progress on the banshee...

[ericjon262]

I beat the snot out of it and went 0-60 in 5 seconds (4.5 according to crappy GPS app) with ~2 seconds of wheelspin.





I've made the decision to build an engine, I have a ton of parts, including a set of forged rods and pistons and rods. I haven't decided what route to go yet, I looked at the pistons that I pulled from the engine I built way back when, and it shows that they've been thrown around a bit over the past 8 years of moving across the country, I might order a new set for the existing rods, and talk to total seal about a set of gapless top rings, I need to measure bores first though. I have two blocks to choose from, one virgin, one that I think may have been bored .010 over, I don't remember.

[ericjon262]

well, digging through this thread for other nonsense, I see this, where I claimed to have cut open the oil filter from this engine... so if that's the filter, what was the one I cut open yesterday? also, that first sentence is something I don't like or remember, but makes more sense about the engine being noisy.

at one point in time or another, the oil pan had been unbolted and at least partially removed from the new engine, I can tell by 2 things, a bolt is missing, and there was quite a bit of RTV on it near the timing cover. upon pulling the heads, everything looked fairly clean, crosshatch was still visible in the bores. I also did something a smarter person would have done at the junkyard, cut open the oil filter and inspected the inside for bearing material, thankfully, I didn't find any, just a little sludge. the lighting makes it look sparkly, it definitely isn't.

yesterday, I swapped the HV oil pump in, while under the car, I saw on the shelf under my workbench, the oil filter that was on this engine when I bought it, so I decided to do what I should have done before I brought it home.





next to a newer, cleaner filter



can't see it in the pictures, mainly because the oil in that filter is so loaded with trash but there's metal in that one too. I'm going to keep driving it, if the weather is nice tomorrow, I'm going to take my oil filter cutter, and the tools to pull valve covers to the junkyard and see what I can find

[ericjon262]

I've been heavily debating what to do, I have two blocks in the garage, and two crankshafts, one of which is definitely good, the other that is most likely ok if I had it turned. I have one complete set of pistons and rods, and one set of 5.

one of my blocks has been bored over, I don't remember the exact number, but I suspect .030, the other should be a virgin block. the complete set of pistons and rods are from the original build, the Diamond Forged pistons, and the Eagle H beams. the Eagles have about .020 more side clearance than stock, I knew there was more, but for some reason I thought it was way more than .020, and the pistons are for a .030 overbore. I had looked at several options,

1. just throw the old engine back together and run it, there appears to be nothing crazy wrong with any of the parts
2. get a set of stock rods and pistons for it, clean everything up, and throw it together
3. get a set of Ti LF4 3.6 rods (available pretty cheap on ebay) and custom pistons to make a potent, lightweight assembly. this would require the crank to be ground

the 3rd option is by far the most expensive, but would probably net the best result, the rods would have the correct side clearance, the pistons would be nice and new, and everything should be plenty strong.

option 2 costs slightly more than option 1, because I don't have a complete set of stock rods and pistons on hand, and would have to get a set from somewhere. it's also probably as strong as I need it to be, but not quite as trick.

option 1 only has the issue of the side clearance, I already have a full set of rods and pistons, balanced to a crank, it probably wouldn't be a bad idea to get a new set of rings.

I've spent more time debating these three options than I really want to think further on, so I decided that I would go with option 1. later this week, I'm going to drop the virgin block off at a machine shop, have them tank it, bore it, install new cam bearings, and prep it for assembly. I have 2 sets of heads on the shelf, that I may send off to have ported as well, making the swap between the engine in the car, and the "new" engine much faster. the new engine will have about 10.5:1 compression ratio, which should make it pretty snappy, and come onto boost harder. I also plan to use my old, more aggressive cam, I miss the gnarly sound it had.

in other news, I picked up a new shifter, I plan to rebuild it and remove 100% of the slop from it, my current shifter is super loose, and my 2-3, and 4-5 shifts tend to suffer alot because of it, I also picked up another front suspension that I plan to use to take measurements off of so that I might begin development of a front suspension as we are discussing in the tech forum.

[ericjon262]

Since I'm building an engine, and the cam I am going to use is out of the engine, I'm considering taking the time to fix a issue I have so that I can gain full sequential injection.

this is a diagram of a LS1's 24x reluctor wheel, and it's cam position sensor signal, the stock LX9 uses this same sensor arrangement.


