Skip to main content

Alright, solved my pesky overheating problem by finding the paper towel sucked up into the cooling fan so it's time to take things apart and, ahem, fix them.

As those of you who followed my build know, I wanted the distributor space to put an AC compressor, so I installed a Speeduino ECU, crankshaft position sensor and MAP sensor to drive a 2000 Golf electronic ignition coil. It's been working great. 

However, as those of you who live on small volcanic islands where you drive from sea level to over 10,000 feet know, carbs suck at handling big differences in air pressure. So, it's time to go to phase II of the Speeduino saga and replace my trusty (up to 5000 feet above sea level) Kadrons with elevation proof fuel injection.

I looked around at the throttle body options and decided to go with individual throttle body  (ITB) replicas of the Weber IDF with the fuel injectors integrated into the base of the throttle bodies. The other tempting option was to locate the injectors into the intake manifolds, but I wasn't convinced that it would leave enough space for access to the spark plugs etc. CB Performance has some very nice pieces, but they want to sell you the base ITBs and then charge extra for the throttle position sensor, fuel rails, etc. 

I found a set of fully equipped IDF style ITBs from VW Speedshop over in Great Britain that I liked. They came with fuel rails and a throttle position sensor and had lots of options for linkage, etc. 

Paul set me up with a pair of his ITBs and I picked up a set of ITF air cleaners and linkage off of fleabay, as well a 4 regular size Accel 150117 17 lb/hr injectors. I've got Panchito heads so I ordered CB Performance intake manifolds already ported to fit the heads.

The throttle bodies from VW Speed Shop

IMG_20200628_085848IMG_20200628_085918

The throttle bodies with the fuel injectors and fuel rails installed. Note the throttle position sensor that came preinstalled.

IMG_20200628_090047

IDF air filter and manifold installed

IMG_20200628_091453

Test fitting to make sure there's room for fuel lines and injectors

IMG_20200628_091800IMG_20200628_093731IMG_20200628_100640

The  next update will cover setting up the high pressure fuel pump, regulator, fuel return and vacuum compensation. 

Cheers!

Mike

Attachments

Images (7)
  • IMG_20200628_085848
  • IMG_20200628_085918
  • IMG_20200628_090047
  • IMG_20200628_091453
  • IMG_20200628_091800
  • IMG_20200628_093731
  • IMG_20200628_100640
Original Post

Replies sorted oldest to newest

Those look nicer than the CB units(I have a set for a future project). I like the huge vacuum port in between the throats, should make for an easy vacuum tap for the MAP sensor. 

So, batch fire for injectors and wasted spark, correct?

The only way to do sequential injection and spark is with a camshaft sensor, which is difficult but possible on a type1.

I'm very interested to see how hard/easy it is to wire the injectors/pump/regulator and ESPECIALLY to program it.

As you did, I did spark first(Megajolt). I am satisfied with the way it runs though, so I think the IDFs are going to stay a while.

@DannyP posted:

So, batch fire for injectors and wasted spark, correct?

The only way to do sequential injection and spark is with a camshaft sensor, which is difficult but possible on a type1.

I'm very interested to see how hard/easy it is to wire the injectors/pump/regulator and ESPECIALLY to program it.

As you did, I did spark first(Megajolt). I am satisfied with the way it runs though, so I think the IDFs are going to stay a while.

Yep, Danny. Batch fire and wasted spark. I'm not convinced that going sequential is worth the difference for what I'm doing. 

@ALB posted:

Very cool, Mike?  Could you explain why you chose Speeduino? (and if you could keep it in layman's terms it would be very much appreciated- I'm not that knowledgeable when it comes to fuel injection)

Short answer is cost. They're very inexpensive and use the same tuning and logging software that I used on the old turbo 911 (Tunerstudio and Megalogviewer). I used a Megasquirt ECU on the turbo and really liked it, but decided to experiment with the newer, cheaper and less capable Speeduino option because the stakes were less worrisome. Plus, if I ran into a dead end at some point, it would be extremely easy to switch over to a Megasquirt unit. So far I'm very happy with Speeduino. If I wanted to bulletproof the setup, I'd go with the Megasquirt. 

Last edited by Michael Pickett
@Stan Galat posted:

I'm leaning pretty hard in this direction (EFI/crank-fire) for the 2234. Those TBs are a nice bit of kit, Michael.

Got a link?

Paul has a lot of options, but I liked this one because I thought it would have the best chance of fitting easily in our engine bays and I just like the IDF looks. I chose the 45mm throats for my 1776 and added the filter adapter conversion option.

I guess a couple of bonuses are that existing IDF linkage, etc work with them and if you decide to go crazy in the future they're easy to adapt to forced induction.

Here's the link:

https://vwspeedshop.com/produc...p;cat=395&page=1

 

@Stan Galat

https://speeduino.com/home/

US sales:

https://wtmtronics.com/

Looks like $150 for an assembled, soldered board or $100 for a board and solder it yourself. You'll need a case for either version. And possibly a board to drive your coil(s)/coilpacks.

I'm seriously going to have to consider this now, it's dirt cheap.

Edit:

I went to the wtmtronics site and it looks like another $50 will get you the case, end plates, Arduino board, and ignition driver. Wire it up, download the firmware, make a loom, and pay for the software. Still epic cheap.

Last edited by DannyP
@DannyP posted:

Stan, you MUST check out their pulley: dry sump size WITH serpentine belt, 3 different alternator pulley diameters, accommodates crank trigger, and has a LARGER bearing so it won't take a crap on you in NC LOL!

https://vwspeedshop.com/produc...p;cat=485&page=1

I think I threw up in my mouth a little.

I paid about that for a couple of custom crank trigger wheels for my rig.

What a fantastic find, @DannyP and especially @Michael Pickett.

@Sacto Mitch posted:

 

Ordinarily, reading about the scope of this project, Mike, I'd say you're going to shoot your eye out, but it looks like you've thought this through.

The only down side I can see is that you're completely missing out on all the fun of idle jets. I guess every new system has its compromises.

