? about hysteresis

[pierre71]

I have an MSII with v3 board with megatune 2.25. I've setup one of the IAC outputs (using a fast idle valve, no IAC) to control the electric fans. Set at >, 200 with hysteresis of 10, so in theory it is on at 200 and off at 190. Triggered value = 1.

When I first installed it, it worked fine in the driveway. Heat climbs up slowly, thermostat opens, fans trigger, temp drops and cycle starts again. I've driven it back and forth commuting several times but ambient never got hot enough to trigger the fans until recently. Got stuck in stop/go freeway traffic on a hotter day - the temp started climbing past 200 and didn't stop to my shock. By the time I pulled over it was past 240 maybe... got out with the engine still running and checked under the hood, fans were off. I disco'd the relay input from the MS at my junction box and powered it directly from +12. Fans ran as expected and engine cooled back down. Got home, switched it back to MS control. Again it runs just like my initial test, fine in the driveway.

I'm curious, what if in the driveway the fans were able to catch up with the heat, but not when I was in stop/go traffic on the road? Let's say the temp got 10 degrees above the setpoint - would this turn the fan back off because the hysteresis was reached (but in the wrong direction) or should the fans still run? If hysteresis acts in both directions, what can I do to keep the output triggered (and the fans running) in case there is some unusual external condition that causes the temp to spike?

[Matt Cramer]

The hysteresis should not have shut down the fan like that. Could you post a copy of your MSQ? That would help tell if there's a settings issue.

[pierre71]

Matt I'm pretty sure the programming is fine as it worked normally in the driveway. Unless Megatune showed one thing but programmed another?

Did some more tinkering. I detached the CLT sensor from the harness and hooked it to a spare, then hit it with a small butane torch - it pegged at 215deg in megatune as expected and the fans stayed running. When the temp dropped to 190 the fans stopped, so the hysteresis seems to work as intended.

I wonder - the gauge I have in the dash, the sensor is in the top radiator tank. The sensor feeding the megasquirt is in the stock slant six location - just behind the thermostat in the head. Is it possible, given the ambient conditions described, that the top radiator tank reached that high in the upper tank but stayed below 200 in the head? When in the driveway the upper tank seems to be 5-7 deg ahead of the MS sensor. I'll have to take the laptop with me on the next hot day.

[Matt Cramer]

That seems very unlikely as the stock slant six sensor location ought to be hotter than the top of the radiator, although the gauge's reading may be off.

['66fb]

pierre71,

I remember awhile back (may not be true now...) that if you didn't enable the "expanded coolant temp range" in MT when the clt temp hit the limit it would -for lack of a better description - fold back. It would drop the out-of-limit temp and report something like 170F, IIRC.

If MT is doing this then that would shut your fan off...

Try TunerStudio

[pierre71]

Compared to my fluke meter w/thermocouple in the top tank, the gauge was off only by a degree or two. Perhaps the sensor feeding the MS is a bit behind.

It seems to have worked fine since the initial roadside mishap. Every few trips I would let it sit in the parking lot idling to warm up and let the fans cycle. On the latest test it failed. Temp climbed up to 215 on dash gauge before shutting it down. When turning it back on, even before starting, the fans came back on.

I'm confident I have the settings right. I wonder if something hardware wise is flaky. This should be more reliable.

[pierre71]

Some things that popped into mind -

I was re-reading the spare ports section of the manual with regards to using the stepper IAC ports. I have the idle scheme set to valve so that should leave the stepper IAC chip alone. I'm not sure if I'm using the IAC1A or IAC2A port. Manual said some versions of megatune had it reversed, so what I did was set both ports to the same output conditions in megatune. One of the ports is electrically unhooked. Could that be a problem? They are independent so it shouldn't be but just thought I'd throw it out there.

Also I am using a solid state relay to power the fans so the load on the megasquirt is really small. It has active circuitry to limit input current - @5v (min needed to turn on) it is 1.6ma, @32v max its 28ma, so @~12v the ms feeds it is somewhere in the middle. Does the IAC output chip have a minimum required load to function? I could put a small incandescent indicator lamp in parallel with the relay input side. I've been meaning to rig up a "fan indicator" light anyway.

[Matt Cramer]

Is the IAC driving the fan relay directly, or is it using a switching transistor? Let me know exactly how you have it wired up. There are a couple quirky things the IAC outputs do and I want to make sure you haven't wired it so they are interfering with each other.

[pierre71]

Yes the IAC output is driving the solid state relay directly. Per megamanual here they can deliver more then 500ma each so the less then 32ma load should be plenty under that limit. The other 3 IAC outputs are disconnected and isolated, ends of the wire in the harness are taped up.

[Matt Cramer]

Ok. So do you have JS0 wired to the relay coil input and the other end of the relay coil grounded?

[pierre71]

Yes. It was either pin 25 or 29, can't remember which one I used. That went to the relay input positive side, the other side is grounded. I'll have to look through the bom and see what number the iac driver is and read it's datasheet. If it's minimum load spec is higher then the minor input current the sad relay is synching that will be an easy answer.

[pierre71]

Ok so the IAC driver is a UDN2916LB - h-bridge output. I didn't see any minimum load requirements in the datasheet so that theory is out. I have established some consistency - when it is hot enough to trigger the fans, and they don't come on, cycling the MS will get the fans to come on. I wonder if the IAC chip is getting disabled under certain conditions for whatever reason.

Just to clarify last post - I'm using either pin 25 or 29 of the main DB37 connector itself.

[Matt Cramer]

It doesn't have a minimum output current, but its maximum is around 1 amp. The real problem usually tends to come if you try using BOTH sides of the H-bridge to drive a relay, since the outputs reverse instead of turning completely off.

[pierre71]

I'll try to grab the msq and post later this week. I'm not sure what else to look into at this point. Dang, this has been the first major reliability hurdle.

[pierre71]

Here is the msq.