R33 Fuel Pump Wiring , Better Feed Retaining Fpcm ?

[scotty nm35]

Interesting, I think you are correct, and the first reg could be in the tank, similar to returnless fuel setups. The other rail return reg could be switched in for easy starting, and have no hot fuel returning.

[Dobz]

Thread revival!!!!!

I was thinking about how to retain the fpcm and get better current flow to the fuel pump. I'm no electronics guru but don't the solid state relays switch the output voltage in relation to the input voltage? So if there is a 12v input voltage the output will be 12v and if input rises to 13v then output will then allow 13v etc?

If that's the case could the solid state relay input just be the original fuel pump wiring from the fpcm and the output could use direct battery feed for the pump?

It is a loose understand of the relay's operation but i'm happy to be schooled if i'm wrong.

.

[Ben C34]

Thread revival!!!!!

I was thinking about how to retain the fpcm and get better current flow to the fuel pump. I'm no electronics guru but don't the solid state relays switch the output voltage in relation to the input voltage? So if there is a 12v input voltage the output will be 12v and if input rises to 13v then output will then allow 13v etc?

If that's the case could the solid state relay input just be the original fuel pump wiring from the fpcm and the output could use direct battery feed for the pump?

It is a loose understand of the relay's operation but i'm happy to be schooled if i'm wrong.

.

All relays do that.

[Dobz]

But I was under the impression the solid state relays varied the output voltage to match the input where as an electromagnetic relay input doesn't influence the output it just switches on and off. The input being the switching circuit and the output being the switched. Is that any clearer?

[zoomzoom]

Solid state relays would have a very low saturation voltage. I have one which is supposed to be "fully on" at about 3v, but will still pass through 96% of the source voltage at 1.6v.

Its like an extremely high gain amplifier. You would be best using some other amplifier more suited if you were intending to vary the voltage.

An SS relay works very effectively with PWM however. If you have a PWM controller that cannot supply enough current for a particular fuel pump, it can be used as a gate for a SS relay, but be sure to have a good heat sink for the relays to go on. They get quite hot once you start putting 10+ amps through them.

[89CAL]

Solid state relay I think just means it has no moving parts. It still works the same as a normal relay (in that it uses a smaller circuit to switch on a larger circuit) so it will work the same as a standard relay

You need some sort of voltage amplifier, but I feel like it would be more effort than its worth. If you are worried about the fuel getting too hot by pumping it constantly, then just install a fuel cooler

[GTSBoy]

Yuh, SSRs are not transistors, even though they are effectively the same thing when seen from a long way away. The devil is in the details. That's why they're called SSRs and not "mahoosive transistors ".

If you want to switch a lot of current in the way you describe (ie with gain according to input voltage level) then you're wanting to build something with a number of big high current MOSFETs or similar. Which is what zoomzoom just said, by the way.

[Dobz]

Ok that clears things up, thanks.

[joshuaho96]

Massive thread necro here but has anyone considered doing a complete revamp of the FPCM to use a buck converter instead of just increasing the ground resistance? I figure reducing load on the alternator for other items + no resistor heat dissipation to worry about would be a decent benefit for the 6V/10V modes. 

[GTSBoy]

24 minutes ago, joshuaho96 said:

Massive thread necro here but has anyone considered doing a complete revamp of the FPCM to use a buck converter instead of just increasing the ground resistance? I figure reducing load on the alternator for other items + no resistor heat dissipation to worry about would be a decent benefit for the 6V/10V modes. 

Would this not describe just about all the aftermarket variable speed fuel pump controllers out there?

[joshuaho96]

1 minute ago, GTSBoy said:

Would this not describe just about all the aftermarket variable speed fuel pump controllers out there?

Pretty sure, yes. PWMing a brushed motor is effectively equivalent to a buck converter.

I think the goal would be to fit the controller into the stock housing + stock harness though. And uprate the power transistors to handle 70-80A.

[GTSBoy]

I'm not sure there's an advantage to retaining 20-30 yo wiring and plugs and trying to desolder them from the old module and build them onto your new board and be constrained by them needing to be in the same place to poke out through the same holes in the casing.....when some new decent connectors and a proper extruded alloy housing to screw the MOSFETs onto and freedom to lay the board out the way it needs to sound like things worth having.

[Dose Pipe Sutututu]

Just PWM it using 100amp SSRs, because they're high current ones you won't need a heatsink, unless you're running multiple pumps off each SSR.

