i am SICK of spooling to 8-10 psi at part throttle, running .1ish for oxygen sensor voltages, and not having any power, then when my ECU decides it's time to enrichen the mixture i get a surge of power at a quite random time.

i know the J&S knock sensor is "supposely" going to be released soon forcing the ECU into open loop mode when 0 psi is achieved, but i'm not going to wait for it, just like i'm no longer waiting for the ERT downpipe. hell will freeze over before the ERT downpipe is released.

i stole the TPS idea from the J&S thread, and i'm currently working on my own circut to do the same thing. i've started from scratch. i bet someone else out there has done this, but hell, i've searched and searched and i've never found anything.

so far, i've come up with this ghetto schematic:



so what's happening here? i decided to use the motorola mpx4250ap pressure sensor that corbin uses for his "fooling the AFC" mod. +5 volts is supplied to it via a 12 -> 5 volt regulator (powered off of key on power). the output of the pressure sensor happens to be almost exactly 1.8 volts at 14.7 above absolute or 0 psi of boost. it would be nice to be able to hook that directly up to a relay, but there aren't any relays out there that switch exactly at 1.8 volts. so, i'm going to feed the Vout signal from the pressure sensor to some type of amplifier circut (which i have been designing today at work) which will produce no output below a 1.8 volt input, and once that point is reached (and everything above it) it will output 5 volts to trigger the 5 volt relay that i have obtained. this relay will force the ECU to connect to a 1.4k ohm resistor (happens to be the resistance of the TPS at full throttle) instead of the TPS.

a few snags i've noticed so far:

1. the pressure sensor is very sensetive to voltage input. it needs to see EXACTLY 5 volts for the output to be accurate at all.

2. the pressure sensor only uses up 5ma, therefore the output signal must be less than or equal to 5ma. it takes a lot more than 5ma to trip on the relay (72ma specificially) that i have. the relay has an impedance of about 68 ohms... so i'm going to need to design an amplifier circut that uses a FET (Field Effect Transistor) which have very large input impediances (in order to not draw any current from the Vout, drawing current from the Vout will alter the output voltage and that's a no-no)

if i get this thing to work, i might consider selling kits if anyone is interested.

discuss.

What would be the benefits of this?


Nick
97RS

more power at part throttle when spooling around town, and no more surging in boost once the ECU decides to add fuel. hey, maybe it's just my car that stays excessively lean for x seconds in boost unless i give it WOT, i dunno. the moment i give it more than 80% throttle i get .92-.93 for oxygen sensor voltages, forcing the car into open loop. i know my fuel pressure is good enough, if i bring it to 8 psi at part throttle it does the fubar thing, but if i mash it at 8 psi everything's fine. i haven't been able to think of any other solution for this problem.

I have thought of something like this as well. I was not planning on using a separate pressure sensor, but instead a feed off of the stock map sensor. The output ranges of the map sensor can be converted to any range you want via an amp circuit. So when the map sees atmospheric, it puts out 4.6 volts I think. That can be amplified to 5 volts and feed into the TPS signal to the ecu. The variable resistor on the throttle body simply changes a voltage that the ecu reads. You just need to tap into that line and feed the ecu the correct voltage and keep that voltage from going to the sensor. Use a diode so that the voltage can only flow one way, but only at 0 psi boost and above.

I have only thought about it.... I have not actually looked into the physical aspect of it. I know what needs to be done though. If I said anything wrong up there dont kill me.

If nobody else needs something like this, I will be developing something anyways since my car is undergoing boost transformation.

-Jared
96 GS
Melbourne, FL

bumo

*maybe belongs in advance tuning section?

