31 Jul 2016

Here’s “boost creep”: an uncontrolled increase in manifold pressure, seen here as the light blue “MAP” trace at the center of the figure. This is with a 4-lb spring which in a perfect world limits boost to 4 psi, but the manufacturer notes that actual boost will vary engine to engine. In mine it’s about 7 psi (150kpa), which is fine if it stays there. It holds pretty well until about 5000 rpm, but then the exhaust gas volume becomes too much for it to vent off and the turbine speeds up and increases boost. I let off the gas when it hit 13 psi at 6100 rpm because it was trending upward pretty fast (and because I was already going 84 mph). The right way is to do all this on a dyno – maybe after boost is under control.

Parts were (again) ordered to change to a dual 38mm wastegate setup. I figure that worst case, the two don’t do the job and the header has to be redesigned, and both will still be needed in that case so they aren’t a waste. Somewhat related, I found this funny quote which is a pretty good description of what it’s like to live with a turbo engine, where:

… it’s like having a cokehead pornstar girlfriend. There’s going to be unparalleled excitement and thrills, but a lot of unexplainable downtime and a likely violent ending that leaves you broke and insane…

30 July 2016

Back-filled 2011 ; all early posts are under the “Archive” link above.

Disconnected the upper wastegate hose to ensure there’s nothing keeping it from working properly but boost creep persists. I had a talk with a buddy who really knows turbo:

“You increased compression 8:1 to 9:1 (~12% more flow), right?” Yes

“You added cams that flow more air.” Yes

“You switched from a 0.82 AR turbine housing to a 1.06.” Yes

“You’re running E85, which produces a lot more exhaust gas than gasoline.” Yes

“You built an exhaust header with poor flow to the wastegate.” Um, yes

He went on to explain that the 1.06 AR turbine housing is far less restrictive to exhaust flow. That’s a very good thing but means that given the choice, exhaust gas would much rather go through a low-restriction turbine housing instead of making a hard turn into a smallish wastegate. He recommended one or more of the following:

Change the geometry of the header to provide a more direct shot at the wastegate: This is the right way but there’s no room

Add a second 38mm wastegate: There’s room, just

Replace the single 38mm wastegate with a 60mm unit: The “nuclear option.” Builders use these to regulate 1200-hp engines, but much more often, as a Band-Aid to limit boost creep caused by poor wastegate routing. Not an option because chassis tubing blocks the area.

Of the above, adding a second 38mm unit works the best short of starting over. For what it’s worth, two 38mm wastegates have the equivalent valve surface area of a 54mm unit, so it’s not vastly different than going with the 60mm unit.

 

28 Jul 2016

Well that didn’t last long. Woke up this morning realizing I didn’t need some of the parts ordered last night for the exhaust modification, while a couple other parts had been forgotten. Canceled those parts, then found the wastegates are backordered with an unknown delivery date (at least from this vendor) so everything was canceled. It’s not entirely bad as it gives time to do a few more tests, like disconnecting the upper hose from the wastegate (which can keep the wastegate valve shut if pressurized, increasing boost). It “should be” zero but if it isn’t that would do it, but if it’s zero, that leaves wastegate routing and/or the engine changes (compression, cams, and larger turbine) as causing the wastegate to be unable to regulate a setpoint.

It’s too warm and humid to be doing much in the garage so the time was spent back-filling the 2012 blog.

27 Jul 2016

Swapped out wastegate springs for what should have been about 12-13 lbs and it just laughed as it shot past that. Don’t have anyone to blame but myself so now I get to fix it. I know what’s causing it, the near right-angle wastegate take-off angle, but it’s near impossible to do anything better with this compact header layout. I know it can work, it did on the last design, but on that layout, exhaust gases had more of an even split about which path to take. As it is now, exhaust shoots down the primaries straight into the turbine with maybe just a glance off to the side, “what was that?” as it flies by the 90-deg junction.

The hope is that the second unit makes the boost manageable and the parts are already on order. Such is the consequence of endlessly changing things. The plan is to put the brakes on such activities and spend more time driving the car.

