30 Nov 2014

All the dry sump parts arrived last Wednesday except for the pump itself. There’s plenty to do without it though: modify the tank for a -16 scavenge fitting, add a -10 fitting to the dry sump pan for the turbo return line, mount the tank and pan, attach the crank pulley and dry sump mandrel, and fabricate an alternator mount. I got about half of this completed.

The tank and pan were modified without issue. Attaching the pan to the engine however, hit a snag because the manufacturer didn’t leave enough space around the OEM bolts along one flange – unlike all the other flanges – and the bolts jam against the wall of the pan. It’s not a big deal but annoying since the hardware stores with metric Allen head bolts were closed for the holiday.

The ATI crank pulley went on no problem, until offering up the dry sump pulley mandrel that mounts to the front of it – the OEM crank pulley bolt interfered with it and had to be shortened. With that fixed, the mandrel (for the dry sump pulley) was attached to the nose of the crank pulley with three supplied screws and surprise, one had a different thread pitch. Nothing could be found in the documentation to indicate that this was on purpose, so a note was sent to the dry sump supplier, which of course is closed until Monday.

As long as there’s stuff to do, none of the above is a real show-stopper, but it is a little surprising. Perhaps it’s because it is a dry sump that’s the issue; maybe only pros tend to install them and they typically have real shops, do everything custom anyway, and have a good stock of fasteners. Meh, it’ll get done regardless.

Then there was the alternator mount, which consumed a surprising amount of time. As said before, it would have been much easier to use a custom alternator, but it being sole-source was troubling. It seemed wise to sink the time into it – once – so that a Chevy C1500 diesel truck alternator fits in case it goes out in the middle of nowhere. Anyway, it’s done – probably. The concern is the narrow adjustment range due to the alternator’s arc when tightening it down. If I can’t get a 39″ belt it may take another redesign, but like I said, it’s better to have it be painful one time instead of every time the alternator needs replacing.

In other news, Lulu, Midlana’s book printer, is offering another of their year-end discounts. Currently (subject to their whim) you get 35% off the Midlana and Kimini book, so now’s the time to order for Christmas!

23 Nov 2014

The dry sump bits are on the way. For some reason, a dry sump kit doesn’t seem to save money. The least expensive way (at least for the Honda K20) turned out to be to piece together a system and while a bit more work it allowed picking the exact parts wanted instead of settling for pre-selected components.

While the 70 amp Civic alternator fits, a quick check of the power budget found that: radiator fan (20A), lights (10A), water pump (10A), fuel pump 15A*, and everything else probably around 5A, for a total of 60 amps. Granted that’s worst case – at night in stop and go traffic – but it’s a bit close for comfort on the Civic unit. The single-wire unit my brother loaned me is only rated at 60A so that won’t do either. Powermaster has larger units, but they’re expensive and sole-source. That last bit is the main concern, a failed alternator at a track day or somewhere far from home would mean coming home on a flatbed. For these reasons, a 100 amp domestic pick up truck alternator was chosen. It’s far easier to source, is a bit larger, and also about half the cost of the Powermaster unit. It’ll do but will need to be rewired.

The rivets in the engine compartment floor panel tray was removed in order to get the panel off. It makes it far easier to swap out the oil pan in-place without pulling the engine. Even then though, one of the floor diagonals was in the way and had to be removed – it’ll be welded back in and a new panel fabricated. While it seems invasive, the panel and rivets weren’t in great condition after I left a handful of rivet heads in the road after bottoming out several times. (That’s since been corrected by swapping in stiffer springs and stiffening the shocks slightly.)

At the end of the day the oil pan and oil pump were out. The only dry sump part on-hand is the front pulley, ordered early since it’s needed to determine belt routing and alternator placement. For a while there was some thought about making the dry sump pump and alternator mount to one part. However, it doesn’t save any work, material, weight, and plus it means if the alternator choice is ever changed, the entire bracket has to be reworked. Best to have a dedicated alternator bracket.

Lastly, it looks like the dry sump pump will be mounted at the left rear corner of the engine compartment. It’s easier to access there and allows a taller unit rather than placing it up front behind the radiator. I’m not too thrilled about pushing oil all the way to the front of the car then all the way back. Of course, if an oil-to-air cooler ends up being installed in front of the radiator, it’ll be doing just that. In that case the tank would be relocated up front for CG reasons.

