Post Reply 
 
Thread Rating:
  • 1 Votes - 5 Average
  • 1
  • 2
  • 3
  • 4
  • 5
Tuning for Performance
Author Message
jbalat Offline
Member
***

Posts: 280
Joined: Feb 2009
Reputation: 1
Post: #1
Tuning for Performance
Guys this is some good avice from Les who was involved in publishing the D17 Digital efie document. He has many years of experience in tuning and tuning for hho.. Hopefully this will be of some help to me and others who are struggling to get savings or consistent savings...

Note that there is some reference to my propellor board in there which I use to intercept and adjust signals. Les is familiar with the unichip and sounds like he provides some technical support for the volo chip. Just have a read and take what information you need for your own projects...

.......... I think your propeller board will work for everything. I'm a little skeptical about using it for the timing, but I'm not familiar with it on a '94 so you might be good to go. I know on my '01 the unichip's prediction algorhythm kept instigating a missfire code which makes for lot of fun.

On a '94 model I'm going to venture a guess that you do not have cat. sensors. Even if you did it won't matter don't worry about them, all the ECU could do is complain if they are too cold anyway. I sometimes use the cat. signal to rough in the tuning on an engine. I know when the cat. temp drops I'm getting lean enough. Interestingly I get the best power just as the cat. temp drops.

So I'll run you through the ABCs of what you are trying to accomplish. This will be a long Email so grab a cup of joe, or a stiff scotch whichever your pleasure, I'll wait...


OK, I hope it's a tall one. Here goes. I'll start with a bunch you probably know, but I'll skim over just to cover the bases. We want control over AFR, and timing. Your MAF not only tells the ECU "I'm getting 'X' amount of air." it is also your load sensor it tells the ECU how much demand you are making on the engine so it plays a very large role in ignition timing. If you show an artificially low load the engine will not retard timing as quickly as it should. If you show an artificially large load it will retard too quickly. If it were not for this duality, and the way the ECU uses it's tables we could just target a lean AFR on the O2's, more or less match it with an equally lean MAF adjustment, and be done. It doesn't work that way as you have discovered. Most guys on the internet trying to tune engines for HHO don't have a clue what they are doing. The guys at Unichip don't know how to do it either, and they really don't care to know they just want to sell chips, and when they get a call for HHO help they funnel them to me.



First you absolutely must be able to read the PIDs from your ECU in real time as you are driving to tune the engine. Without a scanner that can read live data you are just guessing, and will not ever get it right. I'm not familiar with your board, but I'll run through what you need to have control over, and a bit of how to do it most efficiently. You need O2 sensor control, which I think you have. You can also control them in tandem if you have 2 sensors by simply slicing your grounds together, and putting your voltage offset to the common ground. You then make the exact same offset to both sensors equally, and simulateously, and it works better, and you only use 1 channel. You need MAF control which you said you have. You need to be able to read your load. Using the MAF is OK, but the TPS(throttle position sensor) is much more reliable, and more accurately predicts demand as the MAF can lag, and be more fluctuant for tuning purposes. Plus you are going to modify MAF, and if you are reading what you are modifying you will get lost in the woods. You need to read coolant temp. This is very important. You cannot make lean modifications to a cold engine without causing yourself a lot of headaches. You will never get the performance you want hot, and still be able to start the engine in cold weather. I don't know if your board is capable but you need to have channels that get their input from another channel, and respond to that instead of an external input. Your O2/AFR table will be controlled by the load(TPS), and it needs to be muted by the temp. reading. In other words, when the engine is cold your board makes no changes until you hit about 175-180 deg. F. engine temp. Then the modified signals need to fade in rather smoothly up to operating temp. which should be about 195-210. Also be careful with your temp reading. Study what it runs now because when you tap in to read it you will invoke a lesson in the Heisenburg uncertainty principle. When you measure it, you change it. You will probably need to find a shunt to ground resistor to correct the temp. back to true because the load of your board reading the temp will change it's reading.


Now Your MAF channel will also be controlled by TPS reading, and temp. mute. Same setup. Hopefully you can use just 1 channel for temp. mute, and 1 channel for TPS to control all the other channels. The hysterisis of reading doesn't seem to affect the TPS because the computer assumes starting condition to be closed throttle, and corrects itself. Your timing channel has to step in time with the TPS reading, and temp. mute also. So what we end up with is a system that does nothing until the engine is almost up to operational temp. and then slips in smoothly to start leaning back the right amount, at the right conditions.


