Never ending fuel pump drama

[GTSBoy]

I've just had the world's best bad idea. We have fuel returning back to the tank. Just point some/all of it at the inside of the bulkhead fitting. Direct liquid cooling!

[MBS206]

7 hours ago, Murray_Calavera said:

 

The ring terminal is slightly larger then the stud, I would not describe those two as making good contact with each other. My assumption was, the good contact comes from the 2 washers on either side of the ring terminal making a good connection, and the washers make a good connection with the nut, which makes a good connection with the stud. I'm guessing this was an incorrect assumption? 

This is the entire issue.

The stud should have a ringland on it, such that the stud metal, is also the same piece as attached to the ringland.

 

Think of it this way, take an actual bolt, turn it upside down.

The head of the bolt, is the ringland, and it is all part of the stud, so you put your ring crimp over it, a washer, and bolt it down.

 

Because you've got a stud with not metal landing, all of your real electrical connection, is through the threads in your nut to stud. Add to that, your nut is a nylock, so it isn't even necessarily all metal to metal contact.

But on a 6mm stud, your thread depth is probably 0.5mm probably a 1mm pitch, and being about 4mm of nut height.

That all equates to f**k all quality contact area. Not to mention, all the other joints in that area adding tiny resistances which all add up to bad heat!

 

You're going wire to crimp, crimp to washer, washer to nut, nut to stud, stud to nut, nut to washer, washer to crimp, crimp to wire. 8 joints in a small area, each with a small resistance.

If each joint adds 0.01ohm, that's 0.08ohm (basically immeasurable on most digital multimeters, you need a 4 wire measurement).

At 25amp, 0.08ohm is giving you a 2V drop.

2V, at 25amp, is 50watts of power. Where is all that power going? Heat.

You can now get someone else to go do the mafs on your stud and specific heat, and you'll start to realise why shits melting.

Even if you say I've way overestimated the resistance by a factor of 10, a 5w power input constantly as heat, is pretty high too and that's on 0.008ohms!

 

Get a good stud kit, that has a metal landing. Get good quality crimp with plenty of metal in the ring.

 

PS those JayCar ring terminals have the thinnest little ring. That area that GTSBoy indicates is one of the more meaty areas of the ring terminal!

[Murray_Calavera]

Ok, I've just ordered the high amp flanged kit from EFI Hardware. 

Once it's installed, I'll go for progressively longer/harder drives and I'll check the temp of the terminals with the heat gun. 

If there are any issues with temp, I'll drop in the Walbro 460 which should draw about 5 less amps and I'll repeat the process checking temps. 

Also I'll make sure to take some high res photos of my crimps this time =P

[Dose Pipe Sutututu]

1 hour ago, Murray_Calavera said:

Also I'll make sure to take some high res photos of my crimps this time =P

I want to see crimps better than what would be minimum standard for the now dated International Space Station.

Street GT-Rs now have better wiring than Space X Falcon Heavies

[MBS206]

1 hour ago, Murray_Calavera said:

Ok, I've just ordered the high amp flanged kit from EFI Hardware. 

Once it's installed, I'll go for progressively longer/harder drives and I'll check the temp of the terminals with the heat gun. 

If there are any issues with temp, I'll drop in the Walbro 460 which should draw about 5 less amps and I'll repeat the process checking temps. 

Also I'll make sure to take some high res photos of my crimps this time =P

Put up photos of the crimps you're planning to use. Let us critique them 😛

But seriously post them up before you use them

[Borci88]

Weird that everybody thinks the terminals wouldn't work. I've had this style set up running on a 33 GTR since 2017 and don't have any fuel pump or heat issues, this is with dual relays and dual fuel pumps. I believe I used the EFI Hardware terminals listed above. 

[MBS206]

1 hour ago, Borci88 said:

Weird that everybody thinks the terminals wouldn't work. I've had this style set up running on a 33 GTR since 2017 and don't have any fuel pump or heat issues, this is with dual relays and dual fuel pumps. I believe I used the EFI Hardware terminals listed above. 

Because he's using a different product, designed different, works different. Can create a problem.

[Butters]

I am wondering how much the plastic hat plays into this.  Both of mine melted more opposite the wire crimp. Tried the efihardware ones , double washers , better eye terminals.  The seat of the plastic ring inside the metal hole is very thin 1mm or so and you are putting through a threaded bolt, which can cut the plastic, won't take much for some conductivity to occur in a metal fuel hat that is earthed. 

Amp draw of pump is going to be a variable here too.  many things that impact that. 

[Dose Pipe Sutututu]

I'm going to sound like a broken record now, PWM your pump and you won't have this issue. Unless you're a speed boat and on WOT for minutes on end.

[Murray_Calavera]

2 hours ago, Dose Pipe Sutututu said:

I'm going to sound like a broken record now, PWM your pump and you won't have this issue. Unless you're a speed boat and on WOT for minutes on end.

I promise it will get done as soon as I can, but that is still probably 3 months away. 

[MBS206]

1 hour ago, Murray_Calavera said:

I promise it will get done as soon as I can, but that is still probably 3 months away. 

