Mechanical Fuel Pump Pressure Creep
 

I had a 2-port (non-return line) mechanical fuel pump feeding the tri-power setup on my '66 GTO but was concerned with the wide swing in line pressures I was observing. Basically the needle on my gauge was dancing widely between 5 and 10~11 psi. After some reading on this forum, I decided to add a fuel return line like A/C cars got and switch the pump to a 3-port pump, specifically Carter M4690 (stock '68 Firebird 400 pump) which is listed as a 5-6½ psi pump. Should have been perfect.

The installation is complete and initial fire-up was successful in that the gauge bobbled gently between 6 to 7 psi, but as the engine warmed up and things stabilized (10-15 minutes run time), I noticed the fuel gauge crept up to 9-10 psi. Blipping the throttle a few times settled the gauge back to 6~7 psi but once at idle it again crept up to 9~10. I called Carter Engineering and the fuel pump tech line guy said "sorry, sounds like you got a bad one - return it for an exchange."

Anyone else experience pressure creep like this? I really expected the addition of the vapor/fuel return line to mellow out the pressure pulses and stabilize the overall pressure reading. I guess I'll try another M4690 if my first one is truly just a dud.

Where is your fuel pressure guage? If it is under the hood, the guage will heatup along with the underhood tempratures. An oil dampened sealed guage will read 3 to 5 pound higher when hot ( that is an erroneous reading)

Good point and thanks for the consideration, but I'm still in reassembly mode following an engine rebuild this winter, so there is no front clip on the car yet. Fuel gauge is between the inline canister fuel filter and the tripower inlet junction block on the front carb. I want to resolve this fuel pressure issue before the front end goes back together. Thanks.

I have had regulated fuel pressure creep with electric pumps but that was related to the regulator.

Can't say I've ever experienced the pressure creeping up with a mechanical pump/no regulator.

What I typically see with stock type mechanical pumps, no regulator, is the pressure will usually go down maybe a pound or so after everything gets warmed up.
For instance on the nomad, with an old but accurate Stewart Warner pressure gauge mounted on the cowl, a cold start will have 9 lbs. As it runs and gets warm enough it has 8 lbs and pretty much just stays there.

Another interesting tid bit because we were on this subject a while back, is cold starts, with what some consider too much cranking using mechanical pumps.
This car will hold fuel pressure after I shut it off for days and days, even weeks. When I explained my cars start rather quickly with mechanical pumps even after sitting for weeks I believe this is the reason why. I don't have drain back or fuel leaking issues that I believe must be going on with other examples.

In fact when I jump in and see 8 lbs. on the gauge before I even crank it, I can watch the pressure drop a pound every time I pump the gas as the squirter shoots fuel. As soon as I start cranking it's instantly back up to 9 lbs. and fires off within 3-4 revolutions after sitting for weeks.

Lost track of that thread but anyway thought I'd share.

My memory is suspect, but I recall reports of new aftermarket pumps delivering pressures higher than claimed when I shopped for a pump a couple years back. If true, then you are not alone.

I had fear of pressure then, as I was chasing a flooding issue and didn't trust my needle and seat to hold more than 6 or 7 psi.

Are you sure your symptoms aren't just what happens when fuel flows into your carb bowl(s) when the needle and seats are closed? If you mash it a few times, the inlet needle(s) should open and the pressure should drop a little. When the floats rise and shut it off, the pressure should reach the "max", right?

3 Carter short can pumps listed for 68-72 (M4868, M4690, M4885) all "spec" a max of 6 psi and a minimum of 5 psi, with a minimum flow of 40 gph.

Yours is definitely higher so will be interesting to see if a replacement corrects it. Basically, is yours a manufacturing defect or is it a design that doesn't deliver what is claimed?

As to the 3-line vs 2-line, Carter doesn't spec either one differently. The M4868 with 2 lines is spec'd the same as the M4690 and M4885 with 3. As I understand, the 3-lines don't allow much flow... basically a small orifice to reduce vapor lock. To me, this means I wouldn't expect them to deliver lower or even more steady pressure than a 2-line pump.

We would all not be guessing if Carter or any other brand shared data. I asked for pump curves and was told "no" by Carter, Delphi, and GMB.



I was on that thread and envy this story. What pump do you use that has zero leak-back? I want one!

Mike

Regarding the "small orifice" we've all seen on older 3-port pumps, here's an update for you. The Carter M4690 no longer has the plug in the tubing with a small drilled hole in it. That's gone. The return line port is simply a one-inch-long piece of 1/4" thin-wall tubing welded to the bottom of the main can. Pontiac may have specified a controlled-bleed port back in the day when AC-Delco made these pumps, but that feature is gone from the current Carter pump!

That particular car uses a Holley mechanical pump. Now that I think of it I believe it's the summit version of that. I'd have to go back to my purchase history but almost positive it's the summit branded version.

Wouldn't this significantly reduce flow to the carb? If this pump is rated for a max of 40 gph, it seems more than half would be flowing back to the tank. A high-flow inlet seat orifice in a QJ is about 0.13" diameter... a wide open 1/4" line back to the tank is way less restrictive.

Is it possible the tube extends into the can and ends with a small orifice?

