Yesterday morning I drove Midlana to a “secret” car show, held in a well-off part of town. Not realizing that it was by invitation, I just showed up and parked – it’s on public streets after all. They were very welcoming and there was a little bit of everything; check out more pictures on my Instagram page (#midlana). I finally got an up-close look at an Ariel Atom 3S and while I’m well aware of Ariel, I don’t keep up with their product line, so it was surprising to see a turbo Honda K24 in it. The driver was very friendly and explained that he brought the car while in England. He had a slight accent, which oddly always seems to go together with being able to afford nice things – the 3S starts at $89,750.
That afternoon the boost controller duty cycle map was tuned, so with that working well, closed-loop boost control was reenabled. Next on the list was to better tune the Volumetric Efficiency (VE) table. The tuner was supposed to have done this, but with air/fuel feedback disabled, it showed that the VE values were off. A few runs provided data points for correcting the table, so the new values were entered and the table smoothed. The plan was to load up the new VET table Sunday morning and see how close I was. Alas…
Sunday morning – the car started making a new noise… sigh… Really? What now? The alternator belt started squealing and it turns out that I underestimated the forces on the alternator bracket, which looked as though it fatigue-failed. In hindsight it really needed a stiffener, so one was added, the car reassembled, but by then my car time was up and it was time to visit mom.
The rear suspension rod-ends were upgraded and like the front, the job took all day. Before starting in and with the car on blocks, the tires were pulled back and forth, side-to-side to see how much play there was… there was a lot, maybe 5 mm* or so. Due to how the toe-control links tie into the uprights, any amount of play gets mechanically magnified out to the surface of the tires. This is the prime reason they were replaced, them loosening up so fast. I can’t blame anyone but myself for having used such poor quality parts in the first place. The first picture below shows the enormous difference; not only does the part on the left have specifications, its wider housing is about twice as strong, but also, the wider housing allows a wider Teflon bearing surface which has higher ratings and will also last longer. The difference in price is substantial, but given their task, I no longer trust the cheap ones. For background, the cheap ones were about $3, mid-range U.S.-made parts are about $11, and the good ones are around $26. The highest-quality parts were used selectively in the highest-loaded positions, and the mid-range parts used everywhere else.
I briefly considered selling the used joints to someone who would use them in non-critical applications, but thought better of it and into the trash they went. I don’t need that on my mind.
A few readers pointed out that since I twisted the steering arms, it shortened the steering rack to upright distance – that’s correct – and they wondered whether bumpsteer might now be an issue. I pointed out that bumpsteer, while a quantifiable and measurable value, is subjective to the driver. If the driver can’t tell the difference between “X” and “Y” amount of bumpsteer, it doesn’t matter. (I should have driven the car with the Miata tie-rod ends, but when I saw that there was some bumpsteer, I set out to eliminate it. In hindsight, even though there was some, it may have been fine as it was.)
The car was taken out for a test drive and I couldn’t detect any bumpsteer, but than again, it was Friday traffic and impossible to do any handling tests. During the drive I was heading downhill toward a traffic light… well, let me back up a bit and explain something. When driving, especially sports cars, we program ourselves to be hypersensitive to certain things. Just getting a glimpse of a black-and-white car is enough to make people back off the gas. There are other things as well, like smelling melting plastic, or coolant, or brakes that gets immediate attention. Such was the case slowing down toward the traffic light at the bottom of the hill, a definitely strong smell of really hot brakes. When I pulled the calipers off the rotors, a tiny bit of the brake pad broke off along the bottom edge; apparently the brake kit I used allows the brakes pads to hang slightly over the inside edge of the rotors. Smelling hot brakes made me worry that a piece might have gotten jammed somewhere and was causing the brakes to drag. Then I looked over and there was an old truck pulling a very overloaded trailer full of concrete debris – what a relief… for me, not so much for him!
In the last picture – it’s kind of hard to see – is a bunch of tiny white specs. Those all are dried-up bug goo, or more specifically, dried up locust goo from when I ran over all of them. That stuff is really hard to get off and just getting it wet doesn’t seem to do the trick. If anyone has some wonderful stuff they use, let me know.