26 Jul 2016

When engine “v1.0” was tuned, one of the many ECU settings was a mode called “deceleration fuel-cut.” It made sense to have, cutting fuel when coasting, so it was enabled and not given any more thought. I recently read that a side effect of fuel-cut is that it makes engine braking much more pronounced and wondered, especially after watching the car jerking back and forth in the Streets of Willow skidpad video, whether fuel-cut could have been the cause. I also wondered if it could have contributed to the Turn 9 spin at Willow Springs, where letting off the gas caused enough engine braking that it helped initiate a spin (that’s it, not a lack of driver talent!).

Just for fun I tried disabling it – and it was a night and day difference! Throttle control is much smoother and much more controllable. I asked a buddy about this, who reached through the Interweb and “dope-slapped” me, saying there’s a lot more than just enabling it, that it’s only supposed to be enabled when:

Throttle position is <2%

When the condition persists for more than 0.5 seconds

When MAP is just below idle MAP

That said, after it was disabled I’ve yet to see a downside. The car runs fine and doesn’t flood or stumble, and the spitting, popping and crackling under throttle-lift makes me smile, so for now it stays 🙂

24 Jul 2016

Took the car out last night and found I was playing a bit fast and loose with the tune – and that it pays to have limits set! Checking out boost, the engine hit a rev limit which was unexpected since it was well below max RPM and MAP. Turns out that the many changes in the engine caused the engine to run leaner than the previous engine, which shouldn’t have been a surprise. This morning was spent checking the logs, finding the correction factors, and adjusting the volumetric efficiency table (VE).

The right way of setting up the VE table is to do so on a dyno. Short of that, having a buddy in the passenger seat make the adjustments and issuing instructions on whether more or less RPM or MAP is needed. I just want to get it close then decide whether to take it in for a proper dyno tune.

Related to the above, I read something that I hope doesn’t become an issue. A guy who knows these engines well and has a lot of dynamometer experience tried some “performance” cams. Many are intended for normally aspirated engines and have increased cam overlap (the period of time that both the intake and exhaust valves are open at the same time). For a turbocharged engine this can cause issues. Because the air entering the engine is compressed, it can allow the air to push through the engine and out the exhaust while the intake and exhaust valves are open. As a result, fuel consumption may increase yet the engine doesn’t make any more power. The valve overlap can also affect idle, causing less vacuum than if it had stock cams. While my engine is still being broken in, it’s hard not to notice that idle vacuum is around 32 kpa versus 23 kpa with stock cams. Granted, as the engine breaks in that value should improve, and it pretty much is what it is. The running lean issue is probably a good sign though because it means that the engine has a higher VE than the previous one. It’s a never-ending learning experience.


Test drove the car and the air-fuel ratio is much improved. After some test runs the VE tables was further dialed in and is nearly good to go. The test runs found that the weakest spring produces around 150 kpa (~7 psi) but starting at around 5000 RPM starts creeping, reaching 180 kpa (~12 psi). That’s not great but I wonder if I install a 12 psi spring whether it’ll peak out at the same maximum value. Hope so, else there’ll be some exhaust manifold rework to do – boo, especially in this heat.

Lastly, I recently watched a documentary on a guy who’s company I’d see before and wondered what the heck it was about. It recently relocated to the area so during a test drive I figured it made a good photo op, so there you go. Google “Dr. Bronner’s all one soap” for a look at a very unusual person. In the documentary, it mentioned him telling the guy who picked him up while hitchhiking, “I just escaped from a mental institution.” To say the lest, he had some very unusual views on life, nature, and the universe, while at the same time making a natural soap that’s available just about everywhere.

23 Jul 2016

It’s been over 100 degrees in the garage this week, which takes all the fun out of working on the car. Still, I scaled up the measurements my buddy gave me for the Nissan GTP air scoop to see what it might look like. As said, it completely changes the look of the car. If I saw it and didn’t know better I’d think it was fake or at least pretentious, but it’s 100% functional and necessary so it is what it is. I asked him whether I was being dumb to make it from aluminum instead of composite. He said that for making just one, aluminum is probably a better use of time and it’s also less messy, so aluminum it is.

Also swapped in the weakest wastegate spring; I may take it out for a test drive later this evening after it cools off to find what the actual-versus-rated boost. It’ll also show how much boost creep there is; with the less efficient routing I’m hoping it’s not too bad.

Back to the weather, the air conditioner is sure appreciated, great for sleeping. Some people suggested a swamp cooler but I feel they’re only for really low humidity areas. We generally get several weeks or even months of monsoon-like weather around here, really warm and humid, and the air conditioning is great for both cooling  and lowering the humidity, something a swamp cooler can’t do. I’m shocked by how much water is in the air; opening the bedroom door and letting it work to cool our small house fills a coffee can with condensate in about two hours.

