With Wings As Eagles: Craig P. Steffen's Blog

its just a jump to the left

2008 June 30 23:16

I took the beetle out for a test drive today. The fixes I made to the gear shifter made a huge difference. The shifting is much more precise and it doesn't rattle while it's in gear. Yay! The steering didn't make as much of an improvement. The steering doesn't wander as much, but the vibration at 45 mph is still there.

Since I had the gear shift out anyway

I replaced the shifter block-off plate. The old one is on the right, and the new one on the left:

You have to push the gear shift down to get the beetle to go into reverse. The shoulder at the left of the hole in the block-off plate is what holds the shifter down and keeps it from popping out of reverse. The shoulder on the plate on the right is worn:

Before, my car had slipped out of reverse from time to time, so I thought maybe replacing the plate would fix it. It's just fine now.


whose is biggest?

2008 June 30 21:49

I was looking up some information on the new Ares V launch vehicle today, and I ran across an intersting page listing heavy-lift launch vehicles of the world. "Heavy Lift" here is defined as being able to life more than 20,000 kg into low orbit.

It's interesting to see the whole list. For some reason I'd thought that the Soviets had a rocket that could lift more than the Saturn V, but apparently not. The big Soviet/Russian workhorse is the Proton at 21,600 kg to orbit (The shuttle is 24,400 kg, but I presume part of that is people). The Soyuz-FG rocket that launches the Soyuz spacecraft is a smaller vehicle and thus isn't on this list.

So the Saturn V rocket built for Apollo stands as the largest launch vehicle ever made, at 118,000 kg to low orbit, and it was retired with a perfect launch record, 12 out of 12 successful. Here's hoping the Ares rockets are as successful.


no, there is too much--let me sum up

2008 June 30 00:31

I didn't test drive the beetle this afternoon because I was having an odd electrical instrumentation problem. I suspect that I caused that problem by touching a live connection while bolting the windshield wiper assembly into the car, having accidentally left the key on while I did so. (I think maybe the current surges associated with the sparking have buggered up my ammeter.)

I took the wiper assembly out to get at the back of the brake warning light which hasn't worked since I got the car. Since I had the wiper assembly out anyway, I decided to go ahead and take it apart and clean and lubricate it before re-installing it. That was moderately complicated to do the first time, so I was working on it this week.

I was working on the brake warning light and thus the wiper system because the part I ordered for hte steering didn't come in until Tuesday. Tuesday the idler arm bushing came in, along with some other parts for the gear shift. So I finished with the steering, worked on and finished the gear shift, but then I wanted to have the wiper system back in to drive the car, so I've now finished the wiper system so that I can test drive the car to see the effects of the steering and gear shift fixes.

I ordered the part for the steering a little over a month ago because when I took the car out for the test drive after the winter of fixing the wheel bearings and brakes, I felt that the steering was looser than I wanted, and the car vibrated at around 45mph (has since I bought it). The test drive leading to the steering fixes was the culmination of tests that started when I realized last October that the car was hemmoraging brake fluid.

So once I fix the wierd electrical instrumentation problem (or decide to ignore it which is becoming increasingly likely), I will do another test drive which is essentially another test to test things that began last October. If that test drive goes well, and nothing else rears its head, then I will go back to working on the problem that I found earlier last summer, which was that I have a valve stretching which pretty much means that I need to remove the engine and replace that cylinder head.

If this kind of thing would annoy you, DON'T buy a vintage car, Volkswagen or otherwise. This is the kind of that happens all the time on a car that old.

There probably won't be a lot of blogging over the next couple of weeks. I may spend some time working on write-ups on my beetle page. I think that for the next while, I'm not going to blog in quite such detail, but instead put my effort into the repair write-up pages. The problem with doing everything here is that I have to write it again for the archive pages.

But I will highlight things occasionally. Here's the windshield wiper assembly just out of the car:

And taken all apart:


bushings in their natural habitat

2008 June 27 00:30

I have to get up early tomorrow, so a quick post.

The idler arm bushing is installed in the front end of my beetle:

and the car is back on its wheels. This will hopefully tighten up the steering and help to eliminate the front end vibration at around 45 mph.

I'm also working on replacing the shift rod bushing to tighten up the shifting and eliminate the gearshift rattle that happens at certain rpms. I got it installed tonight

I'll try to get the rod and the gear shift hooked back up so I can go test drive it by Saturday.

By the way, a bushing is a sleeve of some material to allow some part of the car to move within another piece, and the bushing takes the wear instead of permanent parts. I think perhaps the disinction between a bearing and a bushing is that a bushing takes the load of a part that doesn't move very fast or continuously, wheras the item that a bearinng supppors moves constantly (like the wheels or parts of the engine).

