Printable Version of Topic

Posted by: itsmagic Aug 8 2014, 10:46 AM

My 1972 1.7 914 recently decided to not start. I quickly noticed that the fuel pump is not running briefly like it should after the ignition key is turned. I verified that the relays on the relay board (Master, Fuel Pump, Blower) are all good by trying them on the headlight motors. I then verified that the two fuses on the relay board are good. Next, I popped the cap on the fuel pump relay and pressed it... the fuel pump runs when I do this. (The car will then start and run for a few seconds if I start it). Sooooo, I am at a loss... The fuel pump is fine, relay board is fine, relays are fine... what could be causing the fuel pump to not activate? Thank you for any advice. Wayne

Posted by: silverbullet75 Aug 8 2014, 11:01 AM

It's been a while since I've seen this, but depending on your Fuel Injection system, if the flapper is not moving past a certain point, it will not trigger the fuel pump.

I may be off on this though... it's been a couple decades, and I may be thinking about 1.8L.

Posted by: itsmagic Aug 8 2014, 11:05 AM

Where/what is 'the flapper' ? Thanks.

Posted by: Cap'n Krusty Aug 8 2014, 11:09 AM

Your car doesn't have one. 914 with L-jet only. The computer controls the fuel pump. The main relay supplies power to the computer, the computer triggers the fuel pump relay. Computer failures are EXTREMELY rare, but they do happen.

The Cap'n

Posted by: sfrenck Aug 8 2014, 11:43 AM

There is a nice flowchart for troubleshooting fuel pumps at this website: http://members.rennlist.com/pbanders/FPChecklist_frame.htm

There is a toggle at the bottom of the webpage that steps you through the whole thing (website works on my phone, but not on my work computer for some reason).

Posted by: Tom Aug 8 2014, 11:58 AM

To give a little more help, here is a more detailed explanation of how the fuel pump circuit works.
Sounds like your relays are good, however they still may not be making good contact in the relay sockets. Try spreading the male " prongs" slightly. this will make them fit tighter in the relay sockets. Turn them upside down and look at the prongs, they look like a round pie cut into 4 pieces. Spread very slightly. Also, if you have a very small wire brush, clean the relay sockets lightly.
The circuit for the fuel pump follows:
The ECU (computer) gets power when the power relay ( # 74) gets energized when the key switch is turned to ON. The power relay gives power ( pin #30 to pin #87 ) then via trace on the relay board to pin 86 of the fuel pump relay ( #75 ) and pin #1 of the 4 pin connector to the ECU. If the fuel pump relay is getting a ground from the ECU ( computer ), then the fuel pump relay will close ( energize ) and give power from fuel pump relay contacts 30 to 87, giving power to the fuel pump and the AAR via the traces to pin #12 of the 12 pin connector for the AAR and traces to pin #13 of the 14 pin connector for the fuel pump.
Picture of the relay panel below.
Tom


Attached image(s)

Posted by: pt_700 Aug 8 2014, 12:05 PM

my d-jet was one of those rare computer failures.

shop spent hours checking the basics and finally swapped one from another customer's car and, voila!

Posted by: silverbullet75 Aug 8 2014, 12:10 PM

Sorry for the wrong/misleading info!

Posted by: rhodyguy Aug 9 2014, 11:30 AM

so it dies when you stop pressing on the relay? swap out a KNOWN good one from one of the headlight motors and see if the problem persists.

Posted by: itsmagic Aug 24 2014, 11:43 AM

Follow up: Thanks for everyones advice. I carefully wire brushed the relay sockets, then bent the relay pin's out slightly as suggested... worked perfectly. The relay board and relays sure are finicky. ;-)
So now out for a spirited summer time drive....

-Wayne

Posted by: The Cabinetmaker Aug 24 2014, 12:55 PM

Great job Wayne. And thanks for checking back and letting us know. So many folks get a problem fixed and we never hear back.

Posted by: Chris Pincetich Sep 14 2014, 09:01 PM

I AM GOING THROUGH THE EXACT SAME THING as Wayne, who started this thread.

I thought it was a bad relay board, and have tried several other boards without finding my fix (Thanks Bruce for sending them!).

I just learned about and tried the exposed relay test, and it did work, which means my relay board should be fine. With the relay innards exposed, I turned the key in the ignition to ON then pressed the relay and the fuel pump started running. I have also cleaned all the relay connections on the board, cleaned the relays, tried several working relays, and bending the relay prongs for a tighter fit. Still no power to the fuel pump!

Does this mean it's time to swap out and try a new ECU brain??

I've been driving the 914 for over a week with the fuel pump wired directly to the battery. I have to unhook the battery every time I park.

Is using a modified relay, one that is "always on," a bad idea for a temporary fix??

Seems like it would be the same thing as I have now (pump always on when car is going) but would mean I don't have to unhook the battery every time (and I could use the OEM pump electrical connection/wiring).

