Val Snyder

While you're kinda close to the subject, is there a prefered replacement pump (in tank) for TBI application providing more delivery at higher pressure than the standard replacement BW ? Hopefully better quality as well. Thanks, Val Also,Does any one have comment on running an inline to boost pressure after a Holley 60 gph (no pump in tank) for conversion to TBI?

DM91RS

HEY............Rbob's got a Web site

liquidh8

Well, I've been doing the VE learn for a few days now. Seems like every drive the table needs to be changed again, and again. I am using the LM-1 for the VE learn now. My VE table doesn't look pretty. Do yous guys think this could stay like this?? Some cells are say 54, and the next up in the rpm column, same kpa row, would be 69. If I change it to make it smoother, It is changed back again in the VE learn.

On a side note. Been tuning my buddies silverado ss (dyno'ed over 500 RWHP) with HP tuners, the VE table in that truck is super smooth, but it also has long term fuel trim to make changes if it is a little out.

RBob

iTrader: (1)

For in tank a Walbro 190 or a GM TPI pump are decent replacements. If the fuel pressure is above 20 psi (or so) and the engine is making power then a Walbro 225 is a good choice. Be aware that at lower pressure with a stock regulator the 225 can overwhelm the regulator.

I have a couple installs with inline pumps. The hi-po setup uses an MSD pump back by the tank. It is level with the bottom of the tank so it isn't trying to lift the fuel out. Baffling, although added, is inadequate. Car is a capable corner carver and will easily uncover the pickup in sustained curves.

Another with a stock LG4 converted to TBI has one of the F*ord/Bosch pumps inline, get this, under the hood. It is mounted to the inner fender with the input pulling from the fuel line from the tank. Then through an EFI filter to the TBI. I skipped using the block mounted pump altogether as it added a lot of heat to the fuel.

RBob.

RBob

iTrader: (1)

If the changes are in the same direction then I would keep at it. If the corrections are bouncing back and forth I would look at how large the corrections are. Having an IAT helps in compensating for temperature changes.

I don't buy into the smooth VE idea any more. It sounds good, but there seems to be more to it then that. Part of the issue is AE and DE. In the WUD VE Learn they are taken into account. And, as you have found out, smoothing the table and the VE Learn changes it right back.

I think part of the OEMs having such smooth VE tables has to do with items such as cam, intake and exhaust design.

RBob.

liquidh8

OK, I have a IAT. I am getting to the point now where I think they are bouncing back and forth. I was figuring maybe If the are within +/-2 then I would leave them, then do a BLM learn. I don't think I have enough DE, does the WB mess that up? Because I locked the BLM's at 128? In the lower map areas (30 - 40),VE learn wan't to take away -10 to -12. But then it goes right back up and sometimes add's. I guess I am close. It's had to even get the learn into the higher MAP and RPM cells, Guess guesstimation is what happens there.

RBob

iTrader: (1)

Yes, it is coming along. Have screen shots of the WUD screen displays. Along with the full installation guide up. Then a fair amount of information on the EBL setup and code.

I have a lot that I want to put on the site. I need to put a plan together as to what and how it all links together. Then get going on it.

RBob.

RBob

iTrader: (1)

As for DE you can check that with the WB value. Upon a closing throttle the AFR usually dips rich then goes back up. Too much DE will be seen as going too lean.

The lower MAP areas (30 - 40) can sometimes be difficult. The reason is that the PW becomes small. Then a small change in PW makes a greater change in fueling. If you find the LOW MAP areas have a VE% greater the the higher MAP area next to it, the injector bias can be increased.

RBob.

TierAngst

Will have heads, cam, bored TB, intake manifold, headers... basically everything by the end of the year. So I guess I'll be needing all that stuff.

