I purchased this Fiero in December 2016. A mostly bone-stock 1985 SE, with 53k miles. Minimal rust and fairly decent paint; should look good enough (by my standard) with a wash and wax, and some minor fixes (dew wipes, headliner, that sort of thing). It came with the automatic transmission, but that's fine; transmissions can be changed.
Half an hour into the 2 hour drive home, the 2.8 started knocking; the car had to be towed.
The engine's death signaled to me: PROJECT TIME! Time to do the turbo project I've always wanted for a Fiero, get rid of the slushbox, and fix the "incidentals" while I'm at it.
Now that I've been working on the car a little while, I guess it's a good time to make a project thread.
For the transmission, I decided to go with the Muncie 4-speed. Since this is my first real car project, I wanted something simple for a first-timer... I didn't want to pull a speed gear and have a bunch of needles fall all over the place.
I don't know how the Fiero Muncie will cope with more torque than stock, but Archie sold V8 kits to work with the stock transmission, so... I'll see what happens. I made sure to use the ribbed case, not the smooth case.
Anyway, so here I am scavenging pieces from three Muncies, in order to make myself a good one:
I got a few brand-new parts to supplement the scavenged stuff.
Here, I clean the transmission case:
Closing up the transmission. From my collection of scavenged shims, I was able to shim the tapered roller bearings without having to buy or modify any shims. I didn't have the required shifter shaft selective washer on hand, so I just machined the shift shaft to compensate.
I decided to use the complete 1984 M19 gearset.
I thought about using the 3.31/1.95 1st/2nd from the M17, since they're closer together than the 3.53/1.95 1st/2nd of the M19. However, I wanted the deep 1st gear reduction. My previous Fiero had an Isuzu transmission, and I learned to deal with the wide 1st/2nd gap. The M19 is no worse than what I've become used to.
All buttoned up:
Of course, an automatic car needs more than just a transmission! It needs a 3rd pedal:
The reverse interlock roller was actually seized. I worked it free, but I don't think it made a difference that could be felt by the driver.
Having the shifter cleaned up, it's like driving a freshly washed car with shiny paint. Even though it makes no functional difference, it's still more fun to drive. It's psychological.
For the engine, my friend gave me a 3.1 from a Chevrolet Beretta. Aside from the additional ~300 cc of displacement, I wanted it since it comes with a crankshaft position sensor. The crankshaft position sensor provides the best possible angle information to the ECU, compared to using a distributor with slack in the timing chain.
Driving manners are important to me; for the ECU to do a good job, all its inputs have to be clean.
This is the setup I used to polish the #981 crankshaft's main journals. I used the shoelace method for the throws.
I had the block decked to remove pitting that would prevent the head gasket from sealing, particularly since a turbo engine has more cylinder pressure than stock. I'm using the Fel-Pro MLS head gaskets. I got hypereutectic dished pistons, 0.75 mm over. The machinist bored the cylinders to 0.80 mm over, to give me a little more piston-to-wall clearance. From the turbo engine's extra heat, more clearance is needed to avoid seizing the piston in the bore. I also filed the piston rings to have larger-than usual gaps, again to deal with the heat.
The displacement is 3192 cc with a compression ratio of 7.4.
The obligatory top end cover painting. I used single-stage DuPont (Axalta nowadays?) Nason paint.
Here I am inspecting the camshaft's lobe lift as I rotate the crankshaft. The camshaft is a hydraulic flat tappet cam from Crower. I told them about my turbo build, and they specified it for me. It has a 114° LSA. A Cloyes double-roller chain drives the camshaft. I installed a high-volume oil pump.
The clutch is the RAM 9.75" HD unit from the Fiero Store. I know that some people use it with the Cadillac 4.9, so it should hold a decent amount of torque. Eventually, as I crank up the boost pressure, the clutch might slip. Not a big deal, I'll change the clutch if there's an issue.
