LBZ tuning criticism

[g3rdnx]

I’ve been researching and playing with a V2 on my CCSB LBZ. It’s a stock truck besides downpipe and exhaust. The truck has plenty of power already for pulling my camper and trailer. I’ve played with adding more fuel, boost, etc. The power was fun, but I also have a 17yo son that loves to drive it and for now I think keeping components safe is a priority. I’ve been putting together a “stock optimized” tune for better fuel economy and maybe a little power through timing. It runs good but I have nothing to reference it against other than stock.

There is a lot of tuning info out there and I’ve been through a ton of it in the last few months but I’m finding out some of the info is hacky and maybe not the best advice. Honestly, the Starlite manual is the only thing I feel that I can trust, but it’s mainly just basic info and doesn’t go very deep.

Are there any experienced tuners out there that would take a look for some constructive criticism? It would be a huge help and much appreciated.

Not sure if it would be better to upload here or DM the tune?

Thanks, Marty

 

[2004LB7]

Post the tune up. You might have to zip it so the forum will accept the file. There are already a ton of people's tunes posted around so no need to be concerned about posting it in the open. As long as it's yours and not something you purchase then it won't be a problem

To help others, does the truck have any modifications does other then tuning? Lift pump? Larger injectors? Turbo? Or more importantly a built transmission?

If you don't have anything done to the transmission to help it hold more power then I'd keep the power levels conservative.

What's your power goals? Is it a daily driver and want mpg? Towing? Other?

 

[g3rdnx]

The truck is all stock, except for downpipe, up pipe, and 4” exhaust. We use the truck for pulling our camper, long distance family trips, grocery store runs and occasionally pulling a car trailer. Nothing too crazy. I’m mainly wanting the best fuel economy I can manage without adding a ton of power. What I have changed in the tune, mainly consists of the following.

Timing calculator to make timing 70% in cruise range and 50% in all other areas.

Less boost in cruising ranges

Turbo brake

Rumble idle

Smoothing of miscellaneous tables

Here’s a Dropbox link for the tune

Dropbox

www.dropbox.com

 

[2004LB7]

What do you have for cruising range for loaded vs unloaded? I found no benefits for reduced boost while towing. Infact it made things worse for the most part. Less power ment more throttle and more smoke and fuel use. Less efficient burn. So if you have your RPMs for T/H set right you can have it always at a higher RPM and fuel load that keeps boost up. And low fuel/lower RPM for unloaded driving. I actually didn't find any benefits to lower boost even unload. But it did make it less responsive. I eventually gave up on trying that strategy and made it responsive instead

I'd be interesting to see if that helps you with MPG. In theory it should. But didn't for me

 

[g3rdnx]

I logged a couple drives and my unloaded mm3 from 1400-2100rpm is 20-40. I don’t care so much about saving fuel when towing. I was worried that if I had less boost while loaded it would cause higher EGTs. I might go back to stock boost tables if you think it really won’t make a big enough difference to justify the responsiveness loss. Is 70/30 timing safe to run across the board or should I keep it just in the unloaded cruise range?

 

[2004LB7]

You kinda have to see how it runs to know if it's going to work. I've tried the 70/30 for cruising and gained maybe a half an mpg but gained some extra noise/knocking that I didn't care for. Pulled a little timing out and all was good. But each truck seems to run a little different so you'll have to give it a try to know.

There are some areas like idle and low mm3 areas that you can probably get away with 90/10 or even 100% but as more fuel is needed you have to balance it with cylinder pressure. Since most of use can't measure cylinder pressure you generally error on the side of caution. 50/50 is pretty safe for all mm3 levels. And 70/30 for select areas. You'll probably be fine if you don't overdo the mm3

 

[blueshift]

I know exactly what you're talking about regarding the advice and opinions you'll find on the web regarding diesel tuning. I've been getting back into tuning my truck a lot the last 6 months or so, and I've come across plenty of good and bad advice.

With regards to fuel mileage tuning, some of my findings have been dead opposite of some advice I'd found online (often from some top-name tuners, or at least were at the time).

I'd share some of my findings with my LB7, but I'm not confident it would be valid for your LBZ. For example, I think stock timing on the LB7 is north of 70% BTD at cruising, and I think I'm up around 110%-150% depending on mm3. I can't tell any difference in noise from the stock timing. Also keep in mind that, because the pulsewidths are so low in these areas, a single degree can have a pretty drastic affect on the % BTD/ATD. My point being that a degree or two isn't likely to hurt anything at low load, even though it's a big swing in the %.

Here's my universal suggestion though: datalog, datalog, datalog.

I'll explain, but first a disclaimer: it's VERY difficult to discern tuning changes in mpg calculations from tanks of fuel (even hand calculated). There's too many variables over that amount of time that can easily be more significant than your tuning changes (i.e. weather differences, average speed, driving habits, headwinds, exactly when the pump clicks off, etc.).

