Wrenching moments

So winter was about a month late arriving. What did I do with all that time I wasn’t wasting on paragliding? Well, I spent many hours and a small fortune wrenching on my car.
Check out what I did.

August 20th 2005

I finally got off my arse and bought a new engine for my Impreza. I’ve been planning and researching an engine swap for maybe two years now, or maybe since I bought the car new back in 2000.
My Impreza is a bit of a hybrid, being assembled in Taiwan out of parts that are about 90% Japanese in origin. It’s bizarre in that as far as the engine specs and emissions paraphernalia goes, it’s very close to the JDM version. What’s odd about that is that the car is naturally LHD, so I guess Ta Ching must have had some fun shopping through the corporate parts bin at FHI…
After 135,000km of mostly trouble-free motoring, I need more grunt than Ta Ching saw fit to offer. I’ve thought about either blowing the original engine, or swapping in a complete turbo drivetrain, but the first idea is bad for a daily driver and the second is more money than I want to spend. Not to mention that turbo lag sucks donkey balls in Taiwan traffic.

Anyway, what’s problematic about this swap is that the 2.5 liter naturally aspirated engine is almost unknown in Japan, but very common in the US. However, the USDM car has a bunch of emissions type junk on it that the JDM models don’t, like sensors for fuel temperature and tank pressure, and controllers and solenoids for all sorts of purge canisters and recirculation devices, extra catalytic converters and more sensors to check on them. Shocked! So, a big headache for me is that I can get a USDM engine and ECU, but my car lacks all these (useless) peripheral devices and that’s for sure going to have the ECU bugging out and throwing all kind of trouble codes. I don’t know about you, but I hate driving around with the CEL on.
OTOH it’s going to be a PITA trying to figure out which exact year and model JDM ECU will match my mix of peripherals and the engine series I want.
There were some 2.5 liter Foresters and Legacys imported into Taiwan, maybe about 5 in total and as rare as rocking horse poop. No guarantees on use or condition. Confused

I decided to fit the 2.5 engine and run it off my original 1.8 ECU. There’s no way in hell my stock ECU will just magically hit my target AFR running injectors almost 50% larger than it will think are there, so some tweaking is required.


JDM SOHC Phase II EJ25 longblock and timing belt.
USDM EJ25 fuel injectors.
TDM EJ18 intake manifold and throttle body.
TDM EJ18 sensors.
TDM EJ18 ECU + Greddy Emanage piggyback engine management.
Greddy PC support tool for tuning sensor inputs, adding injector pulse width and modifying ignition timing etc etc.
Crossed fingers.

Current mods:

WRX brakes front and rear.
KYB AGX adjustable sports suspension.
STi v.5 springs.
Clarion/Infinity/Image Dynamics audio with two amplifiers and 900W total RMS power.
Improved sound deadening and acoustic damping.

I was a bit dubious that the emanage would work on my car for two reasons. One, there is no application listed in Greddy’s manual for my car. It work and it might not. I found a couple of guys with turbocharged 2.5RSs that have used it and they offered some settings which worked on their cars. Two, the 2000 and up ECUs are ‘smart’ and are known to ‘learn’ around devices like the Apexi S-AFC. Sam assured me the emanage won’t face this problem.

I saw my new engine today, fresh off the pallet from Japan. It’s beautiful, brand spanking new, never been run. There isn’t a speck of soot in the exhaust ports. I got a longblock only, so the water pump, pullies, belt tensioners, intake and all ancilliaries have to come off the original engine. It looks exactly like my original, except it has the telltale ‘EJ25’ engine code, and no engine serial number, since it was never installed in a vehicle.
Hunh? That’s right. Not even intended for that either. Subaru sells engine direct to builders of experimental aircraft, which is what this one was originally destined for.

Rather than test too many things at once, I decided to install the engine management first, get that up and running, and then let my buddy Sam install the engine. I no longer have time to spend a week in the shop hoisting engines… My day was spent on my knees, head in the passenger footwell cutting and splicing into the factory wiring loom at the ECU. There’s a lot of wires in the bundle and it’s easy to make mistakes. I’ve spliced in with extra OEM style connectors so I can rip it out again or isolate particular wiring groups easily should I need to troubleshoot the install. Thumbs up to Taiwan for these kind of supplies being available right off the shelf in electronics alley.
Finally got the new wiring harnesses all labelled, soldered and plugged in after a 7hr slog. The engine started on the first try, which is nice, but now I’ve got some error codes from the emanage and the TCU, so the fat lady hasn’t sung yet. Got to try and fix those issues tomorrow with some firmware updates and some different approaches with the update and parameter setting procedures. I’m an oldskool tune by jets and needles carburettor guy, this tune by laptop stuff is all new to me. New millennium starts next week.