*15 amps for the fuel pump is higher than what’s used now. However, the power budget is being set up now such that a larger pump will drop right in – one intended for ethanol 🙂

16 Nov 2014

A few developments:

Vendors have been surprisingly nonresponsive; in this age of the internet it’s amazing how many don’t answer email. I understand that they get a lot of traffic, but how is that a bad thing or the customer’s problem? During research into dry sumps, three vendors were contacted. One never answered at all. Another didn’t answer email, twice, so they were called. What was wanted was discussed, and a quote sent – which didn’t include what was discussed. Another email was sent, correcting the first quote and requesting a second quote – no reply. It makes one wonder if they’re this bad now, how will support be after they’ve made the sale? (My answer is: how can they make that sale if they don’t answer email?) The third vendor, called on the phone, was very helpful and said to send an email detailing what I wanted and they’d send a quote, it’s been days and no response. What’s that saying, “only one out of five businesses last more than five years?” That makes it sound like it’s random chance – no it isn’t.

The plan over the next 6 months is to upgrade to a dry sump and a better ECU – both require changes to the alternator for very different reasons. Virtually all FWD dry sump pumps go right where aftermarket Honda K20 vendors tend to relocate the alternator to, so it’s got to move – again. The OEM alternator has a four wire plug and the four wires go straight into the ECU. The ECU that I’m considering doesn’t do anything with those wires and the manufacturer said it’ll be fine to just unplug it. Really? I think “fine” is subjective, because it means that the alternator won’t be using its remote sense input, resulting in lower system voltage, but that’s not the ECU maker’s problem… I’ll try it just for fun and see what happens, and likely have to rewire it in spite of their “it’ll be fine” comment.

My brother pointed out a pretty cool website for one-wire alternators, Power Master, which caters mostly to circle track and hot rodders. They make nice small units that put out decent power, but they’re expensive, single-source, and don’t have remote sense inputs on the smaller units. The concern is being somewhere remote and having the alternator fail. If it’s an oddball part it means having the car towed home, so before considering these units, a late model Honda Civic alternator will be checked out. It’s smaller and lighter than the K20 unit and widely available. It also has the more traditional alternator mount instead of bolting straight to the block like the Honda K24 alternator does, making the tensioner unnecessary. The Civic unit is 70 amps while the TSX/RSX alternator is 105 amps, but it should be fine since there’s less electrical load than in the factory car.

The irony is that my re-engineered alternator mount that’s been working great is going away, but so goes progress.

7 Nov 2014

My brother got his LS-3 powered Stalker tuned today: 416 hp and 398 torques as Clarkson would say. Scott claims that on warm tires, it hooks up in second, which should certainly provide hours of entertainment! One more entry about dry sumps then issue will be set aside. I talked to a buddy who pointed out a condition that he feels is the single biggest reason to have one: In an OEM setup, oil in the pan is picked up by the pump and distributed throughout the engine, with the most important destination being the main bearings. The oil isolates the bearing halves from each other and with pressures of many thousands of pounds, it’s important the two never touch else bearing damage and eventual engine failure are soon to follow. The problem starts with the engine being run at high rpm. Air becomes entrained (mixed) into the oil by the spinning crankshaft to such an degree that it becomes an emulsion, a milkshake-like mixture of oil and air, with air constituting up to as much as a third of the volume (in fact, dry sump tanks are recommended to be filled only to ~60% capacity due to oil foaming potentially filling the rest of the tank). (As an aside, consider adding 30-40% volume to the oil mass, which raises the level enough that it’s now constantly being beaten by the crankshaft, which only exacerbates the problem.) The emulsion is sucked up by the pump and sent to the main bearings, where the normally-incompressible oil film has been contaminated with air. The thousands of pounds of pressure is now able to compress the oil/air down to a dangerously thin layer and greatly increases the chance of bearing damage. Even worse, if and when the pump sucks air, oil pressure drops and the entrained air bubbles expand to about double, effectively doubling the amount air in the oil. This situation of having the oil system contaminated by entrained air cannot be fixed by adding an extra quart or an Accusump, but others disagree. So be it.

