No Idle Change with A/C On

[lothian]

'06 Hyundai Azera
145k miles


While t'shooting a no-cooling problem in this car, I notice the idle does not bump with activation of the A/C; meaning, no movement of the tach. The blower comes on immediately when A/C is switched on, which suggests the climate control module works.

I've done the cursory stuff: verified as OK the A/C relay and A/C relay fuse, the two A/C control module fuses, and the A/C pressure transducer. The factory service manual lists under "No engine idle-up when A/C switch ON" two suspects--Engine ECM and Wire Harness--but offers no t'shooting procedure for either, so... not much help there without a diagnostic computer.

I'm stumped.
Thoughts? Suggestions?

[pete_89t2]

Are you saying there is no change in idle speed AT ALL (idles up or down) when you flip on the AC?

If that's true, are you basing your assessment of idle speed solely on what the factory tach is showing you, or is your assessment also factoring in your senses? (i.e., does the engine sound/feel like the idle changed or hiccuped at all when AC engages/disengages?)

Reason I mention this is: (1) Factory tachometers are not known for their stellar accuracy or response time, so if you're expecting to see a short lived 50 to 100RPM fluctuation you probably won't, and (2) A well designed & tuned ECU/engine combo with closed loop idle control in a quiet car like your Hyundai should perform this function so smoothly that it often goes unnoticed in the cabin unless you're listening closely for it.

So unless you're experiencing the idle bogging down excessively when the AC kicks on, or it races too high when AC initially engages or switches off, I wouldn't worry about this too much. My recommendation here is to just pop the hood, run car at idle and have someone flip on the AC while observing the AC compressor. Verify the AC compressor engaging & cycling correctly and the AC still blows cold inside the cabin. You'll likely observe a short lived idle hiccup when the AC engages while observing from under the hood.

[lothian]

Thanks for your response.

So after I processed your comment about the tach, I plugged in my scan tool to get a more accurate reading of RPM. Problem is, this tool reports RPM real-time. And since RPM reading fluctuates constantly, as expected, it's pretty much impossible to discern any effect from toggling the A/C on and off. The tool doesn't average RPM at a "mark"--like the moment i toggle the A/C. (I need a better scanner. ...and one that also features the ability to toggle the field coil on and off.)

<plaintive sigh>

[pete_89t2]

Sounds like your ECU is doing a good job with the closed loop idle control then... It's making those constant adjustments to compensate for loads, but the feedback loop is tuned well enough that it's not noticeable to the driver. Apparently those Hyundai engineers did a good job.

Tuning an ECU to give you a smooth idle under all engine loading conditions can be a major PITA. I've living that PITA tuning an aftermarket ECU (AEM Infinity) that runs my '89 Mazda RX7 turbo. After tuning my setup to net me 380RWHP/330ft-lbs and overall great drive ability on the road, next step was to dial in the AEM's idle controls to work 100% right - cold starts, hot starts, and various engine loads like the AC. Months of trial & error street tuning, data log analysis and I'm STILL working on getting idle controls 100% right.

[bohica2xo]

Does the compressor run?

Does it cool?

If so, close the hood & smile.

[lothian]

Nope. ...er'm... not yet, I mean. Gotta remain optimistic for the missus, lest she think me incompetant.

You may recognize the other phase of this problem here.

[bohica2xo]

I will merge the topics later.

You are chasing your tail.

The old clutch coil may (or may not) have been bad. The first step in chasing electrical issues is checking the failed part in hand & determining the failure mode.

Clutch coils rarely fail. When they do they can simply go open circuit, short to ground or short turn to turn. The average coil runs around 50 watts, and closed up in that hub they get hot. A perfectly serviceable coil will show signs of heat aging, and will be filthy after 60k miles or more.

IF a coil shorts to ground in the right place it can jump to 100+ watts and never blow a 10 amp fuse. Eventually it will melt the potting compound, and die a messy death. Turn to turn shorts can also raise the wattage.

If is simply quit working one day, it may have been a gap issue. It could still be a gap issue.

The fuse is still good. The relay may test ok on a bench but under load it may be a 20 ohm resistor. Connectors, harness, etc. all play a role.

With the engine off, pull the relay & install a jumper from the relay socket Terminal 40 on your drawing to B+ Does the clutch pull in?

If not, push on the plate with your fingers - does it pull in now? If that does not work, pull the connector at the compressor apart & check for voltage at the plug with the B+ jumper installed.

[lothian]

Bleh. I definitely don't like chasing my own tail. Nor do I like firing the parts cannon at a problem.

I eliminated the fuses, the engine control and clutch relays, controls, wiring harness, pressure sensor resistance, and clutch gap first. Not only because these items are (relatively) easy to check. It's mostly because any or all of them are likely culprits in instances where the clutch suddenly stops engaging.

