Fuel Pump speed control

Fuel Pump speed control

Author
Discussion

Max_Torque

13,783 posts

161 months

Monday 13th February 2017
quotequote all
Fuel pressure is controlled by the pump, fuel consumption by the engine!

All the pump controller needs to do is to maintain the desired rail pressure (that may or may not change with speed, and may or may not change with load)

stevieturbo

14,179 posts

191 months

Monday 13th February 2017
quotequote all
Vee 12 said:
Assuming that the pump's capabilities have been tested like in the Bosch chart above; Is there any aftermarket controller that allows you to command, "I want 2.0 Bar rail pressure and 50 LPH at 1200 rpm engine idle" and then command "At 7000 RPM & 2.0 Bar of boost, raise to 9.0 Bar rail pressure and 200 LPH". Does that type of system exist or am I missing some important information that makes this an impossibility?
As already mentioned earlier in the thread.

Vaporworx in the US offer a controller that runs closed loop for pressure. How good or bad it is...dont know, how suitable it will be for you...dont know.
They are made for a lot of modern US cars and they offer retrofit kits too.

But it's probably about the closest thing an aftermarket company offers for what is discussed in this thread

https://www.vaporworx.com/boostworx-controller/

GreenV8S

27,314 posts

228 months

Monday 13th February 2017
quotequote all
Max_Torque said:
All the pump controller needs to do is to maintain the desired rail pressure
Is this assuming a single ended fuel rail where the pressure is controlled entirely by the pump? I expect that's how the electronic fuel pump speed controllers are designed to be used, but it's not the only setup.

Vee 12

11 posts

30 months

Tuesday 14th February 2017
quotequote all
What if you need say, output voltage is higher than the supply, for example you want 15 volts and your supply is only 13.5 V, how could you raise, Series on a Profet IC?

Tony Palo from T1 Race Dev. is making a fuel pump controller that will flow identically in 10v-18v conditions. Its supposed to on sale in April 2017. Sounds expensive.
https://youtu.be/skMIayfOGj8?t=36m31s

GreenV8S

27,314 posts

228 months

Tuesday 14th February 2017
quotequote all
Vee 12 said:
you want 15 volts and your supply is only 13.5 V, how could you raise
My view would be that you've chosen the wrong pump/controller and would be better off choosing one that could do what you wanted with the power you had available. But if you decide you really do need to raise the voltage, you would use a boost or boost-buck converter.

stevieturbo

14,179 posts

191 months

Tuesday 14th February 2017
quotequote all
Vee 12 said:
What if you need say, output voltage is higher than the supply, for example you want 15 volts and your supply is only 13.5 V, how could you raise, Series on a Profet IC?

Tony Palo from T1 Race Dev. is making a fuel pump controller that will flow identically in 10v-18v conditions. Its supposed to on sale in April 2017. Sounds expensive.
https://youtu.be/skMIayfOGj8?t=36m31s
He's clueless if he thinks a Walbro is the biggest in tank pump around. And given how easily their injectors fail....again I'd be shopping elsewhere anyway. So dont see how any pump controller he has to offer will be any game changer

Back to your queries...what exactly are you trying to achieve by using a different voltage ? There are many voltage boosters around.

Max_Torque

13,783 posts

161 months

Tuesday 14th February 2017
quotequote all
A standard PWM motor controller is a BUCK converter By varying the duty cycle of the output signal, the effective output voltage can be reduced from the supply voltage. In this case, the inductance of the motor is used to smooth the Pulse Width Modulated voltage output into a more constant current, and hence motor torque.

By re-arranging the architecture you can build a BUCK-BOOST converter, that can both reduce the output voltage below that of the supply, but also increase the output voltage ABOVE the supply voltage. In this type of architecture, an external (to the motor) inductor is used to store energy (as a magnetic field) that can be transformed to an output voltage on either side of the supply voltage.

Plenty of info on google about buck-boost and other architecture voltage converters.

However, as a typical large fuel pump takes up to 20 amps, you're potentially talking about a 400w converter, which is non-trivial to design and specify, and requires expensive magnetics / power switches.

Vee 12

11 posts

30 months

Tuesday 14th February 2017
quotequote all
stevieturbo said:
He's clueless if he thinks a Walbro is the biggest in tank pump around.
Pardon my ignorance, who has the biggest electric in-tank pump then (brushed or brushless)?

Vee 12

11 posts

30 months

Tuesday 14th February 2017
quotequote all
Max_Torque said:
However, as a typical large fuel pump takes up to 20 amps, you're potentially talking about a 400w converter, which is non-trivial to design and specify, and requires expensive magnetics / power switches.
That sounds like it would dissipate lots of heat...

stevieturbo

14,179 posts

191 months

Wednesday 15th February 2017
quotequote all
Vee 12 said:
Pardon my ignorance, who has the biggest electric in-tank pump then (brushed or brushless)?
They Veyron pump has been around for quite a few years and flows a lot more than the Walbro and also at a fraction of the current. Walbro quote 450lph at free flow which nobody would really use, at working pressures it's down to around 350lph.

http://www.jjcraceandrally.com/race/fuel-pumps/pro...

Holley also released a version recently although as to how tried and tested it is, is anyones guess.