Nick
97RS

yeah... i thought about posting it in there, but i was afraid that it would just get moved into the turbo section...

update:

my dad hooked me up with this guy who's a circut designer, he helped me make my circut and had a better way to do it.





pretty confusing, i know. basically, he suggested that i create a 1.8 volt refrence voltage (adjustable in the future with a variable resistor to determine the "go into open loop" point) using an IC, part number "ADR291", i think it's either an Analog Circuts or Teradyne part number. that, along with the pressure sensor voltage is sent into the compariator IC, part number "LT1011". this outputs a voltage (high or low, still need to find what the output will be) which is fed into a driver transistor "2N3904" that supplies enough current to click on the relay to make things happy. everything in the circut works off of 5 volts, which is supplied by a generic 12 to 5 volt regulator.

Funny...I was researching the same thing a few weeks ago and so was another member (he actually built one).

Read this thread:

http://forums.2gnt.com/dcboard.php?az=show_topic&forum=8&topic_id=49809&mesg_id=49809&listing_type=search

Corbin

'95 ESI-T
HRC+FMIC+ETC...



Gimme fuel...Gimme fire...Gimme that which I desire

yeah, i remember that thread from a few weeks ago... i just forgot to watch it and didn't see that someone figured out the issue. however, i've already completed 2/3 of the work with designing my own apparatus, and i think it's much cooler to make the thing from the ground up.

today was pretty slow at the office, so i worked a little bit more on it. i have the voltage comparator working like it should, although soldering to it was definately a PITA! the whole chip is smaller than my pinky fingernail. (oh yeah, i'm an imagestation n00b. sorry about the broken links above, i think i have it figured out by now)


yeah, it's pretty ghetto and thrown together, but i proved to myself that it works.


and here's the schematic so far:


i still need to design some type of amplifier circut to take the final output voltage (it's either 0 or 1.6ish volts depending on what the pressure sensor is doing) and convert it into useful power to drive the relay. after that, i just need to solder everything together for real and put it in a box.

anyways, just thought i'd update things.

So who wants to build them? Ill buy...
Nick
97RS

just give us the final schematic! :7

yeah, i'll definately post the final schematic, which might even happen as soon as today. i still need to test it on my car though...

i got bored in the lab today, so i spent about another hour and a half finalizing the schematic. as promised, here it is:

i was having issues with the output voltage from the LT1011, it wasn't where i wanted it to be. basically, i wanted it to be either 0 or 5 volts, which would feed the 2N3904 transistor that drives the relay. i wound up putting a 10k pull up resistor beween the 5 volt feed and the output of the LT1011 and everything became happy. oh yeah, there's a diode that goes across the relay. my relay has a diode built in, (you can tell if you reverse the polarity with the trigger winding and see if it still works or not, if it only works 1 way it has a diode) but if anyone wants to put this circut to use and they choose a relay that doesn't have a built-in diode, i suggest that you stick one in there so the inductive kick from the relay doesn't damage the transistor.


what a mess!


the relay, had the ohm meter hooked up to it to make sure it was closing the circut (although i could hear the thing clicking)


my white ass blowing into a tube and watching the ohm meter switch from open circut to short circut


well, that's about it. i proved to myself that it could be done. it wasn't even that hard. all i need to do now is solder everything together for real, (it's sooooo fustercluck right now with all of the clip leads, but it works) put it in a nice box, and install the damn thing.

Not that I need it yet, but you gonna start selling these things?

i've been thinking about it. they're a bitch to make, the voltage comparator is so impossible to solder anything to. i'd have to find one that's more along the lines of a 555 dip timer (with the 8 big pins sticking out instead of the surface mount machine flow soldered chip that i'm currently using) in the lab, it works perfectly and one is able to adjust the "force open loop" point to any boost (or vacuum level) that one desires by use of the 50k ohm variable resistor. however, i still need to box it up and put it in the old 2gnt and see what happens.

if everything goes as planned (i.e., it works) i guess i could be able to make them in my spare time... based on the cost of parts and the PITA of labor, i'd probably sell them for maybe $40 shipped? just a ballpark figures right now.

i'm totally lost with electronics but do you think it'll help my problem?? Read my recent post in this section. if so, i'll take 4. :)


Jon

1997 Eclipse RS

13.887@104.648 5/30/03

Good job man :thumsup


95 Eclipse RS : 5 speed
15.943 @ 86.48mph
Jeep TB writeup - http://www.dimensia.com:81/jimbo/JeepTBfor2gnt.html

Wow,very nice.

