Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

What tools do I need? (P0420 Bank 1 catalytic inefficiency on a 15 year old dual-cat vehicle)

122 views
Skip to first unread message

Arlen Holder

unread,
Aug 20, 2018, 7:13:43 PM8/20/18
to
What tools do I need?

After clearing codes and reliable running the federal test procedure (aka
the "Drive Cycle"), I can set all registers but I'm getting the dreaded
P0420 Bank 1 (cylinders 123) catalytic inefficiency OBDII code on a 15 year
old dual-cat vehicle.

Do you have experience troubleshooting a P0420 cat inefficiency code?
a. Most likely, from what I've read, it's the cat
b. Sometimes, from what I've read, you can clean it
c. Most people, willy nilly, replace the downstream lambda sensor

Both downstream oxygen sensors are original, as is everything else other
than the upstream O2 sensor, both of which were replaced recently to allow
the I/M readiness monitors to be set (it worked for that purpose).

The original problem "may" have been the CCV, which was broken in half, but
that original condition has been repaired.

From what I've read, the P0420 resolution & debugging tools include:
1. Buy an OBDII tool that can show the downstream sensor voltage barely
fluctuating while the upstream sensor fluctuates at warm speeds.
2. Switch the two secondary lambda sensors & run the test above again.
3. Buy a five hundred degree F infrared temperature sensor that can show
the delta between the cat input & output.
4. Use a mallet to hear for flaked off components making noise inside the
cat.
5. Buy a $25 gallon of liquid lacquer thinner (mostly acetone) and pour it
into ten gallons of gas (Scotty Kilmer <https://youtu.be/5icTmYItwiE>).
6. Use laundry detergent and soak the cat (Scotty Kilmer - see above URL).
7. Buy sodium hydroxide (2:1 by weight) and soak the cat (Eric the Car Guy
- https://www.youtube.com/watch?v=6qSMgj5_djE>)
8. Hail Mary play and replace the secondary oxygen sensor.
9. Hack methods include adding $5 spark plug spacers (dunno if it works).
10. ??? any advice what tools I need to properly diagnose & repair ???

Note: The cat appears to be bolted on. It's a 2003 BMW 525i, where
*something* is different in NY or CA since most places won't ship to these
two states but will ship to the other 48 states. Why?

What tools do I need?

Scott Dorsey

unread,
Aug 20, 2018, 8:10:54 PM8/20/18
to
Arlen Holder <arlen...@nospam.net> wrote:
>What tools do I need?

You need a scanner.

>After clearing codes and reliable running the federal test procedure (aka
>the "Drive Cycle"), I can set all registers but I'm getting the dreaded
>P0420 Bank 1 (cylinders 123) catalytic inefficiency OBDII code on a 15 year
>old dual-cat vehicle.

FIRST you need to look at the value coming out of the upstream oxygen
sensor. Is it within range? If so, the system is running locked and
the mixture is good.

If not, then deal with the engine running issue that is the real problem.

If the value coming out of the downstream sensor is totally out of range
or the pointer moves very slowly as the car warms up, then the downstream
sensor is likely bad.

If the value coming out of the downstream senor pretty much tracks the
value coming out of the upstream sensor, then the catalytic converter is
bad. You can gun the engine and watch the two graphs move and they are
almost exactly the same (because the converter is doing nothing and so
the composition of the input and output ARE almost exactly the same).

>Do you have experience troubleshooting a P0420 cat inefficiency code?
>a. Most likely, from what I've read, it's the cat
>b. Sometimes, from what I've read, you can clean it
>c. Most people, willy nilly, replace the downstream lambda sensor

Do actual diagnosis, find the problem and fix it.

If it is the converter, sometimes cleaning it will solve the problem, and
sometimes it won't. If you have more time than money and the vehicle has
the original converter it's worth trying.

Usually cleaning helps when the converter is gunked up with tar from a
previous running problem, but it won't help if the converter is just worn
out.

>Both downstream oxygen sensors are original, as is everything else other
>than the upstream O2 sensor, both of which were replaced recently to allow
>the I/M readiness monitors to be set (it worked for that purpose).
>
>The original problem "may" have been the CCV, which was broken in half, but
>that original condition has been repaired.

This may have gunked up the converter. If that is the case, a lot of highway
driving may clean it out, or cleaning it by hand may help. But the scanner
will tell you for sure when you look at the value coming out of the upstream
sensor and see if it's in range.

