Archives

The autostop on our 2003 HCHI has been a bit iffy lately. It still works most of the time, but every so often it will make a few stops where the car is well warmed up, the environment controls are such that autostop is allowed, SOC is >=80%, and yet no autostop. Ambient temps fairly warm, around 60-65F when this happened. The first time I noticed it was driving at night, which I don't do very often, and thought maybe it was a problem with having the headlights on. So turned them off (just) before stopping at a light, and autostop worked. Next light left after that left the headlights on and autostop was still working, so probably a false correlation. Nothing tricky about the stops either, all are pretty much the same "slow down from 35 mph and stop", no crawling along at 10mph first, which I know can confuse it.

Any ideas what might be causing this?

There are no warning lights and other than that and the $#$#$ CVT clutch judder the car is running fine.

Are you triggering fuel cut during the slowdown before these auto stop events? I haven't as much experience in a CVT Honda Hybrid but my Insight is more likely to drop into AS if the target speed for that feature is approached with the injectors off already. I've also noticed that it can take a fairly purposeful jab of the brakes to get it to drop in as well.

When autostop's not working, is it pretty consistant, ie: from one red light to the next, to the next, consistantly not working?

Also, you don't have ventilation mode to heat/defog do you? That'll prevent it.

When autostop's not working, is it pretty consistant, ie: from one red light to the next, to the next, consistantly not working?


It will miss ~3 in a row, for no obvious reason.

Also, you don't have ventilation mode to heat/defog do you? That'll prevent it.

Nope.

My 2003 HCH I hasn't autostopped in months :(

I did end up having to replace the SLA (12v) battery in the cold January winter last month, so maybe I had a bad battery all this time, and now its too cold to autostop.

Whats the lowest outdoor air temperature an HCH I can autostop in consistently?

I think it's supposed to be somewhere around freezing (32 degrees F). We've had a really cold winter this year, but once my car is warmed up it still auto stops in cold weather.

There is an old thread (which has now disappeared) from greenhybrid regarding the same problem in a Honda Insight. The car was not auto stopping, and it turned out to be due to a bad ambient temperature sensor.

http://www.greenhybrid.com/discuss/f11/auto-stop-not-working-24800/

There's a link in that thread leading to information originally found at InsightCentral. Part of it says:

"...If all of the above conditions are satisfied, the problem is most likely that idle stop isn't functioning due to the exterior air temperature sensor being disconnected. Some Honda dealers are not aware of this sensor, and so haven't been connecting it when they should. This sensor is located behind the grill, in front of the radiator and, if air conditioning is installed, in front of the air conditioning condenser. If its wire is hanging loose, just reconnect it, and this should solve your problem."

It might be worthwhile to check that sensor.

I know with second gen if you have heat/vent mode set to heat/defog that will prevent autostop. There are a several other preventers, but that's a sneaky one I can think of.

Oh dang, I previously posted that. Anyway...

My 2003 HCH I hasn't autostopped in months :(

I did end up having to replace the SLA (12v) battery in the cold January winter last month, so maybe I had a bad battery all this time, and now its too cold to autostop.

Whats the lowest outdoor air temperature an HCH I can autostop in consistently?

My experience echoes Gairwyn's, any lower than the 30's and it doesn't seem to want to auto-stop.

There is an old thread (which has now disappeared) from greenhybrid regarding the same problem in a Honda Insight. The car was not auto stopping, and it turned out to be due to a bad ambient temperature sensor.

http://www.greenhybrid.com/discuss/f11/auto-stop-not-working-24800/



Where is this sensor on the HCHI? The insight central article describes a diagnostic, does that work on the HCHI too or are different actions needed to trigger it?

thanks.

The ambient outside air temperature sensor is located between the inside of the front bumper and the radiator fan. See the following link, and look in the area indicated by the arrow pointing to parts 2 and 3:

http://www.hondaautomotiveparts.com/auto/jsp/mws/prddisplay.jsp?inputstate=5&catcgry1=CIVIC+HYBRID&catcgry2=2003&catcgry3=4DR+MX+%28HYBRID%29&catcgry4=KLCVT&catcgry5=A%2FC+SENSOR

I don't know if the HCH-I has a similar programming sequence like the Insight-I. Logic would say that it does, but I don't know if the programming is different or not. Maybe try asking at the Insight forum or a Civic forum, or maybe someone at a Honda parts department could look it up for you. If I find out anything related to that, I'll let you know.

see next page...

The only thing I've been able to find is a page in the service manual that describes how to retrieve a DTC for the climate control system:

1 Turn the fan switch OFF.
2. Press & hold the recirculation control switch and the ECON switch down.
3. Turn the ignition switch ON (II).
4. Release both switches. The recirc indicator comes on. The recirc indicator goes off 2 seconds later and the A/C indicator comes on, then the self-diagnosis will begin. About 10 seconds later the self-diagnosis will finish and the A/C indicator goes off. The recirc indicator will blink the diagnostic trouble code (DTC) to indicate a faulty component. If no DTCs are found, the indicator will not blink.

