$8D VE cells maxed at 100

While tuning my '90 Corvette's TPI 383, I've run into a bit of a wall. At 2800-4000 RPMs and 70+ kPa, my BLMs are indicating that I should have a VE of over 100% (which I've read some people saying is very unlikely for a N/A engine, and others have said that tuned runners as are found on TPI can give efficiencies over 100%).

I didn't think this was a big problem, but I just noticed that when I commit the changes to the VE table, if I open the table back up I can see that the values over 100 are all rounded down to 99.61, giving me an irreparable lean condition in those cells.

What's my recourse here? Is there a modified XDF that allows larger VEs? Should I be adjusting something else in the BIN to compensate? Or, do BLMs over 100 indicate some other underlying issue?

EDIT:
Looks like decreasing the injector flow rate will work?

yeah you can just drop 10% from both your injector flow rate and your ve table to get 10% more headroom, will work fine.

Quote:

---

Originally Posted by 90Vette

What's my recourse here? Is there a modified XDF that allows larger VEs? Should I be adjusting something else in the BIN to compensate? Or, do BLMs over 100 indicate some other underlying issue?

---

Might also export a log file and run it through the Analyzer at Post #1 of this link: http://www.gearhead-efi.com/Fuel-Inj...8884#post48884.

It will provide details of everything the engine is doing.

Man, you guys here sure have made a lot of great tools. Thanks!

Just passed smog so I'm not too concerned with my tune, but I'm making a long drive this weekend that I'll datalog to make the final tweaks to my VE tables; I'll definitely take a look using your tool too.

Did you have any luck with your tuning? I also have a '90 Corvette and plan on replacing the 350 with a 383 later this summer. How is it working for you? I'm also new to the whole PROM programming.

I'd say so! I've put my final tweaking on hold to figure out a problem with the Ostrich (it causes the car to completely cut out and immediately restart intermittently), but I'm very happy with even the tune I have now. It's as peppy as can be expected with the stock TPI intake on there, it runs smoothly, and as mentioned above passed smog with flying colors.

Tuning really isn't as intimidating as it seems; just be sure to do your research first. Here are a couple good resources to start, and my best shot at a quick summary:

• How BLMs and the integrator work
• Huge compilation of mostly relevant threads
• Intro to 1227730
• Good general info

Hardware and Software

Well, you'll need something to datalog and something to communicate your tune to the ECM. I used a Bluetooth ALDL adapter and a Moates Ostrich 2.0 chip emulator attached via a G1 chip adapter. I haven't read about anyone else having the same problem with the Ostrich I have, so I'd still recommend going with it. It's more expensive than a EEPROM and chip burner, but I'm able to pop the hood and flash a tune in ~10 seconds which is pretty nice when you're constantly updating your tune. That being said, the EEPROM seems like a better long-term solution. However, unless you plan to datalog with ALDLDroid with your phone, I'd suggest using a USB ALDL cable. It's cheaper, and getting the Bluetooth adapter to play nicely with TunerPro was a bit of a hassle.

On that note, TunerPro RT and Trimalyzer are two invaluable pieces of software I recommend.

You need to be sure you have the right ADX and XDF files. You can find $8D.xdf (simply put, $8D is the name of the program this ECM runs) I believe on TunerPro's website, so that's not a problem. However, the ADX was an absolute nightmare to figure out. Something about different systems on the car communicating through the ALDL datalink causes endless data errors without the macro that tells them to shut up. CaptainKawasaki's ADX has this, and is available on this forum. However, for whatever reason that stopped working for me, so I had to tweak the file a bit. Actually, I think I ended up having to copy his macro to a different ADX I found. (It was a huge pain; I'm familiar with programming but there's no documentation out there on this language.) I also adjusted a couple of the histogram settings to make things easier (no idea what, it was quite a while ago). For simplicity's sake, I've attached both files I use. Can't guarantee they'll work for you, but they should.

Fueling Tables

Tuning the VE tables (volumetric efficiency: the ratio of how much air enters the cylinder on the intake stroke to how much air it could theoretically hold; using this number the ECM can figure out how much fuel to add) is pretty dang easy. You basically drive the car, check the BLMs (basically the ECM's corrections to fueling to achieve an optimum AFR), and modify your VE tables. Rinse and repeat. I tried doing this via an Excel spreadsheet I made for a while and didn't get anywhere, but when I started using Trimalyzer (created by a ThirdGen user) everything fell into place pretty quickly. Use Trimalyzer, it's free, works better than doing it manually (at least for novices like us, I can't speak for the pros), and is a heck of a lot easier. What I didn't realize that kept me spinning my wheels is that the ECM interpolates to calculate the BLM it will use - that is, it pulls from table cells above, below, left, and right of the one you're in. Trimalyzer has a good smoothing algorithm, preventing serious headaches. Let me know if you have questions about it; it took me a little while to figure it out (2 tips: use "geometric" method, and click "New VE Layout" to get it to match your logged data tables).

