was trying to covey is that at higher RPMs there should (in theory) not really be a difference between sequential and batch firing because at higher duty cycles the injectors are held open for most of the time either way on both systems ie. a batch system holding the injectors open 100% of the time is going to be running exactly the same as sequential system holding the injectors open 100% of the time.

The performance differences, good, bad, or none at all (in theory) should only occur at lower RPMs. It would not supprise me that to get any advantage at all from such a setup that an engine might need to be designed specifically for it.

BUT that is all just theory and we all know strange things happen in the real world which is why i would love to be able to test one properly (i love finding this sort of information) and also why i envy the resources you have to do such things.

Nothing beats real world testing.

I would also like to know more specifics if they are known on the Megasquirt failures you or anyone else have seen. I am trying to identify the weak points in the system so that i can improve upon them. I have not been able to put enough miles on mine to burn anything out yet and am interested in finding out which components are burning out first. My guess is the injector drivers are most likely to fail but i have yet to see it first hand.


As for the complexity of Megasquirt, i blame a lot of it on the tuning software that is available for it. The Megatune software is pretty poorly laid out and every new feature that has been added to it was just sort of tacked on the the existing interface without any sort of design considerations. I try to avoid using it as much as possible and usually adjust my fuel maps from datalogs using other software. There are people developing other tuning software for it, but as no one is getting paid for any of it who knows when it will be completed. The tuning software avalible for palm pilots is simplified quite a bit and its handheld however you are still dealing with a fuel map rather then an adjustment knob.

Having done hardware and software support for years I can fully understand why you would chose the more user friendly option when you are the one having to hold someones hand through the whole process.

Thats pretty cool, i never thought of using the dual table system in that way. So far 12x12 has done everything i have needed it to, but i might try messing around with a 24x12 table and see if i notice any difference. I do most of my tuning with megalogviewer though and i am not sure if it handles a dual table setup like that.

As for the increased resolution of the MS2 i was refering more to the higher clock speed/sampleing rate of the processor itself. I am not sure if it has any noticeable effect on performance or if it just results in smoother looking log files.

Here are my experiences with MS failures. I had 2 failures with wires coming loose because of solder failures at the DB 37 connectors to the Megasquirt board. Both times I was stranded. I had 1 injector driver go up in smoke on an MS1 board but that was in the garage and I hooked stuff up wrong.

And, I had the king of all failures on Jake's dyno where I could not get a tach signal for some unknown reason. When he said he has the scars from it, he's telling the truth. He got some nasty burns on his hands. He grabbed a 12 gauge wire we ran from the battery that shorted to ground on the dyno. His fast thinking and action may have saved all of us from going up as the wire ran across fuel lines. I hate to remember it, that was a really bad day. I never did look for what caused the problem, cause I never unpacked the MS I shipped back home.

That system is only as good as the builder and installer. I think the the weakness is all the soldering and the hodgepodge of used parts people use.

Since then I have run SDS. It is OK for the most part. I don't like the tuning module/ screen. I miss tuning on the PC. It is simple for sure. I had problems at altitude at Tahoe for WCR where it over compensated for elevation gain and ran lean. I had to bump the fueling by using the tuning knob. I also had starting problems that disappeared when we got back down the hill. The acceleration enrichment has been impossible to tune and I live with a bogging down at really low rpm. I'm not often down there but, on the occasional AX course it is a pain cause it forces 2 extra shifts. I would say it is a pretty good electrical carb.

Dave

I'd rather take a personal injury as burn the dyno cell to the ground.. My primary goal is mission accomplishment, after that I worry about my own safety.

Anyway, Dave's situation was one of the reasons why I never did anything with MS. We are seeing a trend where more and more people are using it and I must directly interface with it in my own car to further understand it and help people know the pros and cons of the system.

Here's my 36 tooth wheel design mounted with AC pulley to drive my 547 fan design.
Click to view attachment
Click to view attachment
Click to view attachment
Click to view attachment

I've only had two issues with my SDS, the first one was exactly the same a yours Dave.Took ages to find, but it was just the P/U point for the MAP vacuum (D-jet plenum). I moved it to a different point and the problem was gone.
BTW the guys at SDS do not like or recommend ITB's.

The other problem was programmer error.

I actually like the pendent programmer, I have it mounted to an old ashtray, with my O2 meter, giving me an almost heads up display. Once you learn how to use the fast scroll function it's OK.

The guys at SDS don't like ITBs because they have experience with them on other engines where a common plenum is more forgiving.

I have exactly the opposite experience, with our engines ITBs prove easier to tune and more forgiving than a plenum, primarily because the TIV engine has such exceptional intake characteristics.