So what I did (and am still doing on a couple things) is load up a DSP5 program with changes to ONE variable to each of the 5 programs. Say timing, if you're at 70% TDC now, leave that tune in position 3, and then add and subtract a degree or two from the other 4 programs. So it'd look like -2, -1, 0, +1, +2 degrees through the 5 programs.

First, do some typical driving and switch between the tunes. Make note of any drivability changes like more/less smoke, more/less responsiveness, more/less noise, EGTS, etc. Of course, stop driving it if a particular tune sounds excessively bad.

Now find a stretch of road that you can drive for a solid 4-5 minutes with the cruise control set, preferably without too much traffic. Some hills are good, because you'll also get some datapoints at different loads.

Find markers for start and stop points (milepost, exit sign, whatever), and you basically want to datalog between those exact points with the cruise set (same speed for every test, obviously). I always datalog the same tune in both directions, just to take winds and net elevation change out of the equation. In general, try to be as consistent as possible between tests. And if possible, run the tests back-to-back to minimize the effect the weather conditions may have.

Repeat this test for all 5 tune positions. At a minimum, I datalog main rate, main timing, main pulse, desired and actual fuel pressure, pilot rate, pilot timing, speed, rpm, and ECT.

Then compare the average values for each log. A decrease in main rate typically means an increase in fuel mileage. I'll also look at the pulsewidth / fuel pressure combo (since those two combined essentially equals fuel flow).

Take note of TRENDS and percent changes, and not necessarily exact values. In the timing example above, if the main rate increases a bit at -1, increases a bit more at -2, while decreases a bit at +1, and decreases a bit more +2, then that generally indicates a bit more timing will help your fuel mileage. If it increases at -1, -2, but doesn't change at +1, +2, then adding timing isn't likely to help.

Once you've done that with a few different variables, combine them all together and run a tank or two of fuel through, and hand calculate the mileage. See if it's noticeably better.

In my testing, some things make very little, if any, difference (fuel pressure), some things make a slight but noticeable difference (main timing), and some things can make a huge difference (pilot timing).

Hope this helps. It sounds daunting, but it's really not that bad. There's really only a few variables that affect the fuel mileage, and a couple hours datalogging can offer a lot of clarity.

Oh, last point: people lie about their fuel mileage. Don't be disappointed if you're not getting 25 mpg, because the folks online claiming it aren't getting it either. At 70 mph on a stock height, stock tire CCSB, you won't do better than 18-20 on a consistent basis. So if you're in that range, you've fairly well optimized it, and more time spent on it isn't likely to net significant gains.

 

[g3rdnx]

I know exactly what you're talking about regarding the advice and opinions you'll find on the web regarding diesel tuning. I've been getting back into tuning my truck a lot the last 6 months or so, and I've come across plenty of good and bad advice.

With regards to fuel mileage tuning, some of my findings have been dead opposite of some advice I'd found online (often from some top-name tuners, or at least were at the time).

I'd share some of my findings with my LB7, but I'm not confident it would be valid for your LBZ. For example, I think stock timing on the LB7 is north of 70% BTD at cruising, and I think I'm up around 110%-150% depending on mm3. I can't tell any difference in noise from the stock timing. Also keep in mind that, because the pulsewidths are so low in these areas, a single degree can have a pretty drastic affect on the % BTD/ATD. My point being that a degree or two isn't likely to hurt anything at low load, even though it's a big swing in the %.

Here's my universal suggestion though: datalog, datalog, datalog.

I'll explain, but first a disclaimer: it's VERY difficult to discern tuning changes in mpg calculations from tanks of fuel (even hand calculated). There's too many variables over that amount of time that can easily be more significant than your tuning changes (i.e. weather differences, average speed, driving habits, headwinds, exactly when the pump clicks off, etc.).

So what I did (and am still doing on a couple things) is load up a DSP5 program with changes to ONE variable to each of the 5 programs. Say timing, if you're at 70% TDC now, leave that tune in position 3, and then add and subtract a degree or two from the other 4 programs. So it'd look like -2, -1, 0, +1, +2 degrees through the 5 programs.

First, do some typical driving and switch between the tunes. Make note of any drivability changes like more/less smoke, more/less responsiveness, more/less noise, EGTS, etc. Of course, stop driving it if a particular tune sounds excessively bad.

Now find a stretch of road that you can drive for a solid 4-5 minutes with the cruise control set, preferably without too much traffic. Some hills are good, because you'll also get some datapoints at different loads.