>From what I've read, the P0420 resolution & debugging tools include:
>1. Buy an OBDII tool that can show the downstream sensor voltage barely
>fluctuating while the upstream sensor fluctuates at warm speeds.
>2. Switch the two secondary lambda sensors & run the test above again.

That's not a bad plan, but before doing that you should check the upstream
sensors. This is a good way to isolate whether it's a sensor issue or a
converter issue, but if there is an underlying problem you'll know.

>3. Buy a five hundred degree F infrared temperature sensor that can show
>the delta between the cat input & output.

This is a lot more work than just using the scanner and looking at the
sensor values, and again it doesn't tell you if your exhaust is clean or
not.

>4. Use a mallet to hear for flaked off components making noise inside the
>cat.
>5. Buy a $25 gallon of liquid lacquer thinner (mostly acetone) and pour it
>into ten gallons of gas (Scotty Kilmer <https://youtu.be/5icTmYItwiE>).
>6. Use laundry detergent and soak the cat (Scotty Kilmer - see above URL).
>7. Buy sodium hydroxide (2:1 by weight) and soak the cat (Eric the Car Guy
>- https://www.youtube.com/watch?v=6qSMgj5_djE>)
>8. Hail Mary play and replace the secondary oxygen sensor.
>9. Hack methods include adding $5 spark plug spacers (dunno if it works).
>10. ??? any advice what tools I need to properly diagnose & repair ???

These mostly look silly. Cleaning the converter may be a good plan, if
it's a problem that was caused by earlier running too rich. But first make
sure it's not still running too rich and that the upstream sensors are good.

>Note: The cat appears to be bolted on. It's a 2003 BMW 525i, where
>*something* is different in NY or CA since most places won't ship to these
>two states but will ship to the other 48 states. Why?
>
>What tools do I need?

The cheap aftermarket converters will not make you happy. If it's bad, get an
OEM one from Bavarian Motorsport. You will need kroil, an acetylene torch,
an asbestos mat to put behind it while you heat up the bolts, break-free,
a breaker bar and a large vocabulary of curse words.

You MAY find it's easier to drop the whole exhaust system and take it apart
outside of the car, if you're working in a driveway.
--scott

--
"C'est un Nagra. C'est suisse, et tres, tres precis."

Clare Snyder

unread,
Aug 20, 2018, 9:29:42 PM8/20/18
to
Long term you MIGHT find it easier to trade it on somethung without a
3 letter name - that doesn't have a Kraut accent -- -

Arlen Holder

unread,
Aug 20, 2018, 11:50:02 PM8/20/18
to
On 20 Aug 2018 17:10:50 GMT, Scott Dorsey wrote:

> FIRST you need to look at the value coming out of the upstream oxygen
> sensor. Is it within range? If so, the system is running locked and
> the mixture is good.
>
> If not, then deal with the engine running issue that is the real problem.
>
> If the value coming out of the downstream sensor is totally out of range
> or the pointer moves very slowly as the car warms up, then the downstream
> sensor is likely bad.
>
> If the value coming out of the downstream senor pretty much tracks the
> value coming out of the upstream sensor, then the catalytic converter is
> bad. You can gun the engine and watch the two graphs move and they are
> almost exactly the same (because the converter is doing nothing and so
> the composition of the input and output ARE almost exactly the same).

Here are the values coming out of the scanner just now, with the engine
warm, and with my foot on the throttle with steady pressure.

