Index Home About Blog
From: (Bob Haar)
Subject: Re: OBD II computer system?
Date: 13 Oct 1995 15:16:40 GMT

In article <>, (George Gogis)

>I read about this a little and noticed that the 1996 model cars and
>trucks have this new OBD II system. I noticed that there are now three
>oxygen sensors instead of one.  I guess its to make sure the converter
>is not tampered with beyond this I don't understand why they are going
>to this new system. Can someone please write about this system or where
>to read about how it works since it now on all new cars. 

OBD II = OnBoard Diagnostics, level two.

This is being mandated by government regulations, first in California and
eventually across the entire U.S. It is being phased in starting with '96
models (actual some '95 vehicles have it, but only for part of the model

OBD II is a product of California Air Resource Board. Parts of the
regulations were developed in concert with the auto industry through
the SAE, other parts were pretty much dictated by CARB.

OBD II focuses on having the vehicle control systems monitor and
report information related to emissions controls. There are several
pertinent parts of this.

1 - OBD-II vehicles will have a "standardized" interface for off-board
diagnostic tools. This includes the connector itself, electrical
signal specifications, communication protocols and diagnostic messages.
There are some allowed variations so the definition doesn't proscribe
a unique interface. Part of this is the intent to make the diagnostic
information accessable by a "inexpensive" (under $1k) hand-held scan
tool that will operate with all OBD-II compliant vehicles without 
adaptors, etc. 

Note that standardization applies to only those messages and parameters
required by OBD-II. Outside of that, manufacturers can do anything.
However, their are reasons to use the same communications scheme for
networking other parts of the vehicle, but the message formats, etc.
are not covered by OBD-II.

The two main protocols involved are SAE J1850 and ISO 9141. Neither
on of these is directly compatible with RS-232 serial communications
so and add-on interfaces for PC's will require some additional hardware.

2- OBD-II requires additional sensing to detect emissions problems. Some
of these sensors will have no use in normal operations other than to
satisfy OBD-II requirements.

3- OBD-II vehicles will be monitoring and recording "emissions events"
when the vehicle is operating in ways that might violate emissions
regulations. These stored event records can be downloaded through the
OBD-II interface when your car is being serviced or getting a periodic
emission test. It is unclear what happens if your vehicle is running
outside the allowed limts for 6 months and you get it fixed the day
before you have an emission check. 

All this has the potential for making the diagnostic information more
accessable and useful.

There are a number of SAE conference papers and information reports
available. If you have access to a major engineering library, they
are likely to have copies. Otherwise, SAE will sell you the papers.
Unfortunately, there is no net access to these reports that I know of.

  	Robert Haar               InterNet : 
	Computer Science Dept., G.M. R & D Center
DISCLAIMER: Nothing in this note is an official statement of General Motors
Corp. AT BEST, it is personal opinion based on experience. TANSTAAFL

From: John De Armond
Newsgroups: rec.outdoors.rv-travel
Subject: Re: Gas Cap Thats Faulty
Message-ID: <>
Date: Mon, 08 Apr 2002 23:45:53 -0400

On Tue, 09 Apr 2002 00:11:31 GMT, "aubrey and carol" <> wrote:

>I have a 99 Winnebago Adventurer 35 ft with the Ford V10 Engine.
>My problem:
>The "Check Engine Light"  comes on every time I fill the gas tank. This
>happens about 20/50 miles down the road.  My way of resetting the computer
>is to stop the rig, cut the motor and than disconnect the chassis battery
>for about 5 min.   Re- connect battery, wiggle the gas cap and start up and
>proceed down the road, the problem doesn't return and I can travel for 3/400
>miles without the check engine light coming on.
>However, refuel, and the light comes on again.  Has anybody had this
>problem?  How can I cure the problem.?  Ford tell me they don't have any new
>filler caps.  Napa and AutoZone have them but, they don't fit.

Yep.  You're getting an OBDII code, probably "P1443  Evap emission control
system purge valve Circuit malfunction." or something in that range.  The most
common cause of this code is a leaking gas cap.  The obvious solution is to
tighten your gas cap more carefully.

While disconnecting your battery lead will reset the code, that's like using a
sledge hammer to kill a fly sitting on a crystal bowl.  There is collateral
damage in the form of lots of lost diagnostic information plus the learned
behavior (where your engine tunes itself and then remembers the tuning).  At
best, the loss of the learned parameters costs you mileage for awhile.  At
worst, you've deprived your mechanic of valuable diagnostic information.

>Any suggestions.? Why does the check engine light come on? I feel that it
>may have something to do with a build up of pressure in the tank or
>something, and there could be some sort of sensor that  reports a gas vapor
>leak  to the computer. Am I on the right track?

In very simple terms, the OBDII computer tests the integrity of the vapor
recovery system by monitoring certain parameters while the charcoal canister
is purges.  One parameter is the rate of change of pressure in the gas tank.
This test detect any tampering with the system and also detects a loose gas
cap, failed hoses and similar problems.  There are several codes in this class
so when the code is read with a scanner, you or your mechanic will have much
more information for looking for the problem.