 

Heh, I didn't miss out on having a full collection of little brass thingies, er, I mean idle and main jets. I just got tired of changing them!

@DannyP posted:

@Stan Galat

https://speeduino.com/home/

US sales:

https://wtmtronics.com/

Looks like $150 for an assembled, soldered board or $100 for a board and solder it yourself. You'll need a case for either version. And possibly a board to drive your coil(s)/coilpacks.

I'm seriously going to have to consider this now, it's dirt cheap.

Edit:

I went to the wtmtronics site and it looks like another $50 will get you the case, end plates, Arduino board, and ignition driver. Wire it up, download the firmware, make a loom, and pay for the software. Still epic cheap.

Good points, Danny. I bought my Speeduino from Weaver Markel (WTMTronics). I went ahead and let him build my board and it was a top notch assembly job. I did buy the case and Arduino from him, too. I used the Bosch ignition coil 032905106F cross ref 0986221048 - fits 2000 Golf GTI 2.0L engine - uses LS/LT1 terminals. It uses a low amperage 5  volt trigger so the Speeduino can drive it directly (doesn't need the extra driver circuitry).

The tricky part is getting the connections to the board out of the case and into a wiring harness. I'll give an update on what I did a little later. If any of you are itching to buy a Speeduino now, I'd give Weaver a call and get his advice.

Last edited by Michael Pickett

I'm betting Speeduino could be done soup to nuts, injectors, pump, TB, regulator, hoses, sensors, wiring, electronics, coils, trigger wheel, for less than $1000. Worst case, $1500.

In my case, I bartered my CB throttle bodies, injectors, TPS, rails, pump, and regulator(About $4-500 new). I already have a trigger wheel, VR sensor, and coilpack due to my $160 Megajolt install.

I might spend another $300 to buy the brain, case, and build a harness. Add a few sensors: IAT, engine temp, maybe an idle air control valve(and a driver board).

CB's unit is easily almost $3k. Homemade is half the cost, or less.

Heh, I didn't miss out on having a full collection of little brass thingies, er, I mean idle and main jets. I just got tired of changing them!

I have a set each of 40s, 45s, and 48s DLRA Dellortos, as well as a set of 48 tri-jets. I have 4-8 of every idle jet CB makes for Dellortos, and 4 of quite nearly every main, and 6 sets of 4 air correction jets. I'm not very moderate when I get into something.

I'm ready to try EFI, I think.

Today's update: built out the fuel pump system from the gas tank to the engine bay; built both of the throttle body/manifold assemblies; tweaked the throttle bodies to work with the CB Performance manifolds; ported and gasket matched the manifolds to give clearance for the injector spray.

Here's the gas tank side. The tank feeds a brass T connected to the first gas filter and the fuel return fitting on the regulator. The first filter connects to a Powerco high pressure inline fuel pump (around $50 from Amazon). The pump feeds the pressure regulator which feeds the second gas filter. The second filter connects to an AN6 steel covered fuel line that runs to a firewall AN6 bulkhead T that provides 2 AN6 fittings in the engine bay.

Tank side

IMG_20200713_054406

Short run to brass T then filter and pumpIMG_20200713_054351

Brass T with fuel return from regulator

IMG_20200713_054453

High pressure pump

IMG_20200713_054439

Regulator

IMG_20200713_054507

2nd filter and feed to the engine bay and vacuum line from manifolds to the pressure regulatorIMG_20200713_054514

Gap between firewall and backseat. 6AN bulkhead fuel fitting and Speeduino mounted on back of the firewall. Blue hose is manifold vacuum

IMG_20200713_095909

Speeduino connections broken out into 6 pin waterproof connectors

IMG_20200713_100139

Splash guard installed underneath Speeduino

IMG_20200713_100310

GM intake air temperature sensor tied to the stake inside passenger side air filter. Wires run through 3/16" hole filled with black RTV.

IMG_20200713_111717

Gasket matched to throttle body

IMG_20200713_121000

Gasket used as template for porting manifold for injector

IMG_20200713_122428

IMG_20200713_122843IMG_20200713_122856

Tomorrow I'll build out the fuel hoses in the engine bay and finish wiring up the injectors, throttle position sensor and intake air temperature sensor.

Attachments

Images (14)
  • IMG_20200713_054406
  • IMG_20200713_054351
  • IMG_20200713_054439
  • IMG_20200713_054453
  • IMG_20200713_054507
  • IMG_20200713_054514
  • IMG_20200713_095909
  • IMG_20200713_100139
  • IMG_20200713_100310
  • IMG_20200713_111717
  • IMG_20200713_121000
  • IMG_20200713_122428
  • IMG_20200713_122843
  • IMG_20200713_122856
@Stan Galat posted:

I have a set each of 40s, 45s, and 48s DLRA Dellortos, as well as a set of 48 tri-jets. I have 4-8 of every idle jet CB makes for Dellortos, and 4 of quite nearly every main, and 6 sets of 4 air correction jets. I'm not very moderate when I get into something.

I'm ready to try EFI, I think.

Shoot, Stan, next time I need something, I'll get in touch with you. The local place I used to source parts just moved to Blackfoot, ID.  Which is actually one day further away from DellortoShopUK now.

 

Dang, I just realized I should have snagged a variety of ORings when last I saw him. I had to buy a new set of throttle plate screws when I changed my venturis.

es 

Last edited by dlearl476
@dlearl476 posted:

Dang, I just realized I should have snagged a variety of ORings when last I saw him. I had to buy a new set of throttle plate screws when I changed my venturis.

Doesn't the venturi come out the top when you change it? All you need to remove is the carb top IIRC. I did rebuild a set of Dellortos once. I don't remember removing the throttle plate or the throttle shaft.

I do remember Webers have a single set screw for each venturi.

Mike: keep the pics coming, you're teaching me how it's done!

I was thinking of installing a swirl pot though for pressure feed and return from regulator, back in the engine area. Basically use the existing fuel feed and low pressure pump to feed the swirl pot. If not that, then put a return into the top of the fuel tank(probably smarter and simpler). I've always seen the pressure regulator right at the fuel rail, shorter vacuum line that way.