Making sure that the pressure plate won't rub the inside of the bellhousing:
Cylinder heads are stock. I didn't port the heads, as I didn't want to risk dinging the valve seats and needing a valve job (more expenditures). I replaced the exhaust valves (pitted contact faces), and lapped all 12 valves. Valve springs are new; they came with the camshaft. I have 1.52 roller-tipped rockers for these heads.
Long block (mostly) complete:
[This message has been edited by pmbrunelle (edited 01-03-2019).]
In a standard 90s GM setup, the crankshaft position sensor is wired to the DIS brick.
I don't want a DIS; I want my Fiero to have the classic distributor look. Hence, in this setup, the crank sensor is read "directly" by the ECU; no DIS brick in between.
I did, however, make a circuit to convert the crank sensor's weak analog signal into a robust digital signal. It uses a Maxim MAX9924 integrated circuit to do the job. The case is made from CNC-machined nylon 6/6.
The converter circuit is located as close as possible to the crank sensor, to reduce the chance of picking up noise in the leads.
The converter also provides threaded mounting holes for the GM knock sensor module, which goes right on top of it:
The modules are potted in urethane for sealing out the elements. I made three of them, since if the module ever dies, I can't just walk into NAPA and get a replacement. I left one module not potted; in case I discover a design flaw that led to the failure of the first two, I'd be able to make some changes before potting and installation into the car.
Here is a test I did, turning over the engine with the starter. A tooth on the crankshaft is passing by the crankshaft sensor at the moment of this oscilloscope capture.
Top trace: analog signal from crankshaft position sensor Bottom trace: digital signal sent to ECU
[This message has been edited by pmbrunelle (edited 01-03-2019).]
Since the battery tray was falling apart, I decided to replace it with a new Fiero Store unit.
However, the root of (I think) the battery tray rust problem, common to many Fieros, is that the battery acid overflows and leaks onto the metal below.
Therefore, I decided to make a polypropylene battery tray. It is designed to channel any liquids to a drain fitting, on which a drain line can be installed to direct any liquids to the ground.
I also didn't like factory method of gripping the battery by the two lips at the bottom. I don't find the factory method secure enough, especially when the rust bug bites.
So, I decided to hold down the battery more positively, with a flatbar over the top and rods keeping the flatbar in its place.
Since I want to put the turbo above the transmission, I moved the ignition coil over to the trunk wall to get it away from the heat.
As this engine won't have a working HEI module to fire the ignition coil, I installed a Bosch ignition module. This particular Bosch module came from an 80s Volvo.
In the Volvo application, the module is mounted on its own heatsink, which is then screwed onto the fender. I decided to use this entire module + heatsink assembly. There were other heatsinks used with Bosch modules, but I liked the shape of this one for my Fiero.
[This message has been edited by pmbrunelle (edited 01-03-2019).]
Well that just about sums up the major chunks of work completed so far.
As other blocks of work are completed, I'll post them. The next block of work to complete is the water injection system, followed by the distributor sensor, and everything else after that.
Wow man! Wow! I am totally speechless, that's why I have to type this as I am speechless. What wonderful and innovative work you are doing PMB. You are making Canadian Fieros proud! Bravo!!
In a standard 90s GM setup, the crankshaft position sensor is wired to the DIS brick.
I don't want a DIS; I want my Fiero to have the classic distributor look. Hence, in this setup, the crank sensor is read "directly" by the ECU; no DIS brick in between.
I did, however, make a circuit to convert the crank sensor's weak analog signal into a robust digital signal. It uses a Maxim MAX9924 integrated circuit to do the job. The case is made from CNC-machined nylon 6/6.
Actually Raydar, I think I got the idea of the RAM HD clutch from one of your posts, where you said you were happy using it with your 4.9 Cadillac.
So I decided to buy the RAM HD clutch.
Afterwards, I saw a thread of yours where you mentioned that your RAM clutch blew up I guess I'll see what happens!
Actually, that was a comedy of errors. (Ya' learn... ya' know?)
In the first place, the HTOB Getrag really requires a Cavalier clutch assembly. The Fiero clutch assembly is ~3/8" too tall (flywheel surface to tips of release fingers) to provide for adequate clutch wear.
