Secondary Air Pump - '95 E420 - PeachParts Mercedes-Benz Forum

Syntax26 · #1 05-30-2008, 10:41 PM

I've been attempting to diagnose CEL issue (and codes) for a couple weeks now. Full story - all codes - parts replaced - can be read at

'95 E420 - DM Codes 3, 4 & 25 - What to do next?

But, I just have a quick question or two about the secondary air injection pump. I'm trying to figure out if the pump is defective, or simply not being turned on/activated.

At cold start:

1. If pump runs...you can tell this by looking at the triangular plate in the front, correct? (the clutch engages inside the otherwise free-wheeling pulley) (is this possibly the dumbest question ever asked on here? LOL!)

2. If pump is defective and not running...wouldn't you still get voltage reading at the harness connection to the pump? And wouldn't it be 12 volts, more or less?

3. If you don't get voltage at the harness...is it possible there is a relay not operating? (in other words - the pump might be ok, but it is simply not getting energized)

4. Finally, can I test the pump directly by using another 12 volt source and connecting it to the pump?

If anyone can answer advise me on this....I'd greatly appreciate it. I'm obviously trying to figure out if it is the pump itself that is bad before I buy the replacement. Or, if it the problem is really that the pump is not getting turned.

By the way, the reason I'm focusing on this issue is that after reading codes for week, I've figured out that that first two codes that return (meaning, before any lambda or knock codes) are the following two:

Pin 4 - Code 11 = secondary air injection system, open or short circuit
Pin 19 - Code 4 - air injection system faulty

Oh...and also by the way - I've just put a new 02 sensor in.

Thanks!

Jim

Arthur Dalton · #2 05-31-2008, 04:28 PM

Well, your first 4 questions tell me that, before anything, you should check DA FUSE !!!!

#7

Is the pump running or isn't it???????????????

Syntax26 · #3 05-31-2008, 04:53 PM

Mr. Dalton,

Thanks for your reply.

I had checked fuse 7 before - it is good.

This morning I applied an alternate 12 volt source to the pump - and the pump ran. Which leads me to believe the problem is not that the pump is defective, but instead that it is not be energized (turned on).

I was trying to find a schematic, but even though I own 3 MB repair manuals, I can't find a schematic for this. I would assume there is a relay for the power to the pump.

Can you tell me if there is one?

Thanks!

Jim

Arthur Dalton · #4 05-31-2008, 05:08 PM

K17 is the relay. It is feed from fuse 7.
It is in the relay box behind the fuse box

Syntax26 · #5 06-01-2008, 07:19 AM

Arthur,

Thanks for the info on the relay.

I'll check it today to see if it is functioning.

I'll be sure to let you know if I have success or not.

Thanks!

Jim

Syntax26 · #6 06-02-2008, 10:07 PM

Arthur,

Kind of a unique problem here. I say this because I've researched this code quite a bit...and in no case have I seen the issue be the relay. Nonetheless...

It is either the relay malfunctioning, or, the relay not receiving a signal to close the contact and send 12 volts to the pump.

I know this because I pried the cover off the relay and closed the contact by hand...and the pump comes on (by which I mean the electro-mechanical clutch engages and the pump runs).

Unfortunately, I've yet to be able to obtain a replacement relay. The local MB dealership parts department insists the part number I give them (126 820 20 26 = stamped on the relay itself) is a wiper relay - so they tell me they aren't able to help me. This part number isn't recognized (let alone stocked) by the vast majority of web-based MB parts specialists. I have found 1 company in Idaho that stocks it...so I guess I'll order a unit from them.

I thought you might appreciate hearing what I've learned, since despite researching this issue on 3 separate MB technical forums, I've never seen the relay to be the issue. Always either the pump is broken, or, it's a clog issue in the head ports or other valves and piping.