'
This is a composite log of my crank and cam position sensor output currently:



if you carefully resize the images, you'll find the crank sensor signals are identical, however, the cam position is different, this is because my cam is designed for a earlier 3100 or 3400, which uses a slightly different signal than the LS1 or stock LX9 camshaft. The LS1 wheel decoder program for the MS3 requires the LS1 style 175 degree signal and will not go full sequential without it.

my first step to fixing this, is to determine where the reluctor starts and finishes on the camshaft, and what event it marks. to do this, I looked at the block and determined which lifter were for which bores, and then a head to determine which individual lifter on each bore was an intake or exhaust lifter.



it's fairly easy to tell here that lifter bores 1 and 2 are on cylinder one, 3 and 4, cylinder two, 5 and 6 on three, ect.

looking at a head, we see the valves are E-I-E-I-I-E



combining this data, it's fairly easy to see which lobes on the cam are what,

next I laid the cam down, with the reluctor facing up, and looked for a cylinder on TDC, keeping in mind that the lifter are at an angle to the camshaft. I then determined that the notch represents cylinders 1 and 4 TDC, one TDC exhaust stroke, four TDC compression stroke. I also compared this to a stock camshaft, and found the reluctors to be in the same place.

if I machine off the single tooth, I can turn the diameter to a uniform diameter, and then clamp on a 175 degree reluctor like an LS1 has. The only thing stopping me is that I don't want damage the cam, and I need to know beforehand how perfect the placement of the reluctor needs to be in order to function, if being off by .25 degrees is enough to cause a sync loss, then it's not worth messing with, but if I have a swing of a few degrees to work with, I can make that work.

[The Dark Side of Will]

Use a flat tip set screw (or three) to hold the new reluctor in place, you can adjust it as required. The crank-to-cam reference angle is important, but is in the realm of 2-3 degrees, IIRC. I went through that with my modified cam trigger sprocket on the Northstar.

[pmbrunelle]

In computers, "to poll" an input means to take an instantaneous measurement (a sample) of its value. The computer polls a given input as directed by the program it is executing.

When the computer is not in the midst of sampling a given input, said input could vary wildly and that wouldn't change anything; the computer is blind to this input when it is not polling it.

********************************************************************************

As opposed to a polled input, some computer inputs are hardware-configured to trigger interrupts. Interrupts are usually triggered by edges; an edge being a transition from 1 to 0, or 0 to 1.

An interrupt is a special computer program that interrupts execution of the normal program to be itself executed in priority. Once the interrupt has been executed, execution of the normal program can resume.

An interrupt can be useful if you want to do something immediately in response to some external event.

********************************************************************************

Dropping back into the Fiero world now...

I'm 99% sure the 24x MS setup with the half-moon is using a polled cam. You should confirm with the msextra forum.

With a polled cam sensor, the role of the cam sensor is to provide the ECU with the correct engine phase (a 0 or a 1) when the ECU polls it. The rest of the time, the sensor's output doesn't matter.

You need to find out at what crankshaft angles the ECU polls the cam sensor, and then arrange to present the correct value to the ECU at the instants it polls the cam sensor.

Reading through my MS hardware manual, the polling angles for the 24x setup are not defined in the manual (poor documentation). Therefore, you should ask the msextra forum. Hopefully, James will answer...

********************************************************************************

In communication, if you want to confirm that someone understood your instructions, it is good to ask them to re-explain your instructions back to you in their own words.

Once you get James on the msextra forum to sort you out, then I recommend re-iterating your understanding of things to confirm that there are no communication errors.

For the polled cam with the GM7X crankshaft, I drew a sketch, which James confirmed:
https://www.msextra.com/forums/viewtopi ... 01#p527001

********************************************************************************

If the ECU is only polling the cam twice per camshaft revolution, then in ideal conditions (no other errors), the half-moon could be off by +/-179.99° (camshaft degrees) and the system would still work correctly, because you would still have the proper 0 or 1 when the ECU is doing its polling.

Normally though, the ECU should be polling in the middle of a state, so there is some margin in case the edges move some.

On my Fiero's dizzy, the magnet is adjustable in 30° (camshaft degrees) increments, because it was good enough for the polled cam to work. When you're dealing with a state that is 180° (camshaft degrees) wide, a 30° (camshaft degrees) offset is not significant.

