Joe Perez

iTrader: (8)

Well, very slightly anyway. After two evenings' worth of casual soldering and a trip to Fry's to buy a USB-RS232 adapter, my MS-I came to life and took its first tentative step into the world by telling me its battery voltage (11.81) as per Step 41.



Soon, she will Squirt!

paul

iTrader: (21)

Congrats.

Bryce

iTrader: (24)

Joe Perez just brought another life into this world!

AbeFM

iTrader: (3)

What good is a new born baby?

What are you going to name it? And what's that tiny little black thing where the MS-II belongs?

cjernigan

iTrader: (13)

Looks like a CPU with firmware ready to easily support fuel, spark, EBC, Launch, knock, WI and all other kinds of goodness. I could be wrong though.

elesjuan

iTrader: (11)

Nice power supply Joe.

reddroptop

iTrader: (4)

Iomega USB-RS232 adapter?

I paid something like 50$ for mine :(

Ben

iTrader: (33)

So it was a 2 beer project?

The_Pipefather

iTrader: (5)

Nice! I'm gonna be starting soon as I'm done with my ******' thesis. Will be putting the whole shebang inside the stock ECU case for SCCA STS2-legality.

Joe Perez

iTrader: (8)

A CPU whose operating software is written in a language I can actually understand, rather than this new-fangled "C" stuff. Programming with words? Fad. It'll never catch on.

I think it's an IOGear. No brand-name on the adapter, but it's got P/N GUC232A on the label, and when I googled it I came up with the IOGear website. Since it didn't come with a driver (open box special) I D/L'ed theirs and it works (That tablet PC is running Win2K.) So far, Megatune seems happy with it. I haven't loaded MSnS-E yet, just using the base B&G code at the moment.

Thanks. It's an old HP 6237B, rated for one whole amp at 20V and 500ma at +/- 20V. Gets the job done, though I'll need to find something beefier when it comes time for injector testing.

Beer comes in bottles. Them's Diet Coke cans. And actually they've both been sitting on the bench for quite a while as I'm too lazy to throw them into the bin three feet away.

So far, the official beers of this project have been Red Hook ESB and Road Dog Porter. Though Abe has strongly insinuated that if I come down to his place we will decant some fine homebrew, possibly to include a Milk Stout. (hint, hint)

I've been using this as a diversion from the law school applications process. Which I need to get off my *** and get done with, since the deadlines are fast approaching.

TurboTim

iTrader: (9)

I smell some electronics burning. Better go check that.

A+

cjernigan

iTrader: (13)

Eventually you're going to have to let everyone else know why you have left the heatsink empty of all those transistors the rest of us install so diligently.

AbeFM

iTrader: (3)

Probably onto something there. Not!
Man, I miss the 80s.... 'Course, even then it was all C. Anyway, at least this way if you wire it wrong and fry the chip, you won't lose anything of value.

You gotta check your hard drive power supplies. They are generally reasonably well regulated, and supply amps upon amps. Even the little one I'm using is 2amps. I could give you one of my old ones from a palmtop or something...

'Cept when Abe comes around. Then it magically comes in cans. Well, to start.

Anytime. Gotta get me some supplies though, and get that big red project out of the garage, make room for the finer things in life.


Weight Savings. I'd think it was obvious.

Joe Perez

iTrader: (8)

Because all I've got done so far is the power supply, the serial I/O, and the clock. I needed to reach a milestone before I had to put the whole thing aside and forget about it for two weeks, and I'm going on vacation tomorrow.

The only heatsunk devices I'm currently planning to omit in the finished build are Q9 and Q12 (INJ PWM) and possibly Q16 (High-current IGN) which will free up space on the heatsink for the EBC and WI FETs.

Joe Perez

iTrader: (8)

I think I've got a "regulated" 10A 13.8V Radio Shack supply somewhere. Worst-case scenario I'll just throw an old ATX supply at it.


Sorry 'bout that. Vern is a heathen- his idea of "good beer" is Bud Select, so the fridge over in that garage gets stocked with whatever is on sale. I like to keep the good stuff for myself rather than pour it on the great unwashed so normally when folks come over I just toss 'em a can of whatever's in the bar fridge. But hey, I realized the error of my ways.

AbeFM

iTrader: (3)

I felt really bad about recycling Q16 for boost control (but I hope you noticed how I did it, it really came out slick if I do say so myself) - I like knowing I can run bare coils whenever I want.

Then again, as Braineak has been quick-and-right to point out, converting to toyota coil-on-plug is cheaper than getting plain jane coils and throws a hell of a spark.

reddroptop

iTrader: (4)

If it is the real "IOGear" chipset, it is one of the best. Works with everything.