August 21th 2005

Today I spent a few hours trying different versions of both the fuel computer firmware and tuning software that programs it. Finally I hit upon a combination that allowed me to at least download the existing unit settings and some other neat stuff. Now I can see what the engine is doing in real time, injector duty cycles, ignition timing, throttle position and so on, and do data logging.
I had a transmission light show up as soon as I moved the car out of the shop, which I guess is related to the emanage. The transmission takes some load and speed signals from the ECU to figure when to shift etc, so I figured the now modified signals must be causing the TCU to freak out. Some changes to the engine parameters and map scaling got that to go away, which is nice since it tells me that I did get my wiring right after all.
Yesterday there were times I felt like I was in one of those lousy movies where there’s 20 seconds left on a timer, and you have to cut either the red wire or the green one, or something really bad happens. Red or green. 19 seconds, 18…

August 29th 2005

We had the engine all plugged and plumbed on Saturday. I filled her with dino oil for break-in and unplugged the ignitor so we could spin the engine without it firing, priming the oil pump and getting some lube to the bearings. Plugged back in, the engine started on the first turn of the key.

Idle is perfectly smooth at 800rpm cold and dropping quickly to 600rpm as the beast warms. We waited, listening for any strange noises and looking for leaks. After 10 minutes or so we put the car in the air to look underneath to find a heater hose was weeping a little water, so we had to shut her down to fix that.
With that taken care of we had to bore a hole in the headers and weld in a bung for the wideband O2 sensor we’ll use in tuning.

Roadtest went off without a hitch, though as soon as the throttle was released our old friend the transmission warning code was back to greet us. We made another attempt to tweak the wiring for my Emanage fuel/timing controller, but no go.

Another check to see that all the hoses and gaskets were tight and I took the car home. No time to do any tuning or even check the AFR, but the car was running so smooth it should be fine.

Yesterday I took her for a maiden voyage up to Green Bay, arriving with about 200km on the trip odo and the engine starting to loosen up a little. We still haven’t done any tuning, so I’m running pretty rich in closed loop, some bogging letting off the gas if not just snapping straight off into decel. Low end torque is still amazing. More throttle will kick the ECU into open loop and we’re either way lean there, or way too advanced with timing. Det rears it’s ugly head, so have to back off.

On the way I could easily cruise at 130kmh without leaving closed loop mode, and without breaking 4000rpm, pass anything I wanted to do. 450km now and the new engine spins much more freely.

She goes back for tuning today, and to fix a couple of issues related to the install.

1. The shimming under the engine mounts to get clearance for the pan has caused one of the bolts on the power steering reservoir to contact the underside of the hood.

2. Something is interfering with the left front tire on full left lock as I have a rubbing sound now which I never did before.

We may have to modify the crossmember some to clear the oil pan and allow us to get the engine back at the stock height…

August 30th 2005

We got a baseline tune done today. The emanage has 16×16 cell maps for various functions like additional injector pulse width, ignition advance/retard, the control of extra injectors, and the old standby, lying to the ECU by manipulation of the engine’s load sensor. All of the maps have 16 columns for rpm and then 5 or 6 scales on the rows, from throttle position, manifold vacuum relative and absolute, to the actual voltage on each sensor output. The emanage tracks the cells on the 2D maps in real time and you can make corrections on the fly.

Lacking a dyno to work with, we have been out tuning on Taichung’s outskirts up what hills we can find to create some load without having to break any speed limits. Well…
Tuning is accomplished by driving the car against the load (our hill) at varying throttle settings and rev bands to see what air/fuel ratios we see on the wideband O2 sensor, while listening for detonation on the headphones which monitor the engine with sensitive microphones.
At the moment we are adjusting to get a ballpark AFR around stoich at part-throttle loads and letting it go as rich as 12.5:1 under heavy load / full throttle. We had to make some pretty drastic fuel trimming to hit those targets, so it’s nice to know I have lots of overhead on my injectors if I do further mods later to increase airflow.