Which numbers look abnormal to you?
=========================
<http://www.bild.me/bild.php?file=6121688obdcode00.jpg>
P0132=O2 Sensor Circuit High Voltage BAnk 1 Sensor 1
P0430=Catalyst System Efficiency Below Threshhold Bank 2
=========================
<http://www.bild.me/bild.php?file=9033313obdcode01.jpg>
MIS=OK
FUE=OK
CCM=OK
CAT=OK
EVAP=NOT OK
AIR=OK
O2S=OK
HRT=OK
=========================
<http://www.bild.me/bild.php?file=8572593obd01.jpg>
RPM held steady at approximately 2000
O2B1S1(V)=1.275
O2B1S2(V)=0.760
O2B2S1(V)=0.665?
O2B2S2(V)=0.635G
=========================
<http://www.bild.me/bild.php?file=2946125obd02.jpg>
RPM held steady at approximately 2000
O2B1S1(V)=1.275
O2B1S2(V)=0.750
O2B2S1(V)=0.720?
O2B2S2(V)=0.635G
=========================
<http://www.bild.me/bild.php?file=5523137obd03.jpg>
RPM held steady at approximately 2000
O2B1S1(V)=1.275
O2B1S2(V)=0.740
O2B2S1(V)=0.150?
O2B2S2(V)=0.635G
=========================
<http://www.bild.me/bild.php?file=5766361obd04.jpg>
RPM held steady at approximately 2000
O2B1S1(V)=1.275
O2B1S2(V)=0.740
O2B2S1(V)=0.720?
O2B2S2(V)=0.685G
=========================
<http://www.bild.me/bild.php?file=9733573obd05.jpg>
RPM held steady at approximately 2000
O2B1S1(V)=1.275
O2B1S2(V)=0.740
O2B2S1(V)=0.110?
O2B2S2(V)=0.660G
=========================
I realized I should probably record actual RPM so I added it to the mix
=========================
<http://www.bild.me/bild.php?file=1876737obd06.jpg>
RPM=3227
O2SLOC B1S12--B2S12--
O2B1S1(V)=1.275?
O2B1S2(V)=0.670G
O2B2S1(V)=0.630?
O2B2S2(V)=0.710
=========================
<http://www.bild.me/bild.php?file=9131116obd07.jpg>
RPM=3244
O2SLOC B1S12--B2S12--
O2B1S1(V)=1.275?
O2B1S2(V)=0.645G
O2B2S1(V)=0.680
O2B2S2(V)=0.570
=========================
<http://www.bild.me/bild.php?file=8766702obd08.jpg>
RPM=2226
O2SLOC B1S12--B2S12--
O2B1S1(V)=1.275?
O2B1S2(V)=0.780G
O2B2S1(V)=0.795
O2B2S2(V)=0.750
=========================
<http://www.bild.me/bild.php?file=9305972obd09.jpg>
RPM=2178
O2SLOC B1S12--B2S12--
O2B1S1(V)=1.275?
O2B1S2(V)=0.685G
O2B2S1(V)=0.285
O2B2S2(V)=0.640
=========================
<http://www.bild.me/bild.php?file=3745618obd10.jpg>
RPM=778
O2SLOC B1S12--B2S12--
O2B1S1(V)=1.275?
O2B1S2(V)=0.850G
O2B2S1(V)=0.695
O2B2S2(V)=0.765
=========================
<http://www.bild.me/bild.php?file=2533891obd11.jpg>
RPM=763
O2SLOC B1S12--B2S12--
O2B1S1(V)=1.275?
O2B1S2(V)=0.850G
O2B2S1(V)=0.120
O2B2S2(V)=0.725
=========================
<http://www.bild.me/bild.php?file=8547378obd12.jpg>
RPM=751
O2SLOC B1S12--B2S12--
O2B1S1(V)=1.275?
O2B1S2(V)=0.845G
O2B2S1(V)=0.650
O2B2S2(V)=0.745
=========================
Does anyone here know how to help me interpret those numbers?

Scott Dorsey

unread,
Aug 21, 2018, 10:56:56 AM8/21/18
to
Arlen Holder <arlen...@nospam.net> wrote:
><http://www.bild.me/bild.php?file=8572593obd01.jpg>
>RPM held steady at approximately 2000
>O2B1S1(V)=1.275
>O2B1S2(V)=0.760
>O2B2S1(V)=0.665?
>O2B2S2(V)=0.635G

This is the data you want to see, but if possible you want to see it as
graphs, because a lot of what you care about are trends and not just the
stock voltages. Your software also should be able to put lines on the graph
showing the nominal values.

However, since it's a BMW, you can guess that the nominal value is somewhere
around 1.0. High voltages indicate something running too rich, low voltages
indicate something running too lean.

So.. we look at bank one sensor one... it's running kind of rich, which is
bad... and then we look at bank two sensor one.... it is running super lean
if that voltage is to believed. So we have a problem.

Downstream on sensor two the value on bank 1 is kind of high, so you may be
getting the error being set due to that, but that's not your problem. Your
PROBLEM is that half the engine is running too rich and the other half is
running too lean (probably because the ECU is trying to compensate).

The FIRST thing that I would do would be to swap the upstream sensors from
side to side and see if the readings move from side to side. If they do,
you have a bad oxygen sensor. If they don't, you need to start looking on
the engine side for vacuum leaks and bad injectors.

