Oddjob

With the changes above, you are now at a voltage drop of 0.3vdc at the pump. This is really pretty good, and I will bet it represents a considerable improvement over the stock wiring. I think you will see a further improvement when you run the star-ground lead you described from the firewall to the taillight ground buss. With any luck, all of this will fix your problem. If not, at least you have eliminated the prime candidate. Good luck.

Well, here's the science of it: There are two reasons for your relay to be hot. The coil that holds the relay closed will dissipate some heat. This is normal, and will have no effect at all on your fuel pump. The other reason is resistance across the contacts. A tiny amount is normal, but as you say you have a 0.2vdc voltage drop across the contacts, this could indicate a problem such as pitted contacts. The repair is to file the contacts smooth, and then burnish with emery cloth, assuming you can get inside the case. A new, good quality relay should show an unmeasurable voltage drop across the contacts, at least with a hobby-quality VOM. What is of more interest is whether your pump is hot. The lower the voltage you supply to the pump, the higher the current (amperage) will go, creating more heat. The voltage drop to your pump is governed mostly by resistance, and that is chiefly a function of distance and wire size. The stock wire run to the rear is almost certainly too small a guage for a performance application. I used 10AWG on mine (a Holley red-top), without a relay, from the 12v source all the way to the pump. Another problem is the return path (ground). I measured the resistance across the stock return path and found it around 36 ohms. This results in a substantial voltage drop at the pump. Since I have star-grounded most of my car, the pump now has a dedicated 10AWG run back to the ground bus near the battery (resistance = 0 ohms). Most of the details of all this are posted elsewhere on Hybrid Z. My initial reaction is that your pump is heating up after being run awhile, due to voltage drop. Since the Bosch 044 is air-cooled you can easly check it after a run on the freeway on a warm day. When a pump becomes heat-soaked, its performance changes for the worse.

Here's some information and pictures of the installation on my 1971. The information starts at Post #21, on Page 2. Rose's Ride Fuel Pump Installation

Umm... you haven't said what year your Z is. The switch has changed over time. My remarks above are specific to the early Z (mine is a '71). Sorry for jumping to the conclusion yours was, as well....

No, this won't work at all, unless you are going to rewire much of the car. A simple DPDT center off switch dosn't have anything like the number of poles you would need, nor does it have any possibility of recreating the switching logic of the original. I took a quick look at the wiring diagram I have, and it would take something like 13 poles to recreate the switch (just for the turn signals), and probably more when you start to do the logic. For example, I set out to replace the emergency flasher switch, which you'd think could be a DPDT. It wasn't. Take a look at the eventual solution here for a preview of what you would have to do to "replace" that switch. Scroll down to posts 18, 19, and especially 20 (open the PDFs in 19 & 20). It took 15 connections worth of switching to do, and is much simpler than the combination switch. I don't want to discourage you, it can be done, but you will have to do some circuit design. It won't be a half-hour project.

You are quite right, the tank shown is different than the earlier tanks. Do you have any idea what years this would cover? Yours is a '77, I gather. You might post as much info as you can for the benefit of the 280 guys (dimensions, sources etc.).

Try searching on "star ground" which will yield some interesting results. The idea of cleaning up un-used grounds sounds attractive at first, but you may want to consider the rule of unanticipated consequences, which is likely to bite you. The Z is unibody construction, not a frame, as you probably know. I have measured the grounds to various points, and have found readings from a few ohms up to about 250 ohms. I ended up using a star ground, which measures below the meter resolution (<1 ohm) to each connected load. Works great.

There is a sticky on this around here somewhere. Since they moved the stickies, I'm not sure where. In any case, here's an explanation about the "non-linearity" that you have observed, and why you can't fix it with resistors, different senders, etc. The stock Z sender is designed to output in a non-linear fashion. The stock Z guage is linear. The Autometer sender is linear. The autometer guage is near linear, but compensated. The problem is the stock Z gas tank. If you look at it, it contains much more fuel in the bottom third of height than the top two-thirds. What that means is that the tank fills in a non-linear way. Bending the float and/or putting in resistors will only change the F and E indications, but not the indications in-between. The stock Z sender was designed to be non-linear to compensate for the differing amount of fuel this oddly-shaped tank contains at a given height. Remember, the float indicates a height, not a capacity. In a regularly-shaped tank (a rectangle), the fuel height is directly proportional to a volume. In the Z tank, it isn't. The solution I am using is to use the Autometer sender and guage combination, adjust the F and E points using the float arm, then fill the tank a gallon at a time, re-marking the guage at the accurate capacities for 1/8 through 7/8. In effect, I am re-calibrating the guage to read in a non-linear, but accurate fashion.

Whether or not it is legal, it is a really bad idea for about a dozen reasons. Search on this topic here on Hybrid Z, and you will find a great deal of information, including those dozen reasons.

The tank shown in this post can be filled with fuel cell foam (by the builder), as an option. It would be about $700, I think.

Bodywork on the car has started. It is being done in phases, as the car can't be off the street for an extended length of time. The first phase was a group of small, annoying modifications, including new outside mirrors, new marker lights, new rear bumper, better fit for the antenna, and sealing the badge holes. Note that these are not repairs, but hybrid solutions to various common problems. All of these are detailed below with pictures and parts lists: Unusual Body Details

Here's the high end of stock tanks. It doesn't seem to need either a sump or a surge tank, doesn't require any cutting, and is outside the passenger compartment. "Stock" tank It was $590.

Your problem is almost certainly a bad or missing ground somewhere. When two seeming unrelated things happen like this, that is almost always what does it. To start troubleshooting, and get a sense of why this could happen, go to this discussion: Hazard Switch, and look at the drawings in post #1 and post #19. Good hunting.

Thanks for the information. My first order of business will be the hatch lock. I'm going to go with the PT1500. I'm also replacing the stock licence plate light while I'm in the neighborhood. The plastic on the stocker is failing structurally.

Really nice work! A few questions: Any downside to sealing the headlight bucket seams, such as maintenance issues, access to the headlights, etc? What did you use for the door mirrors? Are the turn signal lenses the ones MSA offers? What is the source of the electric latch for the hatch? I know good ideas when I see them, and I'm going to shamelessly borrow a few of yours.