But I do have a question for you - maybe theoretical at this point. If the relay is not defective...this would mean the problem is the relay is not getting a signal from the ignition control module. Which would mean either the ignition control module is not signaling the relay to engage, or, the signal is not getting to the relay (due to a wiring issue).

Does my logic seem correct to you? Or can you think of another reason for the relay contact not to close that I am not aware of.

Well, anyway - thank you very much for your help is letting me know the relay in question was K17.

I'll let you know the results when I obtain a replacement relay and test my theory.

Jim

Arthur Dalton · #7 06-02-2008, 10:20 PM

If the pump engages when you manually close the relay, then the relays coil side is not getting power to pull the relay IN..that is why I had you check the fuse # 7...that fuse does not feed the pump clutch, it feeds the signal to the relays PRIMARY COIL. That is where you have to check.
So, IS the coil contacts at the relay getting voltage from fuse 7 ??? ..or is the coil in the relay OPEN ???
Those are the questions you have to be addressing. The test for those 2 questions is an ohm meter reading across the relay coil [ 85/86] for coil verification and a voltage reading at those terminals to energize the coil.

No coil reading equals NO PUMP
No Voltage at coil terminals equals NO PUMP.
Bad or poor connection at fuse equals NO PUMP.

By closing the relay manually, you have eliminated the LOAD side of the circuit and verified that circuit/ part to be OK..so we know you have a relay coil feed/sensor side problem. [ unless the pump IS coming ON...that is why I asked " IS the pump coming ON ???"]

There are criteria to get coil feed , but we don't want to go there if you do see the pump ever run. If the pump runs, then the problem is in the valves and piping.
So, answer my questions and I can then give you the next question to help track it down.
You seem to be confused about the pump...it only RUNS when the clutch is energized , so when I ask is pump RUNNING, I don't mean anything else except "Is the PUMP CLUTCH being pulled IN...
I think you have that understanding, so before getting a relay, pull it and confirm the coil windings w/ohm check....you may have never seen or read of a relay to be the problem, but I have seen many relays with OPEN coils...so we go by coil ohm readings , not by never seen one do that before............
If you jumper the relay pins for clutch , then the AIR system plumbing and Switch over Valves can be checked by monitoring the 02 sensor...but check the relays coil first... . and also recheck that fuse ..they will fool you and look fine..I use a test lamp on both sides to be sure.

Syntax26 · #8 06-02-2008, 11:00 PM

Arthur,

Thanks for your quick response. (I was reading some other posts after posting mine - and then I saw you had responded).

You're correct - I was confused about the pump operating or not - but that confusion passed about 2 days ago - and I guess I didn't clarify this.

Originally, not having any experience with this kind of pump - I wasn't 100% sure "what an operating pump looked like". (Not having any history to compare operating versus non-operating). As you know - the pulley is always turning due to the belt - but the triangular face plate at the front of the pump was stationary (on my pump - when the engine was first turned on). I deduced the pump had to have a clutch, and I believed this plate would be seen turning IF the pump was running. But I only figured this out using logic - I didn't know it for a fact.

But once I came up with the idea of putting 12 volts directly to the pump - and saw the change (i.e. saw the triangular plate turn - thus showing me I was right in thinking there had to be a clutch mechanism - and heard the pump operating noise - and then found the suction intake) - then I now had the hands-on personal experience to know the difference between an operating pump and a non-operating pump.

And then I knew 2 things:

1. The pump wasn't running before, and...

2. The pump wasn't defective, i.e. it WOULD run if it got power.

That is what led me to ask you about the relay. And when the local dealership couldn't get identify the part number (meaning I couldn't get and test a replacement relay today), I then came up with the idea of taking the cover off the relay and manually engaging the contact.

As I engage the relay contact - the pump comes on. Clear action - reaction.

So this definitely means the pump, and the wiring harness between the relay and the pump, are ok. It also means 12 volts are getting to the relay (at least one post of the relay, anyway).

