01 Mazda Millenia Remaining Bugs

At this point, the Mazda project has become tremendous success. The near free $100 car that turned into the Money Pit took a dramatic turn from boon to feather in cap upon the successful installation of Felpro multilayer steel head gasket. Doubtful this would have happened were it not for the excellent technical support I received from a select crew of automotive brainiacs at DMS, and my ultimate decision to pay careful attention to their suggestions. This is an excellent example of how a member can leverage expertise at DMS to drive an ambitious project onward to glory.

All in all, the car runs superbly. It is quiet, powerful, and I believe all systems are 100% go on the engine and mechanical components. The manifold pressure is correct. The car starts powerfully, and it shifts and drives like new. The suspension feels solid, and the steering wheel points straight ahead. The brakes are powerful, and all driving systems work as Mazda intended.

But I am writing for a reason, so lets get into my final set of problems I am facing with this car, as hopefully I can get a quick resolution to them so I can quickly return to my native art of electronics. Ironically, it is electronic problems that remain with the Mazda, and here is the breakdown:

  1. Cruise control does not work, shows no sign of life. It is actuated by steering wheel based buttons, and I cannot even get the light to come on. I suspect the buttons themselves. I think that they are disconnected, or perhaps the cruise control module is dead.

  2. AC and heat work in the cabin. But the system has a digital temperature set point and operates thermostatically. The temperature is stuck on 77 degrees. Fortunately, this is good enough to use AC on a hot day, and kick in heat on a cold day. But I want to be able to control this. And I believe it may be a problem that relates to other problems with the car, so fix it I shall.

  3. The check engine light does not come on. I have driven for 200 or so miles since last reset, and no check engine light. I have ELM327 OBDII connecting via Bluetooth to torque lite. When I run state inspection check, it says all items are ‘incomplete’. I have never seen torque lite deviate from this. When I take to DMS and use the Actron OBDII, the check engine light comes on, and the Actron tells me I have 1521 and 1522 which are codes for the two VRIS solenoid selectors (variable resonance induction system). Explained simply, the air intake has two butterfly valves that the ECU can control that can change the length, and thusly the resonant frequencies, of the air intake. I read online that when the engine comes on, one of them should come on. I have yet to see either of them come on, but I have checked the solenoids and they check out. When I unplug the Actron, the check engine light disappears. This problem screams of electrical short. Or maybe it has to do with the fact that I cannot dial in a Mazda on the Actron? One last important detail is that when I look at ready/not ready items in the state check list of Actron, the list looks the same as what I got from state inspection place. This points me more toward actron as good solution, and torque lite is somehow lacking. The check engine light and ghost error code points me toward Actron faulty, or my ECU is not handling the connection to this device well at all.

  4. For the first time, the car died unexpectedly at idle yesterday, while I was sitting waiting to pull out. I restarted and all was good. Short?

Now I will say this: I had a set of problems like this a few weeks back, and I got lucky and stumbled upon a blurb on some forum suggesting to replace this fuse as it caused one of the problems. I checked it, and yes it was out. When I replaced it, about 6-7 problems I had noticed went away. There are many items on each fuse, so its entirely possible that a single fuse or bad ground could be the source of the whole set of issues.

Any words of wisdom from the DMS experts? I would like to get these figured out quickly where possible, so I can get properly inspected and enjoy the car.

Your cruise control maybe due to a bad clock spring. As for the inspection there is no set mileage. The conditions have to be right. Look up the required drive cycles & that may help

1 Like

Cruise control bug is not a bug. I just didn’t see the switch to turn it on over to the side. And I discovered that all buttons do work on the steering wheel as cruise works, and the radio control buttons work now that I turned the radio on.

So only remaining bugs are temp stuck on 77 and VRIS code/state inspection debacle.

Not a bug. It is an undocumented feature… :slight_smile:

3 Likes

Feature indeed. Turns out that this is the default temp setting applied upon reset. I have ordered a replacement controller and amplifier ($30, ebay rocks, like always).

