Megasquirt - The Advanced Thread -For ADVANCED Discussions

[Hybrid_From_Hell]

As I understand it the Req Fuel is more a coarse setting which is then trimmed within the VE table.

Lower Req Fuels will often see higher VE values etc, and vica versa.

For the V8s I have always used BBCs calcs which seemed to pop the corase settings pretty much in the right bands, there is a risk that you might get the Fuel Req Settings too far out for the VE to then trim and tune, but I have never spent any serious time learning either where theses points may be, or the knock on consequences.

I have thus always used 20 and 10 vs your 12.4 and 6.2, this would then explain the higher than I would expect to see on a V8 VE Values, but would still not explain your LC1 and its lean values

FWIW I would suggest you consider 20 / 10 as for no other reason than its what many here now use for a V8, and again I would suggest 20 / 10 and a NB and a datalog / remap, and see what it produces ?

HTH, prop doesn't

Nige

[FridgeFreezer]

REQ_FUEL is just a constant used in the fuel calculation the MS does every squirt, to be really picky it's the value that would be the ideal amount of fuel if your engine was 100% efficient, which the MS then multiples by the VE (Volumetric Efficiency) table numbers to arrive at the real value for that point - the VE numbers are basically % efficiency numbers - hence on a non-turbo engine you should never have a VE value of 100 or more in your fuel table (in reality probably never much more than 65% given the real VE of an engine).

However, it does not matter massively what the value is as long as it leads to a reasonably sensible collection of numbers in the fuel map. Since you don't know the exact VE of your engine even if you calculate it to be exactly 16.1 or whatever you're probably just as far off as if you choose 10, 15, 20 etc.

I think the reason it's 20 is that when TSD and myself first MS'd my truck we used the stock fuel table that comes in the MS and just kept trying REQ_FUEL values in increments of 5 until it fired as changing the REQ_FUEL value is an easy way of effectively increasing or decreasing the amount of fuel across the entire map without changing the map.

The calculation the MS does goes something very roughly like this:

REQ_FUEL x VE x Warmup (coolant) correction x Air density (air temp) correction x O2 correction + Accel enrichment etc. = Squirt

If you want to know the real maths behind it all, and gain a new found confusion or admiration for the guys that programmed all this, you can read the full explanation of how MS works in the MegaManual here.

[steve_d]

When MS is using a table and is between boxes it interpolates a figure to use from the boxes either side.

But, if you go off the table does it try and interpolate from where it left the table or just use the last figure where it left the table.

I'm sure I have seen it somewhere but can't find it again and can't remember which it does.

Reason for the question may seem strange but let me explain.

Example 1.

Take a spark table...Compared to a dizzy you would expect all the mechanical advance to be in by 3000rpm and that would equate to the 100KPa line. All the KPa values down the 3000 column would then increase with vacuum advance ending with about another 10 degrees by the time you get down to 15-20KPa. Any rev columns above 3000 would be the same and a waste of space if MS uses the last box value in a table when you go off the table. If you created a table that only went up to 3000 you could create a more detailed and smoother table with the extra columns.

Example 2.

The VE & AFR tables I have seen include 15 and 20KPa rows but my tickover has been up at about 38-42 (not tuned out yet) so anything below that will be over-run. Having set one row below your tickover value to what you want on over-run then any rows below that are again a waste and could be used to fine tune the areas higher up.

Any thoughts?

Steve

[FridgeFreezer]

From the MegaManual:

Beyond these values, MegaSquirt will use the last value from the “edge” of the table. (It does not “shut down” by substituting zero values.)

[steve_d]

From the MegaManual:

That's it. Looked all over for it. How did you find it and where exactly is it?

What are your thoughts on what I am proposing.

Would not be surprised at a 'not worth the effort' or 'making more of this than necessary' type response.

Steve

[Hybrid_From_Hell]

Steve,

Too knackered for the long reply in me mind, heres the short version

with the 100 KPA line your correct re all advance in at WOT but, for lower KPAs and economy ( OK some fuel savings ) then you'll add in advance that may be different at the variois KPA points between say 3000 and 5000+ ish, it is true also that some race enegines have no additional advance and as such all 12 lines would read the same- but the fuel consumption is hideous, often vacumn addy is left off from race engines partly due to setting hard and solid Adavnce and patly due to having no vacumn addy to screw it up or jam / fail etc

A 12 x 12 table IMHO gives more than enough for a well mapped engine that will beat the std MAnufactuers set up hands down, I have now tweaked mine so that I have the KPA tickover and the main areas it performs in mapped more and the overrun and areas it spends little time in reduced as of little importnance, but I only started this once I was sure that the rest of what I had was solid and safe, so a well spread 12 x 12 VE and spark datalogged with MLV and a NB sensor for some hours will give you this, and also get you familiar with where you engine / driving stylee / power / tweaks needed are, and thats the key, no to engines of the same type seem anywhere near the same, auto also changes VE MASSIBVELY vs manual, so 12 x 12 and log for 10 hours over a period of time and then see how you can tweak it, but I warn you once you start doing this its infectious and you'll be hooked

HTH ?