Whilst not in-tank, Fuelab also make some fairly big electric pumps that have been around for some time too.

Vee 12

11 posts

30 months

Thursday 16th February 2017
quotequote all
A buck-boost converter won't be required if PWM is used. A boost converter will be.
I found a Intersil 2-Phase Boost Controller "ISL78229" can work as its designed for start-stop electronics in vehicles. I looked at the eval board and appears to be using a SER2915H-472KL high current inductor. I wonder how if it would still require the BTN8982 to keep current/voltage/undercurrent protection?

@Max_Torque I'd like to make a boost converter PWM fuel pump controller, how much time does it take to set-up the hardware for software programming/testing?

GreenV8S

27,314 posts

228 months

Thursday 16th February 2017
quotequote all
Vee 12 said:
A buck-boost converter won't be required if PWM is used. A boost converter will be.
This is true, but you may find a boost/buck converter working as a variable voltage power supply is cheaper than a boost converter plus another layer of power electronics for the PWM driver. At these power levels the PWM driver isn't going to be trivial, and you've already got most of the hardware in a boost converter.

AlexjC4

3 posts

27 months

Thursday 11th May 2017
quotequote all
I thought I'd contribute to this thread as I'd returned to it a few times while investigating aftermarket PWM fuel pump control options while using OEM systems.

I've come across this controller that looks like a newer version of the Jaguar control mentioned previously on this thread.

Aston Martin / Range Rover p/n 7H42-9D372-AA

Jaguar (XK/XF 2009-2014) p/n 7W83-9D372-AA

These come on a variety of engines up to 6.9 V12 and 5.0V8 supercharged variants so should support fuel delivery for >500BHP. These are single fuel pump systems (not separate low and high pressure pumps) and not returnless (I think). They are marked both VDO/Seimens/Continental and FoMoCo, the case is a large die cast heat sink. I've opened the case on both part numbers and the components are identical, although the microcontroller may be programmed differently.

Readily available on ebay from £30-40

wiring diagrams for Jag/RR and AM applications are available online, and I believe the pinout is;
1 Gnd
2 Monitor signal to ECM
3 Fuel pump -ve
4 +ve from fuel pump relay
5 PWM control signal from ECM
6 Fuel pump +ve

This uses an esoteric Yazeki YESC connector with 2.8mm blade terminals p/n 7283-5577-10

This connector is available cheap, slow and in bulk from china ($15 for 5) or fast and expensive from NZEFI, if you ask nicely. Or simply use insulated 2.8mm blade terminals. NZEFI also supply a kit with a refurbished fuel pump control module and connector - but charge a heavy premium imo.

The module itself uses the following components:
MLX16201 Automotive microcontroller and "relay driver". This receives the low Hz PWM signal from the ECM, driver the power FET and transmits the "monitor" signal back.
B2545G Shotky diode. Good current rating. Pretty good heat sinking too.
BTS282Z Infineon "TEMPFET" Automotive Low Side FET with temp switch. Looks beefy enough.

The back side of the board just has a couple of big chunky inductors and some quality electrolytic caps. As far as I can (not an electronic engineer) I have verified the circuit layout matches the pinouts I've found.

I'll be bench testing an Astom Martin unit with a rigged up PWM source to verify the behaviour. From what I've read I'm guessing very likely a ~100Hz input PWM control signal from ECM, 20kHz PWM to pump, probably a control range of 10%-50% duty cycle on the input PWM, with perhaps a magic 75% duty cycle signal to turn the motor off.

Any thoughts, especially from MaxTorque?

Steve_D

Original Poster:

12,484 posts

202 months

Thursday 11th May 2017
quotequote all
Watching and waiting.
Looks like all I need is to understand what output I need from my ECU and how to map that output.

Steve

stevieturbo

14,179 posts

191 months

Thursday 11th May 2017
quotequote all
I bought a Crydom SSR the other day to do this on my own car.....although ran into a hiccup when the same output also feeds the bloody oil scavenge pump for the turbos lol.

So need to isolate the two before I can get it running properly as the normal relay for the oil pump didnt like the PWM signal !

I used a 60A Crydom to run 2 pumps and in testing ( engine off ) it all seemed to be ok. This SSR is rated for 1000Hz max though.
I also see they list a specific DC range of SSR's for this type of thing, although again it's 1000Hz max. This isnt what I got, but specs seem almost identical.

http://www.crydom.com/en/products/catalog/power-pl...

AlexjC4

3 posts

27 months

Monday 29th May 2017
quotequote all
https://youtu.be/zz6rnG9VznM

Finally got a chance to hook up the fuel pump controller on the bench. Just using a 12v 21w bulb as a dummy load so the output can be visually checked as it responds to the input pwm signal. All seems to work as expected.

The input duty cycle of around 30-70% seems to corespond to an ouput of 0-100%. I'll check on the scope to confirm, that can be done on the car.

Looking at the design of the controller, I guess no external diode is required across the pump. Generally the quality of components is high and the power FET, large capacitors and inductors on the back side of the board are all heat sunk to the case. The FET also has a temperature feedback ouput to the driver IC to prevent overheating. Everything points to the controller handling 20amps of load without trouble. The lowest rated component is the connector plug.

Edited by AlexjC4 on Monday 29th May 12:39