If you've been getting a check engine light since the rig was new, there is
another possible cause.  That is, the volume of the fuel tank, filler neck and
associated hardware may be so much higher than what the OBDII computer expects
that the pressure doesn't drop fast enough even though the system is actually
intact.  If that is the case, the only remedy is to have the computer properly
programmed.  Either Ford or your coach mfr should be able to do that for you.
The mechanism would be for either Ford or the coach manufacturer to prepare a
patch file that they send to a dealer for actual uploading to your computer.

I'm going to bet that the problem is much simpler.  I suspect that your
problem is with the gas cap - either you're not tightening it properly or else
the cap is faulty.

I'm not sure what you mean about Ford not having a cap, considering that
federal law requires them to stock spare parts for 10 years.  Is this a dealer
telling you that?  I'd find another dealer.  I had a chevy light duty truck
dealer tell me that my OEM shifter on my RV wasn't factory so I went to the
area Chevy commercial truck dealer where they knew what a school bus console
shifter was.

It might be simpler to just find one at the junkyard.  Before I got really
wound up on a search-and-destroy mission, I'd do a simple test.  I'd seal off
the filler neck securely with something known to be air tight - say a rubber
ball.  I'd smear some grease around where the ball meets the filler to make it
absolutely air tight.  Then drive the rig and see what happens.

Note that the persistent code which lights the check engine light will not be
set until the system fails more than one test.  With a scanner hooked up,
however, you can see the transient code every time it is set.

Which brings up the next subject.  If you're going to fix this yourself,
you're going to have to have an OBDII scanner.  They range from simple
handheld units that read out numeric codes to dedicated hand-held computers to
the one I have that uses a PC as the host.  I use this one:

This consists of software he calls an OBDII Browser (you can download the
browser and play with it without paying) and a protocol translation cable that
plugs into the OBDII connector and the serial port on your computer.  This
browser will not only dump the diagnostics, it can also reprogram the ECU and
it can act as a flight recorder, recording any desired parameters to a file
that can be played back and analyzed.  Great for catching transient problems.

Disclaimer - I'm far from an expert on OBDII and I know chevy much better than

Claimer - I diagnosed this very same problem on my mother's Ford van a few
weeks ago after various mechanics had clipped her about $500 for everything
from changing the PCV valve (completely different diagnostic code) to changing
the throttle body.  What I actually diagnosed was my father's not tightening
the gas cap very well :-)

To summarize: quit dumping all the diagnostics by pulling the battery lead and
either get a scanner or find someone who has one AND knows how to use it.  And
make SURE you're tightening the gas cap properly and its gasket is intact.


From: John De Armond
Newsgroups: rec.outdoors.rv-travel
Subject: Re: Right to repair
Date: Wed, 03 Dec 2003 22:16:56 -0500
Message-ID: <>

The controversy centers on access to manufacturer-specific diagnostic codes
and protocols.  EPA requires a standard set of emissions-related coded in the
OBD-II spec.  The diagnostic port that the mechanic plugs his scanner into is
now as of OBD-II actually a network interface to the Car Area Network (CAN).
Many other non-emissions devices can and do reside on the bus.  Things like
ABS, airbags, internal environmental controls, entertainment and so on.  The
various OEMs have tried to keep these codes secret for the obvious purpose of
funneling repair work to the dealerships who have the necessary diagnostic

One of the Fed agencies (FTC I think) stepped in a couple of years ago and
threatened to sue to force the disclosure of all diagnostic codes.  The OEMs
reached an agreement and signed a consent decree last year.  This decree isn't
perfect in that it allows the OEMs to charge for access to the codes.  But
it's a good start.  Supposedly one can now log onto a GM website, for
instance, and gain access to all their codes.  I don't know the URL so don't

I'm not at all familiar with this bill being promoted but I'd assume it would
force more openness than the decree achieved.  Knowing washington, it's also
probably loaded up with pork.  I do know that the Automotive Aftermarket
Industry Association is a lobbying group for the manufacturers who make 3rd
party diagnostic instruments.

The situation wasn't as bad as that organization made it sound.  Many of the
protocols have already been hacked and have been incorporated into aftermarket
diagnostic hardware and software.  But it WILL be a lot nicer not to have to
spend the energy reverse engineering every variant of every controller on
every car.


On Thu, 04 Dec 2003 00:06:19 GMT, "Gil" <> wrote:

>Has anyone seen this?  Truth or hoax?
>I searched and couldn't find out one way or the other.

From: John De Armond
Newsgroups: rec.outdoors.rv-travel
Subject: Re: 1999 - Ford E450 - Service engine Soon
Date: Fri, 18 Mar 2005 02:58:38 -0500
Message-ID: <>

On Fri, 18 Mar 2005 02:47:29 GMT, "Traveling"
<> wrote:

>My 1999 Ford E450 Van Chassis has just illuminated the "Service Engine
>Soon" warning light after turning 80,000 miles. I've checked the items
>listed in the manual, including checking the dip stick to insure it's
>seated. The motorhome runs normally.
>I'll have the motorhome serviced soon but is there a way to turn off
>this light manually in the meantime?

