As per my conversation with Beav...very informative!!!

I lightly glimpsed at the last thread of the turbo thread and saw you discussing the esm and map sensor voltages. It sounds like its a basic voltage clamp that limits the map sensors output to hide the boost from the ecu. The missing link and check valves attempt to accomplish the same things, they both mechanically hide boost or bleed boost off from the map sensor. Then the fmu is responsible for adding fuel. What have you done for timing control? Some timing has to be pulled in boost if you dont want detonation to hurt your motor, but with the coil on plug ignition MSD BTMs and the normal solutions are out the window. If you are still throwing codes then your voltage clamp must not be doing the job, you could try reading the voltage at the map sensors output with the ignition on but the car not running. That is the voltage its going to see at atmospheric pressure right before boost comes on so anything much higher than that is too much for the stock ecu.

The FMU setup is not going to be the best solution, transition into boost is going to tend be crappy and the tunability sucks, reliabilty and consistancy sucks etc. The proper way is going to use bigger properly sized injectors at normal pressure with something more tuneable. The next step up us civic guys used years ago was the SAFC hacks. That involves swapping the stock civic 240cc injectors for 450s out of DSMs. Then you use an SAFC to take out 35%-40% fuel. It all worked out with no clamps or checkvalves because since the way the safc is taking out fuel is it is cutting down the voltage from the map sensor to the ecu to accomplish this so the ecu would not even think it was in boost till somewhere around 10psi where the stock map maxes out anyways. This resulted in more accurate fuel control and good boost transistion but the bigger problem was timing control, with the voltage cut down to compensate the ecu would think the car is in vaccuum while it may be in low boost. The problem with this is the ecu runs much more timing advance the further into vaccuum you go in the timing maps and it could cause severe detonation at low boost. People then found a way to correct this with early civic chips that corrected the maps in this area to work with the SAFC so you used a special chip designed for the SAFC hack. The only remaining issue is some light slight bucking at very light throttle because the map value gets to low and the ecu cuts fuel untill you give it a little more throttle. This safc/chip solution should be possible now with what I have found if you get injectors that can flow enough fuel without an FMU. You would set the SAFC at a negative correction low enough the ECU no longer sees boost, then have a chip custom tuned with my editor. The only problem then is the pressure values at the top of the columns would not line up and could make it tricky to tune unless you figure out what columns match up with that pressure. The best solution would be to maybe build our own simple circuit that cuts the map sensor voltage %50 or some fixed value but keeps the minimum voltage high enough to avoid bucking from the fuel cut. Then I could modify my editors columns values to match the fixed correction and you can tune away just like if the ecu supports boost. This is kinda how some of the obd0 civic editors support boost except for they cut the map value in the ecus code. We would be manually manipulating the signal outside of the ecu.

Should I post this in the turbo thread as well?

 

Also following up on what I have told you so far, If you or anybody plans on sticking with a FMU setup I can at least make a chip with some timing pulled at the last NA columns so you are safer in boost. You will loose power at 0psi and low boost because of having less advance then you could normally run in these areas but it will run good and safe once the turbo fully spools. You would not loose any power in the rest of the partial throttle vaccuum areas of the timing map because you could leave all of that alone. If its a smaller turbo that would work good and if its bigger you could adjust the timing to match the rate the turbo spools at too which also work okay, a lot safter than nothing!

 

Also following up on what I have told you so far, If you or anybody plans on sticking with a FMU setup I can at least make a chip with some timing pulled at the last NA columns so you are safer in boost. You will loose power at 0psi and low boost because of having less advance then you could normally run in these areas but it will run good and safe once the turbo fully spools. You would not loose any power in the rest of the partial throttle vaccuum areas of the timing map because you could leave all of that alone. If its a smaller turbo that would work good and if its bigger you could adjust the timing to match the rate the turbo spools at too which also work okay, a lot safter than nothing!

