Megaswitch v.megasquirttesting

[RobertISaar]

the exact FET I was using was a STP65NF06, which is also a N-channel MOSFET and has a body diode the same way a IRFZ44 does. I thought the same thing until I burnt myself, did some google-fu and ended up here:

http://electronics.stackexchange.com...-or-transistor

the right diagram essentially has the body diode of the FET accounted for and is listed why it doesn't quite work(unless it's a zener). the STP unit's datasheet says the body diode functions as an avalanche, but I either did something very wrong or it isn't sufficient. the solenoid was also very non-linear in response, which is the opposite of what I expected. added in a 1N4007(as shown in the left diagram) I had from my diode selection and suddenly the solenoid operates with almost complete linearity from ~20% to 80% duty cycles and no more scorched fingers. without the diode, the solenoid was buzzing like crazy when commanding a duty cycle between the ranges of fully closed and fully open(which was something like 50 to 70%).

something to test on the bench, if you really wanted to, I was much happier finding this out before calling a project finished.

[Six_Shooter]

Interesting, but I think the zener diode is making the valve more linear by changing the way the load is seen by the source. I'll have to think about that for a bit, maybe even do some testing A flyback diode is supposed to be invisible to the source and only be conduction when the coil is discharging immediately after current is taken away from the circuit .

[Six_Shooter]

So, since RS232 is pretty antiquated I wanted to get around the need to use a USB to Serial adapter, especially since every off the shelf adapter I've bought seems to either not work at all or be flakey.

This is where that earlier modification of socketing the MAX232 IC (U6) comes into play. My plan was to replace U6 with a second DIP socket with the correct four pins jumpered to pass a TTL signal straight through from the db9 connector to the MS uC. I have an USB to TTL signal converter from Spark Fun, based on the FTDI FT232R IC.

Initial testing of the idea:


3 wire connection, ground, TX and RX:


Test jumpers inside the MS:


Replacement DIP socket to pass TTL coms straight through:


Adapter DIP installed:


While initial testing proved that it works for communications between the MS and the computer, there's an issue with early MSII daughter cards (which mine is) in that if there is no 5V present on the RX pin of the MS II daughter card at power up it goes into bootloader mode. This means that either the breakout board I have always has to be powered up, or a modification needs to be made to apply 5V to the RX pin when it's not powered up. Which is not a problem if I leave the breakout board in the MS all the time, but that was not the plan. I use this board for other devices as well, and I'm too cheap right now to buy another or several. lol So my idea of bypassing the MAX232 IC in the way I wanted was not going to happen. I came up with a few different ways to apply 5V to the RX pin without the breakboard installed, and then disabled once the breakout board was then plugged in, but this would be more proprietary than I wanted. ONLY my adapter would then work with the MS and that will usually not be a problem, but I didn't want to take any chances, so I went with an alternative, I created my own USB to RS232 adapter... well I've breadboarded it so far. I still need to make a more permanent version using some proto board.

My adapter is the same USB to TTL breakout board I used earlier but I'm using another MAX232 IC to convert the TTL levels to RS232 levels that the MAX232 IC inside the MS can then convert back to TTL levels. yeah I know it seems odd to convert up just to convert back down again, but this way if my adapter isn't available for whatever reason and someone else's is, it will still work just as if it were normal. I also have a Blue Tooth TTL board that I used in a school project that I'd like to get working, but need to reprogram the baud rate before I can get that working and I need to look up again how to do that. lol

the overall USB to RS232 adapter breadboarded and communicating:


Just the guts of the main adapter circuit:


I've had this circuit working on my bench for several hours how and it's worked flawless, zero drop outs with my main PC and only a couple drop outs with my laptop, but that's far less than with the USB to serial adapter I tested earlier today which is a Prolific based adapter, which are always flaky at best. I had drop outs every couple of minutes with the Prolific adapter, just monitoring basic coms.

[3400tZ]

Yep, always use FTDI USB chip (which I believe this is what your breakout board is now ?!). Prolific drivers are not so great on Windows and even worse on Android :)

You seems to be having fun with that MS Chris. Also looks like its getting closer and closer to make it to your car :D

[Six_Shooter]

Yep, my experience is the same, FTDI or nothin' It just seems that people, myself included end up with Prolific junk. Oh well.

I've just been playing with it on the bench to get used to the software, and figure out what makes the MS tick, and make sure everything works the way I expect.

I've even upgraded the firmware. When I got this MSII it had FW version 3.0.3u, so, it's OLD, that FW is from 2009 if my findings are correct. I loaded the latest (why wouldn't I use the latest? lol) 3.4.0.

I still need to register TunerStudio, I mostly want the custom dashes/gauge layout. I don't really care about auto tune, though I might as well use it with the registered version...

[Six_Shooter]

I figured I should update this a bit, since I have something to update.