In the second place, the depression in the center of the LSC flywheel was too small in diameter for that particular disk. The rivets around the edge of the hub actually made contact with the friction surface of the flywheel. Eventually, as the clutch wore, the rivets stopped the disk from pressing cleanly against the flywheel. These were NOT the rivets that held the lining to the disk. The disk was probably only about 10% worn, when the rivets started making themselves known. I can send you links to pics of the "rivet" thing, if you'd like.
Nothing ever blew up. The clutch was fine. The application was... lacking.
After taking the shifter apart, I sandblasted everything. Fast way to clean the spaces between the coils of the spring and other small places.
The sandblast takes off everything and leaves a rough surface that is good for paint to stick onto.
The sandblasted finish is also good at picking up other types of stains, and good at rusting. I don't like the feel of sandblasted threads either.
So, for these other surfaces, I like to wirebrush them smooth, to make the rough sandblasted finish shiny.
Then, to protect (sort of) the unpainted surfaces, I rubbed motor oil on them, but since the shifter is in the cabin under the center console, I don't think the oil will be washed/rubbed away so easily.
From my time in the Militia, I know that guns require regular applications of oil to keep the rust at bay.
In this case, I would expect the metal to eventually become rusty, as I don't plan on regularly reapplying the oil... but since it took 30 years for the original shifter (which didn't have oil) to get light cosmetic rust, I think this shifter will be good for quite a while. Functional, at least.
After taking the shifter apart, I sandblasted everything. Fast way to clean the spaces between the coils of the spring and other small places.
The sandblast takes off everything and leaves a rough surface that is good for paint to stick onto.
The sandblasted finish is also good at picking up other types of stains, and good at rusting. I don't like the feel of sandblasted threads either.
So, for these other surfaces, I like to wirebrush them smooth, to make the rough sandblasted finish shiny.
Then, to protect (sort of) the unpainted surfaces, I rubbed motor oil on them, but since the shifter is in the cabin under the center console, I don't think the oil will be washed/rubbed away so easily.
From my time in the Militia, I know that guns require regular applications of oil to keep the rust at bay.
In this case, I would expect the metal to eventually become rusty, as I don't plan on regularly reapplying the oil... but since it took 30 years for the original shifter (which didn't have oil) to get light cosmetic rust, I think this shifter will be good for quite a while. Functional, at least.
Afterwards, I saw a thread of yours where you mentioned that your RAM clutch blew up
Wonderful work pmbrunelle...
If the RAM clutch fails, consider Canadian made Bully Clutch. I have the same one in both my cars. 2 others here in Calgary have also bought them. (one is 3800 S/C, the other a 3400/Turbo)
I'll keep the Bully in mind. When I bought the RAM, I didn't do much research, I just picked an obvious choice.
Auto modification is a pretty expensive (at least the way I do it) hobby to have, and as most parts vendors are in the USA, I'm sending a lot of money outside of our country.
I kind of feel some remorse over what I'm doing, but not enough to stop auto modification.
For my next Fiero part, I have a quotation from an online Netherlands-based CNC machining service https://www.3dhubs.com ready to go. However, I did send an RFQ to a local Grand-Mère machine shop, so if they want the job, I'll probably order from them.
Since I want to run sequential fuel injection, just the crank sensor alone is not enough. When the computer reads the double-notch on the crankshaft, that corresponds with TDC compression #1 AND TDC compression #4. How to tell the two TDCs apart?
Following the double-notch, the computer will poll (read) the value from the distributor sensor, which shall be different for TDC #1 vs TDC #4.
For the electronics, I started by gutting some el-cheapo ignition modules. I just wanted them for their cases.
Modding the cases slightly so they can accept a PCB on top. Someone who knows Fieros might notice that something's not stock when they see the coil connector on the left is missing.
Drawing the PCB on the computer:
A view of the module's interior, before closing it up:
Mounted on top of the PCB in through-hole fashion is the Allegro A1250 Hall-effect bipolar latch. The legs are splayed to improve vibration resistance, and to make soldering easier. Depending on the polarity (North or South) of the magnetic field going through the sensor, it will output zero or 5 volts.
