Relay weirdness

[jsmcortina]

Have you moved the 12V feed for the Megasquirt yet?

James

[Keithg]

Have you moved the 12V feed for the Megasquirt yet?

James

Nope. I know it sounds like I am being obstinate, but what I have works for MS1 and MS2 and has for a long time. I am trying to do this methodically, one step at a time. Also, it is how a lot of the Bosch based wiring harnesses are (re: Paul_VR6). If that is the 'only way' I'll do it and add an inline fuse on that line, but I want to see if there is another solution which allows me to use the fuse already in the fuse box.

And BTW, MS3 cranks/starts faster than MS2...

KeithG

[jsmcortina]

Keith,
your obstinate nature is why we asked you to beta-test....

I do think the wiring is wrong, but at the same time Ken and I have some code ideas that _might_ work.

James

[Keithg]

Another possible reason to *not* use the same feed for the MS and the injectors is the potential for noise that may cause.

KeithG

[Keithg]

Another piece of info. I have yet to get this behavior when I have a stall. It seems to only happen when I key off.

KeithG

[Paul_VR6]

Here is my wiring diagram, there's some left off of it but it's the basics.

MS3V3 VR6 PnP.jpg

As you can see the ign switch directly powers the MS, all relay terminal 30's are off of battery (which is why that's called terminal 30 incidentally). Fuel pump relay, and the ECU main relay (for injector/isv power) are switched on by the MS. The heated O2 relay is jumpered in my install, so the wideband runs any time the fuel pump does. This defeats a 'feature' when I run the stock ecu it would shut off at 4200rpm to keep the stock 02 cool or something. The electric water pump relay is powered by the ecm relay, a gen purpose output to turn it on (stock it's used for the smog pump relay).

The only thing that I can think happens on key off is MS loses power first and the relay outputs somehow say low enough to keep them latched. The last time it happened I did check the ISV power wire (shared with injectors) and it still had 12.2v. I can't easily get to the injectors on my car unfortunately.

I *think* I just want to put a 1n4001 between the FP output pin and S12 on the board, much like the IAC mod has.

[Keithg]

Paul,

I do not think a 1n4001 will help there. That, basically, is what I had previously when I had the 1n4001 running to the injector flyback.

I need to check this in the car, but I am guessing that constant 12V (+30)across one of the relays or devices is actually partially powering up the MS when it is turned off by the switch. Look at where 12V comes from on the board (V3 board schematic pg 6). If you turn off the 12V supply on pin 28 and have some voltage bleeding across a device (relay coil, transistor) , the 5V regulator still may power up. If it is partially up, it will power the micro. That I have a pulldown on the base of the FPR transistor and it still does this leads me to believe that a back feed is what is going on. I do not know, yet, if this is what is happening as I am 'down' awaiting a coil pack nor do I have a plan as to how to fix it.

KeithG

[jsmcortina]

As you can see the ign switch directly powers the MS

I tried that a few weeks ago on an MS2 install. Initally it worked, but once we switched to low-z injectors the engine kept running. When I changed to having the Megasquirt power downstream of the main relay, all was ok.

James

[Paul_VR6]

Not sure what's going on here. Only change to before is that I have the relay for the elec waterpump jumpered and it has yet to 'hang.'

[muythaibxr]

I had odd issues like that on the corolla on ms2. It happened whether the MS was attached to the relay that powers the injectors or the main relay.

Basically I would crank but the engine would not turn over. Then when I keyed off, the e-fan would come on. I added an extra engine ground to the chassis, and both the crank problem and the efan coming on (turning on the ms) went away.

The point is that many things can cause odd electrical issues so you're better off following the manual.

Ken

[ashford]

i have an idea about your hanging problem. you said in another post that you jumpered your waterpump relay. that tells me that the relay for the water pump had ign switch power to it.
if that is the case the pump is a dc motor and that means it also acts as a generator.when the ign is turned off the pump acts a a generator and produces current as it slows down voltage drops. when it has dropped enough voltage relays drop out and if the waterpump isnt the first to drop and holds just slightly longer than another when that other shuts off the voltage on the ign line rises with the lesser current just wnough to power up the ms again and turn on the relay that fell before. and if you have something leaking voltage from relayed power(fuel pump or ecm) to ign switch side(coil of main relay) this would cause an on off oscilation.

[jsmcortina]

Here's what I had happen when my relays were wired incorrectly:
http://www.youtube.com/watch?v=LnSmqk63DBQ

James

[Paul_VR6]

Hasn't happened since I've kept the waterpump relay out. I'll get back into it when I get back from vacation on the 26th.

[BowlingSuperior]

I have noted in my install that I have had weird relay behavior. I do have some odd board mods (you can read why in my other thread) and can go back at any time to Orig. Basically, instead of D4 as a zener to ground, I have a 1n4001 going from pin 3 of the ztx450 to pin 1 of Q3 (flyback for the injectors). I am not running the injector drivers on the board. I am running 4 injector drivers off the MS3X port.

OK, just getting to some posting here - I haven't read the entire post yet so this may already be identified and fixed....

Many, many years ago it was discovered here was a return path for +12V on the fuel pump relay on the V2.2 boards. Look at the schematic, page 3, at the fuel pump and the diode D7:

http://www.bgsoflex.com/v22/megasquirt_ShemV2.2.pdf

Because of D7 connecting directly to the +12V internal feed to the MS board the diode is always forward biased whenever the MS +12 in is switched off. The current flow path is the +12V on the relay coil side, thru the relay coil, and back up to the V12 source. This was the reason we indicated that the power source to all external relay coils should be switched by the ECU relay (which this operates directly from the ignition key) - kill this relay and all of the +12V sources for all other external relay coils also go away, and hence the feedback path. The Relay board is also wired up this way internally.

This is why we replaced the circuit with a 22V zener on the relay/IDL drives on the V3 board. This eliminates this feedback path. And since the FP is only activated once (or at a really slow duty for error modes) the zener works fine for flyback suppression. If you use this transistor as a direct drive for IDL solenoids with PWM the zener will get hot and most likely cook.

If one wants to use a 1N4001 diode across the relay then the +12V return should not go to the +12V on the MS board... feed this with a constant 12V source outside of the MS. In fact, often the diode is placed directly across the relay terminals itself - some VF4 relay sockets have the diode built-in.

- Bruce

[Peter Florance]

fundamentally speaking, you are asking for problems if your injector and idle valve +12 is not connected to pin 28 (unless you've modified flyback to brink it out to a different pin). If you are powering relays that are not connected to pin 28, you should use Bruce's zener diode method.

If you want to use 1n4001 for relay, they need to go on the relay pins 85-86 to avoid a sneak path. Where I've had to power up main relays, I've used zeners to gnd instead of 1n4001 to avoid the problem of latching on.