| |
 |
 |
Welcome, Guest. Please login or register.
September 03, 2010, 03:26:59 PM
|
 |
|
|
|
 |
News: New members: Due to spambot activity, new membership is not granted automatically. To become a member, please complete the registration process AND send an email to admin@22rte-trucks.com. The email can be brief, and should demonstrate you are a real person interested in these trucks and not a spammer. Your privacy is respected, and your email address will never be shared. See this topic for more info.
New members MUST follow through with an email or their membership will not be approved.
Really: No email, no membership!
|
 |
|
|
Pages: [1] 2 3 4
|
 |
|
Author
|
Topic: DIYPNP Magasquirt kit in my 22R-TE truck (Read 2987 times)
|
Ronmar
Two cylinders
Posts: 417
|
Ok, I have begun the Mega Squirt journey with one of DIY Autotunes DIYPNP kits for the Nippon Denso 42 pin connector. http://diyautotune.com/diypnp I ordered the kit on Aug 28th and had it on Monday the first of September. I did the board assembly that evening after work, which took me about 4-5 hours with a break for dinner. I have been working on electronics as a profession for 27 years now and as a hobby since I was a kid, so I know my way around a circuit board. That being said, the assembly is not that difficult and there are not many parts and the component packaging and labeling is first rate. The complex micro squirt module is already built except for the connectors that join it to the main board. You do want to use a low wattage soldering iron with a fine point as some of the smaller components such as the transistors have pins that are pretty close together, and of course you don’t want to overheat them. I made a few suggestions to Matt at DIY that might also make it a little easier for someone who is not used to this kind of work.
I reviewed all the soldering connections under a magnifying glass on the following afternoon and loaded the MS2 extra 2.1.0 firmware. This went very well and everything connected together and talked correctly the first go-around. I then configured the ignition settings the same as a Corrolla/MR2 that uses the Toyota VAST ignition system. This information came from DIY. There is a bit of info on MS and the VAST ignition, much of it written by Darin, but the DIY info was for this circuit card. As a test I took the ECU out and plugged it into my truck(just the electrical connections, no MAP hose) to see if I had ignition/spark. I was very surprised when the thing tried to start on the remaining default MS information that loaded with the 2.1.0 firmware(info for a chevy 350 I think). All this took about 3 hours.
On Wed, 03 Sep after work, I went over all the setup info in Megatune and used the VE table generator to construct a base map around the factory HP and torque specs for the 22Rtec engine. I generated a spark map from advance information in the shop manual. I setup the crank and cold start enrichments using the info in the MS2 manual and plugged the unit into my truck again, this time with a MAP hose connected. I configured the TPS and Thermistors using the TPS setup tool and resistance information from the factory shop manual for coolant and air temp sensors. I turned the key to start and it started up and idled first try:) This took about 3 hours. Total, about 10 hours of actual work over 6 days from order to idle… There was also about 20+ hours of serious reading over the 2 weeks prior to the kit being delivered. I have been reading about MS for over a year. READ THE MS2 MANUALS! All this info is available in the file download from DIY for the DIYPNP kit found in the kit documentation section at DIY http://diyautotune.com/diypnp/docs.html
If using all the factory sensors, like I am doing initially, the installation is brutally simple once the build and firmware load is accomplished. In the engine compartment, I ran a vacuum line from the plenum/manifold, thru the firewall to the passenger side floor area where the factory ECU is located. I disconnected the cold start injector connector from the cold start injector. In the cab at the factory ECU, I disconnected the three connectors from the factory ECU and plugged them into the new ECU. I connected the end of the vacuum line to the MAP sensor on the new ECU. That’s it, it takes longer to type about it than it actually takes to do it:) Because there are no other hard modifications, it is just that fast to bail out of a bad tuning experience and revert to the factory ECU.
I have run a few logs on it and the Megatune generated map is as expected VERY rich. So far I have been tuning the low end area using Mega log viewer and the factory narrow band O2. I will keep the upper end as it is(safe) until I get a wide band O2. Once it is a little more tuned up, the DIYPNP will occupy the same spot as the factory ECU does. It will only take me about 10 minutes to permanently mount it in the same location… As I am trying for DIY's $125 rebate to put towards a wide band O2, here is a link to video of my truck running on the old and new ECU's .
http://www.youtube.com/watch?v=nPwmDf9-t3s
Here is the setup file for the interconnections used to configure the kit for a 22R-TEC engine.
|
|
|
|
« Last Edit: September 12, 2009, 10:28:09 PM by ronmar »
|
Logged
|
Ron It ain't broke till I can't make parts for it... My gallery |
|
|
mosk
Pooh Bah
Administrator
Three cylinders
Posts: 707
Been there, done that
|
Awesome write up! Stickied for everyone's benefit.