*I’m making an effort to start using metric values. I’m already getting heat for not using metric values in the drawings, so I have to start somewhere!
I had put off upgrading the poor-quality rod-ends, guessing it was going to take a lot of time – it did. It took all day to get the front suspension done. What made it take longer was also twisting the steering arms to move the steering rod ends away from the brake discs. When Midlana was first assembled, I ground back the rod end housings to make them fit, but it always bothered me about how much they may have been weakened. I really wanted to use unaltered rod ends for the safety factor if nothing else.
It’s been years since I used the gas welder. The last time, the acetylene regulator failed, so bending the arms didn’t happen. With the regulator fixed and some time to myself, it went okay bending the first one. (Using a large rose bud tip with high gas flow is intimidating… it’s a ton of energy to be pointing about!) Twisted the arm to where it seemed about right, let the upright cool in still air for hours, and when reassembled, found it had been twisted a bit more than necessary, but it’s perfectly usable.
Heated up the second upright and got it twisted nearly to where it needed to go, and then the gas torch gave me a scare. The blue inner cones denoting the hottest part of the flame started getting smaller, which meant that either oxygen flow was increasing or acetylene flow was decreasing. About the same time I heard a loud pop, looked over, and the oxygen gauge was pegged (I’d set it to about 15 psi, and now it was off the top end of the gauge at over 100 psi). The loud pop was the pressure relief valve, keeping it from rising to tank pressure, which was around 1000 psi! Apparently the oxygen regulator has failed, probably the same way as the acetylene regulator did. I wasn’t bright enough to guess at the time that if one failed, the other probably wasn’t far behind, but no knowing what failed, it wasn’t obvious. Anyway, the good thing is that the arm had been twisted enough that it’s good, though it’s not quite twisted the same as the other side. Oh well.
I haven’t measured bumpsteer but hopefully it won’t be enough to bother messing with. I’ll probably just drive it and see if it’s detectable. By the way, the rust you see is mostly due to doing one track session in the rain. The lesson is to: buy plated rod ends, use stainless parts, paint them, keep they oiled, or don’t drive in the rain!
After last week’s findings, a lot has to be redone. The wastegate spring was changed to 12 psi and the 3-port controller swapped back in, the wastegate spring duty cycle table has to be redone and the volumetric efficiency table checked.
A surprising number of people have asked why the air scoop points aft:
1. When cornering, the inside front tire throws road debris inboard, with some landing in the area of the air filter. With the scoop pointing aft, it shields the filter element from small rocks.
2. If and when the car gets caught in the rain, it eliminates water intrusion (especially when parked).
3. There is a large rotating air “bubble” behind the windscreen. This was made apparent when the oil O-ring failed and oil ended up on the inside of the windscreen. Air flows up over the top edge of the windscreen and circles down toward the engine cover, then forward to the base of the windscreen again. As a result, air is likely flowing into the inlet of the rear-facing scoop.
4. There have been suggestions on cutting open the scoop to fix the issue. While appreciated, cutting back the the scoop any amount exposes the filter to the elements, or opens up a hole into the engine compartment, allowing hot air to get sucked into the filter.
5. At car shows, it keeps troublemakers from thinking it might be funny to pour a drink into the air inlet.
A proper scoop will be made using the moldless-construction technique.
Next week is a short week for me so I hope to check several items off the to-do list. First up is replacing all the low-quality rod ends with higher quality parts. At the same time, I want to heat and bend the steering arms to move them away from being so close to the brake discs.
Lastly, it looks like my brother and I will be signing up for the infield track at Autoclub speedway, one that’s more twisty and more attractive to cars that handle well. Top speed in this configuration is around 130 mph instead of the 150-160 on the big course.
What a bitch… is that crazy or what? In hindsight, it explains virtually everything that happened during the months of fruitless troubleshooting:
1. MAP-drop never happened with the engine cover off (when I did my 12 runs in a row and deemed it fixed due to an “rpm limit issue”).
2. It never happened with the car on the dyno because the engine cover was always off.
3. The scoop collapsing shut caused the air filter to get pushed off its mount – it was never an air filter issue.
4. Last week when run without the air filter element, there’s apparently enough leakage between the air filter and engine cover that it can’t develop enough of a suction to pull the scoop shut.
