Jump to content
SAU Community

Recommended Posts

1 hour ago, Dose Pipe Sutututu said:

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.

No, I have no problem with PWMing the pump. It's just the PWM motor controllers usually use big transistors as the switching units, without the optocouplers and other shenanigans that go into making an SSR. SSRs are black boxes,  but they shouldn't just be treated as black boxes when you push them outside of their design intent.

Fuel Lab, etc, pump controllers would just be a normal MOSFET switched PWM circuit. I'm sure you could find something workable on the net with some searching. You don't need a PWM circuit that generates it's own pulses and has a pot or anything else to adjust the PW. You only need the back half of the circuit that takes the pulses into the input of the big switch and lets the angry pixies flow. The main difference is that the PWM input and the main switched circuit end up having to share some things (like, at the very least, the earth side, I'm guessing) because the circuit would not be completely decoupled the way that and SSR does for you.

i gave this a shot with an SSR, and aborted. WHen you put a scope on the output it makes no sense, and the shape is wrong. weird buzzing that shouldnt happen and just generally misbehaving. 

as mentioned the proper way to do it is with MOSFETS, which i have not done yet. 

ok, just watched it, 12v 30 amps max.    which makes sense now that  i read the title as 3000W. so ok for a Walbro bro. turns out i have already bought one so i will see how shit it is or isnt in real life. 

was planning on just using a switched output to control pump either flat out, or PWM'ing. 

Great info guys - I might buy a moderately cheap oscilloscope to actually see what's happening. Great if someone wants to donate me a lab grade one I used at uni (I did a whole semester of introduction to electrical engineering for fun as an elective, I was the only student from IT) lol

The car has gone through 3x tanks of fuel so far, and the SSRs haven't died yet - but let's see, if they do then I'll fast track the mosfet idea.

42 minutes ago, Ben C34 said:

i have already bought one so i will see how shit it is or isnt in real life.

Are you saying that you've bought one of these motor controllers? Keep in mind what I posted to Johnny. The control input to those controllers is a potentiometer, controlling the pulse widthe coming out of the onboard pulse generator. You don't need that part of it. You need the PWM output from the ECU to drive the output stage of the motor controller directly. If they don't provide you with an easy way to interface with that (and disable the onboard pulse generator) you're going to have trouble using it.

46 minutes ago, GTSBoy said:

Are you saying that you've bought one of these motor controllers? Keep in mind what I posted to Johnny. The control input to those controllers is a potentiometer, controlling the pulse widthe coming out of the onboard pulse generator. You don't need that part of it. You need the PWM output from the ECU to drive the output stage of the motor controller directly. If they don't provide you with an easy way to interface with that (and disable the onboard pulse generator) you're going to have trouble using it.

Yep. One of them.

I wasnt particularly clear before, however I think I will just use a 2 stage setup with a switched output commanding full speed , so the rough idea is figure out where I want the pot set for enough fuel for general use, then replace pot with resistor. Output from ecu to short out resistor for full speed.

 

1 hour ago, GTSBoy said:

What sort of heatsinking do you have on the SSR? It should be bolted to something reasonably big and steel/alloy just based on what the datasheet says.

At the moment none, but I do have the correct heatsinks for them sitting on a shelf, just need a way to fit it all nicely (which hasn't been done, heatsinks were bought as an after thought).

Funny enough, they don't get too hot.

WIP photo (old) but you get the idea.

IMG-20180803-WA0015.thumb.jpg.fd30fb1480ca4b75c88e8535f36da7bc.jpg

 

MOSFETs are great and perfectly acceptable in an automotive environment - if properly protected. The gates must be protected against transient spikes. A few tens of volts can be all it takes to destroy the oxide layer and short the device. Unfortunately cars are terrible with transient spikes. So you'll have to trust that whatever device youre using has been designed by an automotive minded electrical engineer.

Not to mention that high currents controlled by PWM are probably going to create transients of their own. Those kinds of MOSFET controllers are going to switch hard and create a lot of high frequency transients.

I'd be cautious about those devices which are probably just someone's side project - which isn't intended to be a slight, but creating a rugged device for automotive use is going to take a lot of testing.

20 minutes ago, zoomzoom said:

MOSFETs are great and perfectly acceptable in an automotive environment - if properly protected. The gates must be protected against transient spikes. A few tens of volts can be all it takes to destroy the oxide layer and short the device. Unfortunately cars are terrible with transient spikes. So you'll have to trust that whatever device youre using has been designed by an automotive minded electrical engineer.

Not to mention that high currents controlled by PWM are probably going to create transients of their own. Those kinds of MOSFET controllers are going to switch hard and create a lot of high frequency transients.

I'd be cautious about those devices which are probably just someone's side project - which isn't intended to be a slight, but creating a rugged device for automotive use is going to take a lot of testing.

agreed. Which is pretty much why I have not installed it yet.

I would prefer to try out a fan controller from a VF commodore (or heaps of other options). They pwm the cooling fans so must be capable of a fair bit of current.

I'll start hunting for one of them again.... 

Actually this looks like the go.

Chinese versions available, 600w mentioned somewhere. No special plugs required cos you can chop off the existing plugs . Google this part number people  8K0959501G

 

Before everyone jumps on a train that they can't get off, has anyone actually measured the heat changes in the fuel at the tank?

I'd be interested in quantifying just how much heat we are talking about?  What pump type? And under what conditions?

Maybe what is needed for a dedicated track car isn't necessary for a road going version?

I didn't see this data anywhere in the thread BTW.  