Yes, that would be a go! Could you possible adjust timing to give good off-idle response then be further reduced as more boost comes to tap? I know less timing is better for that off-idle power/tq but I dont know where timing goes between idle to inital boost.
I had discussed this idea earlier, and I want to pick your brain about it. Instead of turning the timing down could you take advantage of it from an akly/methanol injection system because the fuel would be less inclined to pre-ignite due to the higher flash point. It was a thought just because I had no idea about my future timing issues or so I thought they would becoem an issue.
Let me know, and ongoing thanks!
Kyle

 

Further discussion between the Beav and I

Quote:
Originally Posted by sr79labrat
That is exactly what the ESM does. I also have a checkvalve so I dont know if I am being counterproductive. Nothing has/is being done for timing. The reason being is for the low boost and lowered a/f ratio. I have no idea if this is hurting anything because I am a beginner in this area. All that I know is on the dyno the number for a/f ratio looked safe. That is probably poor judgement but that is all I had/have. I know that I can adjust the voltage on the ESM, but does it need to go up or down? I can guess that from what you have said it needs to be lowered as to not exceed the limits of the system-correct? The base you probably saw was at 2.94V which is the same as the NSX reading. You are saying that I should first take a reading from the ignition being on while the engine is not running, and then try to match that V reading to the ESM while the engine is running. Does it need to be warmed up first? I am pretty sure the CEL only comes on while the car is in open loop. Would our test be strayed by that factor?

Secondly, getting 450cc injectors would be the better way I am sure...initally I was going for cheap-though blowing an engine is an expensive lesson. So 450cc injectors alone may correct my CEL problem entirely(is this what you were saying)? they actually reduce the voltage going to the sensors? I would love to have a chip if that meant better/easier tuning. I am far from being experienced at any of this stuff, and all advice is appreciated. Let me know if I seem unclear on understanding this...I feel as if you explained it good, but I am going to try a few things when I get home if I actually do understand all of this.

Quick question for the check valves. When they are installed do they attach to the vacumm tubes inbetween the engine/sensor or the sensor and plug? I may have mine incorrectly installed based off of what you just told me.
Thanks again for the help! I really appreciate everything thus far!!!
Kyle

A set of 4 used DSM 450s goes for 50 bucks or less here, then it would not be a bad idea to get 6 of them cleaned and flow tested just to make sure they all flow the same which is usually about 10 per injector. Then it would need a good bit of retuning but it would have to potential with a good tune to run like a stock boosted car if we tried some of the ideas I mentioned earlier.

The check valves need to be installed in a Tee on the map sensor line with one end venting to atmosphere in that direction that if you blow out the air will go out but if you suck in the valve shuts. This will cause the boost to leak out before it gets to the map sensor but in vaccuum the valves will shut and it will run normal. The problem is sometimes you need multiple tees and check valves to bleed enough. You should not need any though if you have the voltage clamp thing. If you measure the voltage with the engine not running you know what the voltage is at 0psi. This is about the same as it would be at full throttle on a NA car or just before the turbo builds boost on a boosted car. The engine has to be off because at idle its pulling a ton of vaccuum. If you can clamp the map at that voltage it should not throw a code.

 

Yeah Kyle, the map sensor will read a voltage similar to that of a WOT throttle reading in an NA and just before boost in a boosted application. The major difference is that most boosted cars use a mass air flow and not a map sensor. This allows for the ECU to view an almost unlimited amount of change in the airflow through the intake. I know they make aftermarket mass air flow sensors. I wonder if it would ever be possible, after tuning these chips are fulfilled to the fullest to add on a hot wire MAF. Man that would be great.

However, no matter what you will be able to get rid of the CEL and compensate for boost using tuning of the fuel maps with the new discovery that Beav has found. He did a great job with that. I believe if you study up on some tuning/have someone that you trust to tune for you with a dyno and a wideband, the only thing holding you back are the internals and the transmission. 2 bars here we come haha. Good luck Kyle

 

Yeah Kyle, the map sensor will read a voltage similar to that of a WOT throttle reading in an NA and just before boost in a boosted application. The major difference is that most boosted cars use a mass air flow and not a map sensor. This allows for the ECU to view an almost unlimited amount of change in the airflow through the intake. I know they make aftermarket mass air flow sensors. I wonder if it would ever be possible, after tuning these chips are fulfilled to the fullest to add on a hot wire MAF. Man that would be great.