You can put a new converter onto this, but it won't stay good for very long
until you sort out the mixture problems.

Scott Dorsey

unread,
Aug 21, 2018, 11:00:07 AM8/21/18
to
Scott Dorsey <klu...@panix.com> wrote:
>
>So.. we look at bank one sensor one... it's running kind of rich, which is
>bad... and then we look at bank two sensor one.... it is running super lean
>if that voltage is to believed. So we have a problem.

Oh, and I am assuming that your software is correct in enumerating them,
that O2B1S1 is upstream bank one and O2B1S2 is downstream bank one. Make
sure you trust your software either by validating it with a mirror or pulling
a wire from one sensor and making sure it shows up as pulled on the software.

What IS that questionmark? What does it indicate? If you had a plot you
could guess.

micky

unread,
Aug 21, 2018, 12:40:27 PM8/21/18
to
In alt.home.repair, on Mon, 20 Aug 2018 21:29:39 -0400, Clare Snyder
<cl...@snyder.on.ca> wrote:

>
>>>Note: The cat appears to be bolted on. It's a 2003 BMW 525i, where..
>>
>>You MAY find it's easier to drop the whole exhaust system and take it apart
>>outside of the car, if you're working in a driveway.
>>--scott
>
> Long term you MIGHT find it easier to trade it on somethung without a
>3 letter name - that doesn't have a Kraut accent -- -

I think I've found something we agree on. I don't like krautmobiles
and it seems you don't either.

And I think they are overpriced.

Arlen Holder

unread,
Aug 21, 2018, 11:07:58 PM8/21/18
to
On 21 Aug 2018 08:00:03 GMT, Scott Dorsey wrote:

> Oh, and I am assuming that your software is correct in enumerating them,
> that O2B1S1 is upstream bank one and O2B1S2 is downstream bank one. Make
> sure you trust your software either by validating it with a mirror or pulling
> a wire from one sensor and making sure it shows up as pulled on the software.
>
> What IS that questionmark? What does it indicate? If you had a plot you
> could guess.

Hi Scott,

Thanks for that advice, which is good advice, and which I only received
just now, so I haven't been able to take you up on that test to confirm
what I would have assumed, which is the assumption that...
* O2B1S1 (I would assume is the upstream sensor for bank 1, cylinders 123)
* O2B1S2 (I would assume is the downstream sensor for cylinders 123)

* O2B2S1 (I would assume is the upstream sensor for bank 2, cylinders 456)
* O2B2S2 (I would assume is the downstream sensor for cylinders 456)

I'm not at all sure what you mean by using a "mirror" but it's easy enough
to disconnect a sensor, especially for the two upstream sensors which are
tucked away but accessible from the passenger side of the engine bay.

I have the owners manual for the Cen-Tech Harbor Freight item 60693 OBDII
scan tool, which I had borrowed from a friend. The "G" indicates a "graph"
is available, while the "?" indicates "help" is available.

Arlen Holder

unread,
Aug 21, 2018, 11:07:58 PM8/21/18
to
On 21 Aug 2018 07:56:52 GMT, Scott Dorsey wrote:

> This is the data you want to see, but if possible you want to see it as
> graphs, because a lot of what you care about are trends and not just the
> stock voltages. Your software also should be able to put lines on the graph
> showing the nominal values.

I agree. It seems that the Cen-Tech item 60693 can do graphs so I'll look
at them tomorrow when I drive the vehicle warm again.

> However, since it's a BMW, you can guess that the nominal value is somewhere
> around 1.0. High voltages indicate something running too rich, low voltages
> indicate something running too lean.

Apparently, from my googling of how the oxygen sensors work, as you said,
the more "reducing agent" there is (e.g., a rich mixture), the more oxygen
diffuses from ambient to the exhaust gas, hence, as you said, that
generates a higher voltage.

If there aren't many hydrocarbons in the mixture (e.g., a lean mixture),
then there aren't any reducing agents for the oxygen to diffuse from the
outside air into the inside of the oxygen sensor, so it barely generates a
voltage.

> So.. we look at bank one sensor one... it's running kind of rich, which is
> bad... and then we look at bank two sensor one.... it is running super lean
> if that voltage is to believed. So we have a problem.

I agree with you that bank 1 sensor 1 is pegged at a high non-fluctuating
voltage of 1.275 volts, while bank 2 sensor 1 is fluctuating from about
0.110 volts to about 0.795 volts.