3 Nov 2014

I recieved my new throttle position sensor (TPS) from K-Tuned. I’ve used a number of their products on the Honda drivetrain and this one will take care of a problem common with Honda K-series engine swaps, which is that the factory TPS tends to fail fairly often. Unfortunately the TPS isn’t sold separately, so owners are forced to buy a complete throttle body assembly from Honda. K-tuned developed a Hall Effect type TPS, which is a solid state device that changes its output based upon a magnetic field. The huge advantage is that because it’s a magnetic field, there’s no contacts to wear out, no potentiometer substraight to crack. While my TPS hasn’t failed yet, when it starts to go, this unit’s going right in and I never expect to have to replace it again.

I have a solution about the dry sump being labelled as an unnecessary addition. I want everyone who says that to send me a check for $6000. I’ll pin the checks to my wall and promise to tear them up the day I sell the car – unless the engine fails, then I’ll cash one of those checks in order to repair the car. How many people do you suppose will send in checks? About zero, because everyone’s very free with advice when it’s not their money on the line…

2 Nov 2014

What do you do when you’re told by people that a dry sump is a silly idea – not a bad idea, just an unnecessary and expensive addition? Of course, other than cost, these same people also agree that it’s a great idea, so their actual complaint isn’t with what it does, but how much it costs to achieve it. (The irony is that several of those who said so run dry sumps on their cars.) I’m told that there are better ways:

– “Add an extra quart of oil.”
– “You don’t need one on a street car.”
– “You don’t have space for the tank.”
– “Unless the engine’s sucking air, it’s not necessary.”
– “Get an Accusump, they work practically as good for one tenth the cost.”

My opinion:
Extra quart: yes that works, to a degree, it’s just that no one knows to what degree.
Street car: Define “street car.” Please explain then why new Corvettes have dry sumps.
Not enough space: Says someone who hasn’t built a Midlana…
It’s not sucking air now: True, until it does, then I have $$$$ damage which I have to pay for, no one else.
Accusump: This takes a bit longer to explain –
I had one and didn’t care for it. It’s a tank of oil that’s pressurized to the same pressure as the engine’s oil system, so an equilibrium is reached where as long as the engine’s oil pressure doesn’t change (up or down), oil flows neither to nor from the tank. The idea is that if the oil sloshes away from the oil pump inlet, oil pressure will drop and the now-higher pressure Accusump will push its contents into the engine, “stepping in for” the engine’s failed oil system.

I don’t have an issue with the above; it’s the details that bug me. As the engine slows, engine oil pressure drops, so the Accusump starts pushing unneeded oil into the sump, which gets whipped by the crank until the engine speeds up again. As it speeds up, oil pressure begins increasing, now the engine’s oil pump is assumed to have the extra capacity to not only fully lubricate the engine, but also recharge the Accusump tank at the same time. What rubbed me the wrong was the salesman assuring me that their system has a special valve that charges the tank slowly, but releases the oil quickly – a one-way valve.

No. It. Doesn’t.

Lastly if the pump starts picking up air bubbles, they’ll pass through the pump and get compressed along with the oil. Because the air/oil mixture is compressed to standard oil pressure, the oil gauge (and Accusump) will not see any problem, yet the engine is still being starved for oil.

So as a reward for reading this far, here’s a YouTube video of Midlana and friends driving through San Diego’s mountains after a rain. (I should add that the coolant coming from the Evo was simply an overflow tank that… overflowed, no actual problem.) It was a lot of fun, and reminds me to get Midlana back on-track!.

1 Nov 2014

Yesterday the boost-by-gear was fine-tuned and then the car parked. Today the weather was about 10 deg F cooler and now it keeps hitting the boost limit. It’s not surprising: the air is cooler and more dense, the ECU adds more fuel, which produces more exhaust, and up goes the boost right into the limiter. Sigh… KPro doesn’t have the ability to do closed-loop boost control. That is, the driver programs boost control valve values and hopes that boost ends up somewhere near the target. A more capable ECU works the other way round: target boost is specified and the ECU adjusts the wastegate to make it so. The beauty is that it self-compensates for engine temperature, weather, and altitude. Now I know.