Ok we know what we need to do to get there, but you still don't know where we're going. Tuning priority is #1 AFR, #2 fuel trim correction, #3 timing, #4 fine tune. I hope your board can extrapolate between load/speed points or you have a very tedious process ahead. With the Unichip I just tune for about a half dozen points, and then tell the software to extrapolate what all the other points in the map would be if I followed the same reasoning, and it does it instantly. If the software for the propeller board doesn't extrapolate you will have to tune lots of points, and then manually extrapolate by doing the math yourself. Because you can't tune every single load/speed point live, it would take far too long.



AFR: The common myth in tuning for economy is that we need to castrate the engine to get better economy. Not true, and in fact counter-intuitive if you think about it. A gasoline engine is basically an air compressor in reverse. Instead of rotating it to store energy as compressed gases, we are expanding gases to harness the rotary power. Either way volumetric efficiency is the name of the game. When we have the best volumetric efficiency we will simultaneously have best power, and best economy. The difference between an economy tune, and a power tune is the adjusting of your foot away from the floorboard on the throttle. The only change to this with HHO is that when we are at low load cruise we can tune ultra lean, and use less fuel maintaining a speed. Other than that tune for best power, just like a hot rod. Your first step will be to tune for best power at all load/speed sites, just as if you were tuning for speed. Efficiency is efficiency where it lights the tires best you are most efficient. Make note, in most instances you will make better power at a leaner condition than factory setting because the ECU is designed to waste fuel to keep the cat. hot. The ratio numbers do not matter, listen to the engine, and give it what it wants. If it runs best at 16:1 at a certain RPM, and load point so be it, whether it makes logical sense or not that is what the engine responds to, don't argue with it. This is old school tuning 101, and most mechanics have forgotten about it because it isn't taught anymore. 14.7:1 is a federal emissions mandate to keep the converters hot, and has nothing to do with engine efficiency.



Fuel trim(MAF) Step #1, and #2 are done more or less together. As you tune the AFR up or down watch your fuel trims, and try to keep them close to zero. On your system it may be called the integrator, and 0% fuel trim will be about 25 on the integrator. I would research it to make sure I'm correct. I haven't done ODB I in a long, long time. OBD I should be very forgiving on fuel trims though so close will count. Use your MAF control to make the load readings match up with what the computer is reading from the O2. The MAF, and O2 need to "agree." This is reflected in your fuel trim. A positive trim means the ECU is adding fuel to come up to what the O2s say it needs. A negative trim means the computer is taking fuel away to get to the AFR the O2s are targetting. So if you adjust the AFR lean on the O2 you will see a negative fuel trim, as you adjust the MAF down it will go less negative, and then positive. If the trim goes way positive you went too far with the MAF offset. Once you get the AFR set for best power at all points, and your fuel trims stay low through every point too, you will need to correct timing.



Timing: On my Montero I directly intercept the crank position sensor, duplicate it, and send a manufactured signal to the ECU. The Unichip has these fascilities built in so it responds to all the other inputs, and maps simultaneously. It can advance timing because it predicts the CKP wave after it learns it, and can tell the ECU a timing event is happening before it actually happens. Pretty wicked tech. You need to be able to possibly advance timing because you have to show an artificially reduced load to make the intake calcs match lambda, or incoming calcs match actual burn. This reduced load from actual means the engine's timing curves can be scewed a little on the retard side in places, and you may have to advance it to get good performance, also a lot of factory tables are too far retarded for emissions reasons. I found my Montero ran much better about 2-3 deg more advance in most places, and almost 4 deg more advance under moderate to heavier acceleration. Even with HHO which generally needs a little more retard due to faster burn rates. Like I said there are so many variables involved the only thing that can tell you what you need is the engine. Again set your timing for best power at all load points. The timing needs to be controlled by TPS, and temp. mute. If you get missfire codes, God have mercy on your soul. This will be your biggest challenge.



Fine Tuning: you will find that when you correct your timing you can go back, and get a little more power at a lot of points by leaning back on the AFR again. You usually don't need to compensate on the MAF again for these small changes. The ECU has enough "slack" in it's programming to compensate. So you can go back, and forth with the timing, and AFR setting until it's perfect after it is roughed in with steps 1-3. As I said software extrapolation will save you hours at this.