However, no matter what you will be able to get rid of the CEL and compensate for boost using tuning of the fuel maps with the new discovery that Beav has found. He did a great job with that. I believe if you study up on some tuning/have someone that you trust to tune for you with a dyno and a wideband, the only thing holding you back are the internals and the transmission. 2 bars here we come haha. Good luck Kyle

Maybe I did this wrong, but last night I tried lowering the V that the ESm read. I first did the test where I had the ignition on but no running of the engine. That read 2.94V which was the same as the reading when the car was running. Maybe I need to take the reading off the sensor before turning over the engine and taking the reading on the ESM... matbe I am missunderstanding this if this is a sure fire way to fix the problem. I did take the V down, but maybe it needs to go down further. WHat do you guys think?
Thanks so much for the help!!
Kyle

 

2.94 sounds right with the engine off, honda map sensors officially should put out 2.9x at atmospheric pressure. However if you test it with the engine idling the engine should be pulling upwards 20 in/hg of vaccuum and the voltage should be under one volt. If its not then something is not right, make sure the line you have it hooked up has vaccuum and try measuring voltage before and after your ESM. I am not sure exactly when the ecu will throw a code but it wont be much after 3,0 volts. You may have to try hooking up a voltmeter and driving the car to see what it is doing.

 

2.94 sounds right with the engine off, honda map sensors officially should put out 2.9x at atmospheric pressure. However if you test it with the engine idling the engine should be pulling upwards 20 in/hg of vaccuum and the voltage should be under one volt. If its not then something is not right, make sure the line you have it hooked up has vaccuum and try measuring voltage before and after your ESM. I am not sure exactly when the ecu will throw a code but it wont be much after 3,0 volts. You may have to try hooking up a voltmeter and driving the car to see what it is doing.

Are you saying to turn the ESM V up or down? I was thinking WOT=2.94V. Both measurment were off of the ESM...as I understand the ESM-it maintain that current V reading all of the time, and does not adjust proportionately to boost. Lowering the V rating on the ESM seemed to make the car run a bit more through boost but still not all the way.
Thanks for the input!
Kyle

 

Okay, so you said that when the car is cold and in open loop it does not throw a CEL underboost correct? The reason why is because while the car is cold it is running on fuel maps which are based on RPM as well as ECT and some other sensors because the car needs to warm up. In other words it produces a rich condition for the purpose of warming the car up.

After it warms up, it is then in closed loop, or in other words monitoring all sensors to achieve the leanest condition possible with the most spark advance possible and the ECU monitors it's progress through the use of Knock sensors and O2 sensors. Well as you know, you can't have this because #1 you need more fuel under boost, and #2 boost will be out of your map sensor's parameters and will throw a CEL.

When your ESM sends max in range voltage to the ECU, it will think that it is at WOT. While at WOT your computer should do something similar to when it is warming up. It should stop monitoring O2 sensors and simply dump fuel based on pre programmed look up tables that go by RPM etc.

A few questions real quick: Are you simply trying to get rid of the CEL or are you not satisfied with the amount of fuel that the ECU is dumping at idle and cruise? Because, the reason why it might be throwing a CEL is that at idle, say the ECU reads 2.94V but yet it reads a very low voltage at the Throttle position sensor or even low RPM and it compares them hence throwing a code because of this reason. So the main question is do you need more fuel or to get rid of the CEL? When I know that I think I might have a solution.

I will check my manual tonight to see what exactly codes 3 and 5 say. I have a guy I can ask about this later today as well and get his opinion. He knows alot about tuning. Also, the only thing those check valves are doing at this point, since you have the ESM is protecting the map sensor itself from experiencing boost. Otherwise it won't affect the end result, which is the signal from the ESM to the ECU. I'll let ya know what I find out later.

 

Are you saying to turn the ESM V up or down? I was thinking WOT=2.94V. Both measurment were off of the ESM...as I understand the ESM-it maintain that current V reading all of the time, and does not adjust proportionately to boost. Lowering the V rating on the ESM seemed to make the car run a bit more through boost but still not all the way.
Thanks for the input!
Kyle