Well, I did it on my mom's van (after doing a scan to make sure
nothing was wrong that mattered) with a piece of black electrical

Through the OBD-2 port one can reset all the non-continuous test
results which will extinguish the MIL (malfunction indicator light).
Depending on what is wrong, the MIL will return in anywhere from a few
minutes to as long as 4 or 5 days.  Certain non-continuous test
results will clear after a specified number of successful runs.  The
classic one is the evaporative emissions system leak code caused by
leaving the gas cap loose.  It'll reset in a day or two (actually a
certain number of qualifying "trips") after you tighten the cap.

I highly recommend that anyone who owns a 96 and above vehicle acquire
one of the inexpensive OBD-2 scan tools that are now available.  I
think I saw one at Wallyworld the other day in the $40 range.  These
tools only display the OBD code (Pxxxx for powertrain problems, for
example) but there are many places on the net to look up the meaning.

Even if you never plan on turning a wrench yourself, the scan tool
allows you to protect yourself from the many thieves who pretend to be
mechanics and service writers.

This whole concept of charging a "diagnostic fee" for doing a scan is
just so much smoke and mirrors to take advantage of the uninformed.
Does the mechanic charge an extra fee for hooking up a DVM to find a
bad switch?  Noooo.  A separate fee for using a feeler gauge for
setting valve clearance?  Nooooo.  Those are merely the tools needed
to perform the job that they're already being paid for.  Just like the
scan tool is necessary to diagnose the computer systems.

Funny this question would arise now.  I had an exquisite moment
yesterday regarding this same thing.

My manager recently bought a used late model Lincoln from the local
dealer.  Both EPA and OEM warranties still apply.  Her MIL came on a
couple of days ago.  Today I agreed to follow her to the dealer and
give her a ride back home after she dropped off the car.

So I'm standing there listening to this service writer give her the
typical BS that females get.  Even though the EPA 100kmile emission
warranty precludes such practices, he told her that she'd have to pay
an $85 "diagnostic fee" to "hook the car to the computer".  I was
about to jump in when I had a better idea.

I recalled that I had my laptop bag in the car.  I have a trimode
network gateway from that turns the laptop into
a scan tool.  A very good scan tool.  Without saying a word, I got the
laptop out, connected the gateway, retrieved the trouble code and
turned the screen around to the service writer.  I suggested that he
write down the code so that he'd not have to use his expensive
computer.  Then for good measure I reset the MIL.  It'll come back on
in a couple of days but that's beside the point :-)  The fault code is
now clear so he has no justification at all to hook up his machine and
still try to charge her.

I wish I'd had a camera handy.  The look on that guy's face and the
smirk on my manager's face was absolutely priceless.

Before I disconnected the gateway I scanned all the other computers on
the bus (airbomb, ABS, climate control, etc) for codes (nadda) just to
make sure they didn't "discover" any other non-warranty problems to
nick her with.  I made a big ceremony of saving the dataset to disc
while telling my manager in a fairly loud voice that we now have a
record of the car's state as it entered the shop.

That service writer was pissed.....

One of the things I can discover during the scan is the elapsed time
since the MIL was reset.  Thus I can see if they just reset the thing
to get her to go away instead of doing the repair.  I also know what
parts to look at to see if they're new.

These petty thieves that work for car dealers just don't have any idea
how technology can be used against them.  Hehehe.


From: John De Armond
Newsgroups: rec.outdoors.rv-travel
Subject: Re: Service engine Soon (What About Brake Lights)
Date: Fri, 18 Mar 2005 13:13:32 -0500
Message-ID: <>

On Fri, 18 Mar 2005 06:16:37 -0500, "Steve Wolf" <>

>These are a must and the tool cited works great.  You can also hook the
>thing up to your computer and drive around while monitoring many functions
>like fuel pressure.
>READER???  I am about 90 percent sure the front right brake sensor on my
>Chevy chassis has bit the dust but Chevy is keeping the codes and how to get
>them a secret.  Pisses me off enough to keep me from every buying a Chevy.
>It looks a lot like I have to pay for them to read their tool.

Check the latest version at  Mr Peper has been
incorporating the non-standard codes into the software as rapidly as
he can.  I used the program last summer to look over the shoulder of
the guy struggling to repair my mom's MH's transmission.  I think I
saw the antilock brake controller listed on the net.

Chevy can't keep the codes secret.  In June 2003 the big three signed
a consent decree with the EPA to make all dealer service information
and equipment available to anyone.  EPA issued regulations imposing
the requirement on all OEMs making more than 1000 cars a year.

Here is the web site:

This site contains links to the Federal Register where the regulations
were issued.  It also contains links to each OEM's compliance site.
You can get a 24 hour subscription for about $20 and download
everything you can during that period.

There are two major flaws with the EPA plan that they recognize and
are accepting comments on.  One, they did not require the OEMs to
provide the info in downloadable format (PDF, etc).  Ford seems to be
doing so anyway.  Can't tell about the others.  The other problem is
they didn't regulate prices.  Since the OEMs are abusing this omission
($6,000 for a handheld scantool (Chrysler), thousands of dollars for
an annual subscription) it appears EPA will remedy the problem,
particularly if they get sufficient comments.