I just finished my second term at Devry so I know how much work goes into designing it.

Congrats dude

If I knew were to get everything I would prolly try making them too.

bump...

i've been simplifing the schematic - i'm trying to get a generic $.99 Shit Shack voltage comparator working with the stock MAP sensor. this eliminates the need to order $4 parts from weird online retaliers and also eliminates the need for the motorola pressure sensor. however, it restricts the adjustability and one would no longer have the option to adjust the open loop point above 0 psi of boost.

i think it would work better if it kicked in around 4-6 psi. Can you post part numbers and a schematic so i can build one? }(

So have you tested it yet???

Great work! Thanks a lot for finding a use for that pessure sensor that I didn't need anymore sinse I decided not to go with a S-AFC. Let us know how it works and I'll bulid one too! Mad ends to you!:thumbsup

thanks guys. i refined the design more today. i decided to use a "quad comparator" (yes i know, it's overkill, but i couldn't find a "single comparator") that i found at radio shack. MUCH easier to solder to than the surface mount type that i used in the prototype that i made last week. i tweaked the values of some of the resistors resulting in a more crisp operation of the whole circut. oh yeah, i also soldered it together for real and put it on a circut board. it looks 100 times better. the prototype was a mess!

the shit attached by the clip leads simulates the stock MAP sensor or the motorola one, it doesn't matter which one you use - you just have to re-calibrate the big variable resistor that's attached directly to the circut board


and my crack soldering skills :P


so that's about it. i'm going to put it in a box and install it in my car. today. finally. i will keep you guys posted.

rock out! :thumbsup :thumbsup :thumbsup

I've been watching this one, keep us posted. :thumbsup

This is great work, but i still dont see the need to go into open loop at 0 psi. If this happens, then low boost power will suffer, since now the ECU will not adjsut A/F ratio anymore. Its been proven by many of us by datalogging, that our igniton adavance goes down right away into boost, so no worries there. IMO, closed loop boosting should be perfectly safe to around 8 psi. Now i know i have a hard time keeping the ECU into closed loop at that pressure. This will be great to try out, and see how it really affects low boost power... I really dont see the need to worry about how it is now though..unless you can keep your ECU into closed loop at higher boost..

"really dont see the need to worry about how it is now though..unless you can keep your ECU into closed loop at higher boost.."

i had a hard time getting 9 and trying to get 12psi of boost in closed loop, if i did it was for a few seconds which sucks ass. so if we can keep it in close loop so we can read fuel trims at 15-22psi hehe that would be great for me and most of us i would say.
can this be done?

Michael
99gs (broken) Motor being built now
HRC Stage 2 8-Injector Setup
Fully Built by Aug 1,2003 hehe
2GNT Member #32

ok. after getting home from work, i installed this thing.

the good:
it works. part throttle is INSANE. i noticed a power loss at very low boost (like 3 psi and below) but whatever. i'm a tard and i broke my motorola pressure sensor so i'm just using the stock MAP.

the bad:
the thing trips the TPS voltage to 5.0 volts, not the 3.8 that i was trying to achieve. something's amiss. i think the answer lies in the resistors that i'm trying to fool the ECU with...working on a solution.

the ugly:
i couldn't find a good project box so i used an old Bangles casette case to put the project into. it looks WICKED ghetto under the hood, but whatever. it works - to an extent.

after i get the bugs ironed out, i'll take more pictures and put up a newer schematic including inscructions on how to use the stock MAP sensor or the MPX4250AP. hell, i might even sell them.

hey i'm glad you're headed on the right direction, keep up the good work i'm hoping this setup you have is functional soon. although i'm still laughing on you using the bangels cassete. heheheheh!!!:7

keep us posted. even though im not boosting yet, i will be soon and i am sure i will be pissed and looking for solutions for the things you guys are right now. thanks for taking initiative.(sp)

points for you. :thumbsup

bump...

it certinly took me long enough, but i believe that this thing finally works.