[ericjon262]

Use a flat tip set screw (or three) to hold the new reluctor in place, you can adjust it as required. The crank-to-cam reference angle is important, but is in the realm of 2-3 degrees, IIRC. I went through that with my modified cam trigger sprocket on the Northstar.

right, this was more or less my plan, unfortunately, while being adjustable would be awesome, adjusting it would require massive amounts of disassembly of the car. so making it adjustable isn't as helpful as I would like it to be. I actually was intending using set screws, and tack welding them to the reluctor so they cannot back out.

I imagine an OEM PCM is much more sensitive to these changes than my MS3, but the MS3 has surprised me before with it's sensitivity.

In computers, "to poll" an input means to take an instantaneous measurement (a sample) of its value. The computer polls a given input as directed by the program it is executing.

When the computer is not in the midst of sampling a given input, said input could vary wildly and that wouldn't change anything; the computer is blind to this input when it is not polling it.

********************************************************************************

As opposed to a polled input, some computer inputs are hardware-configured to trigger interrupts. Interrupts are usually triggered by edges; an edge being a transition from 1 to 0, or 0 to 1.

An interrupt is a special computer program that interrupts execution of the normal program to be itself executed in priority. Once the interrupt has been executed, execution of the normal program can resume.

An interrupt can be useful if you want to do something immediately in response to some external event.

********************************************************************************

Dropping back into the Fiero world now...

I'm 99% sure the 24x MS setup with the half-moon is using a polled cam. You should confirm with the msextra forum.

With a polled cam sensor, the role of the cam sensor is to provide the ECU with the correct engine phase (a 0 or a 1) when the ECU polls it. The rest of the time, the sensor's output doesn't matter.

You need to find out at what crankshaft angles the ECU polls the cam sensor, and then arrange to present the correct value to the ECU at the instants it polls the cam sensor.

Reading through my MS hardware manual, the polling angles for the 24x setup are not defined in the manual (poor documentation). Therefore, you should ask the msextra forum. Hopefully, James will answer...

********************************************************************************

In communication, if you want to confirm that someone understood your instructions, it is good to ask them to re-explain your instructions back to you in their own words.

Once you get James on the msextra forum to sort you out, then I recommend re-iterating your understanding of things to confirm that there are no communication errors.

For the polled cam with the GM7X crankshaft, I drew a sketch, which James confirmed:
https://www.msextra.com/forums/viewtopi ... 01#p527001

********************************************************************************

If the ECU is only polling the cam twice per camshaft revolution, then in ideal conditions (no other errors), the half-moon could be off by +/-179.99° (camshaft degrees) and the system would still work correctly, because you would still have the proper 0 or 1 when the ECU is doing its polling.

Normally though, the ECU should be polling in the middle of a state, so there is some margin in case the edges move some.

On my Fiero's dizzy, the magnet is adjustable in 30° (camshaft degrees) increments, because it was good enough for the polled cam to work. When you're dealing with a state that is 180° (camshaft degrees) wide, a 30° (camshaft degrees) offset is not significant.

there's alot to digest here, but I think I was able to find my answer on my own, without too much effort.

here's a composite log of the crank and cam sensor signals. green is the cam, teal is the crank, and the purple circled red lines indicate sync loss, the crank and cam not in agreement.



I logged the same data, this time, with the cam sensor unplugged



overlaying the data for both indicates that the sync loss is occuring on the same teeth in either scenario, and nowhere else. that gives some reassurance that I have a fairly significant amount of wiggle room either way, as the ECU is only looking in the middle of the 175 degree range of the sensor. taking the same composite log and applying the firing order and companion cylinders, I can see that the cam position sensor should start in the same place, and end just before cylinder 1 TDC power stroke, which would indicate that I have room to retard the sensor, but not as much to advance it. if I place it in the exact same place as stock, it should read just fine. I'll post these findings on the MSextra forums and see what the consensus is.

[ericjon262]

I "fixed" the cam wheel, originally I wanted it to be a clamp on part, after trying to work out the best way to clamp the wheel, I decided it would be a better idea to carefully weld it on. it's very permanent now.

[The Dark Side of Will]

The older cam is assembled, while the newer one is cast... I thought they were all assembled.

[ericjon262]

The older cam is assembled, while the newer one is cast... I thought they were all assembled.

The OEM camshaft is a press on lobe, assembled camshaft, aftermarket camshafts are cast core.

[pmbrunelle]

I saw your cam sensor / sequential injection thread on msextra.com.