I try to put together a cost list of what I've got in this project. I think Danny's in the right ballpark.

The cost will vary over time, and is dependent on who's hardware, etc. you are using.

For the complete neophytes (like me), it'd be nice to know "what", rather than "how much". What are the capabilities and limitations of Speeduino?

I'm a mechanical guy, and I think in mechanical terms, so the physical components are the first step to understanding for me. I can "see" what is happening inside an engine (with my mind), software logic is a lot harder for me. Once everything is mounted and loomed, and tidy, I'm aware that the fun's just starting - but there are tuners out there who make a living doing just this sort of thing. I may be comfortable tweaking a fuel map someday, but that days is not today. It took me forever to be able to see what was happening with jets (idle, emulsion tubes, air, and main). This stuff could end up occupying me until I'm no longer ambulatory (which is kind've the point of a hobby).

A basic system would be batch-fire and wasted spark, I assume. It'll need a trigger wheel, a crank position sensor and bracket, throttle bodies, a throttle position sensor, fuel rails, injectors, a higher pressure fuel pump, a regulator (which I need more information on), some means of return to the tank for the fuel. Ignition needs a coil pack. People get really, really detailed about stuff like Delphi wiring connectors, so there must be more to that than I think. If I'm not mistaken, that's the basics. If there's more to any of this, I'd like to know. I know there are huge gaps in my understanding of it.

Air temp sensors? Engine temp sensors? O2 sensor? Other things I'm not thinking about? This (like everything else) must have a sweet spot where you have more than "basic", but not so much that it becomes a mass of wire.

Any tips, hints, etc. are welcome - as is as much specificity as possible. You're a brilliant guy @Michael Pickett, and I'd like to sit at your feet.

@DannyP posted:

!I was thinking of installing a swirl pot though for pressure feed and return from regulator, back in the engine area. Basically use the existing fuel feed and low pressure pump to feed the swirl pot. If not that, then put a return into the top of the fuel tank(probably smarter and simpler). I've always seen the pressure regulator right at the fuel rail, shorter vacuum line that way.

When I tested the fuel system with short hoses out in the driveway, with a fire extinguisher at hand, I was shooting for a returnless setup so I only had to have one fuel line going front to back. That didn't work with my regulator and I'd read about people getting bubbles in their gas tank from the return line dumping from the top. I found the CBP installation manual online and saw that they just used a brass T right below the tank for the return. I gave it a try and it worked. That's why my regulator is up front. I figured I'd rather run a vacuum line front to back than a second fuel line.

 

@Stan Galat posted:
 

...A basic system would be batch-fire and wasted spark, ... a trigger wheel, a crank position sensor and bracket, throttle bodies, a throttle position sensor, fuel rails, injectors, a higher pressure fuel pump, a regulator ... some means of return to the tank for the fuel. ...a coil pack....

...Air temp sensors? Engine temp sensors? O2 sensor?...

...If I'm not mistaken, that's the basics...

 

 

You know, fellas, idle jets are starting to sound like not so bad a thing, after all.

 

@Stan Galat posted:
 

...This stuff could end up occupying me until I'm no longer ambulatory...

 

I think Mike went down this rabbit hole because he was no longer ambulatory up around 10,000 feet.

Around here, most roads you want to drive on are under about 6000 feet (and I'm guessing that's true in Peoria, too). There are a few passes at 7000 or 8000, but you're quickly over them and then down the other side (and past the summit, you can coast down, anyway).

My car gets a little pokey at 5000 feet. But then, so do I.

If I'm hanging around up there for more than a day or so, I might tweak the mixture-which-aren't-mixture screws a bit to balance things out again. That's about two minutes gone out of the day.

All things considered, I think I'll be taking a pass on this crank trigger throttle body fuel rail high-speed pump regulator and mapping thing.

Even though it sounds like it would be no trouble at all to get dialed in.

 

 

Stan, you've got it right. Intake air temp, engine temp(coolant or oil) or even just a factory type3 style warmup sensor in the head. O2 controller and wideband Bosch sensor. Batch fire and wasted spark is WAY easier on a type1 and is VERY workable.

All modern EFI systems adjust the fuel pressure with a regulator. It is usually hooked to manifold vacuum. High vacuum(low throttle) lowers the fuel pressure, Low vacuum(wider throttle) increases fuel pressure. The regulator keeps a specific fuel pressure at all times based on vacuum. This will most probably work well for Mike's engine as it doesn't deviate too far from stock with regards to vacuum pulses.

Stan and I are going to have a more difficult time with LARGE valve openings and big valves. At the very least we'll need to tap all 4 intake manifolds and tap them into a large vacuum accumulator. Lenny had the same pulsing vacuum problem on his Raby type4. The solution MAY be to set the regulator pressure and not use a vacuum input, or only give it a narrow range of adjustment.

Proper injector selection is important here to ensure a wide enough pulsewidth range for all running conditions(lean and rich, hot and cold, and load).

Mike, I also thought of using the existing line as a return, and running a new line to feed the fuel rail from the pump. That's easier in a Spyder than a tunnel/pan car.

@DannyP posted:

Stan, and anyone else that wants to "dive off this cliff" with EFI:

http://performancefuelsystems....rface-TechCorner.htm

LOTS of good learning out there to be had.

Thanks for that, Danny. As soon as I started reading, the light went on over my head. I've always been confused why injector pressure needed to vary at all, and wondered even more why vacuum was used to do it. Vacuum seems so... uncivilized for EFI.

As the article says (in so many words), the idea is to maintain a constant pressure in the fuel rail relative to what is occurring in the manifold. If the manifold is under a lot of vacuum, the regulator needs to supply less pressure to keep a constant relative fuel rail pressure. Atmospheric pressure is 14.7 psi absolute (absolute being a perfect 30" vacuum), therefore 2" vacuum roughly equals -1 psi (atmospheric). If the manifold vacuum is 10", then the regulator needs to back off the pressure about 5 psi in order to supply the same relative pressure to the fuel rail.

Brilliant and elegant. I learned something new today. Thanks, @DannyP.