From: John De Armond
Newsgroups: rec.outdoors.rv-travel
Subject: Re: Service engine Soon (What About Brake Lights)
Date: Fri, 18 Mar 2005 20:43:50 -0500
Message-ID: <>

On Fri, 18 Mar 2005 16:59:58 -0500, "Steve Wolf" <>

>It does look as if he added brake codes.  Time to upgrade!  Thank you!

You be welcome.  I've learned to drop in on his site every month or so
to upgrade.  He's a busy lil' beaver.  I just wish he was a better
programmer.  His software runs my laptop flat out.  Cooling fan on
high and about 10 minutes from the battery.  I have to drag out a cord
or inverter when I scan.

>In Ohio the new EPA test involves their hooking up to your OBD port and
>reading any flags.  Well, first, if your check engine light is blinking you
>fail.  They read the port and if you have flags, you fail.  If you don't
>have any flags, you suffer a gas cap pressure test and you're done.  OBD
>tools can pay for themselves in pre-diagnosing problems, fixing them, and
>clearing the codes prior to getting your test done.

Same here.  City officials caved in to the EPA without a fight even
though the data suggests Chattanooga meets the silly new standards.
That the city is going to reap a windfall probably didn't have a thing
to do with it....

I think there might be a good market for a little data wedge that
would sit under the dash and capture that mode 1, pid 1 request that
the "inspection" is based on and return an "everything is wonderful in
lala land" return code.


From: John De Armond
Newsgroups: rec.outdoors.rv-travel
Subject: Re: RV tech forum?
Date: Wed, 23 Mar 2005 14:35:18 -0500
Message-ID: <>

On Wed, 23 Mar 2005 08:30:19 -0700, cew <>

>(Where/) Is there an RV tech forum?
>Looking for an engine running 12v wire in front of the firewall of a '97
>GMC Sierra.

We do a few tech things here in and amongst all the infighting.

Can't help you on your particular model other than to say that GM's
color codes are fairly standard.  Black with a yellow tracer is
usually ignition power while pink is usually accessory power.

Something you and others might find useful is this link:

A few years ago the big 3 and EPA entered into a consent decree
mandating that they make all dealer service info available to
everyone.  EPA then wrote regulations that applied to all OEMs making
1000 cars a year or more.  Each OEM has to have a web site where
service info can be viewed.  They have all done so.  The above link is
from EPA's audit site.  It links to every OEM service web site.

For about $20 you can buy a one day subscription and access everything
GM has for your model vehicle including the service manual.  24 hours
is enough time to download pretty much everything you need.


From: John De Armond
Newsgroups: rec.outdoors.rv-travel
Subject: Re: Ford V10 Brake, ABS Lights On
Date: Fri, 25 Mar 2005 06:54:54 -0500
Message-ID: <>

On Fri, 25 Mar 2005 09:42:33 GMT, "Dennis Vogel" <>

>Thanks for the good advise guys. How will I recognize the speed sensors?
>The lights came on when I was parked in the driveway if that means anything.

You really need to get the fault code dumped, as we're just guessing
at this point.

On the front wheel brake rotor there will be a little toothed wheel
and a sensor near it with wires coming out.  It'll be the only thing
on the spindle assembly that has wires.  Generally on large vehicles
the rear wheel speed is encoded off the transmission output shaft by
the Vehicle Speed Sensor or VSS.  Usually stuck in a hole at the end
of the transmission output housing where the speedo cable would have
gone.  If your vehicle has 4 wheel ABS (usually a sticker somewhere)
then each rear wheel will have a toothed encoder also.

Unless you get lucky and find a wire dangling or the sensor knocked
loose, there really isn't much you can do without knowing what the
fault code is.  You can look at the signal on an O'scope with the
wheel turning but it doesn't matter how good the signal looks if the
control module doesn't like it.

Poke around and look for loose wires but after that it's a matter of
ponying up for a code dumper or scanner or ponying up for a mechanic.
I highly recommend the network gateway and software from here:  Only slightly more expensive than the auto
parts store code dumpers but vastly more capable.


From: John De Armond
Newsgroups: rec.outdoors.rv-travel
Subject: Re: OBD-II  tool
Date: Fri, 10 Mar 2006 01:46:24 -0500
Message-ID: <>

On Thu, 09 Mar 2006 16:17:35 -0700, AJ <> wrote:

>   I'm looking for an OBD-II  scan tool for use on my Jeep Liberty.  I
>want one that uses my laptop as the scanner.  There seem to be a bunch
>of them out there and the prices are all over the map.  Does anyone use
>any of this software/hardware and anyone have any recommendations.

This is the one I use and recommend.

Excellent price and free software updates forever.  I recommend
getting the Tri-CAN, as it will cover present stuff and all the future
CAN-bus stuff.


