wayne_curr

iTrader: (8)

Between the idea of fixing my pooping issues and my new exhaust goodies on their way in the mail, i've got a semi in the pants right now. All thanks to JP and spooky

AbeFM

iTrader: (3)

FYI, I NEVER have that issue on my MS-II. I don't THINK I had it on the oem coils, and I certainly never had it on the toyota COP set up. Except when I play with output test mode. :-)

JustinHoMi

I never had any problems on my MS-I... but I switched to the new one anyways.

Joe Perez

iTrader: (8)

I never had the issue on my MS1, with the stock coils or the Toyo COPs. In fact, I never bothered upgrading my own MS with this circuit- I did the initial work on one that I was building for LostSoul. All that means, however, is that you & I have perfect fuel injectors that never leak a single drop.

Someone (I think from DIY) noted that the MS2 CPU boots more quickly than the MS1, and offered that as a reason why MS2s may not require this mod. They may be faster, however I don't think it's a meaningful difference. The MS1 takes 600ms to initialize the output pins at powerup. Even if the MS2 is a hundred times faster (it isn't), that's still longer than the normal dwell period, and more than enough time for the coils to charge up and fire. The only way MS2s would be immune from this would be if the output pins of the 9S12 naturally float at +5v when the CPU is inactive, and while I don't have an MS2 CPU in-hand to test this theory on, I seriously doubt that it is the case.

Slayer

iTrader: (6)

Thanks for designing this new circuit, Joe.

I had pops at almost every cold start. Injector Rehab had originally cleaned my injectors (which included a leak test), so I emailed Keith (at IR) and asked him if he thought my injectors were leaking. He said he’d be glad to check them again anyway just to make sure. He even paid freight to have me ship them back to him. Here’s what he said-


Here is the video of the injectors not leaking.
YouTube - No Leaks
The only think I am wondering, is that if the ecu fires a wasted spark on
power up, does it ground out an injectors driver for a ms as well, maybe
on power down it does this??
There are no droplets formed, and injectors do not leak. Tested and held
at 5 bar for 2 minutes without a sign of any leaking. Not sure what to
say. I am glad we did perform the test so you can count out our service as
the root of your issue though.
Let me know if you have any questions.

Thanks
-Keith

wayne_curr

iTrader: (8)

I'm glad you did that as I cant afford the down-time to send them my injectors again to test. This is good to know. Very bizzar.

AbeFM

iTrader: (3)

+1

They could have the internal pullup active, and maybe that IS faster? 600 ms is a long time, as is 6. I'm still a fan of a weak pullup (even 20k or more) just to encourage the floating pin not to float. Pull down. Whatever.

Joe Perez

iTrader: (8)

I'm pretty sure it's not squirting at powerup- the design of the injector driver circuit would not cause the injectors to operate while the CPU was inactive. Think about it- you have to disconnect the igniter when you're reflashing to prevent it from burning up, but you don't have to disconnect the injectors to prevent them from flowing continuously.

Now, at power-down there is a real possibility of something like this taking place. At the instant that the key is turned off, a couple of the injectors will be spraying (or will have just finished doing so) and yet said fuel will not have had a chance to be drawn in, compressed, and ignited. I used to have a theory that in cars with super light flywheels, the engine didn't turn through a full cycle after poweroff (which would have the effect of clearing said fuel completely through the engine and out the exhaust) however some folks with stock-weight flywheels were having the problem, too.


I'm having a dickens of a time figuring this out from the 9S12 manual. The pins in question are on PortM (the SPI section) though it doesn't seem to say what their state is at powerup. The only thing I've found so far was in General, A.1.3.1:Not massively helpful...

Perhaps someone with an MS2 and a scope can observe these pins at powerup?

AbeFM

iTrader: (3)

Remind me next time I get a customer's in. I'm so tired of poking at MS stuff the last thing I want to do is pull out my perfectly functioning computer to mess with it. :-)

I do have a spare MS-II chip - works perfect less one GPIO pin. But getting it going would be even more work. So, I expect to be receiving one here soon, and during the process of testing it I'm sure I'll be able to get a scope on there.

Mach929

iTrader: (3)

it can be fun to test this by coasting in gear after turning the car off before coming to a stop everywhere i go

AbeFM

iTrader: (3)

Huh. I like how you think.

There's likely something to it, since it doesn't happen on cold starts, right? Only on warm or luke warm... Honestly, that fuel is good, it helps the motor start up faster, it's the spark which is annoying. go put weak pulldowns on the transistor and be done with it. Is it pull up? I forget the output polarity - maybe it'd be too much for the CPU to sink otherwise.

Joe Perez

iTrader: (8)

Back when I was running a Weber IDF carb on my '71 VW, I had to floor the gas pedal every time I wanted to turn off the engine, or it'd sit there and diesel forever. (Ok, so the car had some cooling problems...)

A similar trick would be a pretty definitive test for folks who are having the pop issue and need to determine whether they're dealing with leftover fuel from the last run or fuel dripping from the injectors after shutdown. Before switching off the key, rev the motor up to about 3,000 RPM or so. That way you're pretty much guaranteed that after the electrical system shuts down, the motor is going to continue turning though at least a couple of full cycles, which should clear all remaining fuel through the engine and out the exhaust. If it still pops after doing that, you've got leaky injectors.

