22 replies to this topic

[lemons]

I've been talking to a guy who has installed a few MegaSquirts. I was asking what it might involve and what professional installation would be....

Josh,

My only concern is that you may have too high expectations for the Megasquirt.
There are people who have successfully turboed an engine and used the MS for
spark control, however these have been done with custom wire looms.  I've sold
5 MegaSquirtnEDIS boxes to Toyota owners which have installed them on their
4ag engines.

You have a couple of options, but both are non-trivial to install.  The
MegaSquirtnSpark works with distributors, but only if they can be advanced
to around 55 degrees static timing.  The other option is MegaSquirtnEDIS,
which requires a complete EDIS-4 ignition, crank wheel, and some fabrication
to mount the sensor, coils, and ignition box.

Having the car on hand, with a wiring schematic would be easier.  It would take
more time to reuse the stock wiring, than to make a new loom.

It would be less work to reuse as much of the stock wiring, but reroute it to
the relay board instead of taking apart a stock ECU.

I have gutted stock ECUs, and it takes considerable time to carefully desolder
the connector and still have a usable connector.  For proper installation, you'd
need a PC board made that would mate with the connector, since wiring the bare
pins would eventually result in fatigue and other failures.

Doing a complete install would take around 5-10 hours to properly wire.  Depending
on how the stock harness is bundled it could take more or less time.  If the stock
wires could be reused and separated easily, then it would save time.  If the wires
are tightly taped and bundled, it's easier and less time to create a complete new
harness and simply reuse the stock connectors (cut a pigtail off the stock harness).

Since you already have it fuel injected, many wires could potentially be reused.
The ignition will require an MSD, Crane, or equivelent aftermarket ignition amplifier
if the distributor is used.

Once the Megasquirt is installed, it must be tuned.  Tuning the engine is no simple
task.  At the very least I recommend that you purchase a Wideband O2 sensor and
controller for datalogging.  With this you can send the output of the WB02 controller
to the Megasquirt and log the true AFR in your datalogs.  You can then accurately
and more easily modify the fuel table, since you know exactly where the AFR is at.
Timing is even more tricky than fueling.  To properly tune your engine for full
performance, without frying pistons, will require considerable time.

I'm not discouraging you, merely trying to educate you of the realities behind such
a conversion.  In your case, you'd be looking at $1000 to $1200 for a full conversion.
That figure is based on the estimated hours at $75 an hour.  I am also available for
tuning, the hourly rate for tuning is $100.

If you did the conversion yourself, you'd probably spend 10-20 hours on the install,
but it would only involve material costs.  The absolute cheapest that you can realistically do a conversion is around $500, at little less on engines already fuel injected.

Most MS conversions run around $750 to $1500, depending on how DIY it's done.

--Perry

So far... megasquirt = pain in the arse?

Pre-Assemblied Pricing
>Megasquirt $240
>Relay Board $75
>Stimulator $65
>MS to RB cable $45, $3 per foot over 3ft

[EPcivic]

I didn't respond to your PM on the Megasquirt board since we're talking over here. I generally agree with everything Perry said. Building, installing, and operating the Megasquirt is not entry level mechanic work. It is quite involved and rather challenging. I'm a mechanical engineer, so I've spent a lot of time learning about things like this, and I still consider it to be a pretty challenging project. Actually, that's one of the main reasons I did it. It was great entertainment during the winter months.

For someone like me with far more time than money, you can do it cheaper than what he quoted, but it all depends on your skill level. I spent around $150 for the Megasquirt and stimulator. It took me around 10 hrs to build. I then spent another 4+ hrs fitting it into a factory ECU case, de-soldering the plugs, and creating a wiring adaptor. All of that is very ghetto and needs to be re-done to have hopes of lasting very long - particularly the connections to the OE plugs. Once I had done that, installation was simply swapping the ECU boxes out and re-wiring one wire to provide the tach signal. After trying to tune it with a standard O2 sensor, I decided I needed to either spend $250 on dyno time or spend it on a WBO2 setup. I decided on a used AEM wideband, which I got for $220. Oh yea, I forgot all about the laptop. Had to get one of those too. Found an old P133 toshiba for $80 on ebay. Looks like I'm pretty close to the minimum $500 price that he mentioned, and I still don't have it tuned as well as I'd like, and I haven't even started trying to control ignition with it yet. I'm still using an MSD 6BTM for ignition retard.

