John Speake

Hi Fred,
Not a problem in your case, as you have fully remapped your LH.

FredR

Hi John,

Did a bit more experimentation this morning. I tried to plot the trajectory through the map from flooring the throttle from tickover and used sharkplotter to calculate those cells at 13.5. Not sure if it was my imagination but the motor felt deinitely crisper on pick up.

We are on holiday for the next week or so and I hope to play around with the timing a bit more. I have pulled back the cells that were a bit more sensitive to the additional advance that was dialled in but I am sure ther eis more to be had. With the sharkplotter it is dead easy to dial more advance and then pull it back when it pings too much.

The only prolbem is that in our location, the modest cooling effect from the monsoon season in the rest of Asia is now over and we are back into the 40's again for the next few weeks. After that it will cool down for the winter season.

Looking at some of the rwhp's the chaps on this list are generating I reckon I definitely need to crack open the piggy bank and put something on the motor to spice it up a bit. Now I am confident I can tune the thing. Any one any idea how much Dave R is catching up on back orders at the moment?

Regards

Fred

John Speake

OK Fred, good work :-)

bd0nalds0n

I'm running the predecessor to the DR twinscrew kits and am very happy with the results and consider myself a twinscrew fan. Still, you might want to research Tim Murphy's centrifugal kits. He also seems to have very satisfied customers, and immediate availability may not be a problem. I think the Stage 3 kits are putting down well north of 500 rwhp, I'd have to check Jim R.'s signature...

edit: okay, not "well north" of 500 rwhp, but slightly north. Still very impressive.

FredR

Brian,

As someone who deals profressionally with compression machines in the 20MW range I tend to appreciate the subtleties of different variants. However, having not experienced the "wow factor" of any blown option personally I can only virtually estimate the pro's/cons of different options. It would be interesting to see your sharplotter profile whilst you are still on regular MAF's. I believe [rightly or wrongly] that the twin screw is better suited to the auto configuration to get "off tickover" torque and minimise the huge disadvantage the auto models have in terms of bogging down at low rpm's and gaps in the pick up range between downshifts [or more to the point the lack of them]. Indeed it is probably better suited to low rpm high torque applications than the manual. The only limitation of the twin screw is that the inlet plenum is ditched so it loses its resonance tuning but then Louie does not have a problem with that and with so much torque on tap, it is probably irrelevant anyway.

My hope is that if I were to "twin screw" the car, it would negate most of the disadvantages of the automatic variant. I was offered a kit to convert to manual last year- maybe I still regret not taking it on but the chap who was selling it, although convinced the gearbox was from a good source, had not in fact turned a cog on that box so that left me a bit nervous in that regard and my job situation was a bit iffy at the time.

I really like John Kuhn's TT kit although the install looks somewhat more complicated and is perhaps pushing the limits of what the 32V engine can handle without serious lower end mods.

Regards

Fred

bd0nalds0n

It was for the reasons you cite that I originally chose the twinscrew over the centrifugal (I had an auto but recently had it converted to 5sp).

At the time, the centrifugals being offered didn't have their own oil supply and for whatever reason I was skittish about the plumbing, which was another detail favoring the twin screw.

It would be interesting to drive a twin screw and then centrifugal to see whether the twin screw really delivered noticably more low end availability. I suspect that the difference may be too subtle to be meaningful. I think Tim has posted or replied to some threads on this topic.

I am running a superMAF although I may not need to. I think I'm pretty close to the limits on the stock MAF, plus the temptation to continue modifying is something I wanted to take into consideration...

DR

Hi Jim,

Wanted to thank you publicly for the SharkPlotter, as I have already told you, it is amazing what you have accomplished and very much appreciated!!

BTW, did the Twin Screw thread get accidently merged with the SharkPlotter thread????

As much as I like to talk about Twin Screws let's please not dilute the great effort Jim has done for all of us by taking this thread off topic.

FredR

Jim,

The sharkplotter is working fine now. Any idea how many data points it takes to get a more reliable iteration for a given cell and what SD we should expect to see? I seem to be getting SD figures in the 0.4 region [but not all]. I am also getting consistent values on different log runs.

The only thing I have noticed since loading the official release version [1.0.011?] is that if I calculate the LH mods, reverse them and ask it to recaclulate it then freezes like the beta version did. The solution is simply to close the project and re-open it and it immediately recalculates the corrections.

It is also apparent that your algorithm is trying to smooth out the cell transitions- very clever.

Regards

Fred

jcorenman

DR, Thanks for the kind words, it's been fun learning about this stuff and I hope I can give a little back.