BING!!! A carb uses changing pressures.  The injection needs a consistent pressure in the fuel rail relative to the manifold because it meters fuel flow using time (how long the injector is opened) and that's easier to do accurately when that pressure differential is constant.  

Once I got that the mapping is telling the injectors how much time to stay open (richer is longer, duh) they started to make real sense when looking at them.  It's fascinating to look at fuel maps, ignition curves and dyno charts together. Then change a value (the shape of the fuel map or Ignition curve) and see the change in the dyno chart, the change in the shape of that curve relative to the others. The first time it clicked was a real Matrix moment....like....I can see the universe man!

I got the pulse timing straight off, but the floating pressure seemed antithetical to the idea of controlling the fuel. All of the sudden, there is was - it's critical to controlling the fuel.

I'm always amazed at how smart people can be when they focus their Jedi minds on a problem, because the vast bulk of humanity generally seems to be driven along by impulse and fear.

Last edited by Stan Galat

The fuel pressure regulator on an EFI is very elegant and simple. It helps to remember that the length of the fuel pulse (i.e. time) is what's being varied by the fuel map, not the size of the nozzle as in a carb with idle and main jets and all.

And 30-45 lbs of static pressure does wonders for your atomization.

As for "capabilities," the arduino/Speeduino system seems to have a lot out of the box, manifold pressure sensitivity way more than needed in a typical NA application, and even fine for boost up to 20 lbs. Just a much broader array of potential situations than a well-tuned Weber can manage. Hence the usefulness in mountain areas.

Hence the ease of start-up.

Hence the better fuel economy.

Hence the better/smoother throttle response.

The Achilles Heel is the idle air control valve for warmup. A mechanically controlled vacuum leak machine that can and will fail. Eliminating this is much of the reason drive-by-wire (as opposed to throttle cable) is the way modern cars go. 

But you can get by without it. And VW hotrodders are all about rough cold-starts...

Given that a pair of new Weber 44s, a fuel pump and filters can easily top $1,000, an Speeduino EFI rig with ITBs at a sub-$1500 price is hard to argue against.

Stan, no problem at all.

Ed, I see you may be coming around......

Speeduino offers the option of using a stepper motor(modern way, just before drive-by-wire throttle) for IAC(idle air control). That's the way I'd go. Truly, you could just use one big throttle body, but what fun would that be?

Four good-sized hose bibs plumbed to an intake air control valve with a mini air filter on the other side. Controlled air leak indeed.

Or, like you say, don't worry about it, let it run a little rough for a few minutes.

I'm glad there's enthusiasm for an "inexpensive" approach to fuel injection and ignition. If I get it working, I'll be glad to share the problems, solutions and limitations encountered along the way. I also very much respect the "if it ain't broke" approach.

The Tunerstudio software that works with Speeduino and the Megasquirt siblings has an autotune feature. You set up a spreadsheet (RPM x Manifold Air Pressure) and fill in the target air fuel ratios you want. Then you just drive around and the software adjusts your fueling to try to achieve the A/F targets. It worked well the last time I did serious tuning (2010).

If I remember correctly, you still need to fiddle with the acceleration enrichment map to get rid of the lag and as Ed noted, the idle air table (think of it as an electronic choke). When I lived in Rhode Island, idle air was a big deal because of the wide temperature ranges. Here, I've brought the wiring out in case I need to add it, but don't think I will.

I've run into some problems with my CBP Panchito manifolds that slowed me down today. They don't fit like the cheap eBay Chinese set I used for mock up, so a lot of grinding is getting done. It proudly says USA on the side, but you'd think they'd clearance the sides so a 13mm nut could actually make a full revolution. Heck, that's behind "one flat at a time" territory.

Ok, enough complaining, I'm still having fun.

IMG_20200714_115705

I spent the morning building out the wiring harnesses for the fuel injectors, intake air temperature sensor and throttle position sensor. I've got the version of Speeduino that has a 40 pin IDC connector on the board. I bought a connector with pigtails to get it out of the case and then I broke up the connections I needed into a number of color coded 6 pin - CHEAP - waterproof connectors. That way I can easily remove the Speedy if I needed to do a swap (I've got a spare) and you can easily break out the connections for troubleshooting using another pair of connectors wired back-to-back ( more than most of you want to know - if you want details PM me). That's how I came up with the location of the spark coils vs the location of the Speedy. It turned out that the coil jammed the missing tooth crank sensing circuitry INSIDE the case, not through the cabling. I put the coil on one side of the shroud and the Speedy on the other side. Fiberglass doesn't shield interference worth a flip :-) 

Spray paint color coded

IMG_20200714_091016

Bag of 5 connectors - Amazon ZYTC 6 pin $12.20MVIMG_20200714_064445

Spartan 2 Wideband O2 controller w/sensor $125 from WTMTronics

IMG_20200714_103738

Attachments

Images (6)
  • IMG_20200714_115705
  • IMG_20200714_105253
  • IMG_20200714_091016
  • IMG_20200714_103738
  • MVIMG_20200714_064445
  • IMG_20200714_054325

You can't spin that nut because everybody figures we all use the special reduced head nuts that cost about a buck per. I've given up and just use them for everything, since nothing fits without them (headers, intakes, etc.). As for the rest - you're a regular Renaissance man, @Michael Pickett.

I can't think of a single guy here who wouldn't be excited about a low-cost EFI solution. I think it'd keep a lot of guys in air-cooled mills if it became more common-place.

Also, I picked up the inlet air temperature sensor from WTMTronics for $15 including the connector and pigtails.  

I got a deal on eBay for my fuel injectors - new in battered boxes 4 Accel 150117 for $115. 

I got a cheap regulator and a few 6AN connectors for $45 from Amazon - Bettercloud Universal EFI regulator.

I bought 20ft of 6AN stainless braided fuel hose and 10 assorted 6AN connectors from Amazon for $60. 

25 ft of vacuum hose from Amazon for $25, misc fittings were in my junk box.

From WTMTronics: Speeduino v04.x with 250 kpa MAP sensor installed and the board prebuilt was $150. I added Bluetooth for $8 and a Hammond case for $20.