Back when I was trying to get sequential injection working with the GM7X wheel, I created a thread there with my questions, and showing what I did.

Exact same story repeated today!

Six Shooter saying that sequential wouldn't work... and then when I did make it work, he doubted my results, and was still grumpy.

[ericjon262]

I saw your cam sensor / sequential injection thread on msextra.com.

Back when I was trying to get sequential injection working with the GM7X wheel, I created a thread there with my questions, and showing what I did.

Exact same story repeated today!

Six Shooter saying that sequential wouldn't work... and then when I did make it work, he doubted my results, and was still grumpy.

The thing that bothers me, is that he is jumping up and down saying there's no chance it will do anything, yet providing no evidence that it won't. if he had said "I think you'll run into problems, I recall a post where the developers said this won't work" and left it at that, it would be fine, instead, he keeps insisting, I'm wrong, when I never said it would work in the first place, I simply said I was going to try it and find out.

I really truly regret ever asking for his help or his opinion on anything now, one: I can clearly see he just spews information without anything to back it up so who knows if any of it was right or not. two: I don't want to feel like I've contributed to his overinflated ego.


++++++++++++++++++++++enough of that++++++++++++++++++++++++

my boost control solenoid appears to be stuck. I was suspecting it wasn't functioning properly based on the erratic operation and control, now with boost control turned off, I'm getting about 7 psi, whereas before with it turned off I was getting about 2 psi. I ordered a new solenoid, it should be here later this week I hope, unfortunately, it's not a plug and play unit, so I make need to make some configuration changes to the lines for it to work.

[ericjon262]

From the MS forums:

you can also run this in 24 no missing tooth configuration. i have done this several times on ls based engines. in your composite logs your "rising" edge is 24 evenly spaced events if you have a cam signal that is not too close to the input capture this will work just fine on a 6 cylinder as well as v8.

this was kinda a mind blower, I never realized this was a thing, but if you look at the pattern, you can see that he's right. I plugged some rough numbers in, and gave it a go, it works! There is one major caveat though, the crank is no longer determining TDC, it is being determined by the cam, so ignition accuracy goes down with this operating mode, but it is still not any less than it would be with a distributor. I will still attempt to use the LS1 wheel settings with the modified camshaft once the new engine is in the car, as it will provide more accurate injection and ignition events.

please excuse the massive derailment in the middle of the thread.

https://www.msextra.com/forums/viewtopi ... 55#p581455

[ericjon262]

ok, so the sequential update has made some interesting changes, unfortunately, it's really hard to quantify how, because my O2 sensor got trashed because I didn't adequately stow the interface...



normally, I have the interface tucked way back behind the timing cover in an area where it can't get hurt, when I took the composite log witht the cam sensor disconnected, I somehow neglected to put the interface back in it's nook... at this point, sometimes it reads values that seem reasonable, other times, it's pegged high or low... I ordered a new interface, when it gets here, I'll put it on.

What I can say has changed is cooling fan bog, before, it used to bog HARD when the cooling fan kicked on, now, it barely changes a perceivable amount.

I also ordered 2 TMAP sensors from Ballenger, the monitor pressure and temperature in one sensor, pretty rad!

https://www.bmotorsports.com/shop/produ ... ts_id/1721

I plan to install one pre intercooler, and one post intercooler, that way I can monitor intercooler performance, by both pressure, and temperature.

[ericjon262]

today, while working on the Gran Damn, I was trying to decide which of my LX9 lower intakes I would use, both had thermostat housings still installed, but to simplify things a little, I was going to use the stock housing because it has a different line coming off of it than the LX9 part. I pulled the housing off the first intake within reach and found this:






I guess it was having overheating problems before I pulled it? oh well, it's another piece of the "why is there bearing material in the oil" puzzle.

[ericjon262]

more news from the dead... when i threw together the engine in the car, I put in lifters from the engine that I pulled out of the car, they were way cleaner and looked like they were in better shape. recently, I saw them in the shop and put them in a old oil bottle, and filled it with gasoline, to in theory, dissolve the gunk inside. if you follow the Gran Damn, you've seen this picture:



note, there's 24 lifters on that pan, the 12 from the Gran Damn, and the 12 from the engine in the car. when I poured out the oil bottle full of gas and lifters, I saw a bunch of metal particulate as well... the 12 lifters from the Gran Damn came apart relatively easily, the 12 from the fiero engine would not come apart... I'm beginning to think this engine was toast long before I got it...

in other news, I replace the o2 sensor interface that was destroyed on the belt/idler pulley a few weeks ago, the car had seemed quite sluggish, and I very quickly found out why... I pulled a ton of fuel out of it, it was cruising at about 10:1, the only change i had made between the O2 sensor failing, and now, was the change to SFI. I have a tone more tuning to do, so we'll see how that goes. on the way home from work tomorrow, I'm going to try and get some injector dead time data to test with and see about dialing them in a bit now that I have something somewhat stable.