WOT = 2.94 or whatever and it should read about the same with the ignition on but the motor not running because in both cases there is relatively no boost or vaccuum in the intake manifold, the intake manifold pressure is sitting right at 0psi. With the engine idling however the voltage should drop below 1 volt because the map sensor should be seeing about its maximum vaccuum. If it does not this a huge problem because the map sensor is the key sensor the ecu uses to detemine how much fuel the engine needs, if the voltage is stuck at 2.94 all of the time then your ecu will will think you are at WOT all the time and will deliver fuel accordingly. The ESM should clamp the signal from going over 3 volts or whatever its set at but should not effect the signal while its under that level. This means that as your motor goes from vaccuum to WOT to boost the voltage will go up but the ESM should stop it at a maximum of 3v so when you are in boost the ecu still only sees 0psi or normal NA WOT. The ecu will keep give the same fuel for normal NA WOT and the FMU will take over adding fuel from there. At 0psi or less when you are out of boost the ecu should see vaccuum and the corresponding lower voltages just like normal so the ecu can handle things like it normally would. If you have any check valves still on the map sensor line take them out, the stock honda map sensors are designed to handle just over 10psi with no problem and the ESM should be capable of clamping the signal to make the ecu happy.

 

this weekend it is my goal to thoroughly test my car until the CEL stops coming on. I will plug the area that is now supporting the checkvalve then turn the V on the ESM back to 2.94 V and rerun this test. I may try multiple V readings to see if this helps.
The help is still appreciated! Thanks!
Kyle

Happy Easter Guys!

 

With the DSM 450cc injectors besides modding the injector seals is there anything else that has to be done to get them functioning properly?

 

I agree with Beav, you can go ahead and take the check valves off. If you were boosting over 10 psi or so then you may need them again.

Also, double check all settings on the ESM; I know you probably got it right, but you never know. Also, since you are altering your map voltage I would recommend a thorough test of the sensor as it will never hurt. That's good that you plan on doing that.

I would check power to the map, make sure it's source voltage. Check the ground wire, with your red lead to the positive post and your black lead to the ground pin on the connector to the map sensor (you should get source voltage there as well). Check signal voltage from map before ESM with key on engine off (you should get WOT signal of course). Check signal voltage at map before ESM at idle (you should get very low voltage like .5 V or so). If all of that checks out, then check signal after the ESM at idle (should still be low voltage).

Lastly if the map is acting weird through all of this testing, you can get a vacuum pump. Turn key on engine off, Hook the vacuum pump to the MAP sensor vacuum line and with your meter leads on battery ground and map signal wire, slowly work the pump up towards 20 inches vacuum. You should get a smooth voltage sweep on your meter all the way from WOT voltage to idle voltage. You should not get any spikes in voltage. Hope I've helped you out a litt, just thought it might make your testing easier to have it all written out like this. Good luck man.

Mike

Happy Easter

 

With the DSM 450cc injectors besides modding the injector seals is there anything else that has to be done to get them functioning properly?

If its anything like other hondas the seals could stand to be bored out and the 95 integra ls taller seals fit a little better but they actually fit somewhat okay even with the stock seals, just a little tight. They will even plug right in if you cut the little key tab out of the inside of the injectors connector. The legend also seems to use low impedance injectors with resistor packs so you are good there since the DSMs are also low impedance. If anybody really wants to know I can go look at my car and detemine if they plug right in.

Is anybody else here turboed? I believe that with my software and a different map sensor like the gm3 bar you could run up to 10psi with no cel and full fuel and timing tunability in boost. New map sensor may not even be required if you do a voltage mod to the stock map sensor, but for now I am pretty sure the GM3bar at around $60 bucks would work well. From the reseach I have done tonight the stock map should read 2.8v at atmospheric pressure and throws a cel at around 3.0v. The gm 3bar runs does not hit 2.795 volts until 10 psi so 10-12psi should be possible. The motorla 2.5 bar which goes for about $20-25 could run up to 6psi before it hits 2.8 volts and 8psi at just over 3v. Anybody whos turbo just needs injectors that supports there power level at a stock fuel rail pressure, new map sensor, and a socketed ecu. Then its up to them to do some wideband tuning and get it running like a factory boosted car.

 

The legend also seems to use low impedance injectors with resistor packs so you are good there since the DSMs are also low impedance. If anybody really wants to know I can go look at my car and detemine if they plug right in.

If you could check it out that would be great. I'm looking to get those injectors asap. I finally put in my new fuel pump and AFPR this week. The injectors would finish off my fuel mods and I could move on to other things.