The idler arm bushing and the shifter bushing have different roles to play, so they're made out of very different materials. The idler arm bushing is bronze, because that takes heavy cross-loads to keep the wheels in alignment in turning. The shift rod bushing is a plastic, because the shift rod doesn't take very heavy forces but it needs to be very tight in its bracket, so it's made of a material that can be compressed when the rod gets pushed through it.


I never took the Kobiashi Maroo test.

2008 June 26 00:18

Classic Volkswagen Superbeetles (like mine) are unfortunately suseptible to a shaking front end, particularly right around 45 miles per hour. I guess the combination of very long tie rods and rubber bushings tend to generate a loose steering system.

The superbeetle has an idler arm (that is, an arm that swings in a circle wherever it's pushed) that has a bushing that wears out. The word on the street is that if this bushing is worn, it can be a major contributor to front end vibration. So you can replace the bushing with a stock one, which has rubber for shock absorption, or you can also buy a solid bronze bushing that is supposed to eliminate the vibration.

A couple of weeks ago, I ordered a bronze idler arm bushing for my car. It came yesterday, with some other stuff for the gear shifter. So I've been sorting out how to remove the idler arm bracket, remove the old bushing, put in the new one, and then put it all back together again.

I'm going to show you where the idler arm and its bushing live. Here's an overview drawing of the front suspension and steering system on a Superbeetle, from roughly the point of view you'd have standing right in front of the left front fender:

I've marked parts by circling them with color-coded lines. Blue circles the idler arm bracket. Over to the right side of the car is th steering box which contain the gears that steer the car, bu the idler just rides there. The idler arm connects to the center tie rod at the tie rod end circled in purple. The side tie rods each connect to the center of the center tie rod, and the other end attaches to the wheels at the red circle. The other major thing that connects the wheel to the central frame is the track control arm which is circled in green.



Now we've walked across the car and we're looking into the right front wheel well. Parts that are visible are circled with the same colors as before. Circled in purple is the bottom of the idler arm. The arm itself goes out of view to our left.



Now we've gone into the wheel well and turned a bit toward the back of the car. The idler arm bracket is obscured, but we can see the bolt where the idler arm sits on the bottom of the bracket.



Completely under the car now. We're looking directly toward the back of the car, so the idler arm is pointing straight at us.



A camera-only shot (I try not to get my head that far under the car when it's on jack stands) looking toward the right side of the car from within the wheel well. The idler arm bracket in its natural habitat.



The official procedure is to remove the idler arm from the tie rod. However, this requires a special tool. However, I was able to unbolt the idler arm from the idler arm shaft and bracket, so I did that instead, leaving the idler arm hanging like this.



Idler arm bracket and shaft on the work bench.



The back (top) side of the bracket. Note the two wings connected to the shaft. The two bolts you see set how far you can turn the steering wheel. When the wing hits the head of the bolt, then the steering won't go any farther.



The bushing was so loose getting it out of the bracket was easy.



Here I've indicated the bushing that we have to replace. You can see the stock bushing's structure here. There's an outer metal sleeve, a rubber middle layer, and a metal inner sleeve.



Here I'm pounding the inner sleeve out of the outer sleeve of the old bushing. The white stuff is liquid wrench. I put a bunch on the bushing before I tried to remove it. It turns out that not only does liquid wrench make things more slippery, it also dissolves rubber! After that, I used a flat punch and a hammer to pound the inner sleeve all the way out.

(By the way, the offical way of doing this job reqires a hydraulic press:

...and lots of jigs and stuff. Partially I didn't want to take the time to get a machine shop to do this, but mostly I just thought that I ought to be able to figure out a good way to do it.



The inner sleeve was easy; it's surrounded by rubber. The outer sleeve you can see here is steel and is a press fit into the Aluminum bracket. I banged on it for a while, trying to deform it enough to get it to pop out. This failed.



Here's my solution. I clamped the bracket to the bench and assembled the hack saw with a good, new metal cutting blade going through the middle of the old bushing. Then I just cut a slot into the bushing along its axis to weaken it enough so that I can pound it out with a hammer and punch.



You can sort of see here the slot that I'm cutting in the bottom of the bushing.



The outer bushing sleeve removed. You can see the slot that I cut in the sleeve.



The bushings are a press fit into the idler arm bracket. To make it easier to insert, I'm heating the aluminum bracket to make the metal expand slightly and make the hole bigger.



Here's the bushing fresh out of the freezer.



The bushing's in. Yay! When I put in this bushing, once had it started a little, I banged on it with the plastic hammer. The problem is that once the cold bronze touches the hot Aluminum, it starts to come to room temp and get bigger.

(greased for idler arm shaft)



My idler arm shaft is slightly wierd. I think it was slightly defective, in that the splined section was quite exactly lined up with the smooth section. I had a dickens of a time getting the shaft into the installed bushing. I absolutely couldn't have pushed it in. However, a few whacks with the plastic hammer got it in no problem.