Kinda sucked forking over the $$$, but never have any issues and never will.

If it was open box unbranded, you got a good deal for sure.

Joe Perez

iTrader: (8)

You're using Q16 (a VB912) as the boost control driver? I hadn't noticed that previously. Very interesting... I'd been planning to use an IFLZ44, but your way is cleaner.

And it begs a question- if you can drive a VB912 directly off the CPU, then what's the deal with U4 needing to be there at all for the main injector drivers?

I'm curious about something- you've removed and jumpered R37 and R38, but it looks like Q14 and Q15 are still there. Looking at just the INJ1 channel for simplicity, my understanding of the schematic is that it is the voltage drop across R37 which causes Q14 to operate and turn off Q1 in the event of an over-current condition. Without the resistor, Q14 might just as well not be there.

AbeFM

iTrader: (3)

No, I'm not that clever. I thought about it, but wasn't sure how it would work - so I went with the chip that DIYA.T. sold me as part of their boost controler kit (IRLZ44Z). What I did was use the circuit on the board. There's good fat traces everywhere you need them (it's an ignition circuit) and it keeps space free on the proto area. I just jumped one "resistor" and subed a 10k for what went in the hole. Check my build thread for details, but it's pretty clean and you don't have to mount the chip sideways to clear the d-37

Oh yeah! I REALLY didn't get the directions when I did that. It said "remove R37, and jumper" or something like that, and I didn't know if it meant to jumper it or remove it. I remember reading and re-reading. This was the very last thing I did before I got the schematics (which I can never find!!!!). I meant to revisit it. I really wanted to replace those, mount them on the back of the board - but I wanted to check into what they were and what they did. They have a "front" and a "back" - are they just resistors, or are they somehow polarized? I was mounting them on the back of the board when I kept changing my mind about how they should be mounted (polar or not) and ended up snapping off the leads on one. So, I can solder to the stub, but then I figured the instructions made it sound like most people don't use them at all.

I really wanted to put them in, and I can't imagine they really need a heatsink. Anyway, without the jumper, Q1 might as well not be there. :-) So, I guess I could take out the useless Q14/15, but I'd really rather just put R37 & 35 back in. As best I can tell, they really just ARE resistors, so in the words of Space Ghost: Now what was the point of sanding my face off?

Joe Perez

iTrader: (8)

I see, you stuffed an IRLZ44 into the hole, and used the existing traces. Slick, very very slick. I see from the schematic (the one scribbled on paper in your other thread) that you figured out how to integrate This Design into the Q16 hole, I thought you were actually using the original Q16 part. I'm definitely gonna copy your idea- that's very clean.


I've been noodling around with a printed copy of the schematic and a pencil for weeks now. Getting very familiar with it.

R37 and R38 are just resistors, and they're non-polarized. Specifically, they are 0.05 ohms, being used as current shunts. Looking at this portion of the schematic (the INJ1 overcurrent protect) here's how it works:



U4 drives pin 7 high when it wants to turn on Q1, which is the primary injector driver. "INJ-1" is the incoming (or outgoing, depending on how you look at it) connection from/to the injector, which has +12 on the other side of it. The injector comes into Q1 at pin 2, and it conducts to ground through pin 3.

As the current flows through R37 on the way to ground, there is a slight voltage drop across the resistor. Notice that the base of Q14 is connected at the junction between the ground side of Q1 and the "hot" side of the resistor. So any voltage drop across the resistor will show up as a positive voltage on the base of Q14.

So long as the current flowing through Q1 (and therefore through R37) is within reason, the voltage drop across R37 will be sufficiently small as to not trigger Q14.

If the injector shorts out however, causing the current through Q1 to rise to an unacceptably high level, then the voltage drop across R37 will increase until the base-emitter saturation voltage of Q14 (about 700-800mv) is achieved. By my calculations, this should happen somewhere in the neighborhood of 14-16 amps. When Q14 turns on, it will pull the trigger pin of Q1 to ground, causing Q1 to turn off and therefore interrupting the short circuit condition. It's a really clever design.


So when note 1 says you can "Remove and Jumper" R37 and R38, they mean just that. If you need to free up some space on the heatsink, you can remove the resistors. In doing so you must place the jumper in their holes so that Q1 and Q5 still have a path to ground to operate the injectors. This defeats the overcurrent protection circuit, since there is no voltage available to trigger Q14 / Q15 anymore.

I'm going to leave those in place, and instead remove Q9 and Q12. Those are the PWM damping circuit, and I don't plan to run PWM injection control since I already have hi-z injectors and I intend to run HR code which doesn't support injector PWM. That will give me space for my WI FET. Of course, the EBC FET will be going in the Q16 hole via your neat little trick