 

I log fuel data, comes from the flex sensor.

Driving around the streets for about 45 minutes brings up the temps to about 35 degrees. Considering the outside air was about 18 degrees (data inside the Temp, Pressure, Fuel) table.

I suspect this will climb quite quickly on the track.

Mind you, the pumps are running on PWM and duty set quite low off boost.

To prove a point, I'll set both pumps to full noise and see what temps we get.

image.thumb.png.975c0044d6f0b4eb1c0b8dfe0ab3dd6e.png

 

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now



  • Similar Content

  • Latest Posts

    • OK, so regardless of whether you did Step 1 - Spill Step 2 - Trans pan removal Step 3 - TCM removal we are on to the clean and refill. First, have a good look at the oil pan. While you might see dirty oil and some carbony build up (I did), what you don't want to see is any metal particles on the magnets, or sparkles in the oil (thankfully not). Give it all a good clean, particularly the magnets, and put the new gasket on if you have one (or, just cross your fingers)
    • One other thing to mention from my car before we reassemble and refill. Per that earlier diagram,   There should be 2x B length (40mm) and 6x C length (54mm). So I had incorrectly removed one extra bolt, which I assume was 40mm, but even so I have 4x B and 5x C.  Either, the factory made an assembly error (very unlikely), or someone had been in there before me. I vote for the latter because the TCM part number doesn't match my build date, I suspect the TCM was changed under warranty. This indeed led to much unbolting, rebolting, checking, measuring and swearing under the car.... In the end I left out 1x B bolt and put in a 54mm M6 bolt I already had to make sure it was all correct
    • A couple of notes about the TCM. Firstly, it is integrated into the valve body. If you need to replace the TCM for any reason you are following the procedure above The seppos say these fail all the time. I haven't seen or heard of one on here or locally, but that doesn't mean it can't happen. Finally, Ecutek are now offering tuning for the 7 speed TCM. It is basically like ECU tuning in that you have to buy a license for the computer, and then known parameters can be reset. This is all very new and at the moment they are focussing on more aggressive gear holding in sports or sports+ mode, 2 gear launches for drag racing etc. It doesn't seem to affect shift speed like you can on some transmissions. Importantly for me, by having controllable shift points you can now raise the shift point as well as the ECU rev limit, together allowing it to rev a little higher when that is useful. In manual mode, my car shifts up automatically regardless of what I do which is good (because I don't have to worry about it) but bad (because I can't choose to rev a little higher when convenient).  TCMs can only be tuned from late 2016 onwards, and mine is apparently not one of those although the car build date was August 2016 (presumably a batch of ADM cars were done together, so this will probably be the situation for most ADM cars). No idea about JDM cars, and I'm looking into importing a later model valve body I can swap in. This is the top of my TCM A couple of numbers but no part number. Amayama can't find my specific car but it does say the following for Asia-RHD (interestingly, all out of stock....): So it looks like programable TCM are probably post September 2018 for "Asia RHD". When I read my part number out from Ecutek it was 31705-75X6D which did not match Amayama for my build date (Aug-2016)
    • OK, Step 3, if you need to remove the valve body, either to replace it, the TCM, or to do a more complete drain.  First, you need to disconnect the TCM input wires, they are about half way up the transmission on the drivers side. One plug and the wires are out of the way, but there is also a spring clip that stops the socket from sliding back into the transmission. On my car the spring clip was easy to get, but the socket was really stuck in the o-ring of the transmission housing and took some.....persuasion. You can see both the plug to remove (first) and the spring clip (second) in this pic Incidentally, right next to the plug, you can see where the casting has allowance for a dispstick/filler which Nissan decided not to provide. there is a cap held on with a 6mm head bolt that you can remove to overfill it (AMS recommend a 1.5l overfill). Final step before the big mess, remove the speed sensor that is clipped to the valve body at the rear of the box.  Then removal of the Valve Body. For this the USDM Q50 workshop manual has a critical diagram: There are a billion bolts visible. Almost all of them do not need to be removed, just the 14 shown on the diagram. Even so, I both removed one extra, and didn't check which length bolt came from which location (more on that later....). Again it is worth undoing the 4 corners first, but leaving them a couple of turns in to hold the unit up....gravity is not your friend here and trans oil will be going everywhere. Once the corners are loose but still in remove all the other 10 bolts, then hold the valve body up with 1 hand while removing the final 4. Then, everything just comes free easily, or like in my case you start swearing because that plug is stuck in the casing. Done, the valve body and TCM are out
    • OK, so if you are either going for the bigger fluid change or are changing the valve body which includes the Transmission Control Module (TCM), first you should have both a new gasket 31397-1XJ0A and a torque wrench that can work down to 8Nm (very low, probably a 1/4 drive one). You can probably get by without either, but I really didn't want to pull it all apart together due to a leak. First, you now need that big oil pan. The transmission pan is 450 long x 350 wide, and it will probably leak on all sides, so get ready for a mess. There are 24x 6mm headed bolts holding the pan on. I undid the 2 rear corners, then screwed those bolts back in a couple of turns to let the pan go low at that end, then removed all the middle bolts on each side. Then, undo the front corner bolts slowly while holding the pan up, and 80% of the fluid will head out the rear. From there, remove the remaining bolts and the pan is off. You can see it is still dripping oil absolutely everywhere...it dripped all night.... I got another couple of litres when I removed the pan, and then another few when I removed the valve body - all up another 4l on top of the 3 already dropped in step 1.
×
×
  • Create New...