Nige

[FridgeFreezer]

I found it by going to "How MS works" and searching for "edge" B)

My thoughts are that a 12x12 table is plenty big enough unless your name is R.Dennis or F.Williams I remember when we used to get by with 8x8 tables and all this were fields

You'll find with the super swanky after market mega money EMS that professional tuners will map a few key points on the massive (256x256 in some cases) fuel map and then hit a button in the software marked "interpolate" and then just do a bit of fine tuning / smoothing out.

It's all too easy with this new-found toy to get carried away, when you find yourself worrying about the last +/-1% in a couple of spots on the fuel table just cast your mind back to carburettors and mechanical distributors and ask yourself if you really need to worry about it

[Hybrid_From_Hell]

Oh so F True.......

As Bill S said to me (when I said I was worrying about something minor on the MSQ), his classic comment was :

"You are because you can, you can also stop you know"

Nige

[steve_d]

My main concern at the moment is the spark table.

Haynes for the 3.9 says initial advance at tickover without vacuum is 4 degrees and the general consensus is 36 degrees all in at 3000 so that sets the 100KPa line.

Megamanual then says add 0.3 for every KPa which pretty much ties in with the 10 degrees vac advance on a dizzy.

I have never seen an advance table posted here that looks anything like that and all the ones I have seen have been vastly different to each other.

If the timing is wrong how can you expect to get the VE right?

Steve

[Hybrid_From_Hell]

If yiour spark is wrong then no the VE can be spot on, and any changes to a spark table affect VE

However Luke, stay away from the dark side yes, the MS Force is strong in you, you must feel and hear the

force and rememebr that on the darkside the clockwork spark generator is limited by mechanical movement it is, and as such the spark adavnce

can only be what it can do limited by mechanics and clockworks bits yes.

On the MS table you can have whatever advance you wish to have that the engine actually really needs, and not just what a Dizzy is able to produce - on mt eales the dizzy timing was set at 3500 rpm and then locked up, this gave me a horrible tickover advance as the dizzy was not able to go low enough - many people can't get their minds around the fact that they shouldn't / don't need to mimick the dizzy advance curve and can have more fredom, hence with mone I have 12 at tickover and huge amounts at the top end.

There are a number of spark tables on the master MS thread mine wass done vai a rolling road from scatch, and others on there are more tweaked downwatds for sanity and safety with less tweaked engines.

Do not over spark the V8, the head design limits the max advances you can run, I have posted much on spark stuff on a thread elsewhere inc the max advances and at what rpm break points per V8 Capicity engines as they vary, bigger = less max advance

Nige

[steve_d]

Agreed the MS table offers scope to tweek the timing in all areas. If I were playing with the timing on a non MS engine I would put it under load at various revs and keep adding more advance until it pinked then back it off a couple of degrees and this is also a method with MS.

Unfortunately mine is an auto so as soon as you try and apply load it changes down.

Can you point me at a table you consider suitable, or a good starting point, for a standard 3.9 serpentine?

Many thanks

Steve

[FridgeFreezer]

Since this is the techy kinda thread, you could always add a knock sensor to the MS which would greatly aid in ignition tuning.

[steve_d]

Since this is the techy kinda thread, you could always add a knock sensor to the MS which would greatly aid in ignition tuning.

May be a good route particularly since doing MS on the RV8 is a precursor to putting it on my 508HP Chevy engine which I have been known to work quite hard and was not cheap to build.

Problem is I have not seen anyone saying that any particular after-market knock system works properly. They all seem to have issues over fine tuning them and in some cases the sensitivity has to be turned down so far you wonder if it actually going to hear anything.

Anyone on here used one?

Steve

[landybehr]

Problem is I have not seen anyone saying that any particular after-market knock system works properly. They all seem to have issues over fine tuning them and in some cases the sensitivity has to be turned down so far you wonder if it actually going to hear anything.

I do, but with very much that problem you describe.

I think the knock sensors "KnockSense" uses are Bosch. The one he took resembles very much the one that goes with the Thor-V8´s, except for the connector.

Looking at http://www.bosch-motorsport.com/content/la...2/html/2883.htm and the datasheets it seems like the sensors are one-size-fits-all. And I my believe that fine-tuning it to the bore ("resonance") could work. But I have no idea about the sensitivity and where to put it correctly. If I follow the spark maps on the Buildup-thread I am igniting fine - still get signals from the sensor. This with having reduced the sensitivity fair amount already - as much as starting to think it is of no help. Above 3000RPM the sensor is "on" all time, still. Sensor is mounted in OEM position. What I feel I´d need is a device to control the knock sensor itself - which itself would obviate a knock sensor for tuning purposes.

[FridgeFreezer]

Knock sensors are a whole ball game in themselves, to be honest I've never got very far with researching them as they always seem like more work than I have time for, but the basics of operation are:

- All engines knock at different frequencies, most knock sensors are therefore tuned to the engine or have a signal conditioning circuit tuned to the engine.