[ericjon262]

I hadn't touched the car in a hot minute, about a month ago, i started working on intercooler plumbing, ran a hose along the gas tank, and used one of the stock heater lines to run back (the line is now plumbed into the passenger coolant tube). I hadn't gotten around to mounting the pump or the tank yet though. today, I figured the best method available to mount the pump was to put a standoff on the underside of the tank, and then put the pump bracket on the standoff.





My aluminum welds still suck... I haven't has much practice though.

I bent a few pieces of aluminum flat stock to make mounting brackets, if I had more foresight, I could have made two identical brackets, but the pump mount on the underside of the tank interfered.



in the top of that picture you can see 2 red power leads, those both go back to the fuse panel installed at the battery. I'm going to attach another smaller aux fuse panel and three relays in front of the master cylinder, one for the pump, and the other two will be high and low beams.

The pump was sent with a pigtail, I elected to make my own, instead to avoid having a splice in the wiring, the pigtail sent had a red and a blue wire, my dumbass assumed the red wire was the positive lead, and the blue was ground, which should have been a fairly safe assumption... it wasn't. I jumped the pump off of a M18 battery, it moved a good bit of water at a pretty decent pressure, but when I swapped the leads, it was probably double the flow! DOH! I haven't decided whether I want to finish the wiring tomorrow, or work on the Gran Damn. For now, I plan to wire the pump relay to a IGN+ source for a trigger, eventually, I'll use a binary output on the microsquirt to control it, probably trigger it to run when ECT >100 or something like that.

[ericjon262]

I mounted a Bluesea fuse block and four 40 amp relays to a piece of aluminum flat stock, originally I was going to use riv nuts to secure the piece to the front bulkhead, but I noticed the studs that would hold the sunroof air deflector and decided to use those instead. Doing this makes the bracket a little long, and bigger than it needs to be, but it also made it easier to install. later I may take it off and trim it down, but it's also aluminum, so it's not like the weight penalty is excessive, and it's not in the way of anything. as of this moment, only the 3rd relay is in use, the first and second are reserved as headlight relays, and the fourth is just there for future growth, maybe EPS? I also have 5 more fuse positions available in the block that don't have a planned use yet, I may use one for a line lock, but otherwise I'm not planning much more expansion.



as per the usual, I said I was going to do one thing, and did something else...

I pulled a few schematics and couldn't find an IGN+ to trigger the pump relay with, so I decided to go ahead and wire it in proper. I'm using the high speed fan pin on the C100(F7) to feed the trigger through, and I set it up to run at any time over 200 RPM. The pump is kinda loud, I can trigger it based on two parameters, so I set it up to run any time the engine is over 1200 RPM -OR- manifold air temp is over 110 F, that way, if it's hot and needs the loop to cool down, it will run, but it also won't run if it's not needed.

[ericjon262]

the intercooler appears to be functioning, but I think there might be an air bubble in it, it MAT is can still get hotter than I would like to see with a quickness, that said they also drop relatively quick too. Peak here is 151F, running on straight water. I might loosen the fitting on the top and try and bleed the system to see if that helps. I'm also noticing the MAT sensor is heat soaking a little bit too.

[Shaun41178(2)]

One thing you didn't show in your log there is how much boost you are running.

However the fact that your iats come back down into the 90s after boosting tells me your system is working just fine. Your prob 10 degrees above ambient which is fine.

If you feel your temps are getting too high, then you are probably overpowering your intercooler. Meaning it's too small. But I'd say temps recovering quickly tells me your heat exchanger is working perfectly.

Post up a screenshot of your same log with it zoomed in on the max 150 degree temp under boost, and your boost level. Also length of pull if you can or just mention what it was.

My temps on a hard third and 4th pull to redline get into the 150 range too I think. I'd have to go back and double check. I am definitely overpowering my intercooler as it's too small. I run e85 so I don't worry as much though