Paging Mueller, Bowlsby, Bleyseng and Jaroen Options

[Demick]

Aw shoot. I had forgotten about this. I finished the drawings and sent them to Jeff Bowlsby, but this was during the week(s) when 914club was down and I never posted it here.

Anyway, I modeled two different diaphragms. One with the larger pleats, and one with the smaller pleats. Both were sent to me by Jeff. I don't know in what MPS's the different versions were used.

I believe that the material used on the two diaphragms may be different. One of the clues to this was that the diaphragm with small pleats was highly discolored, and the one with large pleats was nice and shiny with no sign of corrosion or discoloration. For now, the drawings call out quarter hard beryllium copper, but that can change at any time. I think the initial purpose of this exercise was just to define the shape of the diaphragm for quoting purposes.

PDF's of the two drawings are available here:

MPS diaphragm, Large Pleats

MPS diaphragm, Small Pleats


Demick

[Bleyseng]

I think the large pleat is like the OEM design, the small pleats is the Brett Ins type.

[redshift]

Let's just break into Brett Inst. and steal their stash of diaphrams!

Where is Miles when you need a good plan!

oh dear!


I don't have time to draw one up... find the old one, and change these variables:

Brad=Diaphram
Cops=Employees
Aaron&Matt=Whomever is close

See you at Uncle Jiggly's


M

[ppickerell]

I have a little bit of experience forming beryllium Copper (33 yrs) I could make these, but the form tools would run about $2000 for each of the two designs.

[ptravnic]

I commend the work you gents are doing on this. If/when it comes down to ponying up for the first batch to cover costs I'll donate $50.

[Dr Evil]

Does anyone have the value curves for these? I am supprised no one has hidden a solid state MPS in the old unit that would mimic the correct signals for the given pressures. Has someon mentioned this and I missed it?

I could do it if I had the values that I needed. I may need to set up a circuit to compensate for certain variables, but it seems very doable and would keep your system looking/performing stock.

Anyone wanna work with me on this.....not like we are all in a hurry (IMG:style_emoticons/default/happy11.gif)

[Bleyseng]

Here's a site to look at....
http://news.thomasnet.com/fullstory/453621/2399

We need a vacuum sensor not a pressure sensor, I have no time to delve into this.

[Dr Evil]

That is what I am reffering to in part. My MJLJ has a vaccume MPS that is the size of a quater. I was just looking over the brian schematics and it looks failrly straight forward. I just need the values that I should be shootinf for, or I will take a much longer time and rig up test connectors to a running car, yadda, yadda.

I am not saying I know it all, far from it. I'm just saying that I have not seen a reason not to consider this line of thinking.

[Bleyseng]

I started to look into a just haven't as I have been outta the country workin...


Its all yours baby! (IMG:style_emoticons/default/piratenanner.gif)

[JeffBowlsby]

Dr. Evil Sir...

The values are on Brad Anders fine site. The MPS converts intake manifold vacuum, or rather the micro-dimensional displacement distance on a diaphragm subjected to the change in vacuum pressure...to an electical inductance value. Can you do that?

[Dr Evil]

Maybe, havent tried it yet, but I am willing to give it a try (IMG:style_emoticons/default/happy11.gif)

[r_towle]

i guess that we never got a clear answer about the metal properties...

Hardness and alloy...
If I get that..and take these drawings, at least I can get a price.

Rich

[r_towle]

Maybe, havent tried it yet, but I am willing to give it a try (IMG:style_emoticons/default/happy11.gif)

I looked on thomas.net and alibabi (SP) and I found several screw in ones that measure vacuum, and produce elec....
One of the companies said they would make it any way that was needed....

So, search there...

I could possibly help provide readings...
We can talk about that...but I think I could rig up an inline system to the car to take readings and plop it all on a computer so we can map it out...

Rich

[Demick]

I am not saying I know it all, far from it. I'm just saying that I have not seen a reason not to consider this line of thinking.

There are a bunch of us who have thought about this. Like you point out, a pressure sensor that outputs a porportional voltage (or resistance) is easy to get. A standard MAP sensor is cheap and readily available that would do this (Megasquirt uses a MAP sensor like this).

Unfortunately, to mimmic the MPS, you would need to take this signal and create an inductance value based on the amount of vacuum. A hefty inductance value at that (0 - 1.5H). I know of no voltage controlled variable inductor - especially at that kind of value. So this is where I get stuck. Not a simple problem.