Hall Effect crankshaft position sensor from DIYAutoTune for $38.

CB Performance crank trigger kit $170

@Stan Galat posted:

You can't spin that nut because everybody figures we all use the special reduced head nuts that cost about a buck per. I've given up and just use them for everything, since nothing fits without them (headers, intakes, etc.). As for the rest - you're a regular Renaissance man, @Michael Pickett.

I can't think of a single guy here who wouldn't be excited about a low-cost EFI solution. I think it'd keep a lot of guys in air-cooled mills if it became more common-place.

Back in the 90's when I was living in Denver, I had to get an insurance quote from the Kennedy tool dealer and discovered they were an Au-VeCo dealer. After running all over town looking for two of those 12mm bolts for my 912 in vain, I bought a box of 100 and a box of 100 copper exhaust crimp nuts. Ever since, I've been able to "single use" those exhaust bolts like you're supposed to and I never have to run to the hardware store looking for those unicorn bolts. 

I've  still got about 30 of the copper ones left and 50 of the others. 

Last edited by dlearl476

Yeah, about those nuts.....

I use the brass 11mm nuts for the exhaust. They can be re-used a number of times. I use steel ones for the intake manifolds, backed up by a THICK washer so the manifold doesn't get galled. Blue loctite on the intakes.....no leaks, no loosening. Also, the gaskets: are they mushy or stiff? Mushy gaskets for intake are a failure in the near future.

The crank sensor wiring is I think the only wiring where shielding is really important. Only ground the shield on the ECU end, and run the shield as close to the connector end as you can. 

Mike, keep it coming.

Also, I picked up the inlet air temperature sensor from WTMTronics for $15 including the connector and pigtails.  

I got a deal on eBay for my fuel injectors - new in battered boxes 4 Accel 150117 for $115. 

I got a cheap regulator and a few 6AN connectors for $45 from Amazon - Bettercloud Universal EFI regulator.

I bought 20ft of 6AN stainless braided fuel hose and 10 assorted 6AN connectors from Amazon for $60. 

25 ft of vacuum hose from Amazon for $25, misc fittings were in my junk box.

From WTMTronics: Speeduino v04.x with 250 kpa MAP sensor installed and the board prebuilt was $150. I added Bluetooth for $8 and a Hammond case for $20.

Hall Effect crankshaft position sensor from DIYAutoTune for $38.

CB Performance crank trigger kit $170

0-100 regulator? Where did you get the fuel pump? Anything for engine temperature?

Interesting, EFI pump for Evinrude 40-225hp motors, $35:

https://www.amazon.com/ROP-Sho...728151_t1_B07NQW1XV6

Connector assortment:

https://www.amazon.com/Waterpr...TG9nQ2xpY2s9dHJ1ZQ==

Looks like the ones Mike is using:

https://www.amazon.com/MUYI-Wa...QX2SQSB79DQQYPB1AAPE

I'd buy these, Deutsch style connector kits, many different assortments available:

https://www.amazon.com/dp/B085...TG9nQ2xpY2s9dHJ1ZQ==

A really cool bus-bar pair, especially for fiberglass cars:

https://www.amazon.com/dp/B07C...aWNrPXRydWU&th=1

P.S.: I'm totally happy to help anyone spend their time and money LOL!

Last edited by DannyP

The regulator is a 0-100 psi. Buying a little bit better regulator and pump certainly could be a good idea.

The Deutsch style connectors that Danny found are very nice. I started with just a couple of 6 pins for spark but now have 5 connectors (using 25 out of 30 connections plus a couple of spares). You could reduce the number of connectors and make it cleaner. The last time I did this, I used a pick-and-pull 50 pin behemoth and a wiring bundle the size of Marianne's wrist. I went the other way this time because I wanted it simple. Two 12 pins would work. I'm using one port/wire to turn off my AC compressor at idle. I doubt any of you would need that 25th wire.

As an old electronics tech, I personally would rather build using solder and heat shrink with pigtail connectors than to have to crimp a lot of pins to build the connectors. Just my preference.

For wires, I bought a couple of 10 ft long, 11 color 22g automotive wiring bundles $16 each. https://www.ebay.com/itm/22-aw...10-FEET/292158875171

 

I'm using a version 0.4.3 Speeduino board that uses a 40 pin IDC connector. These evidently are mostly used to make it easy to plug a converter cable that goes to an OEM wiring harness (lots of Miata users out there). Here's a pic, the IDC connector is the black rectangle on the left. IMG_20200715_075533

Doing it again, I'd probably go with the previous version (v0.3.x) that just uses screw down connections to the board. The differences otherwise are negligible.

IMG_2668-1024x701

There are other options now and coming soon, but probably worth talking to Weaver Markel at WTMTronics if you want more insight. 

Back to IDC/Dupont connectors, I bought a big 20 piece pack of female 40 pin IDC shells from eBay via China. I still have 19 of them and would be glad to send you one for free if you need it. If you want to wait for a big pack: https://www.ebay.com/itm/20pcs...a:g:AdkAAOSw7XZXhGru

As I said previously, I hate building crimp connectors so I bought a bunch of precrimped pigtails that I plugged into the IDC connector attached to the Speedy and routed them through a hole drilled in the end of the case that had a rubber grommet. Here's the link to the pigtail jumper wires I used: https://www.amazon.com/gp/aw/d...s_sparkle_scm_asin_0

Like I said earlier, doing it now, I'd probably go with the V0.3.x board with the screw down connections.

Attachments

Images (2)
  • IMG_20200715_075533
  • IMG_2668-1024x701

All of the computing power for the Speedy comes from a standard Arduino CPU board. The Speeduino units plug directly into the Arduino board. Use the Arduino 2560 board to avoid heartache: 

https://wtmtronics.com/product/arduino-mega-2560/

I highly recommend using a Hall Effect crankshaft sensor rather than a Variable Reluctance (VR) sensor because of the vastly reduced susceptibility to ignition interference. If not, you'll need a small daughter board called a VR signal conditioner:

https://wtmtronics.com/product...gnal-conditioner-vr/

Last edited by Michael Pickett

@Stan Galat, for coolant temperature, I use either the VDO 323064 oil relief sensor or the sump plate oil drain VDO 323055 sensor. They seem to be equivalent although you'll get slightly different readings depending on your oiling setup. That's ok, you'll just be using the temperature reading to compensate fueling and that's just a relative relationship, not an absolute thing.

