What is mega squirt?

njansenv wrote: :) I'm really fighting the temptation to MS our early STI swapped Subie. I just KNOW there's power there waiting for me, and it would help me eliminate some of the wierd/hard-to-get Japanese only sensors..... My only hesitation is very cold weather starting. It's been 3 or 4 years since I've worked on that: has it become easier to get it running at -20? This is for a 2.0T STI engine: any reasonable starting MSQ's? Nathan

Not that I know MS much.

But ever since I was born, -20 F was -20F. So it's not going to get any easier- just need more fuel.

(for you outside of the US- close enough would be "But ever since I was born -30 C was -30C....)

alfadriver wrote:

njansenv wrote: :) I'm really fighting the temptation to MS our early STI swapped Subie. I just KNOW there's power there waiting for me, and it would help me eliminate some of the wierd/hard-to-get Japanese only sensors..... My only hesitation is very cold weather starting. It's been 3 or 4 years since I've worked on that: has it become easier to get it running at -20? This is for a 2.0T STI engine: any reasonable starting MSQ's? Nathan

Not that I know MS much. But ever since I was born, -20 F was -20F. So it's not going to get any easier- just need more fuel. (for you outside of the US- close enough would be "But ever since I was born -30 C was -30C....)

Didn't see this earlier. MS has warmup enrichment and cold advance functions that richen the mixture and advance timing based on coolant temp until you're up to operating temp. Both need setup and tuning, but the functionality is there.

^ I know, and it's had that functionality for years...but it was a challenge (some 4 years ago) to get the cold start "right". At the time, there were limited resources to help determine the best way to get it sorted. I understand that both the software and user base has evolved a LOT since then, hence the question.

In reply to njansenv:

The real secret is measurement. If you can measure your exhaust a/f from initiation of crank to fully warm, you will know quite a lot of what you need to know. Even with the best of spark, the relative amount of fuel to run a -20 engine vs. a 70F engine is pretty stunning.

Awesome tip! My last (winter) MS adventure was a fairly heavily cammed 2.4 Neon, and I remember being absolutely floored by the enrichment values. While it never "let me down" in -20ish temps, it was pretty cold blooded (and I never really got around to improving it).

So you know, besides a two week trip dedicated to start the car 3 times a day, every day at -20 (for probably 4 cars), developers also do a couple of weeks in a -20 cell. BTDT. Worst trip ever. Well, for car people. Does not compare with those guys who wait until it's -20 to start digging for gold and diamonds.

If you can get it to work pretty well, pretty quickly- you are doing a great job. Oh, and cranking fuel can be enough that you need repeat injections since a single injection increases the risk to wetting more than a few little ones. Good measurement with a WB sensor, and knowing what you want really shortens the time you need. Or can take advantage of that one day you have a -20 start.

The trick to those really cold starts is more fuel, even way more then you ever thought possible. My 0deg cranking value is 300% req_fuel fwiw.

WB sensor won't help you out most of the time as the sensor will be cold too.. and you don't want it warm or condensation from that first start can shock it into death.

Paul_VR6 wrote: WB sensor won't help you out most of the time as the sensor will be cold too.. and you don't want it warm or condensation from that first start can shock it into death.

yes, it will. Very worth the risk- modern sensors are very robust to liquids, and most of the ones you all can get are fine for a few test cold starts.

We do it every single start. Every second. Every temerpature. Every car.

If you don't do that, you'll never know if that hesitation is becuase you need 400% fuel or 200%

Haha. Timely discussion. I tried to start my self-tuned Haltech ECU'd Turbo II yesterday. No way. I need to add some fuel to the cold start map. It really doesn't like to start when temps dip below about 60. Needless to say, it's just a tad colder than that now.

If megasquirt had been around a tad longer at the time I got my Haltech, I'd probably be running a megasquirt. But the Haltech has been fine.

Its 7 F up in MI today and I had a bad time starting my MS II Turbo protege.... everything else is almost dialled in but still needs a bit of work here, guess I know where to start now...

I do all my cold tuning by feel/experience and it's not rocket science. I don't want to frag a $60+ sensor every time I try and set it up for someone. It even says in MANY wideband manuals to not shock a cold sensor...