I'm not trying to discourage anyone from using the Megasquirt, and I'll be happy to share what I've learned and the program I have developed so far. I've found the whole thing to be a lot of fun, and I have learned a ton about both the electronics and fuel injection theory in general. For the non-geeks out there, there are simply many far easier options that can provide good results without all the extra work associated with starting with no basemap.

-Chris



-Chris

[Grant Bailey]

Chris, you must be the guy in the MSEFI gallery with the red valve cover and the water / air intercooler. I have been considering a Megasquirt for some time now. I was really excited to see that there were other CRX's like ours running them, but I didn't figure I'd actually run into any of those guys.

(digital handshake). Nice to make your acquaintance.

I am waiting on the release of UltraMegasquirt to really get started in earnest on the conversion, both from the standpoint of finances and the features list. When ignition retard, wideband 02 and wastegate control are formally on board it looks like much more of a reasonable proposition.

I was thinking about interfacing it with ignition by hooking it up through a hacked MSD6a. I was just reading the other day on msefi about how on you can set it up with a few resistors and a length of wire such that your igniter will input to the MSD via the magnetic input and then spit the signal out again via the square wave tach output. That can serve as the timing signal for the Megasquirt, which can retard or advance as necessary and then send the modified signal out to the points input on the MSD, which will fire a coil with the modified signal. Didn't know if you had heard of this hack or not.

The rest of my plan: I was going to leave the factory ecu as is to run the dashboard and trip warning lights as necessary. I had figured that the necessary GM sensors for the Megasquirt could be added on, retaining the factory Honda stuff. I would have two 02 sensors, one wideband for the Megasquirt and one narrowband NTK to keep mr. factory ecu happy, or at least use a narrowband emulator like guys use that have two and take one off. (My buddy with a Blazer that has an exhaust on it does this). Snip the injector output wires, hook them up to the Megasquirt, and there you are. The Honda ecu is still getting all the inputs it wants to make your gauges work, but it's not really controlling the injectors.

Maybe it's propaganda, but it sounded like once it is actually hooked up you can put in 5 parameters and your car will run, although kind of crappy, and from there you can use the wideband 02 to get it 90% and a day at a dyno to get it 100%. Is this a pipe dream? Based on your response, it sounds like it might be. I previously had purchased a Greddy e-manage, but once I figured out it wouldn't give me enough injector size correction to run 440cc squirters and it was going to be a royal bitch to interface it with factory ignition, I sold it off again. Megasquirt didn't seem like it was going to be that much more difficult.

Needless to say, I would be very interested to hear more details about your Megasquirt adventure.

Grant

[EPcivic]

Yep, That's my car with the AWIC. I seem to be one of only a few Honda guys using Megasquirt.

You don't need to go through that much effort. The gauges all work independant of the ECU, so there's no reason to keep it. The ignition works independantly also. I have heard of using the Honda ignitor to triger the MSD. I might try that when I get around to adding ignition control. I really like the MSD rev limiter a lot better than what's available from Megasquirt right now, so I'm sure I'll be keeping some sort of MSD box for the rev limit. There are some other ways to drive the Megasquirt ignition control directly from a magnetic trigger.

I used all the Honda temp sensors. There is a table that you use when building it where you can choose the resistor values to use other than GM sensors.