Find markers for start and stop points (milepost, exit sign, whatever), and you basically want to datalog between those exact points with the cruise set (same speed for every test, obviously). I always datalog the same tune in both directions, just to take winds and net elevation change out of the equation. In general, try to be as consistent as possible between tests. And if possible, run the tests back-to-back to minimize the effect the weather conditions may have.

Repeat this test for all 5 tune positions. At a minimum, I datalog main rate, main timing, main pulse, desired and actual fuel pressure, pilot rate, pilot timing, speed, rpm, and ECT.

Then compare the average values for each log. A decrease in main rate typically means an increase in fuel mileage. I'll also look at the pulsewidth / fuel pressure combo (since those two combined essentially equals fuel flow).

Take note of TRENDS and percent changes, and not necessarily exact values. In the timing example above, if the main rate increases a bit at -1, increases a bit more at -2, while decreases a bit at +1, and decreases a bit more +2, then that generally indicates a bit more timing will help your fuel mileage. If it increases at -1, -2, but doesn't change at +1, +2, then adding timing isn't likely to help.

Once you've done that with a few different variables, combine them all together and run a tank or two of fuel through, and hand calculate the mileage. See if it's noticeably better.

In my testing, some things make very little, if any, difference (fuel pressure), some things make a slight but noticeable difference (main timing), and some things can make a huge difference (pilot timing).

Hope this helps. It sounds daunting, but it's really not that bad. There's really only a few variables that affect the fuel mileage, and a couple hours datalogging can offer a lot of clarity.

Oh, last point: people lie about their fuel mileage. Don't be disappointed if you're not getting 25 mpg, because the folks online claiming it aren't getting it either. At 70 mph on a stock height, stock tire CCSB, you won't do better than 18-20 on a consistent basis. So if you're in that range, you've fairly well optimized it, and more time spent on it isn't likely to net significant gains.

Excellent information. Thank you so much for commenting. I have been logging but using your method makes way more sense.

I have some questions on pilot timing. Does advancing the pilot injection help or hinder mileage? Seems like it would hinder mileage but I’m just trying to theorize in my head.

Would a bigger pilot injection smooth out the burn and also add a touch more power?

Thanks for your time

 

[2004LB7]

From reading Fingers posts on it, too much pilot can be bad along with too little. Balancing act. On the LB7 I've found a touch more pilot mm3 seems to make the engine quieter. I use 0.25-0.5 mm3. I think for the later years they are pretty good stock and haven't seen any gains either way

Fingers also mentions that ideally you want the pilot as close to the main as possible without overlapping. The pilot needs to finish burning before the main is injected. This would give the most efficient burn and best MPG. If the two overlap you can get knocking as it acts like a larger shot with advanced timing

It would probably take a ton of revisions to the tune but slowly closing the gap between them until they knock, overlap or other issues and then increase it until it stops.

 

[Dozerboy]

From reading Fingers posts on it, too much pilot can be bad along with too little. Balancing act. On the LB7 I've found a touch more pilot mm3 seems to make the engine quieter. I use 0.25-0.5 mm3. I think for the later years they are pretty good stock and haven't seen any gains either way

Fingers also mentions that ideally you want the pilot as close to the main as possible without overlapping. The pilot needs to finish burning before the main is injected. This would give the most efficient burn and best MPG. If the two overlap you can get knocking as it acts like a larger shot with advanced timing

It would probably take a ton of revisions to the tune but slowly closing the gap between them until they knock, overlap or other issues and then increase it until it stops.

Pilot injection/ timing in relation to main injection is something I haven't wrapped my head around yet. How are you determining how close they are? Are you doing the math cell by cell figuring in pilot US or just taking main injection timing and adding X* to it for the pilot timing?

 

[2004LB7]

Pilot injection/ timing in relation to main injection is something I haven't wrapped my head around yet. How are you determining how close they are? Are you doing the math cell by cell figuring in pilot US or just taking main injection timing and adding X* to it for the pilot timing?

Yes. Haha. Well actually I've made Excel sheets to try and calculate degrees before main and make sense of it all. But without a dyno to run a steady state load and rpm it would be very hard to see any gain. So I've pretty much given up on trying to tune pilot pulses.

But if following fingers advise, it seems you can slowly reduce the time between the pilot and main unit you hear extra noise thand add timing back in until it stops. You'd probably have to do it in blocks of cells that run similar to each other. One cell at a time would probably take forever and the engine really stays in one cell

 

[Dozerboy]

Yes. Haha. Well actually I've made Excel sheets to try and calculate degrees before main and make sense of it all. But without a dyno to run a steady state load and rpm it would be very hard to see any gain. So I've pretty much given up on trying to tune pilot pulses.

But if following fingers advise, it seems you can slowly reduce the time between the pilot and main unit you hear extra noise thand add timing back in until it stops. You'd probably have to do it in blocks of cells that run similar to each other. One cell at a time would probably take forever and the engine really stays in one cell

That's kind of what I figure. I'm more of a wing it kind of guy. I messed with some when I added my 100overs and lost interest over getting the other charges necessary made.