(Assuming you have a capable ECU). The FPCM is garbage, as it's just a BS drop resistor, which is located next to the aerial on R33. 

[GTSBoy]

I'm a little bothered by the suggestion of PWM through SSRs.

Conventionally when doing PWM you use a fixed frequency, something in the order of 500Hz would be common** for simple DC power control. You then vary the width of the pulses to control the delivered "power". But that means that the SSR is literally switching on and off at 500Hz (unless you get it to either 0% or 100% ends of the PWM). And AFAIK, SSRs are not meant to be switched that fast.

Anything high power PWM just sounds to me like the output side should be actual MOSFETs (or other transistory devices that like being switched fast) not SSRs that, whilst they may have MOSFETs inside, don't necessarily like being driven that fast.

You look at any circuit designs for high power DC PWM and you don't see any SSRs. You see FETs. Lots of FETs.

 

**but obviously other PWM applications might like to run at many kHz, and AC PWM control should either be implemented at the same frequency as the supply, or some lower frequency with careful controls in place to prevent DC draw on the mains supply from halfwave switching etc.

 

[Dose Pipe Sutututu]

I've got it set at 100hz though, seems to work well

Had it set at 200hz and everything made an odd hum. The Jaycar SSRs are rated at 0.5ms switch time, so apply some maths and that's smack bang at the 500hz you've mentioned. 

Honestly never looked further than the SSRs, because if modern cars use them to control massive fans and electronic water pumps, then controlling a fuel pump surely should work?

*disclaimer, I am no means an engineer, just have an interest about science and excelled at it in high school. I actually work in IT lol.

[GTSBoy]

2 hours ago, Dose Pipe Sutututu said:

because if modern cars use them to control massive fans and electronic water pumps, then controlling a fuel pump surely should work?

Yeah, but I think there's a difference between using them as relays and just simply switching the big loads on and off at normal rates, vs. banging them on and off hundreds of times per second.

I think the Jaycar SSR switching rate that you mention is not "you can drive it this fast". I think it is "It will change output state this fast after you change the input state".

[Dose Pipe Sutututu]

Are you insinuating they might arc to death? I did have that suspicion too hence I'm running 2x SSRs, one for each 460L Wallahbro pump. Each of the SSRs are rated at 100amp (max load that is) however I'm assuming each pump would only drain a theoretic max of approx. 20amp at the fuel pressure I intent to run them at.

 

[zoomzoom]

A SSR cannot "arc" as part of normal operation.

Without an oscilloscope, you can't characterise any particular SSR, except trusting the datasheet. Or trust the ebay listing details - lol.

From memory, I was able to use those cheap fotek ones up to switching a 1kHz square wave. But I've never used them in anything critical like a fuel pump.

The problem with switching frequency is that you increase the time spent in a transition region between on and off states, where the power dissipated in the switching element is the product of the current and voltage across the device. Eventually it reaches a frequency where the device is floating in a partially on state - which is bad news.

I doubt any ECUs take the PWM frequency that high though, they only seem to go to a few hundred Hz tops, and it should be adjustable.

Also be sure that you SSR is legit. From memory a youtube creator called Big Clive did a tear down of one, only to find a measly 10-15amp rated  TRIAC as the switching element.

Anyway, an oscilloscope is the tool to measure this stuff.

[GTSBoy]

7 minutes ago, zoomzoom said:

A SSR cannot "arc" as part of normal operation.

Yeah, I was gunna say this. I'm not sure what the path to failure would look like for a mistreated SSR. As zoomzoom says above, when they float between on and off they consume a lot of power themselves and will get hot, which could fail gracefully..... or in a fire. I also don't know how much abuse (intensity &/or duration) they can cop. Arguably, switching them at hundreds of Hz for many hours is an enormous number of state changes. 2000 hours at 100Hz is 720 million state changes. It's probably OK.....but it sounds like it adds up quick. It would take a few years of commuting to get 2000 hours. 2000 hours of track use would either come much sooner or never, depending on the type of track usage. A truck would do 2000 hours in ~6 months.

I'm not shitcanning the idea, just mentioning that I think it is not the way that they are intended to be used.

[Dose Pipe Sutututu]

Might have to go back to the drawing board now with what you guys have mentioned

Thought I was being smart by not staging pumps (Not my cup of tea because you get pressure spikes as pumps come online). Not visible on a gauge but definitely visible when you run fuel pressure sensors.

From what I gather I should just bin the whole fuel pumps running on PWM and just have them full noise with a massive fuel cooler to manage heat.