Check out this picture:



See how the float arm offsets the float away from the pump? The float sits outside the bucket to the passenger side of the car (down on the image earlier in the thread).

In the picture above, the bucket wall will pass between the pump body and the arm.

So if you make a box around the bucket, wouldn't there have to be a hole or slot in it so the float could still work? Again, I've never seen a picture of the top part of the baffle, but I don't see how it works with the float.

 

maybe put camera on remote or timer in the tank, then put everything back together to see how much space, etc...

 

But I don't have the top part of the baffle...that's what I'm trying to get a picture of (or a technical drawing, or anything...)

-ch

 

I swear streldoc posted pics at one time but I'll be damned if I can find them. Perhaps they were on another site.

The bucket doesn't interfere with the level sender float because the new walls and lid only extend to the top of the fuel pump (approximately).

 

Damon,

I think it was Gene Felber on www.mantissaconsulting.com that has the photos of modified fuel tank with lid on the plastic baffle. But I think Gene close the website. Does anyone know the new address?

 

I PM'ed Gene a few days ago but he didn't respond...

-ch

 

Here's the site on wayback--but no images!!!

http://web.archive.org/web/200505160...uel_system.htm

-ch

 

How about a second in tank pump but on the opposite side of the tank? I have lots of ideas but none are a simple bolt on solution. Another idea is to put multiple cuts in the bottom of the tank(width) that are as thin as a peice of sheet metal, take a peice of sheet metal and drill a bunch of 10mm holes in it, bend the end of it into an L shape and insert and weld. Do 4 of these in the tank and the sloshing will be cut down drasticly while still retaining a stock looking tank. I should draw an example, but I am at home not at work.

 

Well, here's what I'm thinking: every race team worth their salt has an accumulator (usually in-tank) and a high pressure pump the pulls from the accumulator. The problem with the FD tank is the access panel is too small to put a decent size accumulator in there--and will cost at least $240 for the check valves alone. (This is true of modifying the stock baffle, too.)

An external accumulator (with a high-pressure pump inside it) will cost about the same and be much more reliable. An you don't need to muck around inside the tank (and can leave the stock wiring and inlets in place).

You just need to find a place to put it.

-ch

 

I don't want to add another tank, I am going to try something not sure what yet. I am one of those people who doesn't give up and when I make something if it doesn't work I try again till I get it right. This fuel **** has cost me alot over the last 9 years that I have owned this car.

 

Okay, well, if you're going to stay in the tank, here's an idea that might work:

http://www.airpowersystems.com.au/35...350z-fss09.htm

for reference, a 4" x 4" x 4" box holds about 1L of fluid. (less if some of the displacement is taken by the pump.)

-ch

 

Looked at it but I am not fully understanding. Both pumps are supplying or is one being supplied by the return as well as sucking?

 

Both pumps supplying. They are just in a collector that has holes in it on the sides you can't see. (APS doesn't want to give away their research for free, and I understand.) The return goes back into the collector.

-ch

 

I looked through a bunch of old threads and it looks like the 96+ baffle is 270 through mazdaspeed. I couldn't find any pictures of it. from the looks of the stock one and knowing the the 96+ just has a cover on it, I don't think it would be hard to fabricate something out of Aluminum sheet. If I bought the 96+ and knocked it off, how much would everyone be willing to pay to solve this problem?

 

It's not true of modifying the stock bucket around the pickup. It requires less than $20 of parts!

I read that and thought "No way!" Then did the math and it comes out to just over a liter. Still doesn't seem possible!

 

I built an external surge tank for my GSL-SE racecar which solved all of my starvation problems. There is a very good discription of a simple surge tank like the one I built here: http://www.sdsefi.com/techsurge.htm
This one is much fancier that the one I built. I welded caps on a piece of 2.5" exhaust pipe and then welded small tubes in for the fuel line connections. Since it's all low pressure I only used push on rubber hose with hose clamps. I had an old inline Bosche fuel pump laying around that I used for my aux supply pump. So, I spent no money whatsoever to build it.
You just have to decide if you have:
A) More money than time.
or
B) More time than money.
For me, it always has been B.

 

I had Mitch Piper at Piper Motorsports made a surge tank which sits in the charcoal canister location. Although, I am not installing it/reworking my tank setup until I get my car running and enjoy it for a few months. I would post a pic but I am at work.

 

Okay, just out of idle curiosity...let's say there were an aftermarket baffle cover that worked as well if not better than the $220 Mazda part (if you can even get it). Let's say that the aftermarket part required a bit of assembly but had some distinct advantages. Would you pay $99 for it?

-ch

 

Thanks for that link! it explained the way that system works perfectly

 

Yes.

 

If you buy a new fuel tank from Mazda it will have the updated baffles. I think it's around $300 from Mazda Competition.

There was a definate difference in the baffling between the two but I didn't take any pictures.

 

Did you notice that with the new tank the problem was solved? I would be in for buying a new tank, how do I get it? I thought you have to be a racer to get mazda comp parts?

 

Sorry....but just thinking about this from another perspective (so I may be way off base). It seems we have this pickup problem mostly on high G left turns since our pump is mounted over on the left. And say what will be the longest duration we will need to pick up fuel on on a turn? WOT for 30 secs? As an example, my max fuel will be (2 x 750cc/m prim)+(2 x 1680cc/m sec) x 30 seconds = 2430cc. That's 2.4L! Am I off??? I doubt we'll be consuming that much...so how big a surge tank/baffle does one need to effectively keep fuel located by the pump to prevent starvation?

I ask this because I was thinking what if we consider baffling not from the tank up but from the lid down. A 3-sided trap that just keeps fuel from quickly flowing away from the pump during that turn duration. Is that at all possible? Just my random babbling...

 

Actually, let's talk about what happens. There are two conditions that lead to starvation in the 93-95 tank:

1) long sweeping left hand turns -- at 1g, the fuel surface is at a 45 degree angle in the tank. Much of the fuel that was in the surge bucket flows over the side, and what's left (1/2 * 3x3x10, or 45 cubic inches, or about 2/3 of a liter) is quickly depleted by the motor which consumes about 1L every 15 seconds.

2) heavy braking followed by a left hand turn -- the car dives under braking (3 seconds) throwing the fuel over the front wall of the surge bucket. Then the car turns in, launching more fuel over the right side of the bucket. There is probably less than 1/3 of a L left, so even moderate acceleration will cause starvation.

The latter case was what happened to me in turns 11 and 2 at Laguna. Heavy braking followed by a left. I had starvation at just over 1/2 tank.

The bucket in the tank (shown earlier in the thread) tries to do what you're suggesting: block fuel from three sides. Unfortunately, since the fuel level sensor arm hangs over the bucket side, you can't just make walls that go up to the top of the tank.

The factory approach is a cover that sits across the bucket and below the fuel arm. This provides two advantages:

1) Reduced 'slosh' over the sides of the bucket under braking or sudden turns
2) During long left turns the bucket holds more fuel (although some of it is leaking out through the hole where the pump and return tube go through the cover). My guess is that the capacity of the bucket is roughly doubled with this approach.

I have fabricated my own cover and will be testing it at Laguna in early March.

-ch

 

Pics of what you fabricated?