Once I had it hooked up and in the car, it fired on the first try. After a few minutes of playing with the settings, I had it idling better than before. If I could have someone drive my car on the street while I ran the laptop, it wouldn't take long at all to get it tuned pretty close using the wideband. The 'estimated' program was not too far off, and as long as you're on the rich side, the car is pretty tolerant of a wide range of mixtures. My problem is that my car isn't licensed, insured, or street legal, so I can't drive it anywhere. The easiest option I have it to trailer it 40 minutes to a local 1/4 mile dragstrip on 'test and tune' night and datalog some long 3rd gear pulls. When I get home, I can review the pulls and decide how much to adjust the VE numbers. That works well for full throttle/full boost. My biggest issue right now is the part throttle and transient throttle drivability. This is something that would be pretty easy to figure out on the street, but is difficult when racing. It's not so bad that I can't drive around it, and I've seen tons of carb'd race cars with way worse bogs or stumbles on throttle tip-in.

I think with a wideband and somewhere to drive the car and test, you should be able to get it spot on. If you tune on a dyno, you will make more power on a dyno, but that may not result in the fastest car. My car never sees 4th gear, so a nice long 4th gear dyno pull is great for making high HP numbers, but it doesn't represent how the car behaves in 2nd gear. Many people will tell you that their AFRs are significantly different on the street vs the dyno.

-Chris

[jsgprod]

After having mine partially assembled for most af last year I finally got around to finishing and installing it in February if this year. I wish I'd finished it last year.

I also cannot drive mine on the street so I had to take it to a local chassis dyno three days before a race to work on the VE map. I had it a lot closer than I would have thought possible from just fiddling with it in the garage.

I really like this thing and quite honestly, if you're willing and able to do a lot of the work yourself, you can't beat the price. I've since ordered another one for a spare and the megaview (assembling it now), for possible behind the wheel tuning.

Greg, ya need one of these, I might even share my maps with ya
Jay

[lemons]

The megasquirt seems nice but.... perhaps it's too much hassle for someone of my experance level? It would be great if i had like a few more running cars....

So, for a budget turbo project with under 10psi of boost, what's my best option? I need an opinion guys...

[Dan-O]

Please refer to my other three replies in your other thread(s).


As someone stated earlier, there is no cheap turbo project.

The SMC500 or an apex'i AFC with larger 390 or 450cc injectors is the best 'budget' fuel control combined with an MDS BTM.
I should just cut and paste this instead of typing it.


Have you checked out HomeMadeTurbo.com ?
Tons of info there.
Most people there run the AFC hack, a few run the SMC500 (it's still a new product) and the rest spent the $$ and have new adjustable ECU type fuel and spark management.
The Vortech FMU is not a popular unit on HMT.

[Grant Bailey]

Hot damn, RPR is the best thing ever!

Chris and Jay, thank you so much for your contributions. I am going to copy and save this information for use in my planning.

It would be an immense help if I could take you up on your offer, Chris, of getting a copy of your files as they are, both the baseline maps you're using and the sensor calibration file so that the Honda units work. Jay, if you're running boost I wouldn't mind a look at yours, either.

It really is a shame that Chris doesn't have a buddy to drive his car while he fixes it! Can't somebody help my man out here? I'm getting married to avoid problems like this

So, at this point do you guys feel like the Megasquirt is going to be comparable to the stock ecu as far as the quality of operation? Can it be sufficiently tuned to where it purrs like a kitten in all situations?

Incidentally, Chris, what motivated you to make the custom intake manifold? I kind of figured that under boost the possible flow restriction of the stockie manifold really wouldn't make any difference.

Grant

[jsgprod]

No boost for my car, it's N.A per SCCA rules.

I have to run a stock, unmodified, throttle body with the exception of 1" port matching. Weak, huh? Lot's of other limitations on the motor like CR and cam lift but it's actually kind of fun trying to see what kind of power you can get out of what is close to a stock motor.

So, at this point do you guys feel like the Megasquirt is going to be comparable to the stock ecu as far as the quality of operation? Can it be sufficiently tuned to where it purrs like a kitten in all situations?

I would give that an uneqivocal YES

From the moment I first got mine fired up and idling it was smoother than the Integra computer I have been running for years. Not to mention the fact that it is infinitely more adjustable.

Which will only get better when they (B&G) release the MS II they are currently working on. 12 x 12 VE mapping capabilities and some ignition control among other things.