Hopefully the bracket will be back in tomorrow.


beyond the veil

2008 June 24 00:01

I have some long-term work going on with the transmission linkage in my beetle

but I'm waiting for some parts at the moment, so I can't drive it, but I can't work on the shifter either. So I'm working on one of the very long list of electrical nits that need to be fixed. The brake warning light never has worked in the car. The light is at the top center of the dash board, which means that it's totaly impossible to get at the back of the light because it's obstructed by the combination of the radio and the windshield wiper motor.

So lacking anything constructive to do with the car, I'm try to get at that light, so that I can test it directly and check its connections. I got the wiper assembly loose, but it won't quite come out; the radio is blocking it. Here's the situation:

This photo is taken leaning over the right fender, looking at roughly the center of the dash board at the back of the luggage compartment. On the left is the glove box. Down at the right the gold thing is the outside part of the fuel level sensor which is in the top of the fuel tank. Most of the wiring is off the picture to the right. At the top center of the photo the grey cylinder is the wiper motor. Right below the wiper motor is the radio, which keeps the wiper motor from being removed. So I need to de-wire it and take it out.

I stopped there this evening, because it seems that the radio is spliced in:

Although--it occurec to me that maybe the beetles radios were spliced in by stock, rather than having connectors like the rest of the electrical system? Given that every other blessed connection in the car has a spade terminal, I would very much have thought that stock radios would have them too...but there's no reference to the radio in the wiring diagrams that I can see. So maybe the official installation method involved cutting wires.

So the next time I work on it, I need to document where those wires go and what they connect to, and only THEN take the radio out.


It JUMPED into my cart. Honest!

2008 June 23 00:51

A friend of mine loaned me an air compressor for a while last year. Later in the year, he realized he's missed it, and asked for it back. I obliged. But having gotten used to the convenience of having an air compressor, I've been thinking of buying one.

Last time we went shopping, I decided that was enough of that.

So now we'll have compressed air around again. The specifications were driven by being able to do compression tests on my beetle engine, which is why it's a little higher pressure.

The oil-free pump has less maintainence...but it's loud:

89 decibels standing a couple of feet away from it. If I ever end up using it a lot, I'll want to make a spot to put it that insulates the sound somewhat.

So...after filling the beetle spare tire back to 40psi, I hooked it back up to the washer system. The bigger cable tie that I put on the hose from the reservoir (pressurized by the spare tire) seems to have done the trick:

And it hasn't leaked yet; so I guess four times is a charm?


facinating...

2008 June 19 08:12

I'm not going to claim that I'm the world's best documenter, or even that I'm a good one. However, my scientific training instilled in me a high regard for the integrity of documentation; if you make it, then it should be as close as absolutely possible to whatever it is that you're documenting.

And so it always facinates me when documentation is sort only sort of vaguely resembles what it is that it's describing. A couple of days ago I went looking for the wire that leads to be beetle's brake lights, so that I can measure the total amount of current that they draw. First, I found the relevant part of the wiring in the wiring diagram:

The important thing here is the three-wiring coupler. Generally speaking the wires going up on the diagram are going down in the luggage compartment of the car; most of the wires coming out of the top of that coupler go down to the pressure switches on the brake master cylinder. The wires on the right are black, indicating that they are at battery voltage when the key is on, but not otherwise. The black wire with the yellow tracer goes from the black wire to the horn.

I didn't include the whole diagram, but the black wire with the red tracer goes into the wiring harness and back to the brake lights. I've circled the wire in the diagram. I just need to find this connector, find the black wire with the red tracer on one side of the triple connector on the opposite side and opposite end as the black/yellow wire.

It is said that "theory and practice are closer in theory than in practice". No exception here.

To get oriented, here's where we're working.

This is taken from the right side of the luggage compartment looking across to the left side. At the upper right of the photo you can see the signal light on the left front fender. To the upper left is the brake fluid reservoir. At the bottom the black part is the left edge of the fuel tank. I've circled the triple coupler where it sits on the inside part of the fender. The wiring harness goes up to the left.

Here's a close-up, with my test lead already wired up.

The black/red wire is the center of the three, not one of edges. And interestingly, the black/yellow wire is on the OTHER side of the connector and goes into the wiring harness. I guess at some point they decided to connect to the + horn wire farther up the wiring harness, so it made sense to switch the side that the horn wire connected to, but that change didn't make it to the diagram.

So I found the correct wire, and I used the diagram to do it. However, whatever procedure was used to wire this car, the wiring diagram wasn't taken directly from it. By the way, the wiring diagram photo was taken from Bentley, but I did wander downstairs and double-checked the official dealer wiring diagram; it's the same.