The only engines I have had issues tuning or matching a combo to have been those with a common plenum.

I will through a few thoughts in here for the sake of anyone looking to go with Megasquirt in the future.

Relay board - If you have not done a megasquirt install before, use the relay board. It will make things a lot easier, cleaner, and it has everything already fused in case you miswire something.

On the DB-37 connector/cable - This is in my opinion probably the weakest link hardware wise on a standard Megasquirt install and for that very reason has been eliminated from the boards that I use. It is not at all suited to what it is being used for. The gauge of wire that Megasquirt should be using is much bigger then this connector was designed for. It is a connector used for data transmissions, not passing current for injector circuits. That being said, ways to make it more useable....

Don't use the solder cup style connectors, only USE ONLY CRIMP PIN CONNECTORS and buy wires that have already been machine or professionally crimped if possible. I have no idea why they show it done this way on the Megasquirt site, it’s just a bad idea. This is an automotive application and vibration will kill hard joints at the end of a wire in no time flat. To make the problem worse the size of wire you need to use is to large for the solder cup, and most people are not going to be able to get the proper heat to the metal to form a proper solder joint without melting the plastic in the process.

Which brings me to my next point...

SOLDERING
Really the only soldering you should be doing is on the board itself, as stated before automotive wiring harnesses should not have soldered joints in them. If you are going to attempt to build your own board, DON'T use your typical RadioShack 15w/30w soldering iron to do this. Get a good fully temp adjustable soldering station and multiple tips that are able to apply a lot of heat to a concentrated area fast. This makes a world of difference in terms of the ease of assembly as well as the quality of your joints. As for the joints themselves, you should be melting the solder on the pins/pads when you are soldering not on the tip of the iron. If you can’t get enough heat in the joint to melt the solder on the pins/pads then you will have a bad joint. To much heat can also be a problem for the components, which is why you need a good adjustable iron.

Hopefully anyone building a Megasquirt should already know these things, however as I have seen more then a few bad/scary Megasquirt builds I am posting all this here for the sake of anyone who doesn’t.

I bought as much as possible from DIYautotune, because thats what most of those who are wanting my input from the system plan on doing.

James, lets talk, email info@aircooledtechnology.com and have Dean set you up with me next week, please.. My engine will be ready for my engine dyno at the end of next week and I'd like to touch base prior to that.

The db-37 connector was the main reason I put off buying a ms as I also thought it was a data transmission connector. I recently got a bunch of ms stuff from a ebay seller off of a car he put back to stock to sell. I am going to change the connector but haven't picked a good replacement, it's on a back burner as the ms is mainly a learning tool for a possible universal ecu for my simpler toys. My 75 1.8 with original fi will probably be the test motor on a stand to get started but I also have a good tt suby motor thats not obd2 to also play with. My son thinks it is easier to use a suby ecu with opensource tuning and bypass the ms headaches.

I really like cop ignitions but the parts aren't cheap.

Are those screws in the wheel? Interesting solution!!

YES! 1/4 x 20, Then I cut down the heads of screws down and cut the threaded
part flush with the aluminum wheel.

Why?

It's a great idea, but I guess a lot of work for a $30-50 parts.

Which sensor are you running?

Changing the db-37 out on a stock board is probably going to be more trouble then it is worth. The actual through holes on the board where the db-37 attaches are a lot smaller then you would proabbly want to try to do anything with. If you already have an assembled megasquirt I would just leave it alone, the connector does work, it just might not have been the best design decision and you need to be carefull with it.

The boards i had made up do away with the db-37 header completly, i did not see an easy way of avoiding it with the stock boards.

It's was the only way I could do when running a AC unit and a DTM type fan. I use
a Hall Efect sensor for pick up (3 wires).

It's was the only way I could do when running a AC unit and a DTM type fan. I use
a Hall Efect sensor for pick up (3 wires).
[/quote]


ah, I see.

If I changed the connector my idea was to connect to the leads already in the holes and cut them off the back of the connector and only do the higher amp ones and leave the db-37 for the other original stuff.

It sounds like I need to leave it alone and just see what you come up with. I'm in no rush. I did check out the board and look at the solder joints with a magnyfing glass and it is a ver3. I'm a little concerned that the pot has been turned too far though as I tried to turn it a few turns gently and never felt a stop. I think I remember it may be a 10 turn pot and maybe I just didn't turn it far enough but I didn't want to push my luck until I do a little more research.

For me sequential would only be beneficial on a high reving motor with a plenum and long runners with a egt on each exhaust port. Subys are bad about port 4 I think getting lean, thats the one that has more piston failures, but 3 seems to be the hottest and thats on a watercooled motor.