Is it better to go with a BBC TB with ultimate TBI mods or something like...
eBay Motors: 454/350 GM TBI 50mm SBC tbody-750cfm
?

sbcrules

I tried to keep my VE tables smooth like the OEM bins.
I built my engine based on GM design (350 ramjet cam, vortec heads with a 383 displacement. I thought that the 350 ramjet no smog VE tables would be a good starting point.
I would make a run with the VE learn, adjust the cells around the changed cell and try to keep them smooth.
They were constantly changing back. I also noticed that when I smoothed them out it didn't run as well.
I have made several VE learn runs this week and I let the VE learm decide what the tables should look like, now the engine has a lot better Power band. My 4l60 is slipping from the tourque.
I would think that the gas/air flow would be smooth between cells but that does not seem to be the case.
When I first looked at the bin that was included (ebl.bin) I noticed that the VE tables were anything but smooth, thought that this was done to make us do our own bins. Guess not
Anyway I have done 50+ bins (3 differnt engine mods) and it keeps getting stonger and smoother. 22 PSI 80# inj. Soon will add the wideband

I have no complaints.
Thanks RBob

shawngenung

Rbob (or JPrevost)-- is the learn VE function just an automated way of doing the traditional excel BLM divide and multiply, or are there some routines in there to smooth or weed out strange values?

RBob

iTrader: (1)

The VE Learn in the WUD is more sophisticated then the traditional excel data crunch. With the higher data rate a greater amount of data is collected along with being more precise. Items such as AE, DE, delta TPS and MAP are taken into account. Several other factors are taken into account before a BLM or WB value is collected.

With the high data rate the MAP, RPM, BLM/WB-AFR values are an instant snapshot. Separated by no more then 50 msec. Where the standard 160 baud ALDL has a separation of 1.6 seconds. This provides greater precision in the group of values.

RBob.

shawngenung

awesome -- thats what I was hoping someine would/had put together -- continually amazed

liquidh8

Ok guys, I've had the EBL for some, SOOO MUCH better for me and tuning is awsome. I've had various differnt issues with the WB, IAT, yadda yadda. Other than the tranny being toast, and my exhuast pipes hitting the floor giving me some false knock counts ( I think), I'm getting to the point where I have finer tuning to do. Now the question. I seem to have the VE set up in all the cells I can hit, so much so that ever time I do a WB/VE learn they are either all 0, or not more than +/- 2. I think the AE is close in some areas. But my issue is, when I Stomp on the gas fast, or nail it from a dead stomp, the WB shows a lean condition, 19:1 or so, and it boggs bad. But if I keep increasing the AE tables, whenever I am just driving along, when AE comes in it drops to 10:1. I seem to be unable to find the balance here.

Another question, In the upper MAP/RPM areas in the VE table, If I lover that will that lower my PE AFR too? I can't hit these cells with the VE learn so I am just guessing. They are pretty high now, around 96. Any comments??

Fast355

iTrader: (2)

If you are still running the VAFPR, you may need to adjust the response rate in the ECM.

RBob

iTrader: (1)

Last question first: to tune the higher RPM/MAP areas adjust the VE values so that the WB reported AFR matches the commanded AFR. Use the Analysis screen and scroll to that area of the log. The AFR column is the commanded AFR, and the WB column is the WB reported AFR.

For the AE stuff, most likely you need to increase it at the high delta TPS% values. Go to the Analysis screen and create a dump file. Leave the 'single header' box unchecked. This will create a dump file in the same directory and same prefix name as the log file with a '.ebl' extension.

Open this is Notepad or possibly Wordpad. There will be a lot of data, scroll through it to find the thorttle romp. The last several columns of data are the AE parameters. The dTPS and dMAP columns show the actual delta values that are used for the table lookups.

This shows which rows are being used in the Delta TPS AE table (using dTPS), and in the Delta MAP AE table (using dMAP).

The values next to each of the dMAP & dTPS is the AE in msec. There is also a final AE column which is after the RPM & coolant corrections, and is the actual amount delivered.

While in that area of the log check the injector DC%. At mid-RPM ranges it is easy to run them static from AE.