An interesting aside; I went to a local Home Depot and when I walked in, was hit by a wall of really humid air. Sure enough they use swamp coolers and with them adding humidity to the already high levels, it was actually less comfortable inside the store. There were workers running around with broom squeegees. The combination of high heat, high humidity, and a cool concrete floor meant that – like an iced drink on a table – the cool floor surface was condensing water out of the air… think, hundreds of gallons. I thought they had a water main break but nope. That was pretty amazing.

18 Jul 2016

A number of things:

First off, it’s very important to take time out to make wife-points. I was pleasantly surprised (even after 20+ years) the effect of building a couple garden vine supports had.

Progress getting the car back to normal will take a while because of how much was changed. In a way, it’s like starting all over again, having to set up – or at least confirm as correct – every tunable parameter.

The idle is now good to go. The ECU configuration software has a “feature” where if you click on a field, even if the cursor is not in that box, if the mouse wheel is moved any amount, it changes the value. What’s bad is until you click on a different box, the first box is still “live” even if it’s off the screen. I apparently did that with idle control, unknowingly changing it from “rpm” to “coolant temperature” based calculations. As a result, it was waiting for the engine to get to 1000 degrees C (!) instead of 1000 rpm! While harmless, it’s a reminder how easy it is to mess things up; fortunately I didn’t mess up something that could have caused real harm.

With the idle solved, next up was boost control. Because so much has changed, boost has to be reconfigured from scratch. I messed around with the boost control values, hoping it would be close enough. Nope, so I have to back up two steps and start with just the wastegate spring. My notes show that the 50 kpa spring is in there now but it wasn’t limiting boost at that value. The spring rating rarely matches how it works in a specific setup, so I’ll swap in the weak spring and see where things are. A single 38 mm wastegate worked with the old header but this header’s wastegate flow path isn’t as efficient. I think it’ll be fine still but have to confirm what the boost is with a given spring.

Lastly, work is progressing on the cardboard templates for the air inlet scoop. A buddy gave me numbers off the air intake snorkel from a Nissan GTP car, which isn’t so terribly different in terms of airflow. Scaled up to fit the intercooler, it’ll be a fun project. While making it from composite is probably the right way to do it, depending how the templates turn out I may just go with aluminum. One loose end is decided whether to pick up air to the air filter from the same duct, or run a hose over to the side vent where the intercooler used to be.

Oh, and enabling comments here hasn’t gone well. I think I’ve gotten about half a dozen real comments, and about 25 BS spam ones. If you want to contact me or discuss something, please do so through the Forum (link above).

 

11 July 2016

I forgot to mention yesterday that I did some quick “hand sensor” airflow experiments about where to get air for the intercooler. There are three possible locations: each side of the main roll hoop, above the top edge of the windscreen, and above the main roll hoop.

Each side of the main roll hoop

I was unable to reach back to feel what the airflow is to the sides of the main roll hoop. I would like to retain rear visibility and some might wonder why I care. Try tipping your rear view mirror up in your commuter car and driving to work that way; it may be a little thing but can be a nuisance. Granted, side mirrors can be added to handle losing the rear view mirror, but it would be nice not to. Retaining rear visibility seems to argue for side inlets but there are two strikes against that approach. First, there’s only a couple inches above the intercooler before it intrudes into the rear line of sight. Also, just having side scoops completely block the center mirror’s view of the car’s blind spots, so going with side inlets isn’t an automatic winner. In fact, depending upon the size of the side scoops, they may even block rear visibility of the side mirrors as well. Before writing it off entirely, airflow needs to be measured.

Above the top edge of the windscreen

The airflow is strong and very smooth as expected and is the clear winner airflow-wise but means running a large duct back over my head and down to the intercooler. Visually it’ll be a big deal, but having a wide scoop along the front edge means it could be fairly low profile. I’ll have to do a mockup to see how wide it can be before it gets in the way of getting in and out. Of course, a shift in perspective is to consider it a feature – a roof. Depending upon shape, it could serve to cut down on both wind flow and perhaps even as hard points for mounting future doors. It would take the most fabrication and time though.