Here's an example I found on the web:

https://forums.maxperformanceinc.com...1&d=1741705518

LIQUID-FILLED fuel pressure gauges are NOT TO BE TRUSTED...

...UNLESS you burp the gauge at every temperature change. A liquid-filled gauge will have a rubber plug, you need to pop the plug before taking a gauge reading every time the gauge temperature changes. Sometimes those rubber plugs have a molded-in nipple that you can cut, creating a tiny vent--but now the gauge isn't sealed. Fluid can leak out, dirt and fumes can get in.

This tends to be a problem with fuel pressure gauges especially, but ANY liquid-filled gauge in ANY application will have the issue of inaccuracy as temperature fluctuates, and they ALL need to be burped before they're taken seriously.

There's probably a strong hint in the name what the "vapor return line" is best at. Thinking it will reduce pressure at idle is not its job, and if it did, then the bleed off at higher RPM would greatly reduce fuel flow to the carb.

Here's a few photos of a standard mechanical pump cut up. First is the deep bowl and second is the standard bowl. Both were reduced down to about .067".

Thanks for posting these photos. Is consistent with logic and with the cross section from the HPPP article.

Any chance you have photos of the check valves from those two pumps? I'm still skeptical about the probability of fuel leaking back from the carb through the pump. Do the valve designs appear capable of preventing shutting off reverse flow or do they appear "leaky"?

Mike

I only have one photo showing a check valve. It is basically a rubber or Nitrile flapper valve that should function at 100% when new. As the pump ages with heat cycles the flapper will stiffen up and loose efficiency. I've also found that sitting around off the engine such as down time during a rebuild that the pump is usually garbage when put back on. Anymore I just order out a new pump and save myself time and frustration.

Several updates on my issue.

Inspired by the cutaway shots provided by Lust, I grabbed the second of the two M4690 pumps I had purchased and I probed the return line (which is easy to do since it's a short straight tube, not curved) with numbered drills to see if I could ascertain if it had an orifice feature to it or not. Starting with #70 drill (0.028") it went in till it hit the roof of the chamber, so now I know how far "bottom" is. I worked my way up to #20 drill (0.161") in 10-step increments, with #20 being the largest drill that will go inside the 1/4" steel tube. Every drill probed the same depth, meaning they all went in till they hit the roof of the chamber, clearly no orifice at play.

Next I went back to the car and with it running and warmed up, I put a catch bucket under the pump and pulled the rubber hose off the return port. Fuel came rushing out like it was a garden hose fully open. A quick check of the pressure gauge showed I was at 8 psi. (I'll try to attach a pic.)

With those two things being done, I called Carter Engineering again and I asked if their M4690 pump is SUPPOSED to have an orifice. Answer = yes. I shared my research with them and said I didn't think the pumps I recently bought had one. That and the fact that I was observing 8-10 psi. A very open and candid tech guy (Robert) said that when Carter moved pump production to China right about the start of Covid they encountered lots of problems with pumps not being made correctly. He added they have just returned pump production to Thailand and Mexico and things are expected to correct themselves in time. He suggested I return the pumps I have for refund as they have no way of knowing when there will be a clean break of the inventory in their system.

That being said, back to square one, I looked up the application again to see what other manufacturers might be available and one that came up is Delphi p/n MF0154. If you look at the pics of that pump on the Delphi website, the bottom view clearly shows an orifice in the return line. All amped up, I ordered one via my local Auto Zone who indicated inventory was readily available in their Detroit warehouse. I writing this post having just returned from the store - the pump in the Delphi box (made in China) is identical to the Carter pump in every regard, including having no orifice in the return port. Very frustrating.

Thanks for the update, sorry it was preceded by seriously bad luck.

What are the odds? Apparently very good.... Sad that both brands let what seems to be the same supplier do this to them.

Nice job figuring it out. Looking at photos online, the orifice is quite visible so I understand how you spotted it missing on the Delphi.

https://forums.maxperformanceinc.com...1&d=1742007781

I never even thought about questioning the return line on my Delphi pump. I guess I'd be smart to go look!



Mike

Serious question. If oil filled gauges are not to be trusted then what brand non oil filled gauge doesn’t just bounce around giving you no steady reading at all? I had a Mr Gasket and a Summit gauge before this one and they all just swung around wildly.

I use a summit dry gauge on my regulator and its rock solid, never had any bouncing or problems related to heat. Are you sure the mr G & summit gauges you had were dry gauges? Easy to confuse the 2. Heres what I have in use for 10+ years thats working great.

https://www.summitracing.com/parts/sum-800139

I use Stewart Warner and Autometer dry fuel pressure gauges mounted on the cowl and they are rock steady. I've never used the cheap Summit or Mr gasket brands so can't comment on those but I can't imagine they are of the same quality.

Yeah mine bounced like crazy. Exact same gauge so I replaced it and did the same thing. Liquid filled is steady as a rock

Weird, had my dry summit gauge in use for 10+ years, zero bouncing or heat related high readings.

Im sure they arent as good as high end expensive brands, but for ~$20 its worked great for me. Liquid filled are known to read high when hot so I stayed away from those for an engine bay mounted gauge.

If yours was defective or had problems, a quick call to summit they would have sent you a replacement for free or refunded your money.