Extra step #5 Tuning for HHO: Now you need to tune for the use of your water gas. Your generator needs to be bullet proof. You need at least 1.5 LPM all the time. Fluctuations will kill you. The best way to accomplish this is with a hot electrolyte mix, and a PWM. We use a PWM with a thermistor control so that the cell heats up quickly, and then the PWM backs off on the duty cycle as it reaches target temp. If the reservoir tops it off, and it cools a little the PWM throttles up to regain target temp within a few seconds. We have a provisional patent pending on our generator controls so I can't go into it with you a whole lot, but you have to have very, very tight HHO production control. You also can't pull a lot of amps to make your gas. You should pull no more than 20 amps to get your 1.5 LPM, or you are going to reach a point of diminishing returns quickly. Our cells make 1.5 LPM at about 14 amps, and 2 LPM at about 18 amps. We target a temperature that makes a little over 1.5 LPM, and it holds that temp tenaciously. If your cell is up to snuff. You can now concentrate on tuning to use it. Even at 2 LPM we are only making enough gas to affect combustion at a relatively low load, and consumption rate. At higher revs, and loads we aren't doing much of anything. So we are only going to change the map at low load. Find the load points that you use to just maintain speed. Low cruise in other words. We are making a small map within a map. An ultra lean cruise zone as it were. You will find that the load point for holding 45 mph isn't very far from the load point for holding 65 mph. In this narrow band you are going to set the AFR as lean as you can without causing hesitations, pulsations, or missfires. It's better to be just a tad rich, and not casue a driveability issue. Also don't try to make too wide of a load band too lean. Only go a little above the low load cruise band you define. You also want your map to taper smoothly back into it's power setting fairly quickly above the lean band so when the engine needs to make power efficiently it can. Don't strangle the golden goose. If you make the engine too lean in too wide a band you will not only loose power you will lose economy. When tuned correctly you shouldn't notice any lack of throttle reponse from the engine because as soon as you make a small acceleration demand the engine should be back in it's power map on your board. In other words it only runs lean when just station keeping, as soon as you accelerate a little it's back to making best power.



A few other considerations: However you go about managing your timing be aware that your start up signal must be unaltered or the engine will not start. I would also suggest making no timing changes until after about 1200 RPM. I would also make no fuel changes below 1000 RPM. Leave idle alone. Yes the engine is set up to run very rich at idle, but leaning it out much becomes problematic, and besides the hyper-miler's chant is "To idle is to sin." You can change your differential grease out for less viscous synthetic variants, and actually have better protection as well as greatly reduced rolling resistance. I use Amsoil, and am very happy with it. I also run Amsoil 10W-30 full synthetic in the crank case, and it looks like new oil at 5K-6K miles later. You can air your tires up to about 40-45 psi without ill effect, regardless of what the tire store tells you. In fact the tires last about 10-15% longer, and you'll get about 10% better mileage. Anything above 45 psi doesn't seem to help much more, and starts to ride like a log wagon. I put nearly 50K miles on a set of 35K mile tire that always ran about 42 psi. They wore out very evenly. 10 psi extra will not cause the centers to wear prematurely from over-inflation. If you have some rough terrain to negotiate regularly you might not want to run a higher pressure though as it does reduce off-road traction. If you built your board yourself you should be capable enough to have good success with this project. Dig into the sofware for your board, and become very familiar with it. The interdependency of the maps is very important. they all have to work together, and respond together from the same TPS input, and not modify outputs until the temp. is up. I wouldn't use MAF as an input because you are modifying it, and you will run into hysterisis by modifying your signal according to it's own reading. Good luck. Hope I have helped, and not confused you. If you have any other questions fire away.



Les,

-------------------------------------------------------------

No I don't mind if you pass the info along. I have told lots of people how to do it, but only a few catch on. I'm not hiding anything or being evasive in any way it's just a paradigm of thought about tuning that isn't used much anymore. Old school mechanics that used to build carby hot rods with manual timing etc... get it. The new techy guys don't have a clue because they are taught not to modify the OBD system at all. Like it's a sin almost.