(Manual proceeds to show a time-line for recirc button coming on, then going off 2 seconds later, and a sequence of three 0.3 second blinks followed by a 1.5 second break, then another sequence of three 0.3 second blinks. This particular DTC pattern is for DTC3, indicating an open in the outside air temperature circuit. There is a graph showing DTCs 1-15, each relating to a different component of the system; I've transcribed it below in the next post.)

Canceling the Self-diagnosis Function:
5. Turn the ignition switch OFF to cancel the self-diagnosis function. After completing repair work, run the self-diagnosis function again to make sure that there are no other malfunctions.

ah... what the heck, might as well keep goin...

DTC (Recirc Indicator light blinks):

1=an open in the in-car temp sensor ciucuit
2=a short in same as above
3= an open in the outside air temp sensor circuit
4= a short in same
5=an open in the sunlight sensor circuit
6=a short in same
7=an open in the air mix control motor circuit
8=a short in same
9=problem in air mix control linkage, door or motor
10=open or short in the mode control motor circuit
11=problem in the mode control linkage, doors or motor
12=problem in the blower motor circuit
13=problem in the EEPROM in the climate control unit; unit must be replaced
14=open in the evaporator temp sensor circuit
15=a short in same

(In case of multiple problems, the recirc indicator will indicate only the DTC with the least number of blinks)

Actual problems aside, I had an opportunity to drive a manual
Insight recently and noticed that both autostop and regeneration
wind up essentially disabled in a very sporadic fashion. My own
tendency while gearing down is to blip the throttle to rev-match
*while the clutch is in*, and apparently the Insight really doesn't
like that because whenever I finished entering the lower gear, I
got *no* regeneration. Apparently you have to poke the throttle
when the clutch switch shows "engaged", and then regen comes back,
but in a real-life stopping scenario there's either no time to
do that and/or doing so produces a terrific lurch which to me is
unacceptable.
.
And if you're in neutral and the car is still rolling at an
appreciable speed, letting off the brake [as I usually do while
arriving at a complete stop, again for smoothness] makes the
engine start right back up again for no reason. I had a similar
issue in the Civic.
.
As this was my first time at the helm of an *unmodified* Insight,
it was a fairly disappointing experience and gave me greater
understanding of why Mike D. said "well, F this" way back in the
day and started designing MIMA. Honda really screwed the pooch
on several operational scenarios, and one thing I took away from
that day was that I'm glad I never went down that path. The Prius
control algorithms, for all their funny temperature-related bugs
otherwise, make so much more sense.
.
_H*

Al, the throttle input while in gear before regen kicks in quirk is indeed somewhat annoying but it is something that can be adapted to fairly easily. The blip can be so quick that you don't even see the gauges react fast enough to indicate an out of DFCO condition and it is second nature to me now. For fast stops, I've got the clutch in anyway because I'm conditioned not to kill the engine while stopping.

If you are slowing down in AS, simply push the clutch pedal in and it will stay in AS as long as you like when you release the brake pedal. There is no such option for the CVT version of course. As for entering that state, I find it works most reliably if you get into DFCO first (preferably with regen), then quick disengage the clutch while shifting to neutral -- the engine should simply spin to a stop without using any more fuel.

Maybe the above will lessen your frustration the next time you drive a manual transmission Insight. I personally find its behavior much more predictable than that of a Prius... and having a key available to over-ride the computer's notion of when the engine should be on is a definite plus.

The ambient outside air temperature sensor is located between the inside of the front bumper and the radiator fan.

Hmm, suspicious location. The first owner rolled it into a wall (parking accident) and the front bumper and radiator were replaced. Given the position of that sensor, it should have been squished too. There must still be one there though, as autostop does work most of the time.

I will run the diagnostics you posted and have a look at the sensor this evening when I get home.

Thanks.

Ran the diagnostics - no flashing light. Maybe not surprising, since autostop was working at the time. I will have to remember to do this the next time it starts missing autostop, as there could be an iffy connection/component somewhere.

FYI, while the diagnostics run the climate control system develops a mind of its own. Kind of like there's an invisible gremlin in the car twiddling the controls.

Have you had any luck locating that outside ambient air temperature sensor circuit? It seemed like the person in the Insight thread I showed you was able to visually see that their car's circuit was disconnected.

I have a new HIH-II and don't have any problem with getting autostop to work after the engine is warmed. I usually roll up to a stop lightly pressing the brake pedal then give it a relatively sharp jab to stop completely. I do that because I don't want to engage the brakes until the last second, but I do want to send regenerated power back to the battery. I will say now that the car has been broken in autostop kicks in sooner than when the car was brand new. I swear it would know when it was raining and not autostop regardless of windshield wipers or defrosters.