These magnets are leftovers from work. They are isotropic injection-molded NdFeB with a PPS binder. These particular examples have been diametrically magnetized. One side of the magnet has a North pole; the other side a South pole.
The magnet is located in a groove in an aluminium hat (made by 3dhubs). Silicone potting compound encapsulates the magnet to keep it in place and to protect it from the ozone. The 12 mounting holes allow me to time the hat such that the North pole is centered around TDC #1, and the South pole around TDC #4. Or vice versa; the polarity requirement can be switched in software.
The distributor was gutted and modified to accept the aluminium hat:
Here is the assembly:
All it needs now is some conformal coating to protect the Hall sensor, and some RTV to seal it up. Maybe a dab of RTV behind the Hall sensor too. Last, but not least, these modules need serial numbers.
I figured that an air-to-air intercooler would be tricky to package (at a minimum involving bodywork + sanding), so I felt like trying a water injection system instead. NACA Report No. 756 describes the increase in boost pressure made possible (due to knock suppression) by water injection on a test engine... so why not try to duplicate those results?
This is the water tank I made. It is made of CNC-cut rigid PVC plastic panels glued together (with plumber's cement). The Holley water/meth pump (a re-labeled Shurflo diaphragm pump that I paid too much $$$ for) sucks water from the bottom of the tank.
The tank fits behind the RH quarter panel. No modification to the wheelwell liner is necessary. The top of the water level sending unit is visible; it drops into a hole at the top of the tank. This will allow me to install a water level gauge in the cabin, at a later date.
The first stop after the pump is the filter. From there, the water continues towards the pressure regulator, which is referenced to manifold pressure. A 0 - 5 V pressure transducer will allow the ECU to monitor the water pressure. Pressurized water is supplied to an Aquamist solenoid valve. A stainless braided hose carries the water from the solenoid valve to the engine. Since I'm using a return-type regulator, excess water is sent back to the water tank via the return line.
The connections to the water tank are below the battery tray (hence the battery tray is screwed in place, not welded). On the left is the return from the regulator. On the right, the fill line.
This is what the fill cap looks like from up above. I decided to engrave "ADI", which stands for Anti-Detonant Injection.
Here is a steel mount i made for the filler tube. Like most steel parts I make for this car, I sandblasted and POR-15ed it.
Another post detailing the intake tract mods for the water injection will follow!
There's a rudimentary "pressure regulator" in the oil pump cover (a flat disc and a spring, IIRC), so no, you don't get more maximum oil pressure due to the pump being high-volume.
However, at low RPM, when the oil pump isn't spinning fast, you'll have more oil pressure from the high-volume unit.
I'm using the Fiero 2.8 oil pan, so I decided that I wanted to use an oil pump sold for the Fiero 2.8, so that its pickup tube would be suited for my pan.
When researching whether I should use the standard-volume 2.8 pump or high-volume 2.8 pump, I found out that the high-volume 2.8 pump is the same pump as the standard pump sold for later-model 60° V6 engines, except for the Fiero-style pickup tube.
I figured that I couldn't go wrong with the standard pump for later generations of this engine family. I'm assuming that GM changed the volume to address some sort of reliability problem with the early incarnations of the 60° V6 (I might be wrong, but that's my guess).
I re-purposed the IAC passage in the Fiero intake manifold for the water injection system.
After the machine work, I had the intake plated with electroless nickel (thanks olejoedad for advice) to protect it from corrosion due to the boost juice. Unfortunately, it doesn't seem to hold up to the methanol in windshield washer fluid, so I'll most likely be using distilled water from the pharmacy. I'll have to take some special precautions for winter storage of the car (maybe a purge with 50/50 antifreeze).
Water enters the main gallery, where it then goes out to the six ports.
The IAC passage outlets were drilled and tapped to 1/8" NPT. Just enough material was removed to clear the socket I intend to use for nozzle installation.