Last edited by Michael Pickett

Very good info, Mike. I was thinking the same thing, go with version 3, buy pigtails and tin the wire ends under the screw terminals.

But version 4 and a IDE connector from you would work too. I did find those 6 pin Amazon connector pigtails you used.

If I do this, I'll go VR pickup and signal conditioner, as the wheel is already on my crank pulley and the VR sender is there already.

I was going to use a temp sender in the head, as that more closely follows the transition from cold to warm in my engine. It takes a while to warm the oil, only a few minutes for the heads. There has to be a reason the VW engineers put the temp sender in the heads on type3.

0_IMG_20200401_085832@DannyP, sounds like a good plan. You are absolutely right about cylinder head temperature being more accurate for EFI than oil temp. If you want one or two of those IDC headers just PM me your mailing address. Be careful with the placement of your Speeduino and the location of your coil. I had to put steel between the two no matter how much shielding I had on my VR or Hall Effect crankshaft sensors.

Attachments

Images (1)
  • 0_IMG_20200401_085832
Last edited by Michael Pickett

Thanks Mike. My Megajolt is behind the passenger seat in my Spyder, that's where I'd put the Speeduino. My Ford type coilpack is mounted above the trans/engine junction, so almost 3 feet from ECU.

Is your Hall/VR cable shielded AND twisted pair, shield grounded to ECU ONLY? Is your ECU in a metal case? I used leftover Subaru VR sensor wire for my install. The ground and shield really matters from what I've read. Mine worked out-of-the-box, no troubles or interference.

@DannyP, I really struggled to get a clean crankshaft sensor signal last fall. I used to be a research electronics tech and built probes and amplifiers to measure the signals inside a single kidney cell, so I was careful to shield and ground the VR/Hall wiring properly (to no avail). I finally discovered the problem when I wrapped a wad of aluminum foil around the Speeduino and grounded it. The problem was drastically reduced. Turns out that a small bit of fiberglass does not a Faraday cage make. The end caps of the Speeduino case are plastic and the interference was leaking into the box there. My coil was mounted on the firewall then, so I moved it to the other side of the shroud from Speedy and the problem was solved (switched to resistance ignition wires along the way). This may just be a noisy coil, but I'd make sure that there's some grounded metal between the coil and Speeduino. I never had this problem with the Megasquirt units, but the old 911 was a steel tank. I don't think a 50 caliber Desert Eagle would do much more than dent it.

Making slow but steady progress on the installation. Found out that the left side air cleaner was going to collide with the rear relay/fuse box, so it had to move. IMG_20200719_165215IMG_20200719_165210

Finished building the fuel and vacuum lines, too

IMG_20200719_165142

Passenger side wiring with throttle position sensor harness, fuel injection harness and intake air temperature harness

IMG_20200719_165148IMG_20200719_165154

The throttle bodies equipped with sensors, connectors and vacuum lines ready for installation

IMG_20200719_165130

My able assistant and apprentice

IMG_20200719_165254

Attachments

Images (7)
  • IMG_20200719_165142
  • IMG_20200719_165215
  • IMG_20200719_165210
  • IMG_20200719_165154
  • IMG_20200719_165148
  • IMG_20200719_165130
  • IMG_20200719_165254

More plugging away. I'm still having clearance problems with the CBP Panchito manifolds. I bought a bunch of 12mm exhaust style nuts and there still isn't clearance to put a socket on the short side of the manifold 'Y.' 

IMG_20200720_113125

I decided to make my genuine, made in the galldurn US of A manifold more like the cheap Chinese one I got off of fleabay. Me and my trusty grinder took a smidge off of the 1/4 inch thick manifold wall to allow a socket to just barely fit. Sorry if it sounds like I'm complaining, but to me, there's no excuse for this kind of lousy design/QA on a premium product. 

IMG_20200720_113510

Ok, got that off of my chest, again, and @Stan Galat, I don't want to hear "No, we all use the even SMALLER special nuts!" 

So, now I could actually attach the manifolds and throttle bodies I could see another clearance problem. The type of injector plugs I bought a just a bit too long to clear the engine bay walls. It looks like about 1/4" needs to be shaved or clearance holes need to be cut. As you may remember, I went with the injectors in the throttle body because I was worried that putting them lower in the manifold might make it hard to change the spark plugs. Pay me now or pay me later. I'm still not sure what I'd recommend, but either one will work, both have drawbacks, but they aren't showstoppers.

IMG_20200720_115259IMG_20200720_115115

Time to get some lunch and to decide what kind of fiberglass dust I'm going to make.

Attachments

Images (4)
  • IMG_20200720_113125
  • IMG_20200720_113510
  • IMG_20200720_115259
  • IMG_20200720_115115

I don't see a big deal with those manifolds.  I clearanced mine, too, although mine might have been the CB Perf. “El Cheapo” versions.  And then I went for the super-slim nuts, too.  IIRC, mine take an 11mm wrench.  Three of them will accept a box end on the wrench Or a socket, but the fourth has to be either a Super-slim-wall 11mm socket or an open end stubby wrench (my personal favorite for turning the nut one flat at a time).  It’s part of the price for the fame of having a “racing set-up” like the PCA guys have.

Last edited by Gordon Nichols

I channeled ALB and drilled some 1.5" holes so the injector connectors wouldn't run into the sides of the engine bay. The car's .73 ounces lighter now, so there's that. Perhaps more importantly, the injectors have plenty of clearance. 

IMG_20200720_141440

Just like on the driver's side where the rear relay box was in the way of the air cleaner, I discovered that a previous decision was incompatible with the ITB plan.  I had a 50 amp circuit breaker that protected the alternator from a major short somewhere forward in the chassis. Turns out that the 'test' button was pressed with exquisite precision by the fuel rail hose on the passenger side. 