"Today, 03 February, 2004, 2039:46"


"Today, 03 February, 2004, 2239:32"


and what i ended up doing...


basically, the old design worked, but it switched the TPS voltage to 5.0 volts, which made the PCM freak out... my A/F ratio would be rich for a while (until like 4000 rpm's) then was all over the road.

now it should work perfectly, because i'm switching everything correctly, everything is grounded where it needs to be, and i have a REAL voltage divider this time instead of whatever the hell i had before. man, i must have been smoking something good when i built the old "voltage divider".

the mspaint crap was my scrapbook while i was messing around with new resistors. there is no rhyme or reason to anything on there, i was too lazy to get up and get a pencil and some paper so i just used mspaint. i tried to label my random thought process but i got bored so whatever i guess.

bottom line? if this thing works (i'll know in about a week when i get my cylinder head back on) i'll build and sell them to whoever is interested.

Cool man... if it works like it should, I'm down for one.

put me down for one too Nate.

Now if you'd just build a nice externally powered voltage clamp for the MAP sensor into it, we'd have a nifty all-in-one solution to two problems. :-)

How does this affect ignition timeing? Does it just jump to ~16 digrees of advance once your circut throws it into closed loop?

yeah, it should go to about 16-20 degrees of advance once the circuit is tripped. i think that the MAP sensor has more to do with timing advance than the TPS does, because if i'm at like 30% throttle and manifold pressure goes to atmospheric, timing gets reduced to around 18 degrees, but i guess the ECU will retard timing to the "open loop" value if the TPS sensor goes to 3.8 volts or 100%.

the circuit needs 2 voltage refrences to know when to click the relay... there's the internal one (Vref) that's calibrated by a variable resistor, and the external one (Vin) that accepts a 0 to 5 volt input. this input can be hooked up to any type of MAP sensor that operates anywhere between 0 and 5 volts (or any other voltage output for that matter, hell, you could operate it manually with a button or something provided the voltage is higher than the cal resistor). for example, if i was using the stock MAP sensor i'd turn the key to ON and not start the car and turn the calibration resistor until i hear the relay click.

the circuit logic goes kind of like this:

IF (Vin >= Vref)
{
RELAY = 1;
}
ELSE
{
RELAY = 0;
}

if one was to use any other pressure sensor the calibration would go along the same route... the motorola pressure sensor that Corbin uses in a bunch of his projects can also be used, as long as everything is grounded at the same point. i think the motorola pressure sensor shows about 1.6 volts at atmospheric pressure.

Dino, that's a pretty good idea... building an electronic FCD into this circuit would certinly be doable, it would be nice to integrate everything into one box. however, i'd have to crack open that hunk of junk black box that i got with... um... and do some "research" if you catch my drift... something tells me that it's more than just a zenier diode...

anyways, i tested my creation last night in the lab here at work and it definately works now. in theory, it should work perfectly.

The MAP sensor clamp is pretty simple. I've sourced all the components very cheaply and even talked briefly to wyatt a few years back about building them. I believe I remember getting everything for <$1, but it's been a few years since I last looked at it.

probably also been a few years since you talk to Wyatt too. Actually it's been a few years since he took some people's mone...I mean since he's been on.

This is geat work, but I really don't see the need for this, as to how it works and what it does. Once I reach boost, my timing is automatically pulled to around 16-17 degrees. This is because of the map sensor. If anything, staying in closed loop at lower boost will be BETTER than open loop, because the ECU is still controlling the A/F ratio, which makes the car run a bit leaner, but still makes good power. I can tell the difference when part throttle boosting in closed loop, and open loop boost at the same pressure. Car has a bit more power in closed loop..I personally wouldn't want to run open loop once I see boost. My car has a lot more drivability and power in low boost this way..Don't know, maybe my car is just different :) Running really large injectors, you would want to keep in closed loop at part throttle, lower boost, to help your A/F ratio out a bit..Just my opinion- it would be better to use this device to force the ECU into open loop not at 0 psi, but like 6-8 psi, as I would worry a bit about going to high into boost still in closed loop, and running to lean. I think I have hit 10 psi or so with the ECU still in closed loop, and have gotten light amounts of detonation from running to lean. This is a very rare occuracne though, and only happened in 3rd gear at lower rpm. Better yet, make it adjustable when it forces open loop with a potentiometer, and just use a motorola map sensor for the input. That would be sweet.