Of course it’s not necessary to get values in every cell of every map since it’s not possible to get the engine to redline at 0% throttle, or get the revs to stay at idle with the pedal at 100%. This is the great thing about a piggyback computer. You can leave cells at zero and default to the maps on the stock ECU. With a standalone you have a lot more functions and flexibility, but you have nothing to fall back on. Probably only half the cells in my maps have an adjustment value entered. A lot of the values are quite surprising to me, adjustments being required in areas I had thought would be fine and vice versa. I had thought I would be drowning in excess fuel at idle, but it turned out we had to add a little to avoid stumbling. I thought the top end would be a little rich, but in fact we had to trim fuel in the 80~100% throttle range by up to 50%.

Working with the headphones is amazing. I always enjoyed tuning racing motorcycle engines because they give so much feedback. Induction and exhaust noise give away a lot about the mixture and ignition settings, so it’s much harder working with street cars that are so muffled and insulated. With the ‘phones on it’s like working with a bike again. Too bad they make the engine (any engine) sound like a worn out old tractor.

September 6th 2005

Hmm. I just figured out from a Subaru forum that the approach of tuning fuel via fudging the airflow signal has some side-effects. Apparently the ignition advancer is also referenced to the airflow table in an effort to modulate timing against load. Of course this works like the old diaphragm on distributor system. More load = less vacuum and since more load will require some ignition retard, more load should = less advance. What I’ve been doing in my maps is trying to add fuel in some areas and trim in others.
The emanage can add fuel via two methods. One is increasing the MAP signal voltage to emulate less vacuum (higher load), the other is directly adding pulse width to the injectors, bypassing the ECU altogether. The first method will affect the ECU’s final ignition advance, the latter will not. Sam had tried using a mixture of both methods to hit my AFR targets, when I think we should have stuck to only adding fuel via the additional injection map…. I think I’ll finish running the engine in before I go back and redo the maps.

Sam finally dragged in an emanage ‘expert’ to help diagnose the transmission warning light. After checking all my wiring and pronouncing it correct he went and declared my MAP sensor to be too weak to signal both the emanage and the TCU. Remedy? A Greddy 2bar MAP sensor added to the system, at a cost of another 6k!
But, although it got rid of the warning lights, there were side effects. First, the new MAP sensor has a wider range than the factory unit, and thus less resolution. The idle quality went to hell and so did the tip-in response. Second, removing the factory sensor input from the emanage meant the airflow adjustment map no longer functioned, and we were way rich at high rpm/load again.
As a last-ditch attempt our ‘expert’ dug into my jumper and switch settings. He says that the manufacturers manuals are in fact all f***ing wrong and with a change of a setting or two all the problems disappeared. In theory the current settings (a Subaru hotwire setting on a MAP car) shouldn’t work, but they do… and finding this fixed all the problems at one stroke, and saved me six grand. Cool

My clearance problems have been resolved, though not in the way I hoped. I left my guy with instructions to lower the engine to get clearance from the hood, massaging the crossmember as necessary. Of course he felt it was much easier to bend the tab on the steering pump reservoir and grind the bolt down instead.
The rubbing I had at full lock turned out to be one of the splash guards in the wheel well. It must have gotten a little bent at some point, probably while refitting the powertrain from underneath.

September 14th 2005

Today the car started idling really badly at times, like every third stoplight. It sounds like it’s missing on one cylinder. At times it does the same thing in light acceleration or when cruising around 40kmh. Sam (on the phone) said it was probably the emanage, those things are unreliable.

September 18th 2005

Weather looked better for flying today than I’ve seen it in a long time. But instead I spend the day trying to track down that misfire. It’s gotten so frequent now that I can drive the car in a certain way for 20 minutes and force the problem to appear. I started disabling the emanage maps one by one to see if it was a programming issue. Then I started isolating the different harnesses one by one until the emanage was totally disconnected from the car and powered down. No change. Now I’m thinking either the emanage has damaged some other part of the car’s electrical system, or the problem has nothing to do with the emanage at all. I notice the belts are a little tight and loosen them off some. That seems to make the idle improve a little, but it doesn’t stop the bucking at cruise. I reconnect the emanage as that doesn’t seem to be the problem.

October 15th 2005

Today I drove up to Taoyuan to meet a young STi owner and buy the factory exhaust he no longer wants. He has an HKS turbo back system with a tailpipe big enough to launch puppies from. Sounds badass though. He and his buddy are interested in my swap etc, but I want to get home rather than hang and chat. I drop the car off at Sam’s place so he can graft my new pipes on tomorrow.