RBob.

liquidh8

Actually, I am. I left it at the setting RBob had, 37.stuff. I was moving it around up to fifty and down to 20, But I'm not sure it made a difference. Though I will try it again when I get home. Where do you think I would see the change in the filter for the VAPFR?? In the AFR of the WB? Or maybe in the WUD somewhere. I guess what I am getting at is there another way to see what my changes are affecting other than "feel"?
----------
Thanx RBob, You had said something similar to me or someone else before, just not as complete in one post. Thanx, I'll have to try that one the way home. Will changing the BPC rate filter affect the VE/AE tables?? If I were to change that filter rate %, would it throw the VE/Ae tables out of wack??

RBob

iTrader: (1)

Changes in the BPC has no affect on AE. It does affect the regular fueling PW. So changing the BPC filter % can affect the VE table. This will be seen in the horizontal direction, as the MAP changes.

I've been thinking, with a fuel pressure transducer, one could track the fuel pressure change along with the change in the BPC. This should be enough to get the filter% close to the correct value.

RBob.

liquidh8

If one was looking at the fuel pressure in the wud using the FP Xduxer, how would you compare the two?? Would you look at the injector PW to see your BPC filter changes? IOW, If the fuel pressure is rising fast, then you would make the BPC filter higher, say 80%, and see if the PW tracks the same speed as the FP increase??

RBob

iTrader: (1)

Compare the fuel pressure to the BPC. They would both be available in the Analysis data dump file. Not sure if the BPC is in there now, if not, easy enough to add.

Can do step response tests: at normal cruise hit the pedal about 25% further down and hold it for two or three seconds. In the data log can then see if the BPC lags or leads the change in fuel pressure.

RBob.

liquidh8

I'll have to check to see if the BPC is in there now, I don't think so though. I would also need to see the fuel pressure on the WUD. I was holding off on the FP gauge, but now I see a valid need for it, (Autometer w/xducer) Just need you to add the FP display and the BPC in the log file, :hint: :hint:

liquidh8

You mean vertical, left/right, with the the MAP. "cause RPM is horizontal, up/down.

RBob

iTrader: (1)

Yes, left/right looking at the VE table.

RBob.

liquidh8

ok, kool. On the way home from work today I tried moving around the BPC filter %. Obviously I have it tuned for where it is now, 37.5%, 'cause when I move it either way by more that 10% it runs worst. But since I can't log FP AND the PW at the same time, I guess I am p%$^%$g up a tree. I'll make due with the current tune, just get the VE table in better order @ WOT. I know they have to be off cause if I look at the log from a WOT run, the WB gives me a 9.8 AFR from like 1500 RPM @ 100KPA, to 3800 RPM @ 100 KPA. I gues I should adjust those cells first and scale the othe KPA/RPM cells DOWN from there. What do you guys think??

Val Snyder

IAC needs to be set to 0 before logging for Nalpha tables, will AB on the ALDL (key on) set IAC to 0?,then disconnect IAC? Does TPS voltage need to be reset, as I am sure that the butterflies will need to be cracked further open to get an idle with 0 IAC?, Thanks, Val

TierAngst

My intake is still making the annoying whistle, I've adjusted the IAC steps and the idle screw, what now?

Dominic Sorresso

Guys,

Thought I would throw my $.02 in on tuning with EBL. RBob's stuff has made it much easier to deal with less understood (at least for me) parameters such as the PID controls. I've learned the difference the Prop Gain Controls can make. What I have never taken into account is how these controls really come into play when you're running high FP. Since many of us have pushed beyond the boundries of the TBI envelope, the stock settings for things like Prop Gain can cause havoc. The high FP magnifies issues in these controls. Idle surging, trailing throttle hiccups and surges I have found cause in the PID controls. The Log Analysis from RBob's stuff makes this area and AE settings so much more visible than ever before.

RBob

iTrader: (1)

Is the whistle from the IAC? It may be from a vacuum leak.

With a warm engine and the trans in gear (auto only), A/C off, check the IAC step count. It should be around 5-10 steps. I would be surprised if the IAC would whistle at that low a step count.

On a cold start with the IAC opened up, it does make a lot of noise. A full air cleaner supresses it a lot. Where an open element does not.

