Problem with almost all the ebay speed controllers, is they base the adjustment on a variable resistance...ie the ****.

So not sure how easy they would be to use for such things.

I've one here I bought many years ago to try, but it wants a proper PWM input and is a high current controller for a pair of motors. Needless to say it and a couple of Arduino boards I also bought....are still sitting on my desk lol

 

I planned to just use an arduino Nano and a PID to control a high speed Solid State Relay.

The formula shouldn't be too hard, since I can do whatever rate I want. The PID just needs an input, target, and output.

Input would be fuel pressure from a 0-100 PSI sensor mounted on the rail.
Target could be many things. For example:

basepressure = 60;
Target = basepressure;

or

basepressure = 60;
Target = basepressure x (kpa/100);
This would give me 50% of base pressure at idle (30 PSI) or 200% at 200 kpa (120 PSI).

or go with the normal 1:1 rate:

basepressure = 60;
Target = basepressure + ((kpa-100)/14.5)


Then the PID will output a PWM to control the solid state relay.
The relay would be wired exactly like a standard relay. Battery to relay to fuel pump. Arduino would pulse the relay, the PID would constantly adjust the pulse width to maintain whatever fuel pressure the target is.

A standard 25A SSR should be fine, since the DC-DC SSRs are just a MOSFET with a big heat sink.

My only concern is how much is a fuel pump going to enjoy a 5% duty cycle at idle. The programming and hardware in theory are very simple. Its the details like that which I'm not sure about. Not to mention, the Arduino will only do PWM up to ~500 hz.

 

When I tried messing about PWM'ing my pumps....just going by sound alone, anything under 45% duty and they just didnt sound nice at all.

Wasnt really paying much attention to pressure or current though, was just an audible test.
Lower PWM frequencies just made it all sound worse, so I kept them up near the max of the Crydom SSR I had which stated 1kHz max so I tested around 900Hz

 

Bah lol.

My main reason for wanting to do this was overpowering the factory truck regulator with a pair of AEM 380s.

I could still do a PWM controlled pump on my current setup and set minimum pulse width to 50% which would let me get away with a pair of them running full time with no problems.

 

As said..I'm not saying they were not working below that...they just didnt sound nice. But that's usually a good indicator they're not happy too.

Whether that would be the same for them installed in a different environment I dont know.

Perhaps against a smaller regulator or setup offering more restriction they might be happier at lower duty cycles.

Only way is to test them on yours, and I'm sure that at higher frequencies, they'd maybe be able to run lower duties too.

 

I'm not in a hurry to redo my entire fuel system right now.

I have a pair of 380s and I'm going to try them with the factory truck regulator for now to see how well it is able to control fuel pressure with that much flow. If not, PWM on my current fuel system would take me 30 minutes to get hooked up. I'll just set a minimum duty cycle to 50% to start with.

I imagine a complete system would be easy to do on an N/A setup where you can get away with a higher % duty cycle at idle.

 

I can certainly appreciate the DIY aspect of this, but at the end of the day, why not use the Vaperworxs dual OEM style set-up? By the time you experiment and burn up a couple of good pumps, you might be money ahead...

If you do try the DIY approach, why not look into the various Ford PWM controllers? Ford has been doing this for a long time and I am sure that some of the controllers are not CAN bus, but I do not know that for a fact.

I know on the new GM stuff (e92 ECU) the PWM fuel pump controller communicates with the ECU via CAN, so that would be harder to implement.

Andrew

 

I haven't bothered with the OEM stuff because the arduino does flawless PWM with very little effort/programming. I can have it output based on a genuine (and possibly better than the GM OEM based) PIDs, and its 100% configurable in any way I want.

As far as cheap PWM control goes, this takes the cake. I've used it to PWM dozens of other things without a single issue. This will be my first fuel pump attempt though.

This is on a microsquirt controlled engine and I'm already using the CAN on it to do closed loops idle control, gauges, water/meth, fans, etc. I'm really shooting for a stand-alone PWM based fuel system here.

 

My concern is based not on the source of the PWM signal (in this case the Arduino), but more on the "middle" part, in your case a SSR. I think the OEM's do something more sophisticated for this "middle" part, hence the various PWM "controllers" (like the C6 fan controller I mentioned before).

I've gone down this road a very little bit, because I have a Holley Dominator, which has user programable PWM- and PWM+ outputs. The challenge has always been the interface between the ECU and the electric motor (in your case a fuel pump). As I understand it (I could be way wrong) it takes a special kind of electric motor to be able to operate in a lower voltage environment. Or, it takes a "special" controller that is more than just a SSR.