Bosch said: System measures: - Never switch on sensor heating resp. control unit before engine start. - Delay of sensor heater start or power control of the sensor heater (see diagram in 1.4) as a function of engine and ambient temperature.

That's right from the LSU4.2 sensor docs from Bosch. http://wbo2.com/lsu/Y258K01005e03mar21eng.pdf

In reply to Paul_VR6:

For production and millions of cars, we do that. For testing, we don't.

I've been running WB sensors in the cold for, geez- longer than I can remember. Still haven't broken a sensor. I'm pretty sure I've run a whole lot more sensors than you have, and have run many thousands more tests.

I would not hesitate to turn on a sensor at -20 F before starting the car.

If feel is good enough for your car, that's fine. I'd rather know. that's my experience.

Seriously, share your experience, and don't be vague! How many times can you do that before there's a calibration issue? Are you doing that with std Bosch 4.2 sensors? What about the NGK ones? Does your control circuitry differ then commercial wideband offerings?

If I've been overly cautious, shame on me. It's a tough sell to a customer to possibly kill a sensor, but if it's something impossible to do, it would make things easier.

Now, i don't have megasquirt, but i haven't had a problem with killing WB sensors in cold. Literally run them down to the -20F we were talking about.

Paul_VR6 wrote: Seriously, share your experience, and don't be vague! How many times can you do that before there's a calibration issue? Are you doing that with std Bosch 4.2 sensors? What about the NGK ones? Does your control circuitry differ then commercial wideband offerings? If I've been overly cautious, shame on me. It's a tough sell to a customer to possibly kill a sensor, but if it's something impossible to do, it would make things easier.

Well, as opposed to leaving them totally off.

AFAIK, since nobody complains about re-calibrating their a/f meters, and there are many hundreds in circulation, I don't think that there is a calibration issue, particularly due to running the sensor in the cold. The risk of running liquids into the sensor does not change the calibration- the real risk is cracking the heater. And since we've done that so many times with so few breakages (I honestly have never seen one as the result of cold testing) over so many years- I really don't know the numbers over the past decade, but I'm pretty sure that it reaches well into the thousands- but I can't keep track of every single model and trip and development cycle.

And over the past two decades (almost my tenure here), we've transitioned between various Bosch and NTK sensors, all with the same response- they are very similar. Both work fine. The controllers for the bosch and NTK are not compatable with each other, so there's not much point in comparing them.

Calibration sensors and production sensors are the same, except that the tolerance range in response is allegedly tighter for the calibration sensors.

Is it impossible to break them? Of course not- that's exactly why there are the warnings that Bosch and NTK put into production. But I've never encountered that risk in my time doing the work- covering temps between -28F and 250 F, pretty much all inclusive. Well, if you want that in exhaust temp range- call it -28F to ~2000F.

But this is getting a LONG way from MegaSquirt. It just highlights my first recommendation to anyone looking at programmable ECU's- know what you want, and measure everything. If you can do those- any system is capable.

Since the topic of wideband sensor cost came up, here's a post I put on another forum showing how you can get a replacement Innovate LC1 sensor for $32. I thought it was worth sharing since it might save some of you money:

Hey guys, I recently bought a new sensor for my Innovate LC1 and was able to get it the same day for $32. I thought I would share in case any one else needs one and doesn’t want to pay $70-$80 like Innovate and DIYAutotune want. If you search you already know you can get it for $52. To Save $20 more, just order up the Bosch 17014 sensor from Advance Auto (Autozone has them too, but no coupon code). It will be $51.99 in your cart. Then apply Promo Code ES123 to knock $20 off of that. It will be $31.99 in your cart. Then you just pay tax ($2 for me) and pick it up at your nearest Advance Auto. Hope that helps someone out. It comes with the plug and is PNP for the LC1.

Alfa, thanks that was the exact sort of information I was looking for. I've never run them 'off' in the cold but would always wire them no not turn on until the fuel pump/coil power was applied (engine rpm reading).

That O2 is a good deal.... Now to find a OEM connector.......