October 17th 2005

Sam was too busy yesterday to swap my exhaust. Today he did a less than spectacular job on it, squishing my tailpipe to get clearance with my trailer hitch. Well, it was a busy night tonight. Matthias picked me up at the factory at 6:30. Then we went to Tsaoma to get him a rental car. Then off to drop the Ferrari (Ford Festiva) at his place. Then race over to Sam’s in Beitun. By the time we got out of Sam’s to start work it was 8:00. We spent about an hour and a half to do about a dozen pulls. Then I had to wait about a half hour for Sam to get back from his ride to remove the wideband from the car. It was well after 10pm before I got any dinner.

It’s actually pretty quick work when you have the wideband hooked up. I was in the right AFR range all the way to about 3,200rpm and up to about 45% throttle. So we did a pull at 50% throttle and revving as high as the car would pull to get a baseline. I start pulling fuel out of the cells that were highlighted on the map during the pulls and then we go do another run at 50% to see the results. A couple of pulls later we’re set. Then we repeat this at 70% and 90%. With those in place it’s simplest to just use the emanage’s interpolate function to fill in the cells in the ranges in between. A couple of quick pulls confirm those values are good, and then we move onto WOT runs. That takes a little more time because the runs go further each time and you end up having to coast 5km back down the hill letting the brakes cool off for the next run. I was pretty shocked to find I was trimming as much fuel as the emanage would allow and still running as rich as 12.0:1 at WOT and high revs.

October 30th 2005

My bucking problems were making themselves felt with a vengeance, and my buddy Monie suggested I try a Subaru performance shop for a new perspective, so I gave it a go. They had a dozen fully race prepped GTs, WRXs and STis in there. Drooling does not cover it. The head tech had some ideas about the ignitor and some other parts, and suggested finding me a used coil to try. Of course I agreed to that.

November 2nd 2005

Today the car actually fired up on three cylinders from cold. Although pissed off about it, I took advantage of the situation by pulling the plug leads one by one to see which cylinder was being lazy. It’s #2. Strangely, once I’d held the lead away from the coil for a few seconds, seeing how far the spark would jump, #2 decided to report for duty and the idle smoothed right out. Hmmm, the coil is bad maybe? The terminal for #2 was grubby and corroded, was the solution so simple? When I got home I got some #400 sandpaper out and cleaned it up really nice.

Novmeber 3rd 2005

Rough idle and bucking is back. I take the car to Subaru to see if they have any bright ideas or can pull a trouble code before I go and reset the ECU. There’s a stored code for a MAP signal out of voltage range, but that could be from back before the last round of switch setting changes.
I asked them about the ignitor/coil unit and they had one in stock. Six thousand dollars. I asked them if they were sure that was the problem, and they said no, they couldn’t be sure. Could I borrow a used one to see if it cures it? No. Could I return the new one if it didn’t cure it? No. I decided to try new plug leads first and ordered them as they had none in stock.

November 8th 2005

The plug wires arrived so I swung by Subaru to collect them. Fired the car up and it was nice nice. Drove it for 15 minutes and the bucking came right back. Keep looking for clues.

November 10th 2005

A used GT coil got spliced onto the car and I drove off. In the evening as I drove home, the bucking started again.

November 11th 2005

Kai Chou is looking confused at the car’s behaviour, but not yet troubled. I leave the car there all day for him to test and play with. They go into the timing belt cover to look at the cam and crank position sensors, swap those even, but no change. Then he decided to have a look at the sparkplugs and found they were all really old and most had signs of electrical leakage over the insulators. Why anyone would put worn out old sparkplugs into a brand new engine is beyond me. A set of new plugs seems to make a world of difference, much smoother. Maybe this is the cure.

November 13th 2005

The lumpy idle is back, though it’s nothing like as bad as it was, and it’s happening much less frequently. It’s occuring to me that I might have two different problems that just happen to manifest in a similar way. They may or may not be connected.
Kai Chou decides he wants to try swapping the coil again so I let him do that. On my way home from work the bad idle starts up again. In desperation I swing by downtown and buy a few meters of #8 wire and some ring terminals to fabricate a single-point ground kit. I install that on the car around midnight, sweating in the parking garage and fending off swarms of blood-sucking insects.

November 14th 2005

Back at WAT for further head-scratching. Kai Chou decides to remove the emanage to see what happens. I’ve already done this, but that was before changing the plugs and wires. I drive home and don’t experience any bucking.