I've never thought of the sound from an open IAC as a whistle. More like a low howling sound.

RBob.

RBob

iTrader: (1)

To set up data logging for n-alpha calibration the butterflies need to be left alone. They need to be in the same position as they will be once the calibration is complete and the IAC re-connected.

It may idle a little low but it shoudn't be too bad. Warm the engine up first, then close the IAC and data log.

Yes to shorting AB to close the IAC. Key off, short A & B, then key-on/engine-off. Wait 20-30 seconds and disconnect the IAC. turn the keyy off and remove the AB jumper. Set up data log and drive around. Should also do some traffic crawl. An office park after hours is good for this.

The data from that log will allow the MAP from TPS & RPM table to be populated.

Another data log with the IAC active will provide the data to populate the IAC airflow values.

To get the IAC squared away again: key-off, reconnect the IAC. Start engine and let run for 10 seconds (can use foot on gas pedal if required). Then key-off, wait 10 seconds. IAC is now reset.

The TPS voltage is not critical. It is self adjusting within the code. The max is about 1 volt and the minimum is about 0.2 volts. These are calibration parameters and can be slightly different from what I listed.

If the idle screw setting is changed the n-alpha table calibration will also need to be updated.

RBob.

Dominic Sorresso

The whistle is used by some and is supposed to be an indicator that the Min Air has been set properly.

TierAngst

I've setup the step counts and it doesn't whistle at idle, just in the 1400-1600 rpm range.

RBob

iTrader: (1)

If it comes and goes with chaning the go-pedal position then play with the Throttle Follower stuff. Try setting the gain a little higher or lower, and/or changing the max throttle follower steps.

If it whistles at 1400-1600 RPM no matter the throttle position I'd be more inclined to think that it is a vacuum leak.

RBob.

DM91RS

Mine started the whistle after setting the idle as suggested. It only does it at idle and is only noticed when I have the breather off. Idles with a IAC of 10 or less.

It's not a problem with me, just a curiosity. My idle is better than I have ever had with a modded motor so I am very happy with the setup!

By the way what are the PID controls? The most noticeable driveability issue that I have is the lack of a crisp throttle blip for downshifting. Any suggestions for that?

DM

Fast355

iTrader: (2)

I was initially having trouble with an automatic that would bog between shifts under light throttle acceleration. Not really bad, but enough to be noticeable. I had too much MAP AE. I shifted alot of it to the TPS and it runs much better.

DM91RS

Thanks............I will do some logs just to check out what is going on with the AE in that case.

Val Snyder

Did a log as discussed, log was done with n-a not set, then set in both learn from and to during playback, no change to table that I could find, no doubt I botched something along the way. Should I have enabled n-a during the log? I may be looking under the wrong rug for n-a to solve the light throttle surge anyhow, problem happens when duty cycle is around 5%, 1.5 ms. with EBL going clear up to 2000 rpm, lighter as rpm increases. You guys are making the rusty wheels drift to the PID paper department,(more hints,Please). An indicator as to progress thru the playback (like a bar)in addition to the time would be helpful or a showing of total log time beside the progression window. W hile in the wish list mode (maybe it is already there and I haven't stumbled on to it yet) how bout an indicator for Async or Sync? Thanks for all the time and help, Val

TierAngst

Found something interesting on my last datalog, I wanted to hit some of the upper-ish rpm areas for tuning so I drove around in third gear. I ended up getting better mileage at 80mph in third vs OD by about 7 mpg, I was consistantly getting around 25-30 at around 3600rpm. That was running rich to boot, my blm was sitting right around 111. Also found that the whistle comes back up in that rpm range as well, seems like the two areas where I get the best mileage it whistles.