From: John De Armond
Newsgroups: rec.outdoors.rv-travel
Subject: Re: OBD-II  tool
Date: Sat, 11 Mar 2006 13:28:15 -0500
Message-ID: <>

On Sat, 11 Mar 2006 07:12:14 -0800, Ralph E Lindberg
<> wrote:

>In article <d0pQf.35944$>,
> "Nate" <> wrote:
>> Does this unit allow the user to initiate component tests?  This would
>> require two way communications.  I could not find mention one way or the
>> other.
>  Not that I remember (I use the same tool)

Yes it does.  The network interface (the formal name for the OBD-2
cable) is bi-directional.  He is incorporating the proprietary codes
as they become available.  Thanks to the EPA finally enforcing the law
and forcing disclosure, this is happening rapidly.  I am now able to
do things like energize the cooling fan, actuate transmission
solenoids and the like for vehicles that permit that.

Even without him incorporating the codes into the software, if one has
access to them, one can initiate the function through the SEARCH tab.
One can execute any function that is meaningful anywhere on the bus
with this tab but it is not for the faint of heart and it can be
dangerous.  The wrong code can write trash to a module's flash memory,
rendering that module inert until a service center can re-ert it.

I have used it exactly once to trigger shift solenoids in my mom's
MH's transmission.  The next upgrade in the software included those
codes.  Just my luck :-)

I recently compared this tool to one of those $5k OTC tools that my
neighboring mechanic has and for the car in question (Honda Accord) I
could do everything he could and my diagnostic messages are more

The ultimate will be when he cracks the encryption and enables us to
reprogram PCMs without having to pay a dealer.  If anyone can do it,
IMO, Mr Peper can.


From: John De Armond
Newsgroups: rec.outdoors.rv-travel
Subject: Re: OBD-II  tool
Date: Fri, 17 Mar 2006 17:37:30 -0500
Message-ID: <>

On Fri, 17 Mar 2006 13:57:26 GMT, GaryO < @ . > wrote:

>Do you know if the device at has software for use on a
>PDA or Palm?

I'm pretty sure he doesn't.  He seems hardware oriented with the PC
software just a necessary evil.  The software is ill-behaved.  It
spins for keystrokes instead of halting so the CPU runs at 100% all
the time.  Laptop battery life is awful.  I don't even try.  I have
one of those little 100 watt cig lighter inverters attached to the end
of my laptop's power supply cord.  I plug that into the target car's

>I'd like to be able to do diagnosis, data capture, etc on either a PC
>or a PDA.  The PDA is especially handy for just data capture while
>driving without needing to use the notebook.  Besides, older PDAs like
>the Palm m105 are now available on eBay for $20 or so.

I've thought about the PDA but I don't think the screen is large
enough to be useful.  When I'm trying to figure out a problem that
isn't directly related to a code, I like to have several datalog
parameters on the screen at once.  The type would be so small on a PDA
screen that I couldn't read it, especially when driving.

>I see that supports PDAs, but they charge extra
>if one wants both the PC and PDA capability.  Seems a bit pricy.

I've stayed away from those types of companies.  They seem to be
infected with the same greed that drives the proprietary code reader
companies like OTC where they nickel and dime you (or more like $10
and $100 you) to death with tack-on fees.  My garage buddy across the
street pays several thousand $$$ a year for updates and stuff for his
OTC.  I think I got his attention with mine :-)

>I have found this place which has free
>Palm software for the ELM devices.

It seems to me that ELM has established a de-facto standard for serial
OBD-2.  I'm reasonably sure that other hardware will work with Peper's
software and vice versa.


>A mix of both commercial and free software for the ELM devices can be
>found at:
>       .........gary
>On Thu, 16 Mar 2006 12:44:57 -0500, Neon John <> wrote:
>>Kewl.  Thanks.  I knew about Elm's single protocol chips but I didn't
>>know about this multi-protocol chip.  I'd not be surprised to hear
>>that Peper is using this chip in his Tri-CAN cable.  I'm about to
>>trade in my pre-Tri-CAN cable for one.  Got to stay up to date,
>>dontcha know? :-)
>>On Thu, 16 Mar 2006 12:38:27 GMT, GaryO < @ . > wrote:
>>>Here is a PIC which does: CAN (ISO 15765-4), SAE J1850 PWM, SAE J1850
>>>VPW, ISO 9141-2 and ISO 14230-4.  I've noticed a lot of devices
>>>springing up lately which utilize it, along with both commercial
>>>software and freeware.
>>>      ........gary

From: John De Armond
Newsgroups: rec.outdoors.rv-travel
Subject: Re: Tranny temp gauge - easiest way?
Date: Mon, 25 Jun 2007 16:12:24 -0400
Message-ID: <>

On Mon, 25 Jun 2007 14:27:17 -0400, Steve Wolf <> wrote:

>Hell, these will standard equipment.  What a great idea!  I'm in and
>will report it back to the group.
>Anyone already have one?

It should work fine.  All it does is receive and display OBD-2 data.   There are a
variety of these things available, both in kit and ready-to-go form.  For about $10,
you can buy an OBD-2 to serial converter chip (basically an 8 pin PIC processor
programmed to convert the protocol) and roll your own.

Two things to be aware of.  First, sending data to the OBD-2 port is something the
PCM does when it's otherwise bored and unoccupied.  When the PCM gets busy, data
output can be spotty.  Don't depend on this for real-time data such as for a tach.
Oil, air and transmission temperatures and stuff like that are fine.