So now we are down to the unusual and inconsistent behavior between Actron and Torque Lite. Well no more, because I dug deep and ponied up the $5 for Torque full professional version. And would you believe that it is reflecting the same two 1521, 1522 codes. It says they are confirmed, but there is no check engine light. I don’t understand.

Torque pro has a lotta cool features. It appears to run through the less documented registers and verifies them. I hope I am able to set bits and see valves toggle. Researching that one next. I looked at inspection readiness, and it says incomplete for everything. I think it just does a reedy all or incomplete all. The actron tells us which ones, so there is value in the orange DMS unit.

1 Like

Will have to find my android tablet, (to see which exact version I have installed) but torque with my dongle on several fords definitely shows the system readiness per system.

2 Likes

Torque with ELM327 shows all the readiness criteria headings, maybe 10-15 of them. However, my Mazda is not ready, and so ALL of the headers show as INCOMPLETE. So were you to have a passing situation, all of them would likely show READY and so you may not be aware that the list is not broken down individually?

That being said, I have made advancement in the direction of understanding why the Mazda is not ready. Evidently, there is a mystical mode $06 for the OBDII specification. This mode $06 refers to the ability to bitbang registers within your car, which yield information about readiness and testing as your ECU performs continuous and non-continuous tests. One would need a higher end scanner to interpret a lot of the registers found in mode $06, but torque does interpret many of them for Mazda Millenia. I notice the Actron interprets many of them as well.

In addition, there is the concept of prerequisite requirements for some of the testing to be performed. Satisfactory completion of a given test is required prior to even starting a subsequent test. So a system may get hung on a given test, but that may not be the extent of the fixing required to fix the car. Once the first test is fixed, you may find multiple additional problems nesting nicely beneath the first. Thank you, OBDII. We needed this confusion in our lives… lol…

So my codes, not illuminating check engine light, but preventing successful completion of state testing are 1521 and 1522. Properly interpreted for my vehicle means VRIS solenoid 1 & 2 open or short conditions. I verified by taking the same style solenoid from a different part of engine (that I had previously replaced so I know it is new) and swapping it with one of the VRIS solenoids. When I did this, then carefully reset everything, I only got 1521 code. Well, this ie excellent news given that I have new solenoids arriving by friday. Sadly, this will be just beyond the 2 week date for state recheck, however even if I pay extra $25, I think I may be able to pass inspection with good solenoids installed.

Only catch is: what may be lurking beneath the solenoid errors? What tests are not being run due to the fact that the computer knows that the VRIS butterfly valves are not properly actuating? EGR and purge are in that area… more soon…

1 Like

Also, as far as readiness headers go, Actron seems to be superior in that it breaks down each header as to what passes and what fails. All I know from torque is that something failed, and if I were not more experienced, I would assume that ALL the headers failed from what I am seeing.

Fortunately, I got a printout from the state showing what was ready and what was not. Actron’s list mirrored this list exactly, with the tests not ready likely being the testing rendered impossible because 1521 and 1522 are gating tests for things like EGR and evap testing (still not ready after hundreds of miles…)

@kbraby

OK, well good fortune paid me a visit today when the mailman delivered my new solenoids early. And guess what…

Now Torque shows independent status on inspection readiness. Now, instead of getting all headers designated as INCOMPLETE, I now have some of the headers showing READY, while others show INCOMPLETE. I think that the reason for this is as follows: Torque shows all INCOMPLETE when a prerequisite test gating subsequent testing does not pass. Since the entire battery of testing is not possible, In this situation, Torque simply shows all testing as INCOMPLETE, whereas the Actron tester shows any test that did pass (and was obviously not gated like the others).