Jay

[Greg Gauper]

Greg, ya need one of these, I might even share my maps with ya
Jay

I think that's going to be the winter project! I did a little investigation and looks pretty easy to build and pretty cost effective. And I have my wife's old laptop to use for making changes at the dyno & track (Since I got to spend money on my new engine, she got to spend money on her new laptop....gotta keep the missus happy )

For the runoffs, I might have access to a rare 'Mugen' ECU to replace the OEM ECU.....that is if I can sweet talk Bob into letting me borrow it for the week

[Grant Bailey]

Yes, the MSII. I was on the msefi site today reading up on it a bit, it sounds pretty darn good. I really would like to have the full on UltraMegasquirt, but they are being so close-lipped about a release date on that, I can only presume it's going to be a year or more before it really works. It's a shame, I don't really want to buy a wastegate controller and a wideband controller just to have those turned into an inexpensive kit a month later. At least there are some more reasonably priced ones out there now.

I have heard that Peter Gargano is something of an ass and his Australian diy-wb is to be avoided, very slow to respond also from the reports I've gotten.

The units I would really consider for the purpose are these:

Innovate Motorsports LM-1

PLX Devices M-300

Still, it's a solid $350 regardless, which is steep for my wallet. How is your AEM working out, Chris?

Grant

[lemons]

Sorry Dan-O for bugging you! I'm new to CRX's and even newer to how cars work (less then one year experance). If I have the money to do all this budget turbo project, i just want to make sure i'm doing it right the first time! A $300 mistake is very costly to someone who makes that every 1-2 weeks. So I want to be 100% sure on what i'm using before I use it. I'm still unsure in the fuel tho, SMC or Megasquirt, i'll discuss this with my dad.

But i'm taking a suggestion (by Dan-O i think) to use and MSD 6 BTM for my timing and spark, this will kill two birds with one stone. It's 335.88 from Summit Racing. Here's a link It's also got a rev-limiter built in for 7000 and 8000 RPMS. Looks like i'll need to try that Si-Dx spring combonation for stiffer springs.

And Chris, i'm going to be taking a trip to South Carolina, mabye me and my dad could stop by and check your ride out, mabye me drive and you tune?

I'm surprised how well this post got feedback. But, I promise, which ever route I take, that I will be documenting this so others can follow and improve on my somewhat crude methods. VIVA LA CRX!

[EPcivic]

Man, this thread is getting hard to keep up with. Let me see if I can answer everyone's questions:

Jay - I would definately be interested in seeing your files, particularly your acceleration enrichment bins.

Grant -

It's not a problem finding somebody to drive the car while I tune, I just can't find a place to drive it. Not being street legal sucks sometimes, and not being familiar with the area you live in makes it that much worse trying to find a free, safe, 'controlled' place to test.

My decision to make the intake manifold had more to do with the long cold winter here and having a TIG welder in the garage. It was more of a project just to do for the fun of it. I have no idea if it actually makes more power with it. It sure seems like it should, and I spent a lot of time on it making it right, but I don't have any real data.

The AEM UEGO wideband works great with the Megasquirt. It's a very small setup, with no built-in display and it has a non-linear output that isn't compatable with many ECUs, so they tend to go pretty cheap on e-bay. I got mine for $220.

Lemons -

Why don't you just buy my whole setup? I posted an ad for it a week or so ago. It would definately be a lot easier and quicker to buy it all at one time, and I'm asking less than what each part tends to sell for on e-bay. Here's a link to the ad:
TurboRex kit F/S

-Chris

[Dan-O]

It's cool.
I wasn't being pissy, it's hard to type the tone of your voice.


Chris' set-up would be best. It's been running so there would be no testing or heavy diailing in, basically bolt it on and go!!!


Justy making sure you realize that there is no budget turbo, it an oximoron, kind of like jumbo shrimp.

[rpr]

This is interesting, but I'm going a slightly different way (plus I have an easier engine ECU to deal with) in that I just ordered an EEPROM burner, and I am going to chip my ECU and add datalogging and a wideband O2. Should be interesting. The boys over at pgmfi.org seem to have their shit together pretty well.