 

Beav,

Kyle is the only one that we know of to be boosted and running right now. That's why everyone is so astonished by your finds. Most of the performance guys on this site have just been waiting for something like this. Do you see how important your findings are to us now?

Anyways, I believe you are correct. A little modification to fuel systems and timing with some tuning and we could safely run up to 10 psi.

I don't know anything about that MAP sensor I will have to check that out. Is it something you have heard good things about?

 

Hey guys,

Top of the grab my ankles for the IRS morning to ya! I just wanted to update you with my new test. I pluged the other side of the MAP sensor like you told me to and turned the ESM back to 2.94 V. The car still throws the CEL, and now it seems to come on sooner. Do you think that maybe I should try turning the V down and keeping the MAP pluged in the same place? I may try that just for the reasoning that I can generate boost before WOT. If I can make boost before WOT it would probably be safe to assume that I need to lower that min V reading so that the SAFC can take control of the fuel adjustments. Let me know if this theory makes sense. I would want a custom chiped ECU...I actually think I have an extra ECU on hand. Just let me know if you want to do this.
Thanks again!
Kyle

 

Kyle,

Did you check the signal voltage of the map sensor at idle before and after the ESM. If so, what did you get? That would be my first step, because if you get say .5 V at idle before the ESM and then 2.94V after the ESM at idle, then something is wrong. It may be just how our specific ECU is responding or something that needs to be adjusted on the SAFC II.

 

Hey guys,

I pluged the other side of the MAP sensor like you told me to and turned the ESM back to 2.94 V. The car still throws the CEL, and now it seems to come on sooner. Do you think that maybe I should try turning the V down and keeping the MAP pluged in the same place?

I am not quite sure what you are saying here. Make sure the map sensors vaccuum lines are hooked up like stock, no check valves. Did you try reading the output after the ESM with the engine idling? This is what I really need to know so we can know if its hooked up right. Everything should be hooked up like stock except the ESM should be connected inbetween the signal output wire from the map sensor and the maps sensor input on the ecu. Make sure you dont have it connected to the 5 volt reference wire to the map sensor. That would explain the static voltage as well screw things up with the ecu and make you run lean.

I may try that just for the reasoning that I can generate boost before WOT. If I can make boost before WOT it would probably be safe to assume that I need to lower that min V reading so that the SAFC can take control of the fuel adjustments. Let me know if this theory makes sense. I would want a custom chiped ECU...I actually think I have an extra ECU on hand. Just let me know if you want to do this.
Thanks again!
Kyle

It does not really matter if you can make boost before WOT because the TPS is not primarily used for calculating fuel, the map sensor is and it will see whatever pressure there is regardless of wether or not you are WOT. What kind of corrections are you running on the SAFC? That could also be messing thinga up since it manipulates the map sensor output, if you are adding fuel at all with safc its going to make things worse and make it throw cels even if you were NA. I would avoid making any corrections with the SAFC for now because its just going to make things worse(at least with stock injectors), it will only make make it run leaner out of boost if you cut fuel or make it throw a CEL that much sooner if you add fuel.

I can definately chip your ecu for you, we need to know what route you are going though, ie are you staying with the FMU or are you willing to test something out with bigger injectors.

 

Kyle,

Did you check the signal voltage of the map sensor at idle before and after the ESM. If so, what did you get? That would be my first step, because if you get say .5 V at idle before the ESM and then 2.94V after the ESM at idle, then something is wrong. It may be just how our specific ECU is responding or something that needs to be adjusted on the SAFC II.

Maybe I am misunderstanding you and Beav. Do I make the measurements on the ESM where it has prong holes for V testing/adjustment, or on the wires themselves? I am only getting the reading off of the ESM which only changes if I change it myself manually. I called my familys mechanic and he said it should read 2.79-3V at idle. If that is correct would we be doing this entirely backwards? Because that is the idle V reading it would be in vacumm so WOT would be on the other extreem of the range? Right or am I just confusing myself? Either way I am testing the ESM out with a much lower V reading to see if that helps my problem. This is soooo frustrating....you guys willing to fly southwest?
Thanks for the continued help!
KTW

 

I am not quite sure what you are saying here. Make sure the map sensors vaccuum lines are hooked up like stock, no check valves. Did you try reading the output after the ESM with the engine idling? This is what I really need to know so we can know if its hooked up right. Everything should be hooked up like stock except the ESM should be connected inbetween the signal output wire from the map sensor and the maps sensor input on the ecu. Make sure you dont have it connected to the 5 volt reference wire to the map sensor. That would explain the static voltage as well screw things up with the ecu and make you run lean.