Well, the original plan for the MS was to install it on my LX9 in my Datsun, but since that's been broken since December with no sign of getting back together anytime soon, due to reasons, which are beyond my control right now, I installed the MS on my T-bucket (283+.030" V8) which I converted to EFI about a month ago using GM components, including a '7747 ECM. I used the Delco to start with because I knew that it should be close and could drive it on the stock tune, or tune it a bit better, and then switch to the MS. Both of my widebands were also loaned out in other people's vehicles at the time and didn't want to try tuning the MS blind, so another reason to start with the Delco at the time.

Well there were mechanical issues that needed to be worked out and at this point should be pretty much solved. I got sick of the '7747 losing connection with the AutoProm and lack of reliable datalogging. There might be an issue with the ECM itself, but I was just sick of it. This was last Friday. I went home and started on an adapter to jumper the '7747 harness to the MS. I'll post pictures later.

Sidenote: When I installed the Delco EFI I knew I'd be going to the MS shortly so I added wires at the time for the IAT and a couple extra grounds.

On Sunday (I was lazy on Saturday lol) I finished the adapter harness and ran a vacuum tube for the MAP sensor that is built into the MS ECU. I was able to get the engine to start and run ok, in trying to tune the '7747 I discovered that increasing MAP AE worked very well to get rid of a lean pop through the intake with large throttle changes, so I started there by figuring out how to enable the MAP AE in MS. Turns out that there's a slider in the AE menu that controls the blend between TPS and MAP AE, and you simple slide it towards the MAP AE table...

One thing was always a miss though, above about 3000 RPM, the revving became REALLY sluggish, and no matter what I tried with VE or spark nothing changed it considerably for the better.

I checked base timing and decided to adjust it. I ended up with a weird thing where with the bypass connected the engine would not start, the MS was not getting or at least was not acknowledging any RPM (DRP) signal. Bypass disconnected and the engine would start and run the way that was expected.

In this I decided to check whether the EST signal was good from the MS, which I used my JimStim and connected the EST signal to a spare LED on the JimStim board and observed the LED pulsing... for hours, in fact overnight the EST signal was still triggering in the morning... oopps... lol

I also discovered that the MSExtra manual is wrong where the bypass control circuit is concerned. Using the board modifications and the external connections shown will have the timing control work backwards, as in ECU controlling timing during start, but not during run. *facepalm* This is why the engine seemed lazy above 3000 RPM, because it was running on base timing! In other documentation there is often a relay suggested to control the 5V signal on the bypass line, usually the relay is suggested to be controlled by the ignition switch, so that 5V is provided to the bypass wire only while running, so that base timing is still used at start. diyautotune.com suggests connecting the bypass wire straight to 5V and have the ECU control timing during start up.

I went a different route. The built in settings for the HEI control will turn on the transistor that controls D16 (Q8) 5 seconds after the engine starts to run, or more specifically after the threshold that is set under "Cranking/Startup Settings" called "Cranking RPM" has been exceeded. IIRC stock setting is 700 RPM, I set this to 400 RPM. As I mentioned before the way the MSExtra manual suggests to connect it will not work correctly. I modified the circuit by installing a 2N3906 PNP transistor to provide the 5V on the bypass line that is needed for the ECU to control timing through the HEI module. I have not tested this on the engine yet, hopefully tomorrow I will be able to. On the bench it works it works perfect and I can't see the HEI module needing anywhere near the 200mA rating of the 2N3906 transistor.

I will post pictures and hopefully a schematic tomorrow, once I get a chance to get them onto my computer.

[Six_Shooter]

Just to update this...

I gave up on trying to get the 283 to run right. It was old and tired. I replaced the timing set, a friend of mine played with it for a bit, adjusting rocker arms because the lifters seemed to be pumped up and would not bleed back down making them act like mechanical (solid) lifters. At one point the engine was literally a V-twin, only cylinders 1 and 2 seemed to actually be firing. I gave up and left the car alone for a while.

Last week I had holidays, and my Grandfather's truck had a good running 305 in it, that was destined to be replaced with a Vortec 5.7L that had been installed in the truck years ago but pulled due to a crack in the block for the starter bolt. That's fixed now and using a different starter now that uses different holes seems to have taken care of that issue. So I swapped the 305 out for the 5.7, and then took the 305 and swapped it in place of the 283 in the T-bucket.

So now that it has a good running engine, I reinstalled the MS, and fired it up. It ran ok, then noticed that the bypass wire was still disconnected, I reconnected it and the engine stalled. Tried a few different things, stalled every time. Then used a meter to check the signal on the EST circuit and found there was no output, no wonder the engine would stall when it was looking for that output. Opened up the MS to find that the transistor driving that circuit had failed (small hole in it that had blown out). replaced that and the car runs well. Still needs some tuning but I have driven it everyday since Sunday, because I can and it's fun to do so now, unlike before when the 283 was in it and it would get worse and worse to drive as time went on.

Couple pictures. The wiring will be cleaned up in the winter, it works for now, so it's good until then.