"WTF" screeches the wifey. "It worked this morning. No fuzzy nummies till you restore my comfortable commute."
So motivated, I began t'shooting with obvious fervency.

Things seem OK in the cabin. Let's listen under the hood. ah HA! No click: the clutch isn't engaging. Check the fuse. Check the relay. Check the field coil (FC) connector. Check continuity between the box and the compressor. Check continuity between the FC connector and the chassis. Now we have something. Jump the high-side terminals (40 in the schematic I posted is 30 on the relay) in the junction box--no click. Run a wire from the junction box to the FC connector--no click. Check gap to service manual spec--reads as nominal. (I'm omitting other checks here, but i've made my point.) Tentative conclusion: bad FC.

I completely track with you regarding the unlikelihood of the FC being bad, even while I grudgingly ordered it. Nevertheless, the thing tested bad (refer to other thread) so far as I could tell.

After I installed the new FC, the clutch (also new) actuates in all its clicky yumminess when I jump the high-side circuit in the junction box. It just won't engage when called to do so on demand. I'm reasonably satisfied the original FC burned out, and I have a coincidental and yet unidentified problem elsewhere.

So I continue obstinately, and abstinently, to t'shoot.

[bohica2xo]

A half dozen times each season I watch someone launch in to an extended electrical troubleshoot or the firing of the belt fed parts cannon - over a clutch gap issue. Looks like you actually did your homework.

So the PCM is not commanding a clutch start. Ok, the basic gates for that decision are:

1) Charge level. Not enough refrigerant system pressure will inhibit clutch operation

2) Coolant temperature. Too high on the coolant temp will inhibit the compressor. A bad temp sender on a car with no gauge can be hard to spot

3) Compressor RPM. Some vehicles have speed sensors in the compressor. If the sensor says the pump is not turning the PCM will lock out the compressor. I do NOT believe your vehicle is so equipped, but some are so I include this for future troubleshooters.

4) Ambient air sensor. If the PCM believes it is -40f it will inhibit the compressor.

Some of those gates are modal, and need to be reset. From the diagram you posted you seem to have access to an FSM. Reset procedures are in there someplace...

But start with the basics. Is there enough refrigerant in the system?

I am not sure what sort of pressure sensors your Hyundai uses, but if you see a 3 wire transducer on a refrigerant line refer to the FSM & use a meter - never a test light!

Good luck

[lothian]

First things first...
Thank you. I genuinely appreciate your help. I realize it is frustrating—and mostly futile—lobbing intermittent sans serif advice across this medium at some anonymous ad hoc a/c tech poseur, 'specially when all indications suggest a problem better suited for a laying of the hands by an actual experienced and credentialed tech.


"So the PCM is not commanding a clutch start."
Well... that's my theory at this point. Seems to me I should be able to catch a signal with a DMM across the low-circuit relay terminals in the junction box. I've resisted, though, 'cause there's no procedure for this sort of test in the FSM; and, I'm not keen on the possibility of somehow frying the PCM.

It's clear to me that I could really benefit from the use of a professional diagnostic scan tool, (e.g., Hyundai Global Diagnostic System, Autel Maxisys, Farsight MVDS, etc) with support for secondary systems like A/C. None of what I need to do at this point—signal the PMC to engage the clutch; read each HVAC sensor component input to the PCM real-time; read PCM command output real-time; query control console diagnostic code(s)—is as easily done with a DMM or inference from indirect tests. But that specialty tool costs as much as a new ultra-short-throw laser projector or a Jet Table Saw—either of which I'd much rather have.


"Charge level...Is there enough refrigerant in the system?"
Static high and low pressures are nominal.



"Coolant temperature... not sure what sort of pressure sensors your Hyundai uses..."
The PCM will not engage the clutch based on input from three (maybe four) sensors, one of which is the pressure transducer on the high-pressure line. DMM readings (at line temps registered with an IR thermometer) were inconsistent with those printed in the manual.



"Compressor RPM"
My vehicle is not so equipped. There are ZERO signal leads into or out-of this compressor.


"Ambient air sensor"
I neglected to test this component. I blame the FSM, which doesn't list the 'ambient air sensor' among potential test candidates. Yet under the 'Ambient Air Sensor' heading in the FSM, there's a note that this component can affect engagement of the clutch. (...dumb editors.)



A potential "Item 5" on your list is the evaporator temp sensor. The FSM lists this component as a test candidate when the clutch doesn't engage. I haven't checked the thing yet. Gotta get on this, also.



And finally, the fourth sensor is the photo sensor on the dash that signals to the control console. I haven't bothered looking at this component.


I need you as a neighbor.