Testament

2 hours ago, Roman said:

Also my code is getting pretty bloated again so I've been looking to library-ize various portions of code. 

How does this work though with data that is shared between different parts of my code though?

Like is it best to keep all of the data in a library completely internal? Apart from what you put in, and take back out?
Like for example if I have an array that a library wants to look up, but other things still look at. 
Is it best to keep that array in the library or in the main code?

I guess the whole point of a library is that its quite self contained.
This makes sense but there's a bit of untangling to do as all my code runs through itself a lot haha.
Rookie mistakes.

And I just broke my LCD screen so its a few weeks until another one gets here. Maybe a good time to fix all the code instead haha.

what do you mean dave? rookie never makes mistakes

yup, not to do with lobe pulses but the log. as said runners are basically non existent and #1 is directly infront of the S/C and #2 is very close but #3 and #4 are two 90 deg right turns and 75mm and 150mm of  log away at 45mmish diameter, which is about the same as the port and valve sizes.

The lobe pulses may have some other effects, and the bypass valve is between #2 and #3 so that will have some other effects too.

just all about solving problems no one else ever has because of one else has been silly enough to persist with quite the same setup.

33 minutes ago, Roman said:

It's weird though I would have thought that an uneven intake manifold would only really pose an issue when you're at full throttle and the engine is trying to consume as much as possible.

When you're at part throttle one cylinder doesnt really have a mass flow advantage over another?

What injectors are in there, is this the car with those single pintle 1000cc?

These are xSpurt 1000cc/rebranded ID1000s, not the china 1000's

The pressure drop difference occurs at all rpms,  the blower puts 1.6L of air in every rotation of the engine but its easier for that air to get into cylinder 1 and 2 than 3 or especially 4, so even though only 1 inlet valve is open at a time there will still be a difference, more residual pressure in the manifold than reaching the cylinder in a given time. As rpm increases and there is less time for the air to move, the relative difference may change slightly one way or the other (not sure which way). As pressure increases with boost the relative difference will be less.

On ‎11‎/‎12‎/‎2018 at 10:21, Roman said:

Just thinking about the setup some more 

So you have an air inlet on the supercharger. Supercharger moves a volume of air thats fixed with RPM. 

But then after that a bypass valve lets some of the volume of air back out, essentially in reaction to throttle position.

To deal with varying throttle positions the pre throttle pressure rises until spring pressure of the bypass valve lets air out (?)

Sooooo you have different high pressure situations pre and post throttle body and they both affect VE of the engine.

Efficiency of the throttle is affected by variables of both pre and post throttle changing a lot so maybe not a stable variable for a load axis

What could be interesting would be to hook up a second map sensor between the throttle body and the supercharger.

Because that effectively represents both your throttle position and how well the bypass valve is working. It's really a combination of these two things that affect your map readings in the plenum not just throttle angle itself. Because a mechanical bypass valve will have some latency that isnt represented by throttle angle alone, but pre throttle pressure will change when this happens.

Otherwise if you fit a type of bypass valve that you can control with PWM then you can play with the curve to a really nice linear throttle response and also use that pwm% as the load axis on a compensation table.

so have been making adjustments every morning and afternoon driving to/from work mucking with idle/coldstart and datalogging. slowly getting better. the intake backfire is definitely from lean conditions not ignition. comes and goes with fuelling changes. have got rid of most of it now by blending in some mapDOT acceleration enrichment and adjusting the fuelling.

still needs the full sequential setup though, there is a dull miss (probably from too rich) even at idle which must be from the uneven air fuel ratios, which would also explain why it seems to need a richer mixture than you would think - e.g mid load cruise 75kPa just over 13:1 when you would think 14:1 would be ok even with the S/C.

dashcam for tunnelram blindspot and we're good?

24 minutes ago, sluggy said:

Anyone want to buy an LS3 with 10,000 k's on it ?

 

Hmmmm I could be interested..........