Here, the nozzles are installed. Since I used a gel-like thread sealant, I'll wait until next weekend before starting spray tests with water; I want to be sure that the sealant has cured.
The nozzles look like this (stock photo):
The nozzles use stainless steel construction, and include both a filter and check valve. The check valve will keep the IAC gallery full and primed after injection, ready to spray for the next shot.
Following advice from Blacktree, I selected check valve springs that are stiff enough to maintain a pressure of 35 psid when closed, to prevent the water from boiling off. Unsurprisingly (when you buy things from alibaba), the non-specified stock springs were no good for my application; I ordered three kinds of springs from Century Spring and I picked the ones that worked best.
****************************************
Since I repurposed the IAC passage in the intake manifold for water injection, what to do with the IAC airflow?
I plugged the stock IAC outlet, and by drilling an angled hole, the IAC airflow is re-introduced immediately after the throttle plate. Since the angled hole intersects the throttle body's coolant passage, I also blocked off the coolant ports (my Fiero won't have the throttle body heat feature anymore).
I like the way you blocked of the ports on the TB. Never thought about using brass plugs to clean that up. Do you remember the pitch of those plugs by chance because I would like to steal your idea!
Those are actually the stock brass tube nuts (with their O-ring seals) that come with the factory coolant lines. In my setup, these plugs have to seal well, or else I'll have a boost/vacuum leak.
Since the tube nuts have holes in them, I soldered in a piece of scrap steel rod with plumber's solder and flux. I then faced the ends square and smooth on the lathe. Someone without a lathe could use a hacksaw and file to achieve an equivalent result.
When you look at the modified piece from the end, there's brass on the outside, and steel on the inside. So, I painted the end black to hide the inconsistent materials, and since unpainted steel would rust anyway.
In your case, if you don't have the stock tube nuts anymore, I can think of two alternatives:
1. Try something threaded, such as a bolt. The measurements of the factory tube nut are consistent with that of a 7/16"-20 UNF thread. The under-head length of the factory tube nut is 5/8".
2. Tap a freeze plug into the counterbore, which measures 12 mm in diameter. After a quick search I found these which would probably work: Melling MPC-108 (normal freeze plug) Melling MPC-108V (with small venting hole in the middle)
Well, when I started the dream of having a turbo Fiero some 10-odd years ago, I bought myself an ebay turbo.
Since I plan on installing the turbo soon, I decided to do some last-minute checks:
In this graph I am overlaying the engine's expected working points over the map of my compressor.
Basically, this turbo does not appear to fit whatsover with my engine. When I bought the turbo, I didn't know how to interpret a compressor map.
Now I'm investigating if I can just change compressor wheels (T04B S-trim or T04B V-trim), or maybe I want to change the entire turbo. To avoid throwing away too much hardware, I want to stay with an internally wastegated T3 turbine, and a compressor housing with a 2.75" inlet / 2" discharge. The Garrett GT3071R appears to be a good match, but I don't know if I want to spend for a whole new turbo. On the other hand, being too much of a cheapskate is expensive when things go wrong, so further study is required.
My winter boots have zippers running up each side. That's how I stay focused enough to do car work
I tried blowing compressed air into the runner (an attempt to simulate a running engine) while it was spraying, and it didn't seem to have much effect on the dripping. These nozzles have a cone angle of about 70°; perhaps a narrower cone angle would be better.
Gasoline is injected at pretty much the same location, and it too does partially condense on the walls before being picked up by the airflow... This time delay of fuel lazing around in puddles of liquid before making it into the combustion chamber is why we need acceleration enrichment, or else you get a lean spot on throttle tip-in.
So, since water evaporates less readily than gasoline, I think I should expect similar or greater (water) puddling in the intake ports... and technically speaking, I should also use acceleration enrichment for the water injection, in order to deliver the requested amount of water into the combustion chamber.
[This message has been edited by pmbrunelle (edited 04-09-2019).]
The water is really only going to be injected under acceleration correct? The chances of puddling I think is going to be very low because the motor is going to act like a vacuum cleaner at this point at sucking everything in.