Directly above the main roll hoop:

Airflow was also good though about 6″ higher due to turbulence from the upstream chassis tubes and windscreen. Right below the main roll hoop the airflow was flowing forward which wasn’t surprising, but it was still impressive how airflow could be opposite just inches apart. Anyway, this is the most direct and shortest (and lightest and least labor-intensive).

Regardless where the inlet is placed, a good first step would be to use manometers to measure pressure differential. Take a length of clear tubing, staple it to a board in a U-shape, fill it slightly with water and then run each end of the tube to the test points. Those points could be both sides of the intercooler, or for an even more direct comparison, measure differential pressure between the three inlet points, looking for the highest pressure of the three.

For mock up purposes, cardboard can serve well, cut and taped to the intercooler with plastic drier ducting serving to route inlet air. It’ll look odd but I’m used to that 🙂 Or just go straight to fabrication, making it from aluminum sheet or composite.

What I don’t know is how large the inlet need to be; obviously the bigger the better but at some point it’s a case of decreasing returns. I haven’t found any equations or rules of thumb for how large the inlet should be relative to the frontal area of the intercooler.

Lastly, air to the air filter needs to be picked up from somewhere other than the engine compartment – I just haven’t gotten around to it. As it is now, at the end of the test drive yesterday, inlet air temperature was 46 degrees C (with no boost) so it needs to change. It’ll have to come from either the same ducting feeding the intercooler or a separate inlet to the top of the engine cover.

 

10 July 2016

Good news, after I scared myself this morning.

Took the car out to break in the new engine and all was going well until I slowed for a stoplight. There was a ugly clunking sound coming from the back somewhere only when slowing and only when turning left. Pushing in the clutch showed that it was related to wheel rotation instead of engine speed – a good thing! Pulled over and looked the car over and found nothing. It almost sounded as if I hadn’t reassembled the CV axle bits properly, but the noise only happened when slowing down so that seemed unlikely.

Got the car home and looked over everything again – still nothing. Figuring I may as well start with confirming that the wheel nuts were tight, I first rocked all the tires – nothing. Went around the car with the torque wrench and all was fine, right up until the left rear, where one nut was sort of tight and the other three just hand tight. I’d have bet money that I had checked all the wheels but obviously didn’t – glad that didn’t end differently. I honestly thought I’d grown through such nonsense and was more methodical. It just goes to show that we can be our own worst enemy.

With that out of the way the fuel tank was filled and I kept driving. The engine acts different, partly because of the different parts and also because the now-old tune I worked so hard to have a perfect idle was once again somewhat unstable. Oh well, that’ll get resolved again.

The break-in process consists of numerous gradual accelerations in order to seat the rings against the cylinder walls. The thing is, with such a light car, there isn’t room to make long smooth accelerations, especially now. That is, during summer weekends around here everyone heads to the coast highway, resulting in really heavy stop-and-go traffic. Instead I headed inland, hoping to beat the heat and get clear of the traffic (I succeeded on one count). The car ran fine, no bad noises or leaks, so I kept on going. I figured a good “dyno” for applying a steady load would be a hill – a big hill – and Palomar mountain again summoned. Once again I got stuck behind a slug of a driver. I know it’s bad when I feel like checking the speedometer to confirm it’s not just me – nope, 18 mph is just plain ridiculous. It’s not so much that they were going that slow, but pull over already! After dispatching that car, drove to the top, passing as always, groups of sportbike riders in their full leathers getting ready for their next assault. Every time I see them I just shake my head, wondering how is it that there aren’t cops all over them.

I haven’t decided what to do about the back end styling. I’ll probably use the same mesh as above to  fill the lower opening. All the better to get rid of engine heat.

Speaking of styling, during the drive I passed a car that I had never seen before. Based solely on it being a three-wheeler and seeing the partial name “…..hall” on the side was enough to have The Google cough up “Vanderhall Laguna.” Styling a three-wheeler is really difficult; for one, people “expect” four wheels, so you’re already at a disadvantage. Also, if the manufacturer uses a typical OEM FWD drivetrain, the engine compartment needs to be tall enough to cover it, resulting in the car looking a little odd, a bit taller than styling would prefer. Then there’s the front tires which couldn’t seem to make up their mind whether they were “open wheel” or not, with the odd combination of small fenders and fenderwells. What’s harder to overlook is the four separate side pipes. It’s as if we’re back in the 1970’s era of kit cars where many had the fake exhausts – but I suppose they could say a few things about Midlana as well. At $77,000 I wish them well.

http://vanderhallusa.com/

vanderhall-roadster

9 July 2016

Fully warmed up the car, chased several leaks, put it back on its wheels, cleaned it up and it should be good to go for a test drive tomorrow. Nothing fancy, just breaking in the rings. The existing tune should be “close enough” but I’ll still keep an eye on the AFR, though since the rpm and loads will be kept low there shouldn’t be any issues. Got up to about 90 degrees F around here which took some fun out of being in the garage.

