911SC only starts with alternator disconnected

[rick-l]

Let me see if I understand how this system works. I think key to understanding this is to realize the run switch in the air flow meter opens when the engine is running (kind of bass ackwards). Also the wires going to the alternator just provide an engine ground.

• Turn the key on

  1. Power is applied to terminal 86 (and 87a) of the fuel pump relay (definition from DIN Std. 86 -Start of winding) from circuit 15 (DIN Std. -15 Switched + downstream of battery)
  2. current flows through the relay coil, through the closed air flow switch and through the alternator (eventually) to chassis ground.
  3. There is a momentary voltage spike at the fuel pump till the relay opens through the normally closed contacts of the fuel pump relay.
  4. The relay closes and connects the fuel pump to the yellow wire (control) of the starter. (DIN Std. 50-Starter control (direct))

• Turn the key to start

  1. The yellow wire is energized which turns the starter solenoid on and runs the fuel pump.
  2. The engine sucks air and opens the air flow switch
  3. The fuel pump relay coil is de-energized and the relay switches the fuel pump directly to circuit 15 through the normally closed contact(with ~ 20 mSec dead time)

• The driver realizes the engine has started and releases the key, with no change in fuel pump power state.
• (If) The engine stalls the flow switch closes, the relay energizes (backwards part) and switches the fuel pump to the inactive starter control shutting off the fuel pump

[mysocal911]

Originally Posted by rick-l

Let me see if I understand how this system works. I think key to understanding this is to realize the run switch in the air flow meter opens when the engine is running (kind of bass ackwards). Also the wires going to the alternator just provide an engine ground.

• Turn the key on

  1. Power is applied to terminal 86 (and 87a) of the fuel pump relay (definition from DIN Std. 86 -Start of winding) from circuit 15 (DIN Std. -15 Switched + downstream of battery)
  2. current flows through the relay coil, through the closed air flow switch and through the alternator (eventually) to chassis ground.
  3. There is a momentary voltage spike at the fuel pump till the relay opens through the normally closed contacts of the fuel pump relay.
  4. The relay closes and connects the fuel pump to the yellow wire (control) of the starter. (DIN Std. 50-Starter control (direct))

• Turn the key to start

  1. The yellow wire is energized which turns the starter solenoid on and runs the fuel pump.
  2. The engine sucks air and opens the air flow switch
  3. The fuel pump relay coil is de-energized and the relay switches the fuel pump directly to circuit 15 through the normally closed contact(with ~ 20 mSec dead time)

• The driver realizes the engine has started and releases the key, with no change in fuel pump power state.
• (If) The engine stalls the flow switch opens, the relay energizes (backwards part) and switches the fuel pump to the inactive starter control shutting off the fuel pump

Yes, it's as simple as that!

[boyt911sc]

Originally Posted by rick-l

Let me see if I understand how this system works. I think key to understanding this is to realize the run switch in the air flow meter opens when the engine is running (kind of bass ackwards). Also the wires going to the alternator just provide an engine ground.

• Turn the key on

  1. Power is applied to terminal 86 (and 87a) of the fuel pump relay (definition from DIN Std. 86 -Start of winding) from circuit 15 (DIN Std. -15 Switched + downstream of battery)
  2. current flows through the relay coil, through the closed air flow switch and through the alternator (eventually) to chassis ground.
  3. There is a momentary voltage spike at the fuel pump till the relay opens through the normally closed contacts of the fuel pump relay.
  4. The relay closes and connects the fuel pump to the yellow wire (control) of the starter. (DIN Std. 50-Starter control (direct))

• Turn the key to start

  1. The yellow wire is energized which turns the starter solenoid on and runs the fuel pump.
  2. The engine sucks air and opens the air flow switch
  3. The fuel pump relay coil is de-energized and the relay switches the fuel pump directly to circuit 15 through the normally closed contact(with ~ 20 mSec dead time)

• The driver realizes the engine has started and releases the key, with no change in fuel pump power state.
• (If) The engine stalls the flow switch opens, the relay energizes (backwards part) and switches the fuel pump to the inactive starter control shutting off the fuel pump

Rick,

This is an excellent write except for the flow switch operation. When the engine stalls, the AFS plate drops down and closes the switch, Not open as you have stated. Maybe a typo?