Second, even though the values are presented in engineering units (degrees F, RPM,
etc), the system is NOT calibrated.  The temperature sensors in particular, are very
inexpensive mass-produced thermistors.  They don't need to be calibrated, only
repeatable, for the intended task of powertrain control.  Don't be surprised if you
see some "interesting" readings or that readings change over time and especially if
you ever replace a sensor.


From: John De Armond
Newsgroups: rec.outdoors.rv-travel
Subject: Re: Tranny temp gauge - easiest way?
Date: Wed, 27 Jun 2007 15:28:50 -0400
Message-ID: <>

On Wed, 27 Jun 2007 12:14:25 -0400, Steve Wolf <> wrote:

>You need accuracy in your fuel gauge, I don't.  No problem there.  I
>stopped shopping for the cheapest when I let my motorhome go under an
>eighth of a tank, uncovered the pump and had to screw with replacing
>that over the next few months.  No, I get gas whenever there is a
>possibility I might uncover that pump.

I can understand that.  I have no submersible pumps and I've gone to some trouble to
get that way, removing the offending critter and replacing it with an external Bosch
roller pump.  This is the one that has remained essentially unchanged when it first
appeared here in the 60s on the Volkswagen Type II station wagon.  In over 40 years
of working on cars, I've yet to see one of those fail.  Anyway, side tangent.

>I understand that a cheap meter might create problems.  Let's bring the
>discussion back to the OBD port.  Let's assume that the meter accurately
>reads the data and displays it properly.  Now the issue is the data from
>the OBD port.  Are you suggesting that the data from a typical OBD port
>would be so bad as to preclude it from being useful?

No, of course not.  The sensors are as good and probably better than any aftermarket
kit.  I just don't want one to get overly reliant on the numbers.  If the
transmission oil temperature limit is, say 250 deg (no idea what it is actually) and
the OBD2 instrument is saying 246, don't think "ah, I have another 4 degrees to go
before worry".

The PCM has some very sophisticated algorithms (some that I've "borrowed" for other
projects) to accomodate sensor drift.  Basically, it intercompares readings and fits
them into a model that harmonizes them.  As long as the drift is slow and within
bounds, the model simply applies a new scale factor.

When I was working on the MegaSquirt DIY EFI project, I did a study of GM air and
water temperature sensors to determine uniformity.  I wanted to know if we could
simply hard code sensor constants in the software or if each sensor would have to be
calibrated and the constant inserted during tuning.  I did my survey very simply - I
went around town with my DVM and thermometer and measured every GM sensor on the
shelves of the various parts houses.

They were close enough for engine management work but they did vary significantly. No
surprise that GM did THEIR homework too and specified the loosest manufacturing
tolerance that they could live with.  We were able to leave the hard coded sensor
constants in place.  However, it did open my eyes to trusting literally the data that
comes from the OBD2 bus.  My Peper-code ( displays temperature
and most other units to 1 decimal place.  That last decimal is noise.  The Units
digit is also probably mostly noise - I haven't calculated that yet.

>While I didn't do a study, the data I see on my PC when driving around
>appears to be good.
>Again, I am not interested in knowing the transmission fluid temperature
>to the tenth of a degree.  I don't want to pay for that and wouldn't use it.

Me neither, of course.  It's irrelevant.  The point I want to get across is the one
that was hammered into us nuclear reactor operator trainees back at TVA - "Trust your
instruments - but not too much".


From: John De Armond
Newsgroups: rec.outdoors.rv-travel
Subject: Re: OBD1.5??
Date: Thu, 10 Jan 2008 10:08:54 -0500
Message-ID: <>


There's a grain of truth there.  In 94 and 95 GM went to the OBD-2 connector format
but continued actually using the ALDL electrical interface.  My 94 Caprice 9c1 that
uses the same PCM ("8051") as the Corvettes, is that way.  I had to make up an
adapter to let my ALDL PC interface work with this car.  The common slang for this IS
"obd-1.5" but that's misleading, because only the physical connector is OBD-2.

I'd be afraid to make generalizations, given how much flux there was within GM in
those years but I can indeed read codes by shorting a pair of pins on my Caprice.  I
don't bother since I have a PC interface but I have tried it.

I'd be surprised if shorting any two pins on the diagnostic plug could cause damage.
I've had a variety of GM PCMs apart and have been quite impressed with the protection
built in.

What he might have done was scramble the Flash memory in which the PCM's programming
was stored.  This was a problem back then.  I cross my fingers every time I flash the
memory on my 8051 PCMs.  Re-flashing the PCM with the proper firmware for his model
car would probably have fixed things but back then that was a dealer-only (and us
hackers, of course) option.


From: John De Armond
Newsgroups: rec.outdoors.rv-travel
Subject: Re: OBD1.5??
Date: Thu, 10 Jan 2008 17:52:16 -0500
Message-ID: <>

On Thu, 10 Jan 2008 15:34:34 -0500, HD Matt <nospammbode@multiprintinc.comnospam>

>Is there anyway to rig an adapter so I can use my Scangauge II on the
>"ODB 1.5" connector? Apparently my MH is of that ilk and the wife got
>me a scangaude for Christmas?