Fred, this is all new so I am not sure what the right answer is on the number of points, but in general the more the better. The key is to get lots of samples in as many cells as possible-- motoring down the highway at constant RPM and load is not very helpful. And rapid transitions probably don't produce reliable data, smoothly varying throttle and RPM's over as wide a range as possible seems to produce the best result.

The "SD" is standard deviation, the +/- range that includes about 2/3 of the AFR samples. I've typically seen numbers between 0.4 and 0.6 so you're in the right ballpark there.

Interesting, I haven't seen that here-- I'll dig into that and try to find it. How many log files were open? If you can send the files via email that might be helpful.

It's not smoothing the end-result, but it is doing a weighted average for each cell, according to how many points in the cell and also which cells contributed to a given point's AFR-- trying to run the LH's interpolation in reverse. The trick is to not let a small number of points make large changes, unless they are the only points around (e.g. WOT, where data is often scarce by comparison).

Thanks again for the feedback!

FredR

Jim,

Just to be clear, the minor recalc thing does not always happen but sometimes even on the first attempt it does not calculate- do not rule out my laptop in the equation- I have problems with data logging in sharktuner if I do not load the new map iteration first and then close it down again.- maybe it is a problem my end- I run on XP service pack 3 if that gives any clues- the laptop is a toshiba satellite pro- about 5 years old and somewhat beaten up but it fits nicely in my footwell.

This mornings first run [loaded exactly as sharkplotter calculated] gave results pretty close ot what I was looking for. It did lean off a few cells in the 3900 band a bit and immediately got more pings so clearly in that transition the thing is very sensitive knock wise but sharkplotter did not recommend pulling back the ignition due to the limited retard applied.

The results suggest that whereas I am targetting for 13.5 at high load range, it may be prudent to take into account the deviation recorded and reduce [richen] the AFR a bit to ensure not too many cells are above actual values of 13.5. Thus, if I am showing an average of 13.5, with a std deviation of 0.5, it may be better to set AFR at 13.2 [for instance] to get the majority of cells 13.5 or lower.

On my logging runs I indeed do try to make the transitions smoothly [and constantly] to fill as many cells as possible.

Regards

Fred

aggravation

I want to "fine tune" my fueling at the upper end of the spectrum. At idle and cruise it stays pretty much dead on 14.7 but at WOT or anytime I accelerate quickly just below WOT I run rich in the neighborhood of 10 to 11.5.

My question is, since I'm running the maps Tim produced for the Stage III kit with some A/F and ignition tweaks he designed into the maps, could I start by copying those maps instead of starting with stock base maps, turn of the narrow band O2 sensor in Sharktuner and do some logging and then incorporate the SharkPlotter to adjust the cells in the upper ranges and WOT to a set target ratio of 12.5 or 13 A/F?

My amateur reasoning behind this is thinking I'd be safer if I leave the majority of the maps alone since they are doing what they should in most places and only change the heavy load/WOT stuff to lean it out a touch.

Right now I think I'm plenty safe and running well, quite fast, but can't help wonder just how much power I'm leaving on the table running so rich at WOT.

John Speake

Any time you accelerate quickly or go to any fairly high load value even at med to high rpm, the LH will go open loop.

Copy Tim's maps into a ST bin file, then do lots of datalogs, as per the SP manual, then use the SP program to set the desired A/F ratio for the selected range of cells you require to be retuned.

The datalogging will show you excatly which cells require modification.

FredR

Jim,

Still having some issues with calculating the EZ correction factors. Individual cells with calculate but not the entire matrix. It will also calculate individual selected cell values but if you move the cursor to another cell, the calculated adjustment for the previous cell is wiped out. If I then save the calculated file, close sharkplotter and then open it again the "saved" value is not there and it reverts to the base map only. For the time being I calculate the cells I wan to correct and then trasnfer the correction to worksheet in Excel.

Has anyone else experienced anything similar to date or am I just the lucky one? I did a bit more ignition tuning this week and took it for a blast this afternoon. The engine feels noticeably much more livelier now.

I am trying to get the data log in Sharkplotter such that every higher load range cell has at least one point that is showing some sign of knock retard within the constrained limits specified.

Regards

Fred

dprantl

I guess this question is also important for SharkPlotter use. Has anyone had problems trying to paste an entire map of values into the SharkTuner? I was trying to copy the whole cat fuel map to the non-cat fuel map by clicking the top-left gray cell and pressing Ctrl-C, then selecting the top-left gray cell of the other map and pressing Ctrl-V. I think the SharkTuner gave me an error message dialog saying to select one cell or something.

Dan
'91 928GT S/C 475hp/460lb.ft

John Speake

Hi Dan
IIRC you capture all the cells of the source map with the mouse in the usual way, then go to the new map area and click in the grey area in top left (or maybe the gray area) ;-)