You do realize that in closed loop, no matter what, the ecu is trying to seek on a 14.7:1 AF ratio? You know that 14.7:1 under boost melts things? That's why this is a good idea. it would allow you to no longer have to fight with, or even account for, what the stock ECU is doing with the fuel while in partial throttle boost. You'd have the ability to more accurately dial-in your boosted AF ratio because nothing will be monkeying with it.

Originally posted by DarkOne
You do realize that in closed loop, no matter what, the ecu is trying to seek on a 14.7:1 AF ratio? You know that 14.7:1 under boost melts things? That's why this is a good idea. it would allow you to no longer have to fight with, or even account for, what the stock ECU is doing with the fuel while in partial throttle boost. You'd have the ability to more accurately dial-in your boosted AF ratio because nothing will be monkeying with it.

Right, exactly, which is why I said not to go too high in boost in closed loop. How about this for an example. at 1 hg of vac, your are in closed loop. You give it some throttle, and you build a few pounds of boost. If we all of a sudden go into open loop, we are going to give too much fuel. We are limited to working with the fuel pressure, which just isn't that fine tunable. I'd rather let the ECU figure out the A/F ratio for that few pounds of boost,to make the transistion from vac to low boost more smoothe, and even more powerful. I just think low boost would be very dissapointing if the ECU goes into open loop right away. A few pounds of boost running at a 14.7 A/F ratio, will in no way melt your motor..this is why I suggested forcing open loop a little later,like at 5 psi or so. I have never had a problem under 10 psi, in open loop, with high egts, or detonation. Low boost is very peppy, and very drivable. I just think you will sacrafice this if you go open loop right away. I don't know about everyone else, but i do most of my daily driving at only 5 psi or less..i'm not on the gas like i'm at the track at every stop light :) This is just my theroy, but if you have a way to fine tune your boosted fuel maps, then that would be the best way regardless, and this would not be an issue. Actually, if you had the 8 injector setup, or other fuel map controlling computer, you would prob WANT something like this, as that way you would have complete control of your boosted fuel map. I just think its better for drivability if you run just a rising rate FPR, by letting the ECU figure things out for a few pounds of boost..

Maybe yours is different in some way then. Mine acts like ner947's in that the low boost is a drag until the ECU switches and then I get a surge. I can only assume that moving the open loop point closer to the vac/boost crossover will bring usable power down into the lower boost as well.

Hopefully he'll be able to tell us results soon.

Definitely include me in once you get this thing going. The J & S Safeguard also kicks you into open loop once you go into boost. In my mind, it's most certainly a desireable thing. Although you may benefit from a tad more power with the ecu seeking out that stoich a/f ratio, I would think so long as you tune yourself to a good a/f ratio at a particular rpm/kPa, you should be where you want to be. I mean, if 14.7:1 is what you want, you could tune for it more consistently IMO.

But for me, it's all about safety. I would prefer the richer partial boosted condition.

hopefully i'll get my car back in one piece this weekend... if all goes well i'll "alpha" test it and then i'll front the money and make maybe a dozen of these things... the relay is the most expensive part:

http://us.digikey.com/scripts/dksearch/DKSUS.dll?Detail?Ref=199891&Row=401800&Site=US

running about $10 a piece for the damn thing... i "could" downgrade to a different relay, but i need something that's small, solders into a project board, 3+ pull double throw, operates at 5 volts, and doesn't draw much current because it has to flow through a small 2N3904 transistor.

edit: fixed the link