The other approach is not to mimmic the MPS exactly, but to understand how the ECU interprets this inductance, and create a circuit that does not output an inductance, but rather another type of signal that the MPS would be able to correctly interpret. But this goes way beyond my electrical capabilities.

Demick

[Dr Evil]

Demick, you hit it right on the head. I figured a veriable inductance circuit is gonna be the bain of this whole thing. I figured if it were easy one of you all would have already done it. I just wanted some input and values and such so I can start to ponder about a solution as well. I definetly wish to share what I find with the group. Like Geoff, I am not out to make a fortune on these. I want a solution so I can keep all of my friends onthe road......looking like a stud for solving this problem wouldn't hurt (IMG:style_emoticons/default/wink.gif) But, we shall see. I have 15 years electronics experience (before med school) and am a wiz at figuring out circuits. I bet Jeff Keyzer and Tony Long could help a lot on this problem as they are EEs.

Just dont expect a fast solution....like Geoff, I too have other obligations (IMG:style_emoticons/default/biggrin.gif)

[Bleyseng]

email Brad Anders too Dr Evil as he has studied and understands Djet like no one else. He is also a Electrical Eng working for Intel. He mostly posts on PP or Rennlist as I think there is too much OT chatter for him here.

[lapuwali]

Jeff Keyser and I have spoken about this at length.

The variable inductor in the MPS basically takes a squareish wave in (from the trigger points), and puts a squareish wave out, modified by the manifold pressure and the air temp (which gets fed back in by the ECU). This square wave is then furthur modified by the ECU, and eventually becomes the pulsewidth train for the injectors.

So, no variable inductor is really required. Instead, you use the pulse capture module on one of the many microcontrollers, read a solid-state MAP sensor, read a temp sensor, then calculate the waveform out in a very similar manner to the way normal fuel ECUs do. Generating the waveform from there is easy. 90% of this could happen in software.

However, the marginal cost of building and programming and testing this thing would exceed that of a Megasquirt kit, which solves the problem, as well.

If you were willing to eat all of the R&D costs, the final unit would likely be simpler and cheaper than a Megasquirt, partly because you could use a simpler processor, and because you'd not need a lot of parts the MS uses, like injector drivers (the D-Jet ECU would do that). Indeed, I think all you'd need is a power supply (one $5 IC, a few resistors and caps), a microcontroller (under $10 in single unit quantities), a MAP sensor (about $25), and a bit of signal conditioning (more resistors and caps, about $1). The whole thing could easily be made to fit into a MPS housing with the guts removed. You'd need to do some rewiring, as I don't think the MPS gets a solid +12 or ground connection (you'd need both).

There would probably still be some tuning limitations set by the stock ECU, so while you'd certainly be able to use this electronic MPS for any engine from a stock 1.7 up to a pretty wild 2056, I'd bet you'd run into problems with, say, a 2316 or so. Hard to say for sure.

A Megasquirt would be cheaper, far easier, and would allow more tuning flexibility.

[Demick]

I think the fundamental question here is, are you trying to come up with an MPS replacement, or an MPS and ECU replacement?

An MPS replacement would retain the stock ECU, and be a simple swap out of the MPS for a solid state one. Since the MPS is the least reliable, most expensive, and hard to get item in the fuel injection system, I think this is worthwhile to try and do.

However, an MPS and ECU replacement (which James is talking about above), seems pretty pointless to me. As long as you are essentially replacing the whole FI control system, you might as well just start with Megasquirt or one of the other aftermarket FI systems and go from there. But the end result needs to be a plug-and-play system. And the hurdles for making a plug-and-play system are pretty much the same for a 914 specific custom FI replacement, Megasquirt, or any other aftermarket FI system.

Demick

[lapuwali]

I think coming up with new MPS diaphrams is a good thing, since this really is a 100% plug and play solution, and would suit those who want a 100% stock car.

For those wanting a solid-state replacement, however, you only have two options. Use a solid-state MAP sensor and some amount of analog electronics to emulate the variable inductor. This would be pretty difficult to design, and I'd expect the parts count to be high, so the device would be fairly expensive to make. The other option is to use a microcontroller to do the job digitally, at which point Demick is right in saying you might as well just use Megasquirt, et al.

For a full stock appearance, it wouldn't be at all difficult to fit a Megasquirt board in a stock ECU housing, and fit a solid state MAP sensor into an MPS. This could use a stock FI harness, and would be indistinguishable from a stock setup without disassembling the ECU and the MPS.

[Dr Evil]

Interesting concept. It could more easily do that for sure, but then it would be easy (IMG:style_emoticons/default/happy11.gif)