You need a code scanner, not just a reader. Readers will tell you what codes are in memory in the ECU, and reset them, but that's it. A scanner will read the PIDs or live data from the ECU. The ECU is designed to take up the slack so to speak for it's sensors, It's called stacked tolerance. This is what the fuel trims are showing. If the MAF says I see "X" amount of air, and the ECU delivers a meter amount of fuel to equal "close to 14.7:1, then it looks at the results of this calc by reading the O2 sensor. If the O2 sensor is reading richer than 14.7:1 the ECU assumes the MAF is reading a little generous, and decides to add less fuel for the corresponding air readings, and the fuel trim numbers go down. There is a limit to how much the ECU will do this. If it has to trim too much it throws a code to tell you it has a faulty MAF sensor. That is why it is important to read, and match up these numbers. Your code scanner will need to be compliant with the diagnostics system on your vehicle. Were it an American Montero I could confidently tell you it is ODB I, Being a Pajero, the (well, by my perspective) foreign variant. I would make sure what on board diagnostics protocol you are running. A good parts supply store should be able to fix you up. Or talk to a Mitsubishi dealer.

When you are creating your maps remember that even though you will need to extrapolate linearly between tune points, your tune points will not be linear they will represent more of a bell curve that lets the engine run closer to stock towards the higher RPMs, and load ranges. Except timing a lot of timing correction can be done at higher RPMs, and loads for some pretty nice performance increases. Remember to tune for performance, and resist the urge to go leaner than best power. It truly will not gain you anything. The only exception being low load cruise if, and only if you are making enough HHO gas consistently. Good luck, and keep me posted on your progress, I look forward to hearing some good results.

Les,
05-29-2010 07:57 PM
Visit this user's website Find all posts by this user Quote this message in a reply
thomasbala Offline
Senior Member
****

Posts: 629
Joined: Jun 2008
Reputation: 4
Post: #2
RE: Tuning for Performance
(05-29-2010 07:57 PM)jbalat Wrote:  Guys this is some good avice from Les who was involved in publishing the D17 Digital efie document. -------------------------------------------------------------
Snap! I didn't know it was that easy; all I have to do is make an ECU that controls my vehicle's ECU. Rather than use Propeller can I use say, an Atmel 168 and Arduino?
05-30-2010 11:44 AM
Find all posts by this user Quote this message in a reply
jbalat Offline
Member
***

Posts: 280
Joined: Feb 2009
Reputation: 1
Post: #3
RE: Tuning for Performance
Thomas no one said it was easy Wink
05-30-2010 06:33 PM
Visit this user's website Find all posts by this user Quote this message in a reply
jbalat Offline
Member
***

Posts: 280
Joined: Feb 2009
Reputation: 1
Post: #4
RE: Tuning for Performance
I asked Les about running in Open Loop. This seems a bit easier but I have noted some drift from high 16 AFRs to mid 17AFRs which probably will be an issue. Again Les is spot on the money.

BTW here are some links for anyone dashing with a soldering iron and wants to make their own wideband controller (if so please make me one !!)
<http://www.siliconchip.com.au/cms/A_111598/article.html>
<http://www.siliconchip.com.au/cms/A_111676/article.html>


Anyway Read more below about running with the O2 unplugged.....

Hi Les, one more question since I am finding it very difficult to find a scan gauge that works with my car
If I run with the o2 sensor unplugged from the ECU
- The AFR jumps to 18 + (True I have measured this with my wideband)
- The engine stays in open loop
- I dont need to worry about fuel trims
- I only need to worry about MAF and timing changes

Can you please confirm my observations/perceptions above ?

Thanks,
John B

------>

Yes your observations are accurate. When you are operating in open loop mode the engine is not checking the results of it's fuel calcs with the O2 sensor. Since the sensor is unplugged obviously it can't so it runs in open loop, and simply "trusts" the MAF to be more or less accurate. So theorectically you could operate this way, and use you propeller board to make a custom map that modifies the MAF signal to generate the AFRs you want at the conditions you want.

The only problem is that it doesn't have any tolerance built in so you will have to leave your wideband hooked up to watch your AFRs and occassionally retune as conditions change. How often I don't know, I've never done it this way as it is illegal here, and you can't get an inspection sticker for your car if it is in open loop operation. You will also have to tweak timing a little each time you change your MAF map because MAF has a lot of control in timing decisions.

Les
(This post was last modified: 06-01-2010 12:01 AM by jbalat.)
05-31-2010 08:45 PM
Visit this user's website Find all posts by this user Quote this message in a reply
jbalat Offline
Member
***

Posts: 280
Joined: Feb 2009
Reputation: 1
Post: #5
RE: Tuning for Performance
Just to clear things up about quality of gas and whether the hho generator is robbing the engine of horsepower this is what Les had to say. Hope it helps since it has been discussed several times on this forum lately...