I'll only spray in high boost (whatever "high" turns out to be), in conditions where the 91 pump gas doesn't have enough octane to resist detonation. I don't expect the turbo to spool off-idle either, so there will be some RPMs (maybe 3000) before the spray begins.
I ordered the turbo today, and as getting parts in Canada sometimes goes, I'll get it in a month. That gives me some time to work out the kinks with the water injection system, before I assemble the top end of the engine and fit the turbo.
I didn't want to spend too much for a turbo, but I didn't want a cheap knockoff that cannot be trusted, so I ordered a new Garrett-brand T04E 50 trim / Stage 3 .63 A/R journal bearing turbo. I already bought some intake and exhaust plumbing parts; this turbo is compatible with the hardware I already have.
In my research, I was really impressed with the Borg-Warner EFR line of turbos. To begin with the accessories, they have integrated wastegates and blow-off valves! That's a lot of installation headaches saved! Then, the compressor maps... they just work with a really wide range of pressures and airflows. The EFR turbos start at about twice what I spent for my journal-bearing Garrett. However, I'm sure they're worth every penny. For someone doing a more costly / high-end build than what I'm doing, that's what I'd suggest.
The water injection is coming along; I measured the amount of water sprayed as I varied the electrical pulse width. I'll use this information to help me set up / tune the MS3 once the car is up and running.
It takes about 4 ms for the solenoid to open; pulses shorter than that don't allow any water through.
At higher speeds, the solenoid valve (well, the entire system) isn't fast enough to allow for fine control; it's more of an on/off thing.
At lower speeds where the valve can be better-controlled, I'll be able to cut back on the water flow to avoid drowning the engine, and to reduce water consumption.
I have a small vibration problem to deal with; it appears that the pulsating flow from the water pump can cause a "water hammer" effect in the water lines and regulator. I'm going to try a length of soft hose to remedy the issue.
[This message has been edited by pmbrunelle (edited 04-21-2019).]
The following day, I re-did the curves, but while connected to the alternator of my truck, which was idling for the duration of the test.
I wanted to see the difference between running off a battery vs. an alternator.
With the greater voltage, the valve seems to snap open a bit faster.
The difference is most apparent in the high-RPM curves (6000 and 7000). These curves have more of a linear zone where the amount of water injected varies linearly with the pulse width.
Question for anyone with ideas... so I'll have to use studs to mount my new turbo to the T3 flange on the exhaust. I can't physically insert bolts into the holes on the turbine housing's flange.
The T3 flange on the exhaust has tapped through-holes. What is the proper way to install a stud so that it won't loosen? Anti-seize, special threadlocker? What kind of nut do I need?
I presently have an unbranded eBay blow-off valve with no specifications; like I did with the turbo, my goal is to rid this car of most of its iffy eBay parts.
So now, I'm shopping for a BOV.
Ideally, one that vents to atmosphere, rather than recirculating into the air filter can, for the simple reason of simplifying the plumbing.
What seems to be a common theme among BOV manufacturers is selection of a BOV spring based on your engine's idle vacuum. At idle, the spring should be strong enough to keep the BOV shut despite the vacuum trying to open it.
OK, but then what about during high-RPM closed-throttle deceleration? My Fiero engine which idles with 18 inHg of vacuum now goes up to 26 inHg of vacuum. I asked Turbosmart Tech Support about this, and they suggested that yes, the BOV may open during deceleration.
I don't know if I'm missing something, but it appears that atmospheric BOVs may allow the ingestion of unfiltered air
TiAL also has a BOV spring selection chart that is based on idle vacuum (their tech support hasn't yet returned my inquiry)
I could stick with the recirculating type BOV that sends the air back to the air filter can (sucking from the filter on decel would be OK too), but it's more fabrication work. I could also just stick a K&N filter on the outlet barb fitting of a recirculating BOV, but then that becomes an additional maintenance item.
If I go for a recirculating BOV, I'm considering the 1G DSM unit.
[This message has been edited by pmbrunelle (edited 05-15-2019).]
(their tech support hasn't yet returned my inquiry)