Dominic Sorresso

Val,

Yes to PID controls. Take a look at the Proportional Gains tables. Read RBob's paper carefully, and follow the logic of the Proportional Gains controls in your bin. I dropped mine by 300%. I have been chasing a light throttle surge, particularly in the 16-2000rpm range. Happens when I have a trailing throttle, typically on a slight decline. I get a "stumble". (DFCO is fine) It also shows up while in neutral and raising the idle speed with the accelerator. In the 14-1600rpm range it will surge with the MAP at 30-40kPA. Looking at the WB, I can see it going from 14+ to 19/20 AFR and then back. The log will show the PW in the .8-.9msec range. I have played with the DE and DFCO, but those don't show up as affecting anything in the log. I was setting Min Air Idle with IACs closed, and started to get the stumble with motor just idling. That's when I thought it might be the Prop Gain. The values in the table don't account for FP at 20psi through 80# injectors. I still have some work to do, but dropping the Prop Gain and the Multiplier, the stumble and surging have significantly diminished.
You may also want to disable Asynch mode by setting min Synch PW to 0 .

RBob

iTrader: (1)

The n-alpha isn't auto learn. From the log taken with the IAC closed the MAP vs RPM & TPS table needs to be filled in. Then from a log with the IAC enabled, the IAC flow and RPM multiplier tbales can be filled in (from the difference of the no IAC and IAC logs).

Then enable the n-alpha mode via the option bit. I have been working on an auto calibation for this. Have made some good headway but not yet completed. If you'd like email me the zipped log of the closed IAC. I'll run it through what I have and send the main MAP vs TPS & RPM data back to you.


For the light surging it may be the proportional gains are too high. This is one of the areas Dominic is referring to. The proportional gains are based on a PW adder/subtracter. So any increase in fuel pressure or injector size increases the amount of gain. Usually beyond what is reqiured.

The two tables that control the gain are:

PRP - Gain vs O2 Err
PRP - Gain Multiplier vs Airflow.

To see if excess gain is the issue halve the values in each of these tables.

RBob.

P.S. saved the wish list ideas on the upgrade list.

{edit:

P.P.S. was typing while Dominic was posting. . .

}

Dominic Sorresso

When you drop the PG, you may notice that where the Learn VE was initially taking fuel out, it may now be putting fuel back in. You'll see th pervasive effect of the Prop Gain when you do the next Learn VE with the PG reduced.

Val Snyder

Thanks, Fellers, I'll get right on it. I'm thick enough that it takes about 10 readings of "the paper "for things to start to fizz, so THANKS for gettin my nose in the right direction, now if I can just keep it out of the door.Thanks again , Val

Dominic Sorresso

I read RBob's paper 3 years ago. I wrote RBob just last week that the "light bulb" finally lit up. Albeit dimly.
So don't feel too bad.

Val Snyder

Tried the 1/2 value, and it helped quite a bit so of course took out some more and the light throttle surge is gone as far as I can tell, However, (ain't that always the way) the closed throttle lurch still persists(couldn't beat it with the stock 747 after two years of forehead whacking). Set the A sync to 0, ha! after reviewing Dominic's post I see it should have been the Sync set to 0. Don't understand the relationship/function of sync vs async . More items for the wish list, some sort of a "progress" bar for log playback that could be clicked on and dragged foreward , also an indication as to total length of log. And for us that have trouble with what day of the week it was this morning ,a feature that would show which files are loaded in learn ve.Uh, I don't seem to savvy the how of additional file creation for more logs, can make new folder for each log, but it seems that a person should be able to do just another file, could be that most everyone else can , but I havn't figured out how. Thanks for the help, Val

RBob

iTrader: (1)

For different log files: when the dialog box for the DataLog file name opens, overtype the default name. Can change it to any valid filename. If no extension is used it will default to .DAT. Which if you type an extension I suggest using that.

There are issues with using a drag bar to forward through a playback. If used to skip ahead in the playback then the VE Learn, SA Retard, and trip computer data will be incorrect.

One thing you can do is to drag the playback speed bar to the right. This will cause the WUD to quickly process the data. Note that when the '<- BackUp' button is used the processing of the VE Learn, SA retard, and trip computer are suspended. Once playback catches up, they are then resumed. This keeps the data accurate.

A progress bar can be done. Along with the playback filename being displayed. Need that in the Analysis screen too.