So now I have seen evidence explaining your assertion that the headers are indeed indiviedualized. Mine are showing up this way now as well. Darned VRIS solenoids!.. lol…

So I do not have completed state inspection readiness yet. As of now, I have no codes as 1521 and 1522 are now eliminated as problems. The incomplete state inspection readiness codes are: Catalyst, Evap, O2 sensor, EGR. I will have this data and will advise as soon as I have exercised the engine enough such that the ECU returns these results.

Thanks for the input, all!

I drove to downtown and back via 75 and now the O2 sensor and EGR passed. So now the remaining INCOMPLETES are down to only two:

Evap
Catalyst

If either of these passes before one fails, I can inspect. For year model 01 cars, they allow a single incomplete to arrive at a passing test. Fingers and toes crossed…

Check the o-ring on the gas cap. That can cause evap errors.

1 Like

You are likely driving too fast. Drive down an old country road. I’d recommend going down 455 in Sanger to Pilot Point. It’s a good little drive of about 55-60mph that set your computer.

Here is a link. http://forums.mazdaworld.org/100-millenia/107122-drive-cycle-needed-emergency.html#/topics/107122?_k=p4vh7b

& another https://www.mazdas247.com/forum/showthread.php?123759295-Found-a-successful-Drive-Cycle-formula!

& 1 more
http://milleniatech.net/codes/info/

2 Likes

You put your right foot in, and you shake it all about. Do the hokey-pokey and turn yourself around… thats how you pass inspection…

I really think that its just a matter of driving it enough to exercise the various systems under test as observed by ECU.

I think I might drive the Milly to the auto parts store and get a new gas cap. Might take a spin around Dallas. I reset it just yesterday after I put the new solenoids in. I think I’ll just keep driving around as needed for next couple of days. And if that does not woirk, then I’ll resort to hokey pokey…

Thanks for the info, gents…

I just put on a new gas cap. You could see that the old oring was a little bit dented from pressing against the fuel intake thingamajig.

So now that I have done this, and supposing that it did in fact make a difference, would I arrive at positive evap result faster to reset or rock on with yesterday’s reset?

And in reading the hokey pokey, I have additional questions about performing said hokey poikey:

  1. Does anyone have CID TID register map or know where to find this info? Mythical Mode $06 information specific to Mazda and the 929/Millenia?

  2. The hokey pokey version A refers to using an OBDII scanner to dteremine drive mode completion. I don’t remember seeing anything like this. What is this and how to check on Actron? ELM327-Torque?

Pretty much any obd2 scanner should tell you if the checks are complete. On my Bosch scanner, they show up as I/M monitors. I believe the scanner at DMS should have the same functionality.

1 Like

I think all you ever see is complete, or not compete. You never see % complete, or why not, until it throws a code.

1 Like

So now the problem all of a sudden got easier to solve. I was digging around in Torque, and I ran a set of tests that I had run earlier and gotten passing conditions (all green). But I did notice that some of the tests at the beginning were greyed out. My thought was that perhaps these tests did not function on a Millenia because the ECU did not support them.

Well, I was incorrect. Evidently, the VRIS solenoid errors were gating these tests. Because now when I run them, all of the tests run. These are the tests under ‘test results’ of the professional torque version. I have attached the results that I get now that the VRIS solenoids (all four solenoids of this type have been replaced on this engine, evidently they switch when tested using 12 V in elab, but yet still be worn out…I suppose due to heat and winding insulation melt issues). Which leads me toward what I believe may be the issue with the charcoal canister and its PURGE SOLENOID which is of a slighty different solenoid type (yes its already ordered).

But back on point, the failing test that is gating my catalyst is CID (component ID) $21, TID (Test ID) $01 with current value 396 exceeding Max limit at 365 for rich to lean sensor threshold.

Obviously, I have an O2 sensor giving me a funny value. Anyone understand what is happening better than I?

I plotted my four oxygen sensors wrt time. I wonder could there perhaps be a problem with the green one at the bottom?!?.. lol…

1 Like

I replaced that O2 sensor using a spare I had laying around. Now my waveforms look like this:

2 Likes