November 17th 2005

So far so good! No bucking or stalling or bad idling epsiodes. This is great progress. Next comes the question of what to do with this new knowledge. Maybe my emanage has some hardware problem that’s screwing with the MAP signal. Maybe it’s the settings, or the emanage simply can’t be made to work with my car. I leave the car with WAT and they will try to play with the settings and see if any progress can be made.

November 18th 2005

Kai Chou calls to say he’s installed the emanage, redone the settings and tuned the car. Great news. He also says he’s been driving it all afternoon and had no problems. I should take the car for a longer test.

November 23rd 2005

Still going well, no issues. The car doesn’t have the same low-end torque it did on my tune of 10/17, so I swung by WAT to see Kai Chou and discuss his tune. He’s also unsatisfied with the low rev bark, so agreed to spend more time tuning the car. His maps are completely different to mine, negative shifts in timing rather than positive, much less fuel trimmed

December 26th 2005

10,000km on the new engine now and switching to synthetic oil. The front brake pads were getting pretty thin so I had those swapped out for some new Akebono ceramic items. Kai Chou has made some more modifications to the maps (or so he says) to try and get my low-end response back, but I can’t honestly feel much difference. It does seem a little smoother all around, but the surging at idle and stumbling while accelerating from a standstill are rearing their ugly heads on particularly cold days. I’m going to try the old emanage settings of 3,4,2 and take a stab at the anti-stall map to see if I can stop the MAP voltage from dipping so much it triggers that AT warning.

December 28th 2005

I decided to have a go at tuning that stumble out with either the airflow map or the anti-stall map before going in and changing the hardware settings again. Looking at the airflow input and output voltage I could see quite a large difference at the stumble point, so I started to reduce the fuel trim back toward zero. So far it’s much better and the stumble is almost unnoticeable. I’ll leave it there and see how it goes, still on dipswitch setting 3,6,3.
I also put some more fuel in the area just above the closed/open loop transition as I’d heard some knock there after Kai Chou put some of my timing back. It seems to pull much better in the midrange now, but I need to point it up a steeper hill to be sure the knocking is gone.

January 1st 2006

Had a very fun and fast drive to Saichia and back. Everything is good, and man do I love this car when it runs right. Came back cruising at 140~160kph with the odd excursion to 180+ and the only thing that overtook me was a GT wagon with a dog launcher.

April 10th 2006

The car has been running quite predictably for a while, but very occasionally I’ve heard some pinging under high load / low rpm situations. In order to better diagnose that I spliced an audio output into the ECU wiring so I can listen to the knock sensor through the stereo. Here’s a how-to.

April 30th 2006

I ordered some new audio parts for the next stage of development. The big change is that the current front stage will go active. Up to now I’ve been running an Infinity Reference set with it’s passive crossover. On order is a pair of Phaze Audio MW-64 midbass drivers and a pair of LPG 26NA tweeters. Both have aluminum cones/domes. The MW-64 should be a huge improvement in midbass response over the weak Infinity items and the tweets have been popular with the DIY crowd for years. Also on order is Zed Audio Gladius amplifier which I will use to drive the midbass drivers. My current subwoofer amp will get shifted to driving the tweets and the current front stage amp will get reinstalled behind the rear seats to drive the sub. The aim is to have more headroom on the front stage and have independant control of the crossover points and slopes, plus avoiding the insertion losses of the passives. One amp per pair of speakers.
I spent all of this three day weekend dragging more wiring under carpet and making racks in preparation for the install. New baffles are being fabricated to mount the new mids in the doors as they are quite a bit deeper than the current ones. This time I’m going to try mounting the tweets in the kickpanel area in the hope of equalizing the pathlengths between the two sides. We’ll see how that works out, but it would be nice not to have to drag more wire into the front doors because it’s such a nightmare.

May 14th 2006

Today I finished the MDF baffles and fitted those to the front doors. The outer rings have been sitting around gathering dust and this may have been one of the biggest mistakes I’ve ever made wrt to the system. I finally got some 6mm MDF spacers cut to the shape of the old plastic spacers to clear some metal profile on the door inner skin. Two are glued together and three M6 bolts pass through the whole thing to clamp it firmly to the door. My MW-64s don’t arrive until tomorrow so I refitted the current Infinity drivers in the new baffles. I finished installing the passenger side and then ran the system to be sure everything was okay. Well it more than okay, night and day difference is more like it. Switching the balance back and forth showed how much difference there was in midbass response… like a totally different driver. Still nothing to write home about but a huge improvement in the midbass hole I’ve been looking to fix. I feel so stupid that I had these baffles sitting in a closet for two years when they could have been in the car filling that midbass hole. Doh.