RBob.

Val Snyder

What do you think of trying to slow MAP input? Dominic, I know that an engine can / does set up a pulse wave that can create fueling problems with a carb, depending on intake tract length and exhaust system length and # of bends, diameter+ tightness etc. My understanding is that the wave causes the air to pass the point of fueling several times,especially with ratty cams at small throttle openings, one "cure" was to drill smallish holes thru the butterflies, so it makes me wonder if you might see some relief by running more IAC in that area. Have you tried the N alpha? My persisting problem occurs when decelerating into a turn (slow in town) as engine speed drops at 0tps, I think it is idle entry,drops, then kicks up hard, (It can be overcome by putting a bit of foot into the throttle but that leads pedestrians to think I'm out to get em)should it hit stall saver to enter idle ? Happens open or closed loop,am down to just 1 or 2 IAC when fully warm. Have not tried to disable stall saver (the ole gal will want to kick my butt if it gets any easier to kill),but have messed with various settings,maybe the approach should be narrowing the stall saver range or slowing its exit, hey, how bout a flag for stallsaver active? I'll throw these brain waves out +you decide which bin to put em in, trash or maybe.Thanks, Val

RBob

iTrader: (1)

Shouldn't be bouncing off the stall saver to enter idle. Look at the log file during this time, the Analysis screen is best for this. Check items such as the IAC closing rate, O2 value, and DFCO. It may even be a too small of a PW and the fuel is dropping off. Will see the O2 drop in this case.

If you want, zip up and email me a log that covers this issue. With that I can get a better idea of what is happening.

I wouldn't disable the stall saver as it is just a symptom. Need to find the cause of the RPM drop entering idle. Once that is found fixing it becomes easy.

RBob.

Val Snyder

My logs are a mess, but I am gonna get the hang of it, I can find the areas in them and the O2 is way low(like 0 or 3 MV),then shoots up. at every occurance, pw is from 1.5 to 3, duty cycle is from 3 to 5. The iac seems to be behaving itself. I set the async transition lo to 15.25, low as it will go and still have a value , no effect, I did mess with the Map filters and got some improvement to around 25% reduction, beyond that light throttle surge returns. Same behavior with DFCO disabled.It is a Holley tbi and I suspect that there may be behavior problems with them in general. Your help is greatly appreciated in resolving this, Thank You, Val PS the Analysis tool is really great, so is the whole project.

RBob

iTrader: (1)

If the Holley has the black plastic spherical caps over top of the injectors it is one of the bad ones. They are the old style with terrible injectors. Unfortunately it will never run right.

However, it can be fit with a GM pod/regulator with GM injectors. That is one way to go with it. IIRC Holley now sells their new pod with the good injectors as an upgrade. It is a little pricey though.

Just to see if it helps, increase the VE in the areas where the O2 starts to drop off. Or, the areas right before the O2 drops off. That may help it enough to keep it from stalling.

One of the issues with the old style injectors is that they don't spray a cone of fuel. It is more of a shot straight down out of the injector. At low air flows this causes the fuel to bounce off of the butterfly and back up the TB.

The engine then leans out, and will then go rich as the fuel gathers up and finds its way into the engine.

Of course, you may have the later model Holley. And it just needs more fuel in those areas.

RBob.

liquidh8

Quick Question. Is there a way to steady my WOT PE actual wb02 readings?? I say a sceen shot of somedodies here, and it was steady. Right now when mine in in PE WOT, commanded 12.8, It jumps all over the place. Like 14.0, 12.7, 13.6, etc, in different RPM's throughout the 95/100 KPA areas. If I look at a dump log, between 2 rmp cells, for that second of the log, the AFR can changes for every tenth/second. You know what I am talking about?? Thanks guys.

RBob

iTrader: (1)

With the LM1/LC1 I believe that there is an update rate parameter. When set to a lower value it smooths or filters the ouput? For some reason I'm thinking that a setting of 10 is what most people use (IIRC it was in a thread here someplace).

RBob.