IMG_20200720_145721

Well, that won't work. How many of you have circuit breakers on your alternator anyway. I just relocated the output to the input terminal and bypassed it all. Maybe I'll relocate the breaker later.

So, after tweaking the ignition coil location slightly, I had room for the new ITF linkage. However, having never had ITFs before, I was in learning mode as I installed the cross bar, again, again and again. Three time's the indicator that it's time to take a shower, dress comfortably for the evening and pour a nice glass of whiskey. Life is good.

Thanks to all for your guidance. 

Mike

Attachments

Images (2)
  • IMG_20200720_141440
  • IMG_20200720_145721
Last edited by Michael Pickett
@DannyP posted:

I agree, the casting could be made a little smaller in the area of the flange. I do recommend a thick washer under those small nuts to spread the load and prevent galling the manifolds.

Ok, got that off of my chest, again, and @Stan Galat, I don't want to hear "No, we all use the even SMALLER special nuts!" 

Actually... I didn't post before, because you said you didn't want to hear it, but everything can be solved with different nuts. Smaller socket, no need for a separate washer to drop into the nether regions of your engine when trying to remove the manifolds when you do this again.

From the EMPI catalog:

nuts

or:

nuts 2

They can be purchased almost anywhere that sells EMPI stuff (which is to say, every single VW shop).

Sorry...

PS: The manifolds are always going to be as big as they can possibly be, because they don't cast separate manifolds for every port configuration. The manifolds for the Panchitos are the same casting as those for Super-Pros or Ultra Wedge-Ports, which get a LOT of material cut away inside the runner when port-matched. Lots of hand-porting guys end up welding material onto manifolds, so they can cut the ports even bigger. Your casting is as big as it is to reduce the need for that. It's not a flaw or a QC problem, it's making one part that works for all applications, rather than 15 for each individual port. CB doesn't do everything right, but this is not one of the top 10 "wrong" things.

Grinding a bit on the outside of the manifold so we can use 13 mm (or 11mm, as the case is) headed nuts and a 3/8" drive socket instead of aftermarket 10 mm headed nuts and a 1/4" drive socket doesn't seem (to me) like an excessive amount of modification, especially for a guy building an entire EFI/crank-fire system from scratch.

I know it's frustrating, but there is a reason for it.

Attachments

Images (2)
  • nuts
  • nuts 2
Last edited by Stan Galat
@Stan Galat posted:

Grinding a bit on the outside of the manifold so we can use 13 mm (or 11mm, as the case is) headed nuts and a 3/8" drive socket instead of aftermarket 10 mm headed nuts and a 1/4" drive socket doesn't seem (to me) like an excessive amount of modification, especially for a guy building an entire EFI/crank-fire system from scratch.

I know it's frustrating, but there is a reason for it.

Yeah, I know, I just needed to vent. CB Performance is one of my vendors of choice. I just got grumpy because both the EMPI Kadron manifolds and the cheap Chinese manifolds had the clearancing done already. I was tired of having to redo things and special order nuts felt like a bridge too far. Still, if I were running a business and selling a product like this, I'd probably throw a few of the nuts that would fit them into the box just as a customer service. Wait, I I may not be over this issue yet.

Yeah, I know, I just needed to vent. CB Performance is one of my vendors of choice. I just got grumpy because both the EMPI Kadron manifolds and the cheap Chinese manifolds had the clearancing done already. I was tired of having to redo things and special order nuts felt like a bridge too far. Still, if I were running a business and selling a product like this, I'd probably throw a few of the nuts that would fit them into the box just as a customer service. Wait, I I may not be over this issue yet.

I get it. But look at it like this: you could always have Ed's problems, or Carey's since working on Anand's Spyder.

It's alive! After a day of chasing down rat-holes because I couldn't get any spark, I discovered that a setting in the fuel injection area (paired, semi-sequential, sequential) somehow was blocking the reading of the crankshaft position sensor (or something). In any event, "semi-sequential" was the correct answer and I nearly fell over when the engine immediately started up and ran nicely. I had set up the air/fuel RPM/Map table with the ratios I wanted at various engine loads and speeds and clicked on the Autotune button. It immediately began to change the fuel table cells to match to the target air/fuel. Magic! While I've got the car in the air I'm going to tweak some air conditioning condenser positioning and then it's on the road again. We're planning a "staycation" up on the side of Haleakala near the Kula Lodge. I predict some high elevation motoring!

An important fuel injection note: When the Speeduino is configured in Sequential or Semi-Sequential Injection mode, it fires the injectors in the order noted on the circuit board (1,2,3,4). Since my engine's firing order is 1,4,3,2 I had to swap the wires on the #2 and #4 injectors to get the injectors firing in sync with the engine firing. In my case, I did the swap at the 40 pin IDC connector on the Speeduino board (I'd added the extra sensors and fuel injector wiring into stubbed out harnesses a few months ago when I had the engine out). 

IMG_20200721_132115

It's a miracle!

 

Attachments

Images (1)
  • IMG_20200721_132115
Videos (1)
Fuel Injection 720

So, a couple questions for you, Mike.

What coil pack is that from/fitment? Is it a smart or dumb coil? Speeduino interfaces easier with smart coils, that way you don't need to buy and wire a coil driver.

You have 17 lb./hr injectors. I tried two different online calculators, and then did the math myself. I have a 172hp 2165, and come up with 28 lb./hr injectors on all 3 tries. I don't remember the rate of the ones that came with the CB throttle bodies, but I measured and they are 2 ohm injectors, so I'll swap them out for high Z or high impedance(12-17 ohm) injectors at the proper rate. Speeduino likes the high Z injectors. Looks like $50-65 each without resorting to Chinese crappy ones(8 for $50 and a seriously high failure rate)!

Can you tell I'm really spending some time pondering this? Love your success BTW, Mike.

Last edited by DannyP
@DannyP posted:

So, a couple questions for you, Mike.

What coil pack is that from/fitment? Is it a smart or dumb coil? Speeduino interfaces easier with smart coils, that way you don't need to buy and wire a coil driver.