- Engines make all sorts of noises as they work so discerning what's knock and what's just noise is tricky - a good start (and commonly used strategy) is to only "listen" to the signal from the knock sensor in the tiny period of time after the plug fires when you might reasonably expect a knock to occur. In between times, most circuits "mute" the knock signal so whatever noises happen they do not falsely trigger the knock response.

- Universal sensors & circuits require some tuning & setting up and really you need a knock sensing circuit that takes care of all the signal processing and just feeds a logic pulse to the ECU when it detects knock, the ECU does not have enough spare processing power to do digital signal processing as well as running your engine.

There are plenty of functioning knock sensing circuits & info the Megasquirt-Extra & MS2 forums and the MS Extra manual. The tricky bit is making it work

[landybehr]

Pity is I remember that a guy wrote about a self-made stethoskope by means of which he heard knock pretty well, but I cannot recall where I read that.

Not sure if it was as simple as using a 10mm copper pipe - flattening one end which then would be bolted to the engine and routing the pipe to the drivers seat.

[FridgeFreezer]

You can get more bling than that - you can get microphones that play the engine noise through your car stereo or a set of headphones! None of it is rocket science - it's interpreting what you're picking up that is the skill. There are plenty of working DIY knock sensors for MS out there if you read the forums, it's just that no-one around these parts has bothered to fit one yet.

[landybehr]

Probably most of us browse through the official MS homepages.

Just in case - the following thread is worth looking at:

http://www.msextra.com/viewtopic.php?f=89&...599&start=0

(basically it´s about taking care of the MAT correction table, and, if living in mountaineous areas, for cont.baro as well. The latest MS2x from today has a change that contains a change to the MAT correction).

[steve_d]

Is anyone using closed loop idle control?

I am having issues trying to get the idle set. I can find a nice low and steady idle but have to up it to cope with the auto engaged then up it again if the fans cut in. Net result is the idle is near 1000 when out of gear & fans off.

Have found that MS1-extra & MS2-extra have closed loop so I can just set the revs I want and the PWM idle valve will open enough to keep the idle up.

Steve

[Hybrid_From_Hell]

Is anyone using closed loop idle control?

I am having issues trying to get the idle set. I can find a nice low and steady idle but have to up it to cope with the auto engaged then up it again if the fans cut in. Net result is the idle is near 1000 when out of gear & fans off.

Have found that MS1-extra & MS2-extra have closed loop so I can just set the revs I want and the PWM idle valve will open enough to keep the idle up.

Steve

Yes, I am

And with a hairy cam makes for fun as the settings at certain points can affect each other and seem to fight each other

Post up pics of the settings you have so far and we can laugh discuss

Nige

[steve_d]

Yes, I am

And with a hairy cam makes for fun as the settings at certain points can affect each other and seem to fight each other

Post up pics of the settings you have so far and we can laugh discuss

Nige

No settings yet as I have not made the decision to jump from MS2 to MS2-Extra. Am in the process of reading through all the -Extra manuals to find out what I have to do to convert.

Steve

ETA I have found a few -Extra features that I want.

1. Idle control.

2. The ability to invert the PWM (mine says 100% when fully closed).

3. Feature to switch to a fixed timing so you can tune the trigger wheel.

[evo828]

I am using closed loop idle on MSII/Extra - in the latest beta it works ok even with the stock rover IAC stepper valve. Can send you the setup once you are on MSII/E.

[evo828]

Need advise:

Yesterday I started to get some "popping" sound through intake when cruising - it is also visible in log as small spikes in MAP (and subsequently in PW etc). AFR is around 14.6-14.8 there. Timing is 23deg. The spark plugs (bpr6es) are there for 50km only - so new.

Any suggestions what might cause them and how to get rid of it? Should I decrease the timing in that area?

[Hybrid_From_Hell]

Too Lean IMHO

14.35:1 is often recomended by Rover tunning experts for best cruising value

the main reason being that RV8s run too hot at 14.7:1and the slightly richer 14.35:1

helps reduce heat within the engines, and smootgher running as well this is esp

true of the 4.6s or any tuned V8

Has it always done this or just started, if just started then whats changed

If its popping what are the cell vaslues it goes into, is there a step or hole maybe on

either VE / Spark tables ?

Nige

[evo828]

The only thing changed was spark plugs 2 weeks ago. Have driven with them already last week - absolutely no issues (been logging as well - no spikes there).

Yesterday went out with driving and tuning ve table with a friend who consider MS+Edis for a 3.5V8 RRC. After 2 km I have noticed these light pops - so I have checked the log and they were there. I haven't touched the VE map at all - as we were still warming up. After 3-4 minutes driving it went away (even in the same spots of the maps) - I though it was related to warmup. Did some finetuning of VE driving along, cruising etc - all was fine. As we were retunring home - the pops started again - the log posted is from the "return home" part of drive.

I am surprised that tuner are aiming to cruise at 14.35 for rover v8. Mine was cruising with the stock ECU at 14.7 with no issues - before I put MSII there.