The lines are back to stock. To read after the ESM would I just put the prongs of the DVOM on the sensor wire going to the ECU? I have the ESM hooked up with the + going to the ECU +, the - going to the ECU - , and with the sensor wire (I believe would have been the last wire remaining) inbetween the ESM and sensor and on the other side between the ESM and ECU. I have to confess that I made an assumption that the 1st gen NSX and 2nd gen Legend had the same color of wiring for the sensors wires except for the sensor wire. Maybe this is where I messed up?

It does not really matter if you can make boost before WOT because the TPS is not primarily used for calculating fuel, the map sensor is and it will see whatever pressure there is regardless of wether or not you are WOT. What kind of corrections are you running on the SAFC? That could also be messing thinga up since it manipulates the map sensor output, if you are adding fuel at all with safc its going to make things worse and make it throw cels even if you were NA. I would avoid making any corrections with the SAFC for now because its just going to make things worse(at least with stock injectors), it will only make make it run leaner out of boost if you cut fuel or make it throw a CEL that much sooner if you add fuel.

The SAFC was adjusted on a dyno, and it was leaned out a tad until boost started to come on but then we made it run richer as more power was obtained. So maybe I need to start from scratch on the SAFC to get it back to 0 corrections and then shoot from there.

I can definately chip your ecu for you, we need to know what route you are going though, ie are you staying with the FMU or are you willing to test something out with bigger injectors.

Injectors is the way to go I suppose. Were you saying you have access to some DSM injectors? If you can find some and the ECU I will pay you accordingly. What do you say? Let me see if I can find that ECU for now.
Thanks! We....mainly you are making a paved road for future FI legends!
Kyle

 

You have done alot for it too Kyle. As a matter of fact, what you have done is 50% of the process and Beav is doing the other 50%, as far as having the ability to turbo a stock Legend and run some fun boost.

I'm gonna look at an ESM wiring diagram and in my Legend manual and tell you the exact wire colors to check.

 

Okay so here we go.

The first few tests you should do are assuming that the map sensor is good and the electrical connections at the sensor and pcm are good. If one of these tests fail then we will diagnose if you have something wrong elsewhere. I recommend using a digital multimeter and not an analog voltmeter to protect electrical circuits.

Keep the electrical connector in the map sensor. Back probe the connector at the RED wire with a small sharp needle (don't pierce wire just slide in back of connector) and clip your positive lead to it. Connect your black lead to the negative post of the battery. Turn the key on but leave the engine off. You should get a reading close to WOT or ~3V. Now start the car, you should get a low voltage reading, probably ~.5V or so. If either one of these are dramatically off you could have a problem somewhere else. If they are ok then move on to the ESM.

The modified signal wire from the ESM to the ECU is Green. Connect your positive lead to the end of the Green wire where it is spliced to the Red signal wire that leads to the ECU (but leave the wires connected). Leave black lead on negative battery post. With the key on engine off, you should get the reading which you have set the ESM to (2.93-2.94V). Start the car. You should get a reading similar to the voltage reading you got at the map sensor connector at idle or around .5V (low voltage nonetheless).

If these tests fail then there's definately something wrong and we can work from there. The tests should take you about 10 minutes or less. Let us know what you get and it'll be easier from that point on.

Edit: I believe that the prong for testing on the ESM is just the easy access way to see what you have it set to. I wouldn't trust it for your purpose. Just test at splice between the Green and Red wire.

 

Yea, i would definately be measuring at the wires. 2.8ish should be what you see with the key on and engine not running. With the engine idling if you measure at the wire going back to the ecu you definately should see .7 - 1.0 volts dependind on how much vaccuum your engine pulls.

 

I think I can try that tonight. I will let you guys know what happens after that. I am going to try setting the SAFC back to 0 corrections if all of the tests go well. If not then I will consult you guys before further changes.
Thanks again!
Kyle