The LBZ is easier I would think because your just adjusting the time before main injection event.

On the older trucks don't you actually have to set the timing in degrees?

 

[2004LB7]

That's kind of what I figure. I'm more of a wing it kind of guy. I messed with some when I added my 100overs and lost interest over getting the other charges necessary made.

The LBZ is easier I would think because your just adjusting the time before main injection event.

On the older trucks don't you actually have to set the timing in degrees?

Yeah, LB7 is done with degrees before main. Actually pretty easy as it uses the same measuring units as the main. So if you want 2 degrees, just put 2 in it. Instead of having to calculate the crank angle, time or anything else.

Only gets crazy when you are trying to keep things consistent because 2 degrees is a lot less time at 2k rpms then it is at 800. Probably the reason for the change in the later years

 

[blueshift]

Excellent information. Thank you so much for commenting. I have been logging but using your method makes way more sense.

I have some questions on pilot timing. Does advancing the pilot injection help or hinder mileage? Seems like it would hinder mileage but I’m just trying to theorize in my head.

Would a bigger pilot injection smooth out the burn and also add a touch more power?

Thanks for your time

I'll preface this with the comment that my situation with tweaking the pilot timing is a bit unique. I replaced my stock injectors with 45% over SAC-style injectors, which are both larger and have a different style nozzle from the stock LB7 ones, so a pretty drastic total change from stock. I never messed with the pilot until after doing this, so I can't really speak to how accurate / optimized the factory values are (for my truck, or other generations).

I have also realized that pilot parameters are a bit of black magic. I have a fairly deep engineering background and consider myself reasonably adept at understanding the theory behind most of this tuning stuff, but I absolutely do not have a full grasp on exactly how the pilot timing and quantity should work.

The one thing I have determined fairly conclusively is that it's the time that matters, not crank-angle-degrees (CAD). Again, I'm not familiar with the LBZ and newer stuff, but if you convert the LB7 and LLY timing tables from degrees to microseconds, it's very obvious what GM was doing. The time is pretty consistent (exact in the LLY's case) across the rpm ranges, with only small changes at low rpm and load.

Before I put the 45% SAC injectors in, I regularly got 18-19 mpg cruising at 70, sometimes as high as 20 in perfect conditions. After I put them in, I tweaked the timing and fuel pressures (as per the datalogging procedures I mentioned in my last post), but I was only getting 16-17 mpg. Even with near perfect conditions, it was only maybe 17.5, so I knew something was still off (or the injectors just sucked, which seemed unlikely).

I saw several posts online recommending switching to the LLY pilot timing maps, which is logical since the stock LLY injectors are also SAC nozzles. It has the pilot about 100-300 us further advanced / separated from the main at cruising. I did this and instantly picked up a solid 1.5+ mpg. The last 3 tanks have all been over 19 mpg.

I wrote a DSP5 program with the pilot timing at -400, -200, 0, +200, +400 microseconds of pilot timing, with '0' being the stock LLY map. I have not yet had the time to datalog like I'd mentioned earlier, but I've done a couple less-scientific tests.

- The -200 (close to stock LB7 timing) is very slightly quieter, but with moderate increases in main mm3 (which matches my fuel mileage observations from the LB7 to LLY pilot timing).

- The -400 is slightly quieter still, but with a substantial increase in mm3. It was significant enough that I can actually feel the truck surge forward a bit when changing from -400 to 0.

- The +200 sounds about the same, and shows a very slight decrease in main mm3.

- The +400 is somewhat louder / more rattle-y, and shows a tiny bit less mm3.

I haven't been scientific enough to determine if there's actually a mileage gain at +200 to +400, but if there is, it's small enough that I'll probably choose the quieter tune over a fraction of a mpg gain. Definitely more logging that I need to do.

I also still need to play with pilot quantity. Unfortunately, this isn't a DSP5 parameter with the LB7. I can still log the mileage with single tunes, but it's MUCH easier to judge noise, driveability, smoke, etc. with the instant switch. It IS a DSP2 parameter, I just need to get a switch for it.

It's logical that I'd need to pull some pulse / quantity since I have bigger injectors, but I pulled 15% on an experimental tune a while back, and both part-throttle and WOT smoke got noticeably worse. Opposite what I expected. I've also read some pretty convincing posts on the Cummins forums of less noise, smoke, and fuel consumption with a larger pilot. Again, opposite what I'd think, so I still have some testing to do there.

The bottom line of all this is that it is definitely possible to see increases via pilot tuning, depending on your setup. Then again, it's definitely possible that you won't.

Hence the 'black magic' comment. : )