5. Upgrading the wastegate made no difference.
6. Removing the muffler had little effect – if anything it made it worse.
Before realizing what was happening, it seemed like the air filter element was getting sucked down onto the inlet bell. A K&N filter had been used before, but was upgraded to a 33-2135 unit, designed for the 403 hp late-model Chevy truck. That’s close enough in power that it should do fine, and in addition, a cross-support was added across the airbox below the filter. The K&N is roughly 1″ thinner than the enormous Amsoil unit, so that’ll improve air flow heading into the bell-mouth inlet.
I can’t begin to explain how much of a relief it is to FINALLY have this brought to light. Of the dozens of suggestion I received, not one suggested a possible filter or scoop issue. Needless to say, the air scoop is coming off until a real one is fabricated.
The Second Edition is now live on Lulu! Also, for those who’ve been pestering me, it’s now available in both perfect-bound and coil-bound! For those who have the First Edition, fear not, the changes are not large, and include hundreds of minor textual changes to improve readability. The few technical changes are already known and will be listed on the Midlana forum.
First up, a neighbor has had two enormous Ficas-type trees in his yard for decades and never trimmed them. The thing about trees is that they can be ignored for years, but that always comes at a price. Early Saturday morning, a huge branch came down on his house and two cars. The roof on the house is visibly pushed in, one car had its side mirror removed and a large dent in the roof, and the second car had the branch come through the windshield in addition to dents on the roof and hood. A tree trimmer who sized up the situation several years ago said that trimming just one tree would be well north of $10,000 and refused to even bid. Because about half the branch landed in the street – city property – a city crew promptly arrived and removed that portion of it, leaving the poor owner with the branch still sitting on his roof. Did I mention it’s supposed to rain tomorrow? I drove by today and was surprised that he’s done nothing. I’d have at least bought a chainsaw and cleaned off the roof so a plastic sheet could be dropped onto it, but it’s his house…
The wastegate spring was lowered to 15 psi/200 kpa and more testing done. The MAP drop was back, big surprise, but this time, I heard something the instant power dropped from 200 kpa to 150 kpa; that’s never happened before. Removing the engine cover showed that the air filter element had been partially sucked onto the bell-mouthed intake. I wonder… could this have been it all along? Could it be the air filter element getting sucked onto the intake bell and reducing the filtration area by probably 80% and becoming a restriction?
As a quick test, the air filter was repositioned and another run performed. Again, the element got sucked off one edge and again, MAP dropped off. Next, the filter was removed completely, and guess what, not a single issue for the rest of the day. I even drove out to where some high-speed runs could be made to red line, and it worked just fine. Well isn’t that interesting.
The filter only recently started having one edge pulled off its mount, but the MAP issue has been happening for months. Full disclosure: I learned that I’m using the filter element upside down. That didn’t seem to matter at the time, but maybe it’s a big deal. There’s a screen mesh on one side of the filter – the top in my setup. It’s possible that without the support of the screen, the folder paper element can’t resist the suction and gets pulled onto the face of the inlet tube, becoming a blockage. What’s highly suspicious is that MAP always drops to 150 kpa, as it did today, which makes me suspect it’s the same failure mode each time. The only questionable part of this theory is that the filter doesn’t appear deformed like it’s been pulled away from the screen; the glue on the back is in good condition, but there is a dent from the inlet. Even so, I will add a support and see if this has been it all along. There’s a few ways to fix it, some easier than others. I’ll first try an easy way, running a rod or rib under the element to keep it up off the inlet bell.
What’s a little bothersome is that if that’s happening, I wonder about the fairly small air inlet on the scoop; could it be getting sucked shut? Looking at the air scoop, there’s two suspicious pale white marks along the top edge – I didn’t put them there and they didn’t used to be there. It’s almost like the plastic has been deformed. I’ll add a support brace to that as well, which will open up the inlet a bit more. I don’t want to get overly excited, but this is the first time something’s been found that is directly connected to the problem I’ve been chasing for months. We’ll see.
And finally, expect some Midlana news this week, so keep checking here, the Midlana forum, or Instagram 🙂