Assuming the test drive is uneventful, the next step is to put about 100 miles on it, then schedule a tuning session.

7 July 2016

The exhaust is done. It looks a little odd because of the V-band flange at the end of it. It’s there just in case I ever run the car somewhere that a muffer’s required – it’ll get added on external to the car. The turbo works well enough as a muffler that I don’t run one on the street. The flex joint is to handle the differential expansion between the exhaust pipe proper and the wastegate tubing – stainless expands a lot more than mild steel when it gets hot.

In other news, my brother signed up for the Silver City (Nevada) hill climb – Google it to learn more. He’s been busy prepping his car and it’s all painted now – finally! Anyhow, he had to remove the drivetrain to track down an oil leak and knocking sound. The former was from the front transmission seal leaking – easily fixed, and the latter was from the transmission hitting a chassis tube. What sucks is that after getting it all back in yesterday, it’s dripping fluid from the clutch slave cylinder – he’s not in a good mood. I didn’t sign up for the event since I don’t want to rush breaking in the new engine and getting it tuned.

5 July 2016

I’ve always wondered why I had such a hard time clocking the turbocharger sections. The center section is free to rotate with respect to both the compressor and turbine housing but not just any way I want because – at least with Garrett units – the 1.06 A/R turbine housing interferes with the oil and coolant hoses. Worse, once it’s clocked it can be difficult or even impossible to tighten the turbine housing bolts, and that was when I noticed the big-ass snap-ring on the compressor section. With the suitable snap-ring pliers, the compressor side can be removed entirely which makes it much easier to tighten the turbine housing bolts and the coolant and oil fittings.

4 July 2016

Happy Fourth everyone.

A lot has been happening and it’s become apparent just how much I changed and how much has to be redone. Moving the turbo had all kinds of unintended (read: not thought through) consequences; every hose and tube connected to it changed. A fair amount of today was spent fabricating the seemingly simple intake elbow, but which had to be made just so in order to fit. Because I switched to a different type intake air temperature sensor, it needed a custom mount (thankfully I already measured its resistance curve). The new air sensor has a much faster response time, important in a boosted setup and is a Bosch 0 280 130 085 if you’re interested.

As you can see from the pictures the intake tract is very compact and the only real concern is how much the engine moves around due to torque. Don’t want engine movement to be able to pull a hose off, which means the fore and aft engine mounts need to be fairly stiff. I have a block of Delrin to fabricate a stiffer rear mount if necessary.

Moral of the story is that I need to quit fussing with the car and just drive it. Unfortunately there’s still the big project of fabricate the ductwork to get air to the intercooler. It’s either going to make the car even look more comical – or more bad-ass, not sure which.

While there’s still stuff to do before first start the list is getting shorter. There’s fabricating the wastegate tubing down to the exhaust, lengthening the wires for both the boost controller and the O2 sensor, finishing up the coolant bleed lines, and not much else, so it’s safe to say after checking everything for tightness, it should be running pretty soon.

Oh, almost forgot, that last picture of the wiring, notice how the green wire has backed out of the relay connector. I wonder how long it’s been that way. You find a lot of stuff isn’t as it should be when you take apart a car.

Lastly, if you want to comment on anything posted after today, click on the post’s title, which will almost always be the date, then scroll down to the bottom for the comment field.

2 July 2016

Success, but only after migrating the site myself. Everything seems okay but please let me know if something’s not working or is missing. As said before, the diaries – I mean blogs – will be m0re fully populated over time and is currently back-filled to 2014.

Another thing that will be backfilled are post Categories and Tags, which will go a long way toward finding past posts more easily. The Archive page has several alternate methods for searching posts.

Finally, comments are enabled today forward though I suspect spam might become a problem. Guess I won’t know until I try it.

Now, back to work on the car!