While tuning VE, it's hard to mess up the car. Because it self-corrects with the BLMs and INT, you have to be reaaaaally off on your VE number to damage your engine or cat. However, two caveats here: If the car is in open loop mode (which it runs in before it warms up), it won't correct fueling. Don't drive anywhere in a datalog run until you can see in the TunerPro RT datalog that it entered closed loop, or you could melt your pistons/cook your cat. Secondly, it can't correct a cell without first entering it. So if your tune is very lean in a few high-RPM, high-load cells, you could have a severe, damaging detonation event. Because of this, it's safe practice to raise your VE numbers across the board ~10-15% from your stock tune to start, and also, slowly work your way up the tach (e.g., don't wind the engine up until you've gotten relatively good VE numbers at the lower RPMs).

When you do get to high-load, high-RPM tuning, you need to be mindful of power enrichment (PE) mode. When a couple conditions are met (namely, throttle > some value), the car enrichens the AFR, adds some timing, and effectively goes to open loop mode. In open loop, the BLMs you get are meaningless. Trimalyzer can filter out this trash data, but it's best to just avoid PE while tuning VE and the main spark tables. Once your VE tables are good just below the PE threshold, you can start gradually raising the minimum TPS% for PE until PE only enables at full throttle. This will allow you to get those high load, high RPM cells better.

Main Spark Tables

Tuning spark tables and PE fueling are a whole 'nother beast, and I admittedly don't understand the process as well. Without access to a dyno, it's hard to say definitively if a change you make is an improvement (except at cruise; one of the links in the thirdgen thread addresses that). But you can get acceptably close. Again using Trimalyzer, you can see a graph of where you are getting knock counts based on load and RPM. Wherever you're knocking, drop the timing a couple degrees. Do this until you get acceptably few knock counts. I believe the rule of thumb is that you should shoot for 2 degrees timing less than it takes to make it knock. That may not be the optimum point, but it's the best guess without a dyno.

Tuning the spark tables is where you can really mess up your car. If you add in a lot of timing at once, you could cause a lot of piston-melting detonation. But just go slow and keep an eye on your knock counts, and you'll be good.

Spark affects burn efficiency, so once you nail it down you should revisit PE briefly to make sure your BLMs haven't drifted too far from 128.

Power Enrichment

Ideally to tune this, you'd have a wideband O2 sensor (a stock narrowband sensor is only accurate close to stoich--14.7:1--so to get the right AFR when you want it much different than that you need a more advanced O2 sensor). I don't have one, so I just did the math to shoot for a 12:1 AFR in PE. Anyway, once you set your desired fueling in PE, you play the timing game again using the table for additional PE spark advance. Once you have it so it's not knocking, call it a day. From what I've read, the difference between a good tune and a perfect one is not big enough that the butt dyno can get you from good to perfect, so unless your final tune is knocking/running rough/really disappointing in power/getting horrible mileage that's probably as good as you're gonna get.



A few more pieces of advice/info:

• Find someone (they don't have to know anything about cars, but tech-savviness makes things easier) to help you. Not only does it give peace of mind to know the person in the passenger seat is keeping a watchful eye on knock counts, it is extremely helpful to have someone telling you which histogram cells are missing sample counts, allowing you to be sure to get a filled-out datalog (you'd be surprised at how few cells you spend most of your time in in everyday driving).
• Building on the above point, shoot for a minimum of 10 hits in each BLM cell. Even in constant conditions, your BLM will wander some, so you need a large enough sample count to minimize standard deviation. I'd drive a 20-min loop to get enough data. Figure out a route where there's minimal traffic so you can slow down and speed up at will (which you'll learn to be doing constantly to get all the BLM cells) and where you can put the hammer down safely.
• The 1227727 ECM is just a weatherproof version of the 1227730 ECM. Most resources available are on the 1227730; know that they are completely applicable.
• Some of the older resources will talk a lot about different hex addresses in your BIN--ignore this. Before the wonder that is TunerPro RT, people had to decipher the machine language to tune their cars. Now we can do it with easy-to-use tables in TunerPro. (Note that the general tuning info given by these sources still applies, just not the hex stuff.)

I did my best to avoid any misinformation, but if anyone notices anything incorrect in my summary, please let me know.
If any questions remain after reading through those links above, I'd love to help to the best of my ability.

WOW, thank for taking the time to type up all the great information! I'm using the Moates chip and adapter. I don't have any cat converters and EGR is deleted. I'll try the trimalyzer, maybe it will help. I have a lot of engine blowby and use about a quart of oil every 1000 miles, so I don't expect a lot. looking to learn for the 383 upgrade soon to come. I'll using the stock intake and heads, but larger cam. Did you upgrade your injectors and TB? Thanks again for all your help and suggestions.

No problem, glad to help!

I'm not sure your reasons for disabling EGR, but I'd actually recommend you keep it. It may be labeled as emissions gear, but in tuning I've learned it's actually a pretty neat device. By allowing inert exhaust to re-enter the cylinder, it essentially decreases engine displacement at part throttle--helping you out with mileage--but once you step on it the flow of exhaust gets cut off, removing any effect on power. As shown below, the EGR % Duty Cycle table is tunable however you want. I kept the stock configuration, which keeps the valve shut at idle and 100kPa load (full throttle).
Attachment 11943

Just curious, what cam are you using?

Didn't upgrade my injectors, but I have a (I think) 52mm throttle body. However, I only got a bigger throttle body because my original one was getting worn-out. As I understand it, the stock 22lb injectors are good to at least 300hp (no way you'll get there with just a cammed 383), and I believe I read the stock 48mm (I think that's the size?) throttle body can support a good amount more than that.