Again, I applaud the DIY aspect of this endeavor, and I am "all in" to learn what you learn. I appreciate your willingness to share!

Andrew

 

I'm not varying the voltage. Voltage will be a constant. The SSR, or big MOSFET with a heat sink in this case, will be on the ground side of the fuel pump.

 

So, if I understand this correctly, voltage is constant, but by pumping it you are able to achieve variable output based on the duty cycle that you program?

For my Dominator and C6 fan controller, the PWM signal is set to 128Hz, and then there is a table that adjusts duty cycle based on coolant temp. But again, the output signal from the controller has been found to be WAY higher than 128Hz. This lead people to believe that the C6 fan controller is actually a frequency amplifier. As Stevie said earlier, motors tend to be happy with higher frequency modulation.

Can the Arduino operate at the 10K-20k Hz range? Can the SSR handle this?

Andrew

 

Hi Joe, OK with your good knowledge, I would start with a P = 3, the I = .03, the D = .003 AND the 50% Duty Cycle.

I would Pre-Set to 50/50. (start-up)

NEXT Vary the frequency (scope read) AND TRY a 39 Volt Zener to ground. (IN5366)

You should be able to find the Motors "sweet spot".

I use a lot of Cree SiC products. (Mouser)

Lance

 

Lance,

Bless my heart....I really wish I could understand your posts...FML...(**** my life)...

Andrew

 

10-20khz seems overkill to me. 10 Khz is 100 microsecond pulses, thats 10,000 pulses per second. I've seen people doing PWM fuel pump control anywhere from 30-400 hz with success. The Megasquirt firmware addon for PWM fuel pump control operates at 250hz as a default IIRC.


I'll try those PID settings, lance. PID tuning has always been a PITA for me.
I'll hook up the Zener as well to see if that helps at all.

The car won't be back together until probably March, and it'll be easier to test once its back together. I could always try on the engine dyno but getting more than 4 hours at a time on there isn't always easy.

 

I kinda think the same thing lol

 

Hi Joe, thanks for the KIND words. (your knowledge)

MANY HERE like the word Giga-Hertz because it sounds "kool" ?

My method is the EXACT OPPOSITE : "Run at the SLOWEST required frequency due to the fact that noise is increased by 4x as speed is increased."

The simple task of a count of the FP's Armature Pole Number x Revolution Per Minute WILL PROVIDE the max frequency required.
The "back" EMF is your concern.
The Motor will find its own home with respect to position/speed.

NOW my MORE MODERN method of PWM motor controll is to use a THREE WAY SiC "brick" that allows for the current to flow into/outof the Armature saving energy AND STOPS the heat created by the EMFs.

Post a picture of your Armature ?

Lance

 

I dont think it will be simple for most of us to dismantle the pumps...or to measure pump rpm ?

And I was under the impression higher frequencies meant less nasty back EMF's ?

Although when you PWM a pump ( or any typical DC motor I presume ) you do need to add a flyback diode to help kill the EMF's

 

Hi Steve, fuel pumps have a high Return Rate.
The need to find that reason is my engineering tech for WorldPac.
The most common problem found is dirt with low pump voltage found being the second common cause of pump failure.

I have inspected the AEROMotive (Essex) AF-1001 FP, a common "go to" pump recommended by many here at LS-1 Tech.
This is a Ten Pole Rotor pump of poor design.
I have inspected my AF-90 FP, with a Twelve Pole Rotor and double the brush life, brushes SAFE with "E" fuels.
This pump is similar to the 044 with greater flow.
The common current draw is much lower due to better rotor design, a sealed balanced rotor.
The art of LESS pump current/same fuel volume proves less heating of the fuel as it travels past the rotor.

When Rotor Pole Count in known, pump RPM (Dynamic) is an easy scope read.

Lance

 

Many of us have used the A1000 due to ease of availability, low cost, etc., but it has always had issues. The (relatively) short lifespan, sensitivity to overheating, and noise were just something to deal with. That being said, do you consider it to be a poor design by today's standards only, or by the standard at the time it was originally introduced? It seems that there are a lot more options now than back then, and technology keeps going forward.

 

It's a little funny when you say poor design by todays standards.

The 044 has been around for some 30+ years or thereabouts...and few pumps can match it for both performance...and probably none matched for reliability.