Tony

[anthony]

I think I'm honing in on the issue. I've ran all the tests again and everything checks out.

While running a test on pin 30 of the relay socket with a test lamp and lifting the AFM plate, I noticed that just as I start turning the key (the moment the gauge lights come on), the fuel pump does run. But, it shuts off as I turn the key farther....

The ignition switch!!!

I made a video. At 4 seconds and 11 seconds in, just as I turn the key the fuel pump runs for a brief moment.

https://www.youtube.com/watch?v=0EnVJgug074

This is repeatable.

And I know:

FP is good
FP relay is good
wiring to the FP is good
FP relay socket tests good
ground and wiring to AFM is good
AFM switch is good
everything I've tested checks out

[rick-l]

Have you thought about hooking the test light up to relay pin 30 to monitor what goes on during a start cycle? You could also monitor it at the warm up regulator.

[anthony]

I realize now that the video doesn't show a lot unless you were there. I had the test lamp hooked up the pin 30 and it did confirm that the fuel pump was getting energized for 1/2 a second just as I turned the key. (I should have included the test lamp in the video.)

[boyt911sc]

Originally Posted by rick-l

Have you thought about hooking the test light up to relay pin 30 to monitor what goes on during a start cycle? You could also monitor it at the warm up regulator.

A test light was connected to terminal #30 with the FP relay installed. This was part of a series of tests conducted to determine the intermittent failure of the FP to run. Anthony noticed that by wiggling the ignition switch key between OFF and ON positions he could momentarily make the test light and FP to come on. And he had replicated this several times.

Tony

[rick-l]

Is that normal for a CIS car? That seems like an awful long time for the relay contact to transfer. Long enough to overcome the inductance in the fuel pump motor and get it to spin (to make a noise).

What happens (light indication) when you turn it to start, start the car and release the key?

[mytoy]

I am of the understanding that this is normal.

It's my understanding that when you turn the key to the start position that the fuel pump will be energized for 1 or 2 seconds then be de-energized when the relay switches over to the starter relay circuit.

This was done to prime the system just before start up.

I think that if Rick's point #3 is correct that it confirms this. Also I recall that Tbitz had to modify the megasquirt program to do this same operation when doing the EFI conversion.

I may be totally out to lunch but I thought that Porsche implemented this in their circuit and therefore the one of the reasons for the red relay to allow for this delay in switching.

As I said I may be totally out to lunch on this. If so disregard.

[anthony]

Yes, did that when I replaced the alternator. The car even ran perfectly with no alternator installed. I probably started it 20 times to move it around. After the new alternator was installed the car ran great briefly. Voltage at the battery was good, thus the new alternator was working properly. And then it stopped running again!!! I thought it was related to the alternator so I was led down the wrong path initially.

Tony really helped me step back and test each component step by step. I really appreciate Tony's aid in helping me set up proper tests to check the FP, relay, socket, AFM switch, etc. And now I have this great understanding of the FP, relay and starting circuit.

The car is still not fixed but a new ignition switch has been ordered from Pelican. I heard it's a pain of a job but will be nothing compared to the days and days I've been banging my head against this problem. Hopefully, we'll be up and running next week.

What happens (light indication) when you turn it to start, start the car and release the key?

Is that question for me Rick? What has been happening is that the FP runs with the key in the START position but doesn't run in the ON position. The AFM switch works and 87a has power, and the relay bench tested fine. None of the tests pointed definitively to the problem.

[mysocal911]

Originally Posted by anthony

The car is still not fixed but a new ignition switch has been ordered from Pelican. I heard it's a pain of a job but will be nothing compared to the days and days I've been banging my head against this problem. Hopefully, we'll be up and running next week.



Is that question for me Rick? What has been happening is that the FP runs with the key in the START position but doesn't run in the ON position. The AFM switch works and 87a has power, and the relay bench tested fine. None of the tests pointed definitively to the problem.

