Acts like it wants to die at idle after hard driving on hot day

If I drive my 1996 M Edition Miata hard it will act as if it wants to die at idle, like at a stop sign.  If it does die it is very hard to restart, but if I wait 45min or so if fires right up.  On one occasion the break pedal was very hard when I returned to the car.  This led me to believe it may have been vacuum related.  I replaced all or cut fresh and reattached all vacuum lines.  According to Torque Pro vacuum is 24.4 hg. I pulled the PRC valve, and purge solenoid from my 97 parts car a put on the car based on recommendation from another forum.  OBDII reports no codes.  Coolant temps run 186 driving 55, to 207 sitting at idle.  Of course temps here are unseasonably cool, so I cant re- create the exact situation.  Drove it today for an hour, never out of 4th gear much trying to work the motor, no code, no temp issue.  What's going on with this car.  I dont trust it!

temperature sensors ok ?

Seems to be.......I dont know how to test it.  What I know is there are zero fault codes according to torque pro, car running 55-65mph shows coolant temp of 185f, idling  the temp goes to 206ish f and I can hear the fan cut on.  Of course the one variable I cant recreate is outside temp.  I have noticed the problem when outside temp was 90ish, this morning in Georgia it was 52f and probably 82 for the high. 

I believe it is/was a vacuum issue.....hose expanding when motor hot and sucking air.  Maybe that's not possible, I dont know.  I replaced or recut reattached all vacuum line.  There was a noticable difference in cars performance.  According to Torque Pro vacuum is a little over 24 like 24.4ish.

Also I have put a used PRC valve and a used purge solenoid on the car.

Am I on the right path? I need a 90+ degree day to see what I've got now.

The vacuum reading is probably wrong. I have never seen an engine make 24.4"hg at idle, only on decel from higher rpm with throttle shut. I have also seen a number of scantools invert the reading so that it read 10" of vacuum when it really had 20" and read 29" when WOT etc.  Might want to verify it with a mechanical gauge. 

You're not getting any kind of lifter tick or clatter, are you? It's fairly common for oil to become aerated after extended high rpm which causes the lifters to work poorly for a while afterwards. That can cause poor idle quality but you would usually have a change in engine noise to go with it. 

Ok, I will pick up a mechanical gauge tomorrow, see what I've got.

Like most na8s I get lifter tick at start-up.  Never noticed any tick when I'm driving the car.  I usually let the car idle to warm up 5min or so before I drive it.

How old is the timing belt? Wonder if it's stretching when hot? Have seen this before on other cars with old belts.

Car has 175k, timing belt, water pump, crankshaft seal, camshaft seal, valve cover gasket and plugs were changed at 170k

Ok got a mechanical gauge....was in the green zone and just under 20, I would say.  Torque Pro was showing -23.9, so not that much of a difference if mechanical gauge is negative number, doesnt say it is, doesnt say it isnt.

Are the ecu numbers trustworthy? RPMs seem to be off by 200 or so when running, gauge higher than scan tool, speedometer is off by 3-4 mph,  gauge higher than scantool, Hmmmmm  wonder what else is just a little off.  Is there a way to calibrate, reset, or test the ECU?

Put approximately 100 miles on the car today.....no codes, no issue.  I may be chasing a ghost.

Unless someone has an idea, I'm going to call it good until it gets hot in Georgia again.

Check  for air leaks, not when shut off, but when the engine  is moving side to side durring  acel. And decel.? I'm hot days the rubber boots can get soft, and slip off air intakes

In reply to 222 :

ECU numbers are more right than the gauges and whatnot.  The whole basis of the engine running are based on the calculation that goes into that. 

Unless there are teeth missing on the tone wheel, there's nothing in the computer that will drift.  If the actual speed on the computer changed so that it's not actually measuring time correctly- the whole computer would crash, and it would not run the engine.

Speedometers are regularly high.  You are far better off having it read high than low, so the driver isn't speeding when they think they are not.  And the engine speed is filtered on the tach.

I replaced most of the vacuum lines.  What I did not replace I cut off the end and reattached.  I will check tomorrow for leaks with engine running. 

I would have assumed the computer numbers were more accurate than mechanical gauges.  I will trust my scan tool from now on, and I got a new gauge toy. 

I will post again in the afternoon once I check for air leaks.

Tested it with propane, no sign of vacuume leak.

When you say "most of the vacuum lines" are you including the vacuum to the brake booster?

I'm not sure how best to check it, but the booster itself?

It doesn't not sound vacuum-y (sorry), and I'm harking back to the reference to the brake pedal going hard.

So I'd check everything from the manifold to and including the booster (e.g. check valve, fitting at booster, etc.) if you haven't already.

Yes, I guess most of the vacuum lines means unless I missed one.  I am including the lines from the brake booster.  Only had the hard pedal one time, but that's what led me to believe it was vacuum related possibly, and yes I am including the lines to the booster.  I blew through the check valve, it worked only one way, didnt seem to stick. The line with the check valve was old, maybe original.  It was dry and cracked.  My thought process was that the heat was causing the line to expand this opening the small cracks enough to allow for a vacuum leak on that line with the check valve.  That explained, in my head, the weird brake pedal, and why it corrects itself when it cools down.  Maybe a stretch!  Still no codes! No abnormal temp readings.