Thanks in advance for more advice and tips

Posted by: Chris Pincetich Sep 16 2014, 01:26 PM

Posted by: jcd914 Sep 16 2014, 10:37 PM

The "exposed relay test" does not mean the relay board is good, it only verifies the power circuit to the relay and power circuit out to the fuel pump are good. But considering you have tried a number of relay boards, yours it probably not the cause.


It means it is time to test and determine if the ECU is trying to operate the fuel pump relay.
Using the diagram and info Tom posted you need to verify the fuel pump relay get power to the 85 terminal and then check to see if the ECU grounds terminal 86. If you don't have those at the relay socket then move and test at the next connection in the circuit.


Bad idea. Fuel pumps are controlled by relays or modules for safety. In an accident or if you had a fuel line burst you could have trouble getting to the battery to disconnect it. And if you had a fuel leak, that little spark you get at the battery when you disconnect it, could be the spark the sets your car on fire.


Yes, bad idea. It would be very similar to what you have now, which is why it is a bad idea.

I worked with a mechanic that constantly drill in to me "Test, don't guess". Diagnose the fault in the circuit and then you can determine the best fix. Tom's info should help.

As was said, ECU's don't fail often but they do fail but IMHO the FI harness is probably a more likely culprit.

Jim

Posted by: Cap'n Krusty Sep 16 2014, 11:01 PM

Remember what I said earlier. The main relay powers up the ECU, the ECU powers up the relay. Check the main relay first. The one just forward of the pump relay. Use the heater relay as a test unit, assuming it works. The board could be bad, too.

The Cap'n

Posted by: Chris Pincetich Sep 17 2014, 10:14 AM

thanks for helping!
I know my current fix/hack job is a bad idea, and will prioritize the testing above ASAP

Posted by: Chris Pincetich Sep 20 2014, 07:46 PM

OK, did some testing and ready to do more once I get help figuring it out!

With both power and fuel pump relays off the board I turned the ignition key to ON and there is power to the hole in the board for power relay pin #30.

With the power relay on the board and the fuel pump relay off and the key ON, there is power to the fuel pump relay holes in the board at #30 and #85, negative ground at #87 and #86 is a dud (neither power or grounding). BAD BRAIN!?!?!?

I removed the top covers to both the 12 and 14 pin connectors (thanx for that tip McMark ) to test for power to each of those circuits.

With the key ON, and both relays in the board, there is power at the 12-pin connector at #2 and #7, and in the 14-pin connector at #2, #3, #8, #12, and #14.

Not sure if this is proof of either the relay board being bad or the ECU brain!

Could I do more testing on the relay board? I have many other relay boards I could bench test! I think there is a way to use my multi-meter to test each circuit's integrity??

Thanks in advance for more help

Posted by: Tom Sep 21 2014, 01:39 AM

You can narrow down the search some by doing the following. Using a jumper hooked to ground on one end and with a thin piece of solid wire ( a paperclip would work) on the other, turn on the key switch and put the wire/ paperclip from the jumper to the #3 connection on the 4 pin plug which is located at the lower left of the picture posted above. The ECU provides a ground to pin 86 of the fuel pump relay, allowing it to energize and close the contacts to provide power to the fuel pump.
If the above test is satisfactory, you are losing the ground signal from the ECU. Pull the 4 pin connector and check the terminal on the wire for pin #3. If that is good, you are going to have to remove the connector at the ECU and see why the signal is getting lost, may be a bad terminal on the wire or bad solder connection on the connector at the ECU. You can use your multi-meter on the ohms scale to check from the ECU connector to the 4 pin connector. Should be less than 1 ohm.
Tom

Posted by: Chris Pincetich Sep 22 2014, 11:24 AM

There was no ground signal at the #3 connection of the 4-pin plug. Adding a jumper to complete the ground circuit resulted in a working fuel pump!

I used the multimeter to check conductivity/connectivity to the ECU. From #3 connection on the 4-pin plug to the #19 spot on the ECU input there was a solid signal. The circuit seems to be good.

BAD BRAIN?!?!?!?

What's next?

Posted by: Tom Sep 22 2014, 12:38 PM

Chris,
Yes, If the wires and connections are good as you indicated, then the ECU has failed. It isn't giving the ground path that is needed. Temporarily you can run with a jumper from ground to the #3 pin. Just remember, the pump will run anytime the key is ON.
Tom

Posted by: Chris Pincetich Sep 22 2014, 01:38 PM

Thanks for confirming, Tom.
My 914 needs a new brain

I've learned a lot during this process and REALLY APPRECIATE all the great advice

Posted by: Chris Pincetich Oct 5 2014, 10:49 PM

Installed an ECU brain - mine was an early one with a VW part # ending in "B" and now I have a newer version with the part # ending in "E" - and everything is going great

In fact, I think it runs better overall with the "new" ECU.

I also inspected my TPS circuit board, as I was experiencing a "hiccup" at 2800-3200 RPM. Unplugging the TPS made the hiccup virtually go away, and the TPS board had some deep grooves. So, bent the little arms to get the contact back on good circuit board and out of the grooves, and the "hiccup" is gone.

Time for more driving