Thus the problem would seem to be, by a matter of elimination, that either:

1. The relay is getting the signal from the ICM to close the contact but is not doing so (because the relay is malfunctioning), or

2. The relay is not getting a signal to close the contact.

Now...I'm no expert on relays, but I seem to remember from somewhere that relays are also electro-mechanical. The coil get's enerized, acts like a magnet, and pulls the contact reed to it - thus closing the contact and completing the circuit - and thus sending 12 volts to the pump. Right?

Anyway - I know the next logical step is simply replace the relay. It might not fix the problem, but the relay isn't anywhere near as expensive as a pump - so I'm willing to go this route.

I know it's possible to diagnose more by checking voltage readings in the relay box - in the female sockets. But since I don't have a schematic of this, I'm not sure I'd learn anything.

So I'll wait for a replacement relay. With my luck it won't fix the problem, and then I'll have to bug you for advice again. Hope you don't mind.

So...thanks again for your help.

Jim

Arthur Dalton · #9 06-02-2008, 11:06 PM

>

You can do that if you wish..

I posted that the coil side of the relay was termninals 85 and 86. An ohm meter reading here will verify relay coil...or, simply pull the relay and put 12v to those terminals and see if the relay pulls IN...if it does , the relay is not your problem..if it does not, it is...

Here are the k17 relay terminals and the harness socket #s they fit into

relay terminals 85 [pin 5] and 86 [ pin 4] are the coil.
relay terminals 87 [ pin 1] and 30 [pin 3] are the contacts

When 12v is applied to 85 and 86, the relay closes contacts 87 and 30, allowing 12v to go to Clutch and SOV of AIR system. [ using fuse 7 as voltage feed]...so , if you jumpered 1 and 3 , the pump clutch should pull in if clutch circuit is OK. [ this would be the same as when you manually operated the relay contactor set]

Syntax26 · #10 06-02-2008, 11:59 PM

Hi Arthur,

Let me say that I really do appreciate your assistance. And I beg your patience inexperience here - I'm somewhat mechanically savvy, but I must admit the electrical side of things is not my forte.

Anyway, I THINK I understand some of what you just told me. I did some of what you suggested, but couldn't do all (as I'll explain):

1. In relay box - female terminals 85 & 86 - I get 5 volts.

2. I replace fuse 7 with brand new fuse (again) - still get 5 volts.

3. I confirm that relay coil contact remains OPEN (with engine first turned on) - and pump clutch does NOT engage.

4. I then manually close relay coil contact - and pump clutch engages.

The 2 things you suggested I do, but could not or did not do (tonight) are:

5. Ohms reading on relay posts - I get no reading. HOWEVER, I'm not sure my meter is operating correctly, as I get no response even touching meter prongs.

6. I did NOT apply 12 volts across relay terminals 85 and 86 BECAUSE female terminal readings were 5 volts, and I wasn't sure if I should apply as much as 12 volts. (The relay cover states it is for 12 volts - but I thought this might just apply to the load, and maybe not to the coil voltage).

So, do my partial tests indicate anything? If not, I'll figure out how to get a ohms reading tomorrow.

Thanks again!

Jim

Arthur Dalton · #11 06-03-2008, 12:22 AM

Quote:

Originally Posted by Syntax26

Hi Arthur,

Let me say that I really do appreciate your assistance. And I beg your patience inexperience here - I'm somewhat mechanically savvy, but I must admit the electrical side of things is not my forte.

Anyway, I THINK I understand some of what you just told me. I did some of what you suggested, but couldn't do all (as I'll explain):

1. In relay box - female terminals 85 & 86 - I get 5 volts.
OK..something wrong there....

2. I replace fuse 7 with brand new fuse (again) - still get 5 volts.

3. I confirm that relay coil contact remains OPEN (with engine first turned on) - and pump clutch does NOT engage.

4. I then manually close relay coil contact - and pump clutch engages.
OK .that verifies fuse 7 and clutch, but not SOV..that is a possible

The 2 things you suggested I do, but could not or did not do (tonight) are:

5. Ohms reading on relay posts - I get no reading. HOWEVER, I'm not sure my meter is operating correctly, as I get no response even touching meter prongs.
OK ..your meter is not working ifit does not show continuity with leads shorted..