I'm looking forward to reading about your continued progress with this setup, ronmar.
Thanks for posting this, and congrats on your easy installation!
-Jeff
|
|
|
|
|
Logged
|
Jeff's 22R-TE PageThe vintage HTML ain't pretty(!), but the articles are still worth reading if you are new to the motor
|
|
|
dcg9381
Global Moderator
Three cylinders
Posts: 957
|
That is awesome!
I use a tool called MegaLogViewer. Using your base map, properly setup 02, and data logs, it can tune the fuel map for you after you've done a drive. It's more effective with a wideband, but I have used it with narrowbands for initial tuning.
There is an auto-tune feature in megasquirt, but it's only going to work with a wideband (per my understanding) AND it was still in initial development when I was tuning. It may be functional and helpful at this time.
Let me know if I can help you with tuning. Some things like doing acceleration testing (throttle-tip-in) are a little tricky to get just right, but make a big difference.
Also watch your 02 step function (how fast MS reacts to being lean/rich) - you want MS to adjust very slowing during initial tuning...
|
|
|
|
|
Logged
|
|
|
|
Ronmar
Two cylinders
Posts: 417
|
Thanks Darin I have been using the megalog viewer VE analysis feature and it works great, even with a narrowband. I was limiting my analysis to below 100KPA though. The autotune function in magatune also works with a narrowband, and was tuning in apparently correct values but it is a little clunkier. I used my rebate from DIY toward a wideband, and my new LC-1 kit was delivered today so I know what I will be doing this weekend:).
Having the sensor data under the microscope in the datalogs and megatune showed me some other issues. My thermostat was a litle sticky. The gauge being old and with a natural wide band in the normal zone like most all factory gauges do, never gave a hint of this. My coolant temp sensor also had a little creep in it, but it was 24 years old, so I forgave it and got a new one along with the thermostat. I am now getting correct and consistent temp info into MS with the factory temp-resistance values loaded into the thermistor cal tool. This makes for much more consistent tune results(and less head scratching)...
I played around with the acceleration enrichments, but was having issues with that. Probably due to the upper end of my VE table not being properly tuned yet and way overly rich. I will revisit that once I get the upper end of the table dialed in with the wideband. then I will see if I have any questions or issues with the setup info.
Initially I accidently setup the EGO correction for a wideband algorithim, which disabled the step size setting window. This only allowed me to set the total percentage of VE table correction, which it proceeded to move by at the prescribed number of igntion events. It ran a little rough on the road when it started applying a 10% EGO correction 3-4 times per second:) I finally got it set to the "simple" algorithim which enabled the step size, so now it steps at 1% up to a maximum of 10% . It steps out one way or the other till it gets an O2 crossover and then hops back and forth chasing the O2 crossover point much like the factory ECU did when in closed loop. I have only tuned the table below 2500 RPM and 100 KPA, but it is dialed in to the point that all the correction occurs well within 10% of the VE table. I like the EGO correction gauge in megatune.
More to follow.
|
|
|
|
|
Logged
|
Ron It ain't broke till I can't make parts for it... My gallery |
|
|
|
|
|
|
1project2many
Two cylinders
Posts: 174
|
Ok, yes. It looks like enough data there to extrapolate a pressure curve. The rpm based change should work as is, at least as a starting point.
Thanks.
|
|
|
|
|
Logged
|
|
|
|
Ronmar
Two cylinders
Posts: 417
|
That timing info did work as a starting point, but after todays measurements, I have determined that it is no where near reality for the R-TE's ECU timing output.
I put the factory ECU back in and did a little work with an O-scope after I made a breakout harness to sample the igniter signals that are exchanged with the ECU, "N", "IGT" and "IGF". N is the timing square wave sent to the ECU derived from the distributor timing wheel. I used this as the O-scope trigger. IGT is the trigger signal from ECU to igniter that fires the coil. IGF is a confirmation pulse that the coil has fired, from the igniter back to the ECU. I sampled IGF as a representation of when the spark is firing in relation to the main timing pulse N. IGF and N are also both clean squarewave shapes with relatively fast/vertical edges to measure timing off of.
The engine with baseline timing set at 5 degrees BTDC idles at about 11-12 degrees BTDC. That is pretty much where the similarities between the shop manual timing data and the R-TE ECU end. As you increase the RPM to about 1100, it jumps from 12 to about 30 degrees BTDC. Under no load, at around 3600 RPM, the timing then advances smoothly up to about 40 degrees BTDC at about 5200 RPM. That was as high as I saw the advance go. Using this information, I was able to roughly calibrate what I was seeing on the O-scope to a timing number that I can record while driving under load.