In another thread Gairwyn posted

http://www.cleanmpg.com/forums/showpost.php?p=322938&postcount=7

this discussion of autostop issues

http://www.greenhybrid.com/discuss/f11/auto-stop-not-working-24800/

My tests so far (while autostop is not working):

1. Stop, and immediately run the climate control diagnostics - always passes.

2. The brake switch is working as assayed by the trunk brake light coming on. (A piece of tape was placed on the outside of the glass so that I could see if that lamp was lit or not, day or night). Have to look up the circuit diagram, it might be a multipole switch, with the other part flaking out but the brake light section still working.

3. It doesn't matter if the fan is blowing slightly and ECON is on (dash outlet position) or if the fan is off. A/C and recirc both off.

4. IMA showing a high state of charge (have seen it as high as 95%)

5. Ambient temperature reasonable (65 degrees or above).

6. Car warmed up.

7. Brakes seem to be working normally (no evidence of a gross vacuum leak).

8. Radio on/off doesn't change anything.

9. There are no other electrical devices plugged in (that is, no extraordinary demand on the 12V system).

10. Speed before expected autostop was more than 20 mph.

11. My foot is not on the accelerator.

12. This is a CVT.

13. Car is not on an incline.

I think there is a trend where autostop works early in the trip and then stops working, but it could be my imagination interpreting random bursts of failure.

Half an hour trolling through the service manual turned up section 19-5 = idle stop switch. According to the diagram it is a switch located right next to the brake switch which must be independently adjusted. Sounds promising. If the brake pedal engaged position changes slightly as the car is driven, which it might, and the idle stop switch is just on the hairy edge of being out of adjustment, it might act like what I am seeing. Have to wait until at least tomorrow afternoon before I will have time to look.

Kind of crazy to have a separately adjustable switch that duplicates the logic state of the brake switch. They should just have put another pole on the brake switch. The one they have now is already a 2 pole switch, one side for the brake lights and the other side, apparently, for the cruise control.

That does sound like a promising lead!!! Let us know what you find.

Let us know what you find.

The switches are very easy to get to. Under the steering wheel there is a panel held in place by two rotating clamps. Turn them, pull the panel out, look at the top of the brake pedal, and there they are. Strangely, because they measure the same pedal motion at essentially the same place, the two switches are fairly different in design, with the idle switch having a really solid metal body, and the brake switch being mostly plastic (externally) and a bit wobbly.

I could not detect a fault in either switch. The two pairs of contacts on the brake switch are normally open and shorted correctly when the brake was engaged. The idle switch is normally shorted and opened correctly when the brake engaged. The problem is intermittent. I already know from the brake lights that the brake switch is still working when the auto stop fails. Which means that I really need to check the idle switch action when the auto stop fails. Difficult. The wires are covered with one of those corrugated plastic protectors, and that would only retract a few millimeters, so there isn't enough space to get a tap into the wires next to the plug. There also wasn't enough slack in the cable to pull it out to solder wires into the plug. So I guess I will need to find compatible male and female connectors so I can extend the circuit out to where a voltmeter can be placed safely.

My gut feeling though is that there is nothing wrong with the switch or the mechanical action there. It all felt solid. That would leave a vacuum boost issue (actual leak or sensor wrong), some odd signal from the IMA, or possibly an error signal from the transmission.

Hm. Do the switches have plugs on the back? If so you might find you can sneak a couple of thin wires into the plug and make pinch connections when everything is plugged in. That's what I did with my original clutch switch bypass install.

Do the switches have plugs on the back? If so you might find you can sneak a couple of thin wires into the plug and make pinch connections when everything is plugged in.

Yes, there are plugs, and that approach might work with something really thin like wire wrap wire.

I tried one other test today. Drove the car around for a while unitil it was warmed up and autostop was working. Turned off the car, pulled out the cover, unplugged the idle stop switch, and turned it back on. The car would not autostop with that switch unplugged. After a five minutes of that, turned the car off, plugged the switch back in, and restarted. Autostop worked again at the very next stop.

So autostop does require the ECM (or some other computer) to see the idle stop switch both closed and open (brake pressed). A fully defective idle switch, or wiring to it, would break autostop completely. Here since it is hit and miss it is either an intermittent wiring/switch problem (which can happen, but is not that common on internal wiring as well protected as this) or it is something else.

I used an old telephone wire I had laying around -- the conductors were small enough to do the job but I got them into the socket by feeding them under the clip that holds the plug in. At least for the one on the clutch switch in my car the front of the clip is attached to the face of the plug via two legs leaving the center open... just enough space to feed in two small wires.

It might be worth trying to clean the contacts in that idle stop switch socket -- a little corrosion might be more likely than an actual issue in the switch or wiring?

It might be worth trying to clean the contacts in that idle stop switch socket -- a little corrosion might be more likely than an actual issue in the switch or wiring?

I looked at them closely (with a magnifying glass) and the contacts were clean. Also the resistance measured by poking the probes into the back of the plug was too low to measure - so the contacts all seem to be good, at least at the switch.

IMA battery replaced under warranty once the IMA light came on. This also fixed the iffy autostop. So it was apparently an IMA issue all the time.