The two upstream sensors are brand new, the P0132 code indicates Bank 1
Sensor 1 has a "O2 Sensor Circuit High Voltage", which is true, since it's
steady at 1.275 volts.

If we assume the two upstream sensors are working properly, and if their
input voltages and heater circuits are correct, then that would indicate a
rich mixture - but - there are plenty of codes for rich mixtures - none of
which are indicated.

So I'm not sure what to make of the steady 1.275 volts on the bank 1
upstream lambda sensor - but your suggestion of switching the two upstream
sensors should help isolate whether it's a bad sensor from the factory or
something in the fuel mixture or input electrical circuitry.

> Downstream on sensor two the value on bank 1 is kind of high, so you may be
> getting the error being set due to that, but that's not your problem. Your
> PROBLEM is that half the engine is running too rich and the other half is
> running too lean (probably because the ECU is trying to compensate).

I agree that the downstream sensor output isn't all that useful until I
figure out why the upstream lambda sensors are indicating such wacky
dichotomy of very rich (bank 1) and not very rich (bank 2).

> The FIRST thing that I would do would be to swap the upstream sensors from
> side to side and see if the readings move from side to side.

Yup. I'll do that tomorrow in the daylight.

> If they do,
> you have a bad oxygen sensor. If they don't, you need to start looking on
> the engine side for vacuum leaks and bad injectors.

Yup. Either it's really running rich, or the voltage to the sensor is bad,
or the sensor itself, even though it's brand new, is bad.

Switching the two upstream sensors should help eliminate some of those
choices.

> You can put a new converter onto this, but it won't stay good for very long
> until you sort out the mixture problems.

Agreed. Here are some readings from today which I generated on the highway
before I read your post above (which repeats for today, basically, the
readings that you already commented on from the prior day).
=========================
<http://www.bild.me/bild.php?file=9463821obd001.jpg>
RPM 749
O2B1S1(V) 1.275
O2B2S1(V) 0.680 ?
O2B1S2(V) 0.830 G
O2B2S2(V) 0.715
=========================
<http://www.bild.me/bild.php?file=5010400obd002.jpg>
RPM 1468
O2B1S1(V) 1.275
O2B2S1(V) 0.095 ?
O2B1S2(V) 0.780 G
O2B2S2(V) 0.740
=========================
<http://www.bild.me/bild.php?file=5797969obd003.jpg>
RPM 2104
O2B1S1(V) 1.275
O2B2S1(V) 0.670 ?
O2B1S2(V) 0.755 G
O2B2S2(V) 0.565
=========================
<http://www.bild.me/bild.php?file=3108371obd004.jpg>
RPM 2182
O2B1S1(V) 1.275
O2B2S1(V) 0.675 ?
O2B1S2(V) 0.700 G
O2B2S2(V) 0.615
=========================

Arlen Holder

unread,
Aug 21, 2018, 11:07:58 PM8/21/18
to
On 21 Aug 2018 09:40:22 GMT, micky wrote:

> I think I've found something we agree on. I don't like krautmobiles
> and it seems you don't either.
>
> And I think they are overpriced.

This Krautmobile is so old as to be nearly worthless so it's not overpriced
anymore. :)

I've had American, Japanese, and German (one British but very long ago)
where my summary is roughly
- American is poor quality all around - but the engines are powerful
- Japanese is good quality all around - but nothing is stellar
- German has stellar handling - but everything else is poorly designed

That's only a one-line summary so it's not accurate - but it's a general
assessment of multiple vehicles that I've owned in each brand. For British,
I only owned the midget but it was way back when - and it was the only
British vehicle so I left it out of the summary.

Back on the oxygen sensors, I'm getting the strangest of readings, even out
one of the brand new upstream sensors...

Basically the bank 1 upstream oxygen sensor is locked at 1.250 volts, which
it shouldn't be, even though it's brand new. It could be a bad wire or a
bad sensor or a rich mixture. I don't know yet. The downstream sensor is
half that, which may be fine (I don't know what a normal voltage should be
yet).

The bank 2 upstream sensor reads completely differently, and it fluctuates,
which it's supposed to do (I'm told). It fluctuates from around 0.95VDC to
about 0.675VDC, which may or may not be a good reading. The downstream
sensor isn't reading much off from that - which may indicate that the cat
isn't doing anything. But sometimes it reads a lot different. So it's not
consistent.
0 new messages