May 17th 2006

I was in the basement until 1am this morning installing my new Gladius amplifier and moving the 100Wx2 amp I had on my front stage onto the back of the rear seats where it now powers my subwoofer. I was really impressed with the look and feel of the Zed Gladius when it arrived. It’s bigger than I expected and twice as heavy, feels way overbuilt. I turned the gains down to almost nothing before running it for the first time and still almost jumped out of my skin when it came to life. It’s rated at 75Wx2 but clearly has more power than the amp it’s replacing. I still believe a watt is a watt, but the extra headroom sure does improve SQ.
I should have packed up my tools and gone to bed once I had the Zed in place, but I went on to install the sub amp instead. A sweaty, miserable hour later I had that safely secured on the back of the rear seat on it’s new rack. It looks very tidy back there. Then I started the engine and noticed the alternator whine. Doh. An hour of rerouting RCA and power cables and that was fixed. Tonight I’ll install my new front stage drivers and power up the last amp.

May 18th 2006

I finally got everything installed today. New beefy mdf baffles, close to a pound of modelling clay under/around each baffle. MW64 in the door, 26NA vertical on the kickpanel pointing slightly up and pretty much straight back. I tried my old drivers in the new baffles to see how much difference they made and it’s a lot, so the old adage applies… it’s mostly install. With the new drivers plugged in and the two amps running them I can now shut the sub off and still have bass. Makes it much easier to integrate the sub. I’ve started off with the sub and mids crossed at 50hz on 12db and the mids to tweets equal at 3.5khz at 24db just to try things out and give the mids some break-in time. So far I have to say I’m quite impressed. I don’t know how much is the extra headroom and how much is the drivers, but it sounds pretty darn good so far considering how cheap it all was.
I haven’t had the mic out yet and the fr sounds far from flat, but Mozart sounds very expansive with a stage that extends well past the pillars and the height is up on the dash… maybe just a hint of rainbow, so I think with a little tweaking it will work out. I think the next move will be try level matching with the mic and then try bring the mids to tweets xo point down as much as the tweets will tolerate.

May 24th 2006

Been very busy the last week but managed to do a few things. I’d tried the tweets pointing at each other, parallel with the mids, but didn’t like that as much as straight back. That doesn’t sound 100% either. I tried a quick and dirty angled mount with polystyrene and it sounded much better… for the roughly five minutes they will stay in place for. So I started into making something more permanent. What I hit on was triangular wedges shaped to point the tweeter 45 degrees horizontally into the car and about 30 degrees up. I used that weird green sponge that florists use as a base for an arrangement. It’s really easy to cut to shape and compresses into shape well. I painted the outside of the wedges with a few coats of woodworking adhesive which gives it both a stiff skin and a texture like that of the kickpanels and other plastic interior parts. A quick hit of flat black paint finishes that illusion quite well.
I’ve had the mic in the car once so far and found a few peaks in FR. I moved the mids xover point down to about 3.15khz and left the tweet at 3.5khz. That seems to have taken care of the spike there. I’m still trying to decide which xover point I prefer for the sub, 50 or 80hz. I also have to go back into the passenger door and take care of some buzzing from there. More output means more rattles…

May 23rd 2006

I had to put the car into Subaru today to replace a blown rubber boot on the passenger side front driveshaft. It had exploded on the way down to Saichia on Saturday and made a mess in the engine compartment, not to mention the stink of burning grease everytime I stopped. I had her serviced while she was in there. 20,000km on the new engine now.

May 27th 2006

Today I got dug into the doors hopefully for the last time. I decided to mat over the last access hole into door to seal it completely. The right door was buzzing like hell at about 100hz and it seemed like it was the window mech doing that. The last thing I want to do now is rip the whole door apart to get at that, so I tried using some more modelling clay to try and damp it out. I ended up putting another half kilo of the stuff into all the recessed spots on the panel but it worked great. Then of course I had to go do the driver’s side to match. Nice improvement in midbass.
I think I’m through with the physical install now, just more tuning needed to get it all dialled in. I decided the sub was better crossed at 50hz but the mids seem to prefer 80hz. Now I have the xover on the HU set to 80hz and switched on the amp’s xover and set that at about 60hz. In theory the xovers will cascade and at some point the final slope will be 24dB/oct. It sounds like a better transition now, makes the sub more transparent and harder to locate, but I’m curious what the RTA will say.

This entry was posted in Local scene. Bookmark the permalink.