Assuming the ScanGauge II is an OBD-2 scanner, no, not possible.  The only thing
OBD-2 and so-called OBD-1.5 have in common is the physical connector.

The ALDL protocol is completely different than the OBD-2 protocol.  It is a low speed
(around 8kbps) serial protocol.  Both the baud rate and the voltage levels are
oddball but not odd enough that a PC can't handle it.  My cable has a DB9 on one end
and the ALDL connector on the other and some level shifting electronics in between.
All the level shifting involves is changing the voltage level (5 volts, I think) at
the connector to RS-232 levels.  Look around the net.  There are dozens of DIY

There are all sorts of readers too.  I use a freebie from here:

It may or may not work for your PCM.  GM, like the others, were all over the map on
diagnostic protocols, the main driver in EPA requiring OBD-2.  You can try it, of
course.  It won't do any harm if it doesn't work.  A commercial package that I know
of that can do the old stuff with the proper interface cable is AutoX-ray.  Fairly

Your PCM may be flashable.  GM was still using CALPAKs (eproms) on some PCMs in that
time frame.  Others like my 8051 got Flash memory.  If yours is flashable then much
tuning can be done.  I have software for that too.  It's a commercial package that
uses a personality file for each PCM that it can program.  It works on a memory
image.  I have freebie software to upload and download the flash images from the PCM.

I focused on the Caprice/Vette, the Buick Grand National and the Pontiac Sunbird*
PCMs mainly so I'm not all that "up" on the others.  Lots of info on the net, though.

*that PCM was/is a wonderful universal hotrod PCM, having both speed-density and mass
air flow mixture control, turbocharge wastegate control and knock sensor feedback.
I've used it on several non-GM vehicles including a motorcycle engine on a Formula
SAE race car.

I should say here that there isn't all that much diagnostic info available on the
ALDL interface.  Nothing like OBD-2.  I've seen a caprice 9c1 engine barely running
because of an air leak, broken plug wire, fouled plug, etc and never throw a code.
Unlike OBD-2 that will generally point you to the right subsystem if not the actual
defective part, ALDL is just another tool in the diagnostician's bag of tricks.


From: John De Armond
Newsgroups: rec.outdoors.rv-travel
Subject: Re: CarMD
Date: Sat, 22 Mar 2008 17:06:03 -0400
Message-ID: <>

On Sat, 22 Mar 2008 14:43:16 -0400, bill horne <> wrote:

>tim fm ct wrote:
>> I just read an ad about what seems to be a pretty handy item..It called
>> a CarMD and it checks out engine specs and problems. good Sam offers a
>> $20 discount on it.
>> Does anyone have any experience with the engine checker?
>>                                               Tim fm CT
>While you're waiting for somebody who actually has one:

And as usual, they're wrong about one major point. (They're running neck-and-neck
with Wikipedia, it seems.)


Turning the "check engine" light off. Unlike many other OBD-II diagnostic tools,
CarMD cannot reset a check-engine light, so all problems that activate a permanent
trouble code--even an easy-to-fix problem such as tightening a loose gas cap--require
visiting a service station and potentially paying to have the light reset.

This is wrong and mostly meaningless in OBD-2.  Failures of discontinuous tests such
as the evaporative emission system pressure drop test that infamously signals "loose
gas cap" reset themselves after the test is subsequently passed successfully.  Three
times, in the case of this test.  And if a tool is used to reset the light then it'll
just come back on again after the test fails again.

Resetting the light doesn't work to try and pass emission testing either.  The PCM
reports over the bus the status of all the emissions-critical tests, continuous and
discontinuous.  For the major discontinuous tests, the elapsed time since the last
change of state is reported, as is the time since the last MIL light reset.  The
emission station computer will immediately reject any vehicle that reports a recent
reset OR any discontinuous test that has not run to completion.  The latter state
happens when either the MIL is manually reset or power has been interrupted to the
PCM and insufficient time has passed for the tests to have completed.  Some of the
discontinuous tests take as long as six days to complete (measured in terms of engine

The lack of a MIL reset capability is of absolutely no consequence to this

Having the OBD-2 connector hard connected to the instrument body is the largest
problem I see.  A significantly large number of vehicles have the port hidden to the
driver's side of the center console.  That includes my mom's MH.  The instrument is
useless in those instances.


From: John De Armond
Newsgroups: rec.outdoors.rv-travel
Subject: Re: New Truck mileage?
Date: Sat, 29 Dec 2007 23:09:48 -0500
Message-ID: <>

On Sat, 29 Dec 2007 20:01:19 -0500, bb <> wrote:

>On Sat, 29 Dec 2007 19:04:58 -0500, Neon John <> wrote:
>>On the other side, the engine's PCM computes mileage by keeping track of the mass of
>>fuel injected over time.  The algorithms are quite complex but in general, unless the
>>injectors are highly different from stock (worn or partially plugged) AND in the case
>>of gas engines, the oxygen system is also defective OR tire size has been changed
>>without reprogramming the PCM, the computed mileage is quite accurate -
>How would a clogged or worn injector throw off the mileage
>calculation?  One gallon of fuel moves the vehicle X miles.  If you
>have a bad injector you probably get less miles per gallon.  Shouldn't
>the computer still be able to calculate the actual miles per gallon,
>no matter how poorly the engine is running?