On ‎26‎/‎11‎/‎2018 at 20:47, RUNAMUCK said:

A mate of mine slapped a b&m huffer on his stato. They're smaller, and lower profile. 

rides a scooter and talks of smaller blowers? what has happened to the bart we all used to know????

throwing $4500 at a 4l60/4l65  auto is money down the drain. unless you start with a factory manual t56 car you will need to change the transmission if you don't want to turn one of the factory autos into a more useless lump than it started as. realistic options would be 4L80, T56, TH400 or Powerglide  depending what you want to use the car for as to what to choose. in any case other than using all factory manual bits it will be some amount of custom work for mounting it up, getting speedos and gearshifts to work and getting a suitable driveshaft.

man that really sucks not being allowed roof height velocity stacks.

anyway

if you want a car to drive, be silly fast and fry tyres the factory LSA motor option is hands down #1 as it will just work and be no worries provided you put the right driveline behind it.

if you want the giant shiny blower and toothed belt drive with two toilets sitting on top all well and good but just be aware it will take more work to make it work not as well as the LSA.

swapping to a different Toyota diesel would probably be better than turboing the 3b. but in either case it would be a project if you have not done that exact swap before, or have someone on hand who has. better just taking a chill pill and leaving well alone although chilling out and driving sedately is usually a difficult thing for card carrying gun swinging fleece wearing hunter types around here

nah you are right, logging map ahead of the supercharger could be quite useful too.

the throttle body is on the large size which amy not be helping either, 65mm off an EL falcon or similar which works out for the power level but maybe with the lighter car could do with a more progressive bellcrank, as at low engine speed 1/4 throttle is basically 100%. once throttle opening area is gets close to the 35mm bypass valve area its going to lose vacuum, shut and go straight to more or less full boost.

do you know what clock speed the link ecus run? it's weird their website doesn't seem to list it

the microsquirt runs at 24mhz

bypass valve is a factory thing off a late model supercharged mini. It has a vacuum operated actuator, so purely mechanical. it runs off the vacuum in the supercharger inlet  elbow and closes when that drops as the throttle opens. the bypass loop is entirely downstream of the throttle plate so don't think bleeding off boost is the right way to explain it but whatever.

my feeling is the full sequential will help a bunch, and getting the  air fuels evened out will be mostly a fixed offset that maybe tapers off with boost and gets worse a little with speed. but yeah latency in the bypass valve response and whatnot, I dunno. I think some factory supercharged setups have multiport solenoid valves to supposedly make better driveability but I have also heard its a thing to just rip those off supercharged mr2s and stuff like that too and not even worry about it

11 hours ago, Roman said:

Hows the Fiat tune going? Running good? 

Terrible lol

I need to spend a day with a clear head just tuning it to minimise some of the strange misses and stumbles at small throttle movements as best as I can, but haven't had the luxury with whats been going on at work lately so barely managed to jam it back together to drive up on Saturday.

but pretty sure a bunch of it is down to the log inlet manifold causing uneven airflows between the cylinders, resulting in significantly different air fuel ratios. So actually have bought a monsoon to put on it so I can get full sequential with cylinder trimming, and umming and ahhhing about wire transferring money to Poland to get an ecumaster can-egt box haha.

the other thing I was discussing with rookiedave was going from speed density to dual table alpha-n and speed density boost compensation/reference, which should give more resolution with throttle movement, as with the supercharger running speed density/map only it goes from vacuum to boost so quickly you end up in the same cells all the time even with different throttle openings and driving conditions.

only thing is not sure how to tune that with the supercharger, as you need to tune the alpha-n with no boost first. not so simple as taking the wastegate springs out. maybe I can take the belt off the supercharger and cable tie the bypass valve open or something.

at least the bypass valve works though, inlet temps cruising on the open road have gone from 80+ to around 50 now

yeah will depend on the gauge if its two point or three point thing and how it is intended to be setup. some of them are intended to be setup on a test ring of a known ID three point ones are more likely to be like this. two point ones could be setup with gauge blocks where  you set your gauge block stack and then have anvil pieces that go on either end so you can measure in between.

often there are several ways to do these things depending what tools you have, how much time you can spend and what accuracy is necessary

its isocyanate stuff as bad as 2 part epoxy stuff. I did by jeep interior floor inside ages ago even outside with all the windows open and brushing it on I was not entirely well for a few days after from having my head too close to the stuff when painting. would def use respirator if its anything where you are less than arms length from it while putting it on

the metal ready prep stuff is great though. I have put than on bare steel and nearly 10 years on inside garage with no paint its still fine.

for single pack ive had much better experience with Wattyl killrust. just takes a while to set is all. prepped with metal ready then brushed on jeep bumper only rust is where I have scratched through the paint and even that is naff all on something that gets scratched on all the things in the bush.