As to having the injectors tested by the sexy nurse and getting a clean bill of health, I wonder if heat plays a role? Those little buggers get damn hot during operation (and likely hotter still just after shutdown, when they no longer have nice, cool gasoline flowing through them and soak up all that surplus heat from the head), and I suppose it's not impossible that this may cause them to be more predisposed to leak a bit after shutdown.


It'd be a pullup. The normal circuit is inverting, so you need to turn on the transistors by applying +5 to their base in order to crowbar the output to ground.

Try it. Maybe it'll work great. Just don't exceed the CPU's capability to sink the current in normal operation, and bear in mind that there's a 1k resistor in series between the CPU pins and the transistor bases. If you put the pullup on the transistor side of those resistors, then the damn things will always be on. I still like my circuit better. No finicky voltage division going on, just nice, clean, TTL-level signalling.

AbeFM

iTrader: (3)

My guess would be no, since "damned hot" and "hot enough to matter" might be a bit further apart than you'd think.. But it's something to look at. By dropping a match in the mani. :-) Or manually firing the plugs. There is absorbed fuel on the walls, it might take a while for that to come off.

You know, I think my pullup resistor was a different size than the one everyone uses for the typical mod. I wonder if that makes it just a bit more resistant.

1) I can't try it, I don't have any problems. :-)
2) I think a very weak pullup would work. Put an 18k from the CPU side of the resistor over. The CPU would easily handle it (7 microamps), and the CPU would be dragging it up. I think it'd work easy easy. I guess what's the minimum to activate the transistor, do you know what part they use stock?

I always upgrade them to 2n2222 variants anyway just to better handle more current.

Joe Perez

iTrader: (8)

Stock is a 2N3904, and Fairchild lists IBL as 50na, though you gotta start paying attention to the current gain characteristics at that point. This is small-signals territory.

According to the ON Semi datasheet, at IC=100ma, you're not into saturation until IB=3ma. IB for IC=30ma is only about 0.35ma for saturation, so depending on your pullup, you'll want to make sure you're at least in that region. 18k at 5v will only get you 0.28ma.





I don't see a need for this. The 2N3904 is good for IC=200ma, which would be <27Ω output pullup, much less the 100-330Ω we use in the improved design, and the 1k specified in the tradiditional mod. If 200ma igniter drive isn't enough to put your coil primaries on the fast-track to meltdown, you've got bigger problems.

The 2222, by the way, isn't much different than the 3904 in terms of IB for saturation. It's got a little more gain than the 3904 at the higher IC levels, but nothing to write home about.

AbeFM

iTrader: (3)

ooops - I'm always thinking in terms of +12V since I use the "warm up" LED to drive my radiator fan relay... Then the extra current matters.

So, 10k will work just dandy. Gives you a comfortable factor of 2, and still not at all unreasonable to expect a microcontroller to sink.
yeah, I think I'm using a 330. But I use it for the 800 mA is can flow.

ilostmymindcom

I'm wondering, are you using the setup like DIYautotune says to do it,
How to MegaSquirt your 4g63 powered DSM - Eclipse, Talon, Laser
or are you using it how the MS extra manual says,
http://www.msextra.com/manuals/MS_Ex...anual.htm#5cyl

I have mine setup as DIY says with your mods done. I don't get a pop on startup, but I do get ignition spikes and drops a lot, mostly over 3k. I don't have a scope, so I can't see if I have noise in my signal.
I have a 4g63 in my Conquest, and I'd like to be able to go faster than a 15.8

//After re-reading the topic, my question is about the 4n25 on the input, so it is irrelivent to this topic.
//"the new cap mod" may take care of my problem, since I don't have that done yet.
//Thanks for the awesome output mod!

Joe Perez

iTrader: (8)

Happy first post.

The Audi link that you posted seems to refer to the crank/cam trigger input signals, whereas what we are describing here is the output drivers for the ignition coils. So I'm not entirely sure which part of the system you are asking about.

If you are seeing RPM spikes in the log, that's completely unrelated to the spark output circuit that this thread is about. It's all about the input circuit.

The conventional circuit that most of us have used for getting the trigger inputs is the one described under "Getting the Tach Input Signal" in the DIY link that you posted, however there have been many different circuits used. Take note of the paragraph that begins "The new "cap mod..." for a simple solution that cured tach spikes for many MS1 / Miata users. Many of the MS2 / NB folks are using a completely different circuit, described here, whereas others, like kday and myself, said "**** it" to the whole concept and are using a custom crankwheel and VR sensor.

I'd prefer that we not clutter this sticky further, so I'd ask that you create a new thread if you need help with your CAS input.

wayne_curr

iTrader: (8)

Just completed this mod, and as I had previously feared, it didn't solve the main issue. Dont get me wrong, it doesn't pop at the initial turn of the key anymore, but as soon as I start cranking...it pops and fouls a plug.

I think the best solution for me at this point is going to be new injectors...unfortunately.

Edit: I should mention that this just happened on my first start after the mod. It may not happen again, but i'm 80% sure it will. It did this occasionally before the mod. I'd be really happy when I turned the key and it didn't pop but then it'd just do it when I started cranking :(

AbeFM

iTrader: (3)

fouls a plug on the first turn over? That's not gas. That's oil. Or tar. Do you smoke?

wayne_curr

iTrader: (8)

Unpossible. And no, I dont smoke with any regularity so shouldn't be tar