There is a range of lambda between lean surge and fairly rich where the power output
is about the same.  The mileage delivered can be significantly different, however.

  If the mixture is lean of stoich then more energy than normal goes out the exhaust
as unrecoverable heat (entropic energy) in the excess air.  Another portion of energy
goes out the exhaust in the form of incomplete combustion.  A lean mix is slow
burning.  Combustion isn't complete by the time the exhaust valve starts opening. The
power stroke is effectively over when the exhaust valve cracks so any burning going
on at that time is wasted. That's why lean running engines will often heat the
exhaust manifold red hot.  This applies to gas but not diesel engines, of course.

On the rich side, unrecoverable energy leaves in the form of the potential energy
still remaining in the soot, carbon monoxide and unburned hydrocarbons. This applies
to both gas and diesel engines.

The injectors are characterized in terms of grams of fuel flow per unit time per PSI
(or kPa if you insist :-)  of differential pressure.  This characterization is
contained in a three dimensional lookup table or map in the PCM.  To determine the
total fuel consumed, one only need sum the flow for each injection event as computed
from the open time and the differential pressure across the injector.  The latter is
computed from the fuel pressure and the manifold absolute pressure.

To compute mileage, after a lot of software handwaving, a moving average of the
distance traveled divided by the totalized fuel flow over the interval is calculated.
The distance traveled is computed by totalizing the input from the VSS sensor.

Any variation in the actual injector fuel flow vs the modeled fuel flow will lead to
errors.  Variations can come from clogged injectors (lower flow), worn injectors
(generally higher flow) or variations in fuel pressure when that pressure isn't
actually measured.

That describes open loop operation, what diesels and gas engines in the power regimes
and before the oxygen sensor(s) warm up operate in.  With a gas engine in closed loop
operation - the O2 sensor working and lambda feedback control in operation - the
situation is a bit different and much better from a mileage calculation point of

The PCM receives an air flow signal in grams per second from the mass airflow sensor
(MAF) (or from a lookup table based on manifold pressure, barometric pressure and
engine speed in the now fairly rare speed-density systems).  Based on a number of
parameters including engine speed, the system looks up in a volumetric efficiency map
the correct amount of fuel to inject.  This is initially based on the same injector
model as described above.  But in this case there is feedback.  The oxygen sensor. If
oxygen sensor reports richer or leaner than what the VE table calls for then the PCM
re-computes that point on the VE map.  These corrected values are used to trim the
injector open times and to compute mileage.

If you have an OBD-II reader, you'll see the "public" result of these computations in
the short and long term trim values.  Short term trim is the instantaneous VE trim,
literally computed over each combustion cycle.  Long term trim changes slowly to
bring the short term trim to average near zero.  The short term trim is reported only
once a second per the OBD-II spec but in modern engines it is computed on a
per-exhaust event basis.  This computation is part of the misfire-detection algorithm
that is required as part of OBD-II.

The only really critical instruments in this scheme are the MAF and the oxygen
sensor.  Mainly the MAF.  It has to have sufficient long term accuracy to accurately
report the incoming air mass.  Everything else can be computed.  The O2 sensor is
necessary for trim but it's a bi-stable device in most cars - indicating either lean
or rich but no proportionality.

Closed loop operation occurs at idle, cruise and during mild acceleration, the
regions of operation measured as part of the Federal Test Cycles for emissions
compliance.  At other times, moderate acceleration, wide-open throttle and all
regimes before O2 sensor lightoff, the system runs open loop using strictly look-up
table values pre-programmed into the PCM and historic trim tables.

Open loop operation will naturally yield lower accuracy MPG computations than closed
loop.  Therefore diesels and gas engines that operate under heavy load (open loop) a
majority of the time (say, gas MH engines) will exhibit less accurate fuel mileage
calculations than gas engines operating in light duty vehicles.

FWIW, I've carefully compared the computed mileage vs gas pump mileage on my mom's
late model Lincoln LX.  I did so by using my data logging OBD-II software on my
laptop to track the total fuel injected vs fuel added to the tank.  The values agree
to within a couple percent.

Oh, and when that "miles to go" readout on the dashboard approaches zero, you can bet
you're out of gas.  My late dad proved that over and over :-(

FWIW2, I've never had much interest in diesels (it cranks, it runs, it pulls heavy
loads, it gets good fuel mileage - end of interest :-) so I'm not very much informed
on electronic engine controls for diesels.  My main hands-on experience was during my
year of trucking when I logged via the truck's J1708 interface bus, many parameters
including fuel consumption and mileage.  My company had a J1708 interface to the
Qualcomm on-board satellite system which let the truck report such data back to the
company.  I tracked my miles traveled and fuel purchased in detail.

This was mainly an anticipatory defensive measure, for just in case my company ever
tried to accuse me of going off-route too much, stealing fuel or any of the other
things trucking company management is known to do on occasion.  I did notice that the
J1708 reported mileage and my calculated mileage agreed very well.


Index Home About Blog