Hi John, and Peter,

Your thinking is not way off, just a little off : ). Your goal is not to "make" horsepower with the HHO. The goal is to better catalyze the fuel to extract more energy per gram burned. You are correct that the energy penalty can quickly become too great to have a net gain. I have the PWM on my Montero set at about 15.5 amps max, I get almost 2 LPM at 1psi there. I have also modified my PWM to be thermistor regulated so that it achieves this draw sooner, and holds it. The sweet spot for most engines seems to be between 15-18 amps. Not so much because of the load on the alternator, a 90 amp alternator at full charge only pulls about 2 HP. On a 220 HP engine at 1800 RPM this is almost an insensible loss. The reason is the heat, and inefficiency of current(affordable) cell design. As you get much above 20 amps you start making a lot of water vapor instead of HHO. If you have a thermometer you will notice a significant heat increase above 18 amps to about 20 amps. At around 22 amps you are basically just boiling water. I know water injection sounds like a good idea, but in this case it isn't. The vapor produced is already in a high energy state, and really can't lend much to expansion in the combustion chamber. A water injection system injects cold mechanically nebulized water into the intake, so it is in a low energy state, and explosively becomes vapor in the combustion chamber lending to volumetric efficiency.

Where are you coming up with the data that you need a liter per minute per liter of engine? That isn't really the case, and as you have noticed is not readily achievable. 1.5 - 2 LPM should be sufficient for a 3L engine. My business partner runs a 360 c.i.d. Dodge Ram 1500 on 2 LPM, and gets great gas mileage.If you wanted to get really fancy, and build a really efficient, and expensive cell you could get 3 LPM under 20 amps, but it really doesn't pay off as well as you would think. You might gain an extra 10% maybe 20%, but it would take longer to recouperate your expenses in fuel savings. You can tandem 2 cells with a cooler, and a circulator between them. Run your electrolyte very concentrated, and set the duty cycle on your PWMs very short, and you keep the heat down in the cell, and make LOTS of gas, but you run the risk of thermal run away if your PWMs ever malfunction, or get inadvertantly set too high. It is also a complicated, expensive, and physically large contraption. Just to gain 1LPM at 17-18 amps. I wouldn't try it unless you have a thermal controlled PWM. Unless you like the smell of ozone, and burnt silicone! I speak from experience. In this set up amperage can go from 15 to 40 amps before the fuse can blow, and wipe out your mosfets on the PWM board.

I would set for about 16 amps, tune for best power there. Lean out the cruise zone, and you should have it. Timing is also important. You may actually have to advance the timing in some cases. A lot of vehicles are set extremely retarded on the timing to begin with to combat NoX production, and quench combustion temperature. You are accomplishing this with the HHO, and don't need to handicap your timing anymore. Set for power on timing in all cases. Consistency in gas production is also very important. You can't expect to see gains from an engine tuned for a certain condition, but not getting that condition very often.
Let me know if you guys need any more help.

Les,
(This post was last modified: 06-22-2010 08:45 PM by jbalat.)
06-22-2010 08:45 PM
Visit this user's website Find all posts by this user Quote this message in a reply
shearinfinity Offline
Member
***

Posts: 33
Joined: Jun 2011
Reputation: 0
Post: #6
RE: Tuning for Performance
This is great information, ive recently began some testing in open loop with maf and timing adjustment. I havent tried to go extreme since im installing my wideband o2 and egt probe this weekend. I just finished the wiring of the gauges. I also have a pwm and a few dry cells installed. Ive adv timing a bit and also have cold water injection to take advantage of it.
06-06-2012 10:20 PM
Find all posts by this user Quote this message in a reply
Autotech Offline
Member
***

Posts: 4
Joined: Jul 2011
Reputation: 0
Post: #7
RE: Tuning for Performance
I have found that understanding how each engine sensor works helps with my tuning. Found some usful info at this site. http://engine-sensors.com/

Best way to save fuel is to stay home, but what fun would that be?
Auto Repair Manuals
11-01-2012 06:44 AM
Find all posts by this user Quote this message in a reply
Post Reply 


Forum Jump:


User(s) browsing this thread: 1 Guest(s)