I honestly think you should move it up by whatever priority possibly.

The time you're spending chasing this issue could be better spent doing PWM and resolving it fully.

[Murray_Calavera]

Update time, the Efi Hardware bulkhead terminals are in. The quality of the kit is leaps and bounds better compared to the Efi Solutions/Taarks kit. I was not happy with the Taarks kit, however I am very happy with the Efi Hardware kit so far. Being able to torque down the plastic insulator is a huge help. The longer post and jam nuts also make life easier. 

I didn't bend the eyelet terminals 90 degrees to fit inside the bulkhead this time. I performed some surgery to the bulkhead to fit the terminals without bending them (in case the bend increased resistance in the terminal). 

Unfortunately I didn't have time to drive the car, but letting it idle for about 10 minutes and the power terminal went from about 13 degrees to 16 degrees. 

I'll update again after I've taken the car for a good shakedown to check temps. 

As promised, I have taken some photos so my crimps can be judged by all.  

Crimps were made with Delphi crimper (part number 12085271)
Heat shrink is Raychem RW-200-E recovered at 170 degrees. 

[Dose Pipe Sutututu]

Crimps are looking better, yashimash.

[Murray_Calavera]

Took the car for a hard drive (the same drive that melted the Taark kit) and a couple of casual drives today. The terminal temps appear to be tied to ambient temps. 

After each drive, including the hard drive, the terminal temps were always within 3 degrees of the other metal surfaces in the boot. Everything appears perfect so far. 

I suppose the next real test will be when I drive down for WTA, I'll check the temps at each fuel stop and see if extended drives change anything. 

(Just waiting for, see it was the crimps all along! Pretty crimps now and no issues, must be the crimps! lol)

[Dose Pipe Sutututu]

I blame the crimps 

[MBS206]

What are the ring terminals you used?

[Murray_Calavera]

26 minutes ago, MBS206 said:

What are the ring terminals you used?

The ones that came with the kit - 

https://www.efihardware.com/products/3194/submersible-bulkhead-terminals-hex-head

I told myself a few times, take a photo of the ring terminal with a ruler for scale. But then I didn't. 

Suppose the most important bit is it's all sized appropriately (ring terminal is the same size as the post), this photo is a good example - 

 

[cabramatta]

Might piggy back off this thread since there's been a good discussion about PWM.

 

I purchased a Walbro 460 F90000267 and have been looking into PWM to understand it. It was purchased more of a future proof as my original OEM fuel pump died (r33 below 200kw) and figured I'd just upgrade it to this. That's when it hit me that a pump this big can cause more dramas with such high amperage draw and heating of the fuel pump being on 100% all the time in such a mild setup.

 

I have a Nexus R3 and slowly collecting power mods but I was looking into the PWM feature on the NSP software:

 

Then looking at the graph from the fuel pump:

 

 

Am I safe to assume reducing duty cycle to 50% for example would make the pump run like a 230/lph ? Where does the PWM come in all of this ?

 

I've gone down the rabbit hole of PWM fuel pumps and it seems there are many differing opinions on how to go about it 😅

 

Edited March 28 by cabramatta

[GTSBoy]

PWM quite literally just chops the ful 12V voltage on and off at whatever duty cycle you're running it at.

If you're running it 100%, then it is on full voltage all the time, same as if there was no PWM.

If you run it at 50% duty cycle, it is only seeing the full 12V half of the time. This is broadly equivalent to running it at 6V. But the crucial difference is that motors (and a lot of other loads) don't like being powered at low volts. They will either fail to start rotating, or draw a shit ton of current, or other undesirable things. But if you give them the full volts, and then a short period of no volts, and then some more full volts, then the times when they are seeing power they are seeing all the voltage, and they are happy. But you get the performance out of them as if they were only seeing that fraction of the full voltage.

It is not really easy to answer your question about what flow you will get out of it at 50% duty cycle. I can tell you that it is not as simple as you think. For a start, that 460 L/h pump is not going to flow 460 when you're on boost. When you're on boost you will be somewhere down the sloping part of that red line. If you have 15 psi of boost, then the pump can only deliver about 95 gal/h, which is <380L/h. So your simple 50% on 460 = 230 wasn't going to work anyway.

But also, it won't deliver 50% of 380 either, because when you devolt or PWM trim th epower being fed to the pump, it is not able to deliver flow or pressure in the same way. 50% duty cycle will probably produce <50% of the full voltage flow.

The way to find out what duty cycle you need to run it at at low load (ie, at idle) is to idle it and turn the DC down until you start to lose fuel pressure, then turn it back up above that with some extra for safety. And then you do the same thing at full load, in case it doesn't need anywhere near 100% DC. And if you're careful/cautious/prudent, you will also do it at a couple of loads in between so you can shape the DC map against load. It might not be linear between the two end points.

[cabramatta]

Appreciate the detailed response! It's all new technology to me and your suggestion for reducing DC and finding sweet spots seems like a great approach.

 

I'll chuck a clamp meter on it at the same time and figure out whats happening at different settings to get a good understanding.

 

Will report back with my findings when I get around to it to help anyone in the future that will be in the same situation as me 😁