6. I did NOT apply 12 volts across relay terminals 85 and 86 BECAUSE female terminal readings were 5 volts, and I wasn't sure if I should apply as much as 12 volts. (The relay cover states it is for 12 volts - but I thought this might just apply to the load, and maybe not to the coil voltage).
12 V should be coil rating . Relay Load would have an AMP rating

So, do my partial tests indicate anything? If not, I'll figure out how to get a ohms reading tomorrow.

Thanks again!

Jim

There is Demand Criteria for AIR system to operate..
They are:
Pump engages no more than 110 secs WHEN:
Coolant temp is 50F to 104F
RPM less than 3000rpm
Air Mass Volume spec is not exceeded
Lambda system is in open loop.

All these have to be met for AIR .

The code comes up when the Diagnostic Module does a self test when the engine is in closed loop and temp is normal. It energizes the system and looks at the 02 sensors value to go real lean..if it does not see this lean condition, it brings on the code as AIR not operating.
So, if we have no pump, we have the code , buit we can have pump and maybe the SOV valve that opens the AIR shut off valve is not working..thast gets the same signal as the clutch..see what I mean??
So, check the relay coil and I will see why we have 5 v at relay coil input. I think there is something wrong there ...See if that 5 V goes away when the coolant is up to temp...

Syntax26 · #12 06-03-2008, 12:49 PM

"I posted that the coil side of the relay was termninals 85 and 86. An ohm meter reading here will verify relay coil...or, simply pull the relay and put 12v to those terminals and see if the relay pulls IN...if it does , the relay is not your problem..if it does not, it is..."

Arthur,

I put 12v to terminals 85 and 86 on the relay. NO closing of contact.

Per my understanding from your advice...I conclude the relay is malfunctioning...and I will source a replacement.

With respect to the issue of 5 volts at the female sockets 85 and 86...I wonder if this is the drop-down voltage after the AIR has tested/set the O2 reading. By which I mean - maybe it's 12v at start-up/during test...and drops to 5v (versus dropping to 0v) - adequately insufficient voltage to close the contact. I'll research and figure this out.

Anyway...at the moment I'm concluding the relay is malfunctioning and will get a replacement.

I'll let you know how it works.

Thanks!

Jim

P.S. Your profile says you live in both NH and Florida. Do you by any chance know Steve Ames, the Pontiac guy, up in NH?

Arthur Dalton · #13 06-03-2008, 12:55 PM

Open coil.............. if you have the cover off the relay, you can sometimes see where one end of the coil has disconnected from either 85 or 86..that is usually what happens. Very fine wire.........

Don't know the Steve guy..
I hung at Townlyne Motors...Benz specialist.

Here is a generic inside diagram of a standard relay..you can see the coil and contact set clearly..that is basically what you are working with on your car.

http://members.aol.com/ajdalton7/relay.jpg

Arthur Dalton · #14 06-03-2008, 01:11 PM

This is the test I am waiting for a reply on..............

Syntax26 · #15 06-03-2008, 03:49 PM

Arthur,

I'll check the 5v issue tonight when I get home from work - and I'll let you know.

In the meantime, the reason I asked you about Steve Ames in NH, is I thought you might be into all kinds of cars...not just MB. Steve owns Ames Performance - specializing in Pontiac restoration parts.

I know Steve because thats the business I'm in - collectible/restoration/muscle car shows - Carlisle Events. www.carlisleevents.com

This week we're getting ready for the All-Ford Nationals...the largest Ford specific car show in the country.

Now you know why I like car stuff (even when they're a pain in the rear).

Jim