Above 1100 RPM, it appears all timing is based around this 30 degree BTDC point. Under lighter loads, the timing advances slightly, perhaps as high as 35 degrees BTDC. The lower the manifold pressure, the greater the advance. Using a boost/vac gauge, I noticed that this advance lessened back to 30 degrees BTDC as the manifold pressure/load increased to atmospheric. As the gauge transitioned into the positive pressure area, the timing retarded even further with it setteling to as low as 20 degrees BTDC under full 6PSI of boost. A little more driving tomorrow and I should be able to rough out a reasonably accurate representation of a factory timing map.
|
|
|
|
|
Logged
|
Ron It ain't broke till I can't make parts for it... My gallery |
|
|
1project2many
Two cylinders
Posts: 174
|
What you're doing is what I talked about several months ago. The difference is that I wanted to use DAQ and a laptop to record timing during my 45 minute commute. The main benefit there is more samples over a greater range of conditions.
Are you using a scope capable of storing traces for later review? I have a dual channel portable scope here, but I wondered if I'd be able to drive and effectively calculate spark advance from time differences at the same time. Seemed like a better job for two people.
I spent a few hours last night scouring the web for any info on the RE or RTE ecu. Part of the search was looking for screenshots of Techtom's programming software displaying stock maps. During the search I found some posts which appeared to indicate Toyota's strategy is to use RAM to store a learned or corrected map which is populated over time. The indication was the ecu tended to leave the corrections once stored. It might be worth noting this if inconsistencies begin to appear in measured timing over different drive cycles.
Also keep in mind that EGR operation requires additional spark advance. If a disparity appears such that a trend toward less advance as the engine warms is suddenly broken by a large increase in advance when load is small and throttle is off-idle, it may be compensation for expected EGR operation.
I had an ecm bench set up years ago for GM tuning work. I suppose I could set one up for the Toyota,
|
|
|
|
|
Logged
|
|
|
|
|
|
Ronmar
Two cylinders
Posts: 417
|
OK, well this is one case where one of my hair brained ideas pays off. I don't know if anyone has done this before, but getting my brain around the readings and calculations for timing data collection with an O-scope was giving me a headache. So I thought to myself, this would be a lot easier if I could just look at the advance with a timing light under load. One webcam, a timing light and a few ty-wraps later, I came up with this... In the video, the pulley rotates from right to left The 3 marks on the red bar represent, from left to right, 20 degrees, 29-30(missmarked it a little bit) and 40 degrees of advance. You can just see the last two factory timing marks at 12 and 15 degrees all the way to the left.
The result after several hillclimbs at various RPM and loads is this table. I of course do not have the ability to get to all areas of the table, so the trends heading out into those areas have been smoothed off by me. The bottom line represents the fuel cut area. The top line represents approx 8 PSI of boost and I can't get that high so 13 degrees is a guess. The lowest advance I saw under full boost(approx 6PSI) was 15 degrees BTDC and this was consistent across the RPM scale as high as I was able to comfortably take it. Based on the data points I was able to collect, I believe it to be reasonably accurate. YTMV
|
|
|
|
« Last Edit: September 16, 2009, 07:14:04 AM by ronmar »
|
Logged
|
Ron It ain't broke till I can't make parts for it... My gallery |
|
|
1project2many
Two cylinders
Posts: 174
|
LMAO!!!
I was thinking of doing the same thing while driving in to work yesterday am! Too funny.
Thank you for posting the results. Toyota really ramps up the advance curve near the top of the rpm range. No wonder they cut fuel. It's the only thing saving the engine.
|
|
|
|
|
Logged
|
|
|
|
dcg9381
Global Moderator
Three cylinders
Posts: 957
|
Ronmar,
That's absolutely super. I would have loved to get a factory timing map, but I couldn't come up with a way to keep it on the scope or record the timing and have the truck under load! Outstanding!
Mind if I steal your timing map for my collection?!
|
|
|
|
|
Logged
|
|
|
|
|
|
mosk
Pooh Bah
Administrator
Three cylinders
Posts: 707
Been there, done that
|
Super stuff! Love the web cam idea -- wish I'd thought of that back when I needed this sort of data.
This is great. I am really enjoying following your project. Keep it up!
-Jeff
|
|
|
|
|
Logged
|
Jeff's 22R-TE PageThe vintage HTML ain't pretty(!), but the articles are still worth reading if you are new to the motor
|
|
|
|
Pages: [1] 2 3 4
|
|
|
|
|
|
22RTE-Trucks