You have 17 lb./hr injectors. I tried two different online calculators, and then did the math myself. I have a 172hp 2165, and come up with 28 lb./hr injectors on all 3 tries. I don't remember the rate of the ones that came with the CB throttle bodies, but I measured and they are 2 ohm injectors, so I'll swap them out for high Z or high impedance(12-17 ohm) injectors at the proper rate. Speeduino likes the high Z injectors. Looks like $50-65 each without resorting to Chinese crappy ones(8 for $50 and a seriously high failure rate)!

Can you tell I'm really spending some time pondering this? Love your success BTW, Mike.

Thanks, Danny. Coming from a guy with your knowledge and experience, that's high praise. I'm using a Bosch ignition coil 032 905 106 cross ref 0986221048 - fits 2000 Golf GTI 2.0L engine - uses LS/LT1 terminals. I've got 2 of them, one very cheap eBay special spare and one genuine Bosch ($68 from Amazon). They are slightly different when I measured input impedance, but they seem to fire very much the same. The Bosch one is the one on the car. It is a smart coil and triggers directly off of the Speeduino board outputs without anything special. It is a wasted spark coil so it has two channels, and fires two spark plugs every time it fires. I wouldn't say it's the Porsche of ignition coils, but it's well known in the aftermarket ECU builder community and has been bench tested up to 24,000 rpm by James Cortina, one of the Megasquirt luminaries. 

I think your injector size estimates are very reasonable for your horsepower. If I were doing it again, I'd still get the VW Speed Shop ITBs, but I'd spring for the mini-injectors rather than sourcing the full size Accels I picked up off of eBay because of the tight space in the engine bay. I also totally agree with you on getting quality injectors. 

Come on in, the water's fine!

Here are the results from the first test drive with new fuel injection system installed. I drove for about 45 minutes. It ran smoothly from startup all the way through full temp. Right out of the box without autotuning it was a major improvement over my best efforts using the Kadrons. 

After it was warmed up, I switched on the autotune setting on Tunerstudio, which was running on my laptop and connecting to the Speeduino via Bluetooth. I set the autotune to be aggressive in changing the VE table cells (volumetric efficiency, the table that dictates how much fuel to squirt). Autotune looks at the AFR reported by the wideband and compares it to the AFR target table (RPM on the x-axis and MAP on the y-axis). It adjusts the VE table cells on the fly and send the new setting to the Speeduino. I went fast, slow, lugged the engine a little, downshifted, partial throttled, etc trying to hit the important engine speed and load conditions you'd normally encounter. It made lots of changes and the car runs great. I turned off autotune and finalized the VE table (and made a copy!!!). I had been logging the Tunerstudio sensor outputs and when I got home I took a look at the acceleration enrichment. It looks like the default configuration is just a little bit conservative (I could see spikes in the AFR immediately after giving it throttle) so tomorrow I'll increase the sensitivity to throttle events and smooth that part out. Even though I could see it in the logs, it was not particularly noticeable while driving.

Phew, it was a very good day and I have to say the results exceeded my expectations. I'll share settings to anyone who wants them. More as I get more road time!

Cheers!

Mike

Congratulations on your success. I'm sure my mind will explode if I ever get to the point of programming! The Autotune feature has really got me, it simplifies things so much, and makes it accessible. And with less head-scratching and wrench-throwing!

Shortly after I posted, I found that coil. I also found the pigtail/plug for that coil. And I can use the injectors I have, if I run 25 watt 7.5 ohm resistors in series to each injector. $0.89 each at Newark Electronics! This limits the current running through the Speeduino while using the low-Z injectors.

My existing coilpack is Ford dumb coil to work with Megajolt, mounted over the bellhousing and using Ford Escort wires. I'll change the coil side ends, make a new mount and that's that. 

I'm very lucky in the spacing/packaging arena. My engine has CB Space-saver intake manifolds, so the carbs/throttle bodies are reversed. Fuel rails, injectors, and idle air-bleeds on the inside. I've got a 911 shroud so that's not in the way either. Final bonus: fully-opening clamshell instead of Speedster claustrophobic compartment.

I'm still looking at about a $300 investment, plus the coil and some wirework to build a new EFI harness. I've got a spare 4 wire harness(flat tow lights) running up to the fuel tank that I can use for fuel pump power. All the rest of the wires are short, ECU is going behind passenger seat. Really, the only thing I need to run up front is the fuel return line. 

Hmmmmmmm. I can do this.

Last edited by DannyP

I agree.  And let me be the first to say that I am VERY impressed with the speed at which you made the decision to change and then pulled it off in under a month to a running example.  I realize that you had some prior experience Doing something similar with your 911, but still....

Very impressive work and something that can be duplicated by others for a reasonable amount of money.  Any chance we could get a partslist and sources for the major stuff?

Well done, Mike!

"Station Wagon"  --  So write-eth "Ed the Spyder Guy"

"Station Wagon"???????

A flash from the past there, Ed.  I think the last 'Murican "Station Wagon" disappeared in 1973 or so, along with the demise of "stations", and those delightful, but very heavy, "Station Carts" with their steel-rimmed wheels, their beautifully worn and patina-ed Maple stays and deep green paint.  

Station Wagon

But then, you could still drive your 1970 Caprise "Estate Wagon" over to Penn Station in Baltimore to collect relatives from their trips to their winter homes in Florida.  You might want to get a chauffeur outfit like Stan has, too.

Attachments

Images (1)
  • Station Wagon

"Station Wagon"  --  So write-eth "Ed the Spyder Guy"

"Station Wagon"???????

A flash from the past there, Ed.  I think the last 'Murican "Station Wagon" disappeared in 1973 or so

You wouldn't even let your chauffeur drive this:

2020 Dodge Charger Hellcat Widebody Wagon

2020 Dodge Charger Hellcat Widebody Wagon Rendered as Dodge Magnum ...

This is just an idea of what could have been; I think they stopped making them in 2008

Last edited by WNGD

Add Reply

Post Content
×
×
×
×
Link copied to your clipboard.
×
×