I think brunox gets mentioned as another decent one for when you need to paint things in non ideal situations with a single pack. I have some but haven't used it yet.

but when you can 2 pack is definitely the way. once you understand the process and buy one of those 3m respirators it is way better sprayed or brushed on.

durepox black primer for everything now.

On ‎22‎/‎11‎/‎2018 at 13:07, Roman said:

You could offload some of the IO to canbus on the Atom.

Like it might be more cost effective to buy an Atom + Can wideband than a Monsoon + analog wideband. 

Canbus is like the secret back door, on top of the predefined things you can send to ECU over can you can also have 8 virtual analog inputs and 8 virtual digital inputs. 
Then use an Arduino Mega or something to give yourself a whole bunch of 0-5v inputs and switches etc.

Which is an absolute shit load above the 3 analog inputs and 2 digital inputs it comes with standard! 

And that you can output essentially as much as you want from the ECU as well.

 

I dunno, I think if you counted your time it would be easily cheaper to just buy the ecu with more I/O to start with. It's one of those things where when you have don't it and your more or less repeating a setup you have done before you might say yeah I can make a saving this way plug it together and know it will work easily. but when you haven't done it before and you have to spend several full days (when you add up your research time etc.) not a chance. I mean if you want to learn canbus stuff then that's the learning budget so maybe fine. but people sure love to say oh you can do it this way and save so much and 99.99% of the time its a load of bollocks.

On ‎23‎/‎11‎/‎2018 at 18:09, kws said:

One thing i need to ask if youre buying an ECU, is what is the preferred ECU for your tuner? You dont want to end up with an ECU your tuner doesnt want to work with, unless you're prepared to travel or tune it yourself.

my experience is that its not tuners not wanting to tune megasquirts, its tuners not wanting to deal with dodgy installs that have all sorts of other problems when you taken them to the dyno. been there done that and seen mates setups have the same issue.

on top of that, just being less familiar with them it will invariably take more time vs using software or doing troubleshooting on something they work with very often.

to be fair the prices have definitely come closer with the link atom and monsoon now, when you factor in shipping GST and customs plus the additional features the links have in some cases you are talking a few hundred dollars. if that means your car gets started and tuned in half the time that's probably worthwhile.

On ‎24‎/‎11‎/‎2018 at 23:39, peteretep said:

wat, just tune it yourself, its not that hard

this is really the case, provided you are so inclined and interested in learning this kind of stuff.

because of the diy nature of megasquirt its all about the installation being done well and troubleshooted well on setup/startup. which you have to do and fuck up a few things to learn how to do it right really.

there a are a few naff things about them but none that make them dangerous to your engine any more than other potential failure modes

1 degree = 10mm if you measure at an effective wheel diameter of 575mm according to the info sheet

yeah powerglide coolant pump and vent filter

I liked the part where you said this car is getting a powerglide and a transbrake

because old man pants decided to take out the on/off 6 puck solid centre 200mm clutch and 6kg flywheel and swap to a OEM cast iron flywheel, but modified to take an early integrale 8v 230mm push type clutch and pressure plate I found on amazon for a bargain after extensive part number googling

but nothing too simple of course. OEM flywheel from fiat 131 is for a 215mm clutch so the friction face is too small, and all the dowels and bolt holes are in the wrong spots by just a little bit.

so had to machine it down to flat on big lathe at work, then machine a step with 230mm diameter, then setup in the mill and drill and tap for dowels and bolts offset to the original clutch ones, and make some new dowels to suit.

fully detailed CAD model of new sump baffling using davescience software as an upgrade to the low weight low drag baffling of the original fiat design

and boogered up

some bits were still ok

compared some sumps while I was there, left argenta, right 131

commodore LS1

out with the old and busted

1