You should be able to jumper #30 power to the FP fuse, and the engine should start
and run normally. This will bypass the #15 power provided by the ignition switch if it's
really bad and power the FP relay pins.

Since there's just one #15 contact on the ignition switch, no #15 power will be provided
to any other electrical requirement, e.g. the light/radio, if the switch is bad. So before actually
replacing the ignition switch, you need to use your test-light to check the various powers at
the ignition switch, i.e. X (if it has one), #15 (run), #30 (constant power), #50 (start).

[rick-l]

With the test light on the circuit that connects to the fuel pump, pin 30 of the relay or the warm up regulator, what does the light show when you start the car and it stalls?

Does it flash when you turn the key on, come on with the starter and then stay on or go out when you release the key? It sounds like everything works except power to the fuel pump with the flow switch on (open).

Originally Posted by anthony

There came my inspiration. I jumpered the fuel pump and then started the car. The car fired up and ran perfectly.

[anthony]

Originally Posted by rick-l

With the test light on the circuit that connects to the fuel pump, pin 30 of the relay or the warm up regulator, what does the light show when you start the car and it stalls?

Does it flash when you turn the key on, come on with the starter and then stay on or go out when you release the key? It sounds like everything works except power to the fuel pump with the flow switch on (open).

The test lamp flashes briefly and then goes out at the moment I start turning the key towards ON. (AFM lifted) (it happens at the exact moment the gauge lights go on. The TL should stay on.

[rick-l]

Originally Posted by anthony

The test lamp flashes briefly and then goes out at the moment I start turning the key towards ON. (AFM lifted) (it happens at the exact moment the gauge lights go on. The TL should stay on.

When you turn the key from lock to on the test light may flash on but that isn't really relevant.

What happens when you turn the key to start and the engine fires and you release the key?

[wildthing]

OT: if you see a post from someone with an almost meaningless name and posts fewer than 10, ignore and do not quote. Those posts have a hidden image linked to some site, most likely to increase their search rank. (More links increase rank.)

BTTT, I see you are in the Bay Area. I am in San Jose. PM me if you need just an extra set of hands, literally. No tech knowledge whatsoever (unless you count computers).

[anthony]

Originally Posted by wildthing

OT: if you see a post from someone with an almost meaningless name and posts fewer than 10, ignore and do not quote. Those posts have a hidden image linked to some site, most likely to increase their search rank. (More links increase rank.)

BTTT, I see you are in the Bay Area. I am in San Jose. PM me if you need just an extra set of hands, literally. No tech knowledge whatsoever (unless you count computers).

Thanks for the generous offer. I think I'm very close to fixing it.

I deleted my last responding to that guy. I did notice that image link in the quote. Now the guy is just repeating what I wrote. Hard to believe that this helps someone's business.

[mysocal911]

This problem, like many, has become overly complex. It should have taken 5 minutes at
most to solve and not over 75 posts and more than a week of troubleshooting.

[anthony]

Originally Posted by mysocal911

This problem, like many, has become overly complex. It should have taken 5 minutes at
most to solve and not over 75 posts and more than a week of troubleshooting.

Easier said than done, my friend. I had to learn a lot to get where I am with this. Unfortunately, there isn't good troubleshooting guide for each and every one of these components in one place. I looked at the Pelican troubleshooting guide. I've searched the net. People have given instructions in various threads but finding them is like sifting through sand.

[mysocal911]

Originally Posted by anthony

Easier said than done, my friend. I had to learn a lot to get where I am with this. Unfortunately, there isn't good troubleshooting guide for each and every one of these components in one place. I looked at the Pelican troubleshooting guide. I've searched the net. People have given instructions in various threads but finding them is like sifting through sand.

It appears that you have not fully identified your problem, i.e. bypassed certain power
points and properly tested the ignition switch, but you've ordered a new ignition switch.
Hopefully this will not be a waste of time and money. The "shotgun" approach to parts
replacement does work on occasion, but not always.

[anthony]

I agree and I realize I'm taking a $100 chance on an ignition switch. And, if that isn't the case I've gotten a lot better with the test lamp and multimeter. :-)