By the way, the brake lights draw 3.87A at 12.7V. Plenty of overhead for an LED 3rd brake light.


Ruh-roh!

2008 June 18 23:27

Last summer, I realized that one of the valves in my beetle was going out of adjustment much more quickly than it should have been. I've mostly curtailed driving it on long trips since then, since I suspected that one of the things it could mean was something badly wearing out in my engine.

I ran a post by Robert Hoover that confirms my fears. His post says, in part:

Mechanically, a loose valve seat will usually cause the valve's lash to CLOSE and in doing so, offer a significant warning well before the grand finale. Normally, valve lash becomes wider in use. The only times it becomes LESS is when the valve begins to stretch or a loose valve seat is be being hammered into the head, either of which is portent of pending disaster, well worth the effort to drop the engine and sort things out.
Robert Hoover, by the way, is someone who's a long-time VW mechanic, VW airplane builder and flyer. Although his writing style is sometimes abrasive, I would guess that his knowledge of VW engines is second to very few living humans. Basically that failure almost certainly is what I feared it might be. The good news is, from what he's saying, it should be easy to spot once I get the cylinder head off.

So I'll probably be pulling the engine sooner rather than later. Maybe in July, we'll see. With a little luck, maybe this won't be a put-the-car-up-for-three-months repair.


the last of the jedi will you be

2008 June 15 22:54

I've been having a series of grumpy old man thoughts recently. Part of it is just part of being 35 years old, I expect, and closely tied to the fact that I'm technically old enough to be the parent of high school students entering college in the fall.

However, I'm terribly afraid that people are losing the interest, and with it the ability, to build things. Don't get me wrong. I like certain aspects of the super-consumer society. Between Google and Ebay I can figure out a way to buy almost anything that I can possibly conceive of.

But sometimes I want something that I can't buy. No company would build such a thing because the volume would be too small to justify the cost of tooling up to manufacture. And maybe it might be something that only I've thought of. I am interested in spending the time to design, engineer, and build something. If it's something electronic, I have enough expertise that I'll probably be able to build something that does what I want to and works properly. In other areas, maybe not so much. My point is my instinct is that if I can't find something, than I build one.

Owning a vintage car has given me great incentive to excersize this aspect of my personality. My grumpy-old-man feelings lately have been that society is losing this skill. I don't have any direct evidence to back this up, and I worry that my age is driving a knee-jerk reaction to what "kids these days" do or don't do. (They do listen to crappy music. Back in my day, music was good...not like what they listen to now. :-) But I digress.)

My indirect evidence is that I find it harder and harder be able to do that kind of work myself. I recently went shopping for what 20 years ago would have been a part that I could have found at Radio Shack or a small electrical/electronic parts store, but to buy one now I will have to go to one of a couple of very obscure electronic parts houses that mostly sell wholesale. (By the way, I was look for an isolated output DC-DC voltage converter; car voltage in, 5V out).

Computers are another area that are becoming more integrated and less module. 10 years ago, when I shopped for a computer I was rabid about making sure that it had lots of bays for extra drives and slots for expansion cards. The coming of USB has pretty much made it possible to buy a fully functional computer with only a few external connections.

My thoughts on this were stirred up a few weeks ago when I was putting the brake master cylinder back in my beetle. The wall just beyond the driver's feet is two walls of sheet metal separated by about an inch and a half. The brake master cylinder mounts to the layer of the wall farther from the driver, but the bolts go through both layers and the heads of the bolts are behind the brake pedal. There are supposed to be sleeves that fit around the bolts so that the two sheet meter walls aren't squished together. Well guess what--the old master cylinder was installed without the spacers, and so the wall in my beetle is bent inwards rather alarmingly.

I decided to try to manufacture something to act as the spacers, but also to make a metal plate that would re-enforce the wall on the driver's side and keep it from bending further. However, bending a shaping metal isn't something that I really have the capability to do myself. So I went and hooked up with a friend with the proper tools and materials and we made the proper pieces and I installed them in the car.

Here's the conceptual drawing of what I needed:

It's basically just a triangle of material that matches the panel behind the pedal cluster in the beetle. The big hole is where the rod goes through from the brake pedal to the brake master cylinder. The smaller holes beside it are where the bolts go through the wall to grab on to the master cylinder.

Here's my cardboard template to locate the bolt holes in relation to each other.

I don't have the larger hole marked on the template. Since only the outer part of the hole is defined by a rubber dust cap, its location isn't absolutely critical. The slot is necessary to slide the eventual piece of metal around the brake actuating rod to put the final plates into position.

And here's the following piece in the car:

The camera is right at the tip of the brake pedal looking down and away from the driver. The metal rod that goes away from the camera goes into the center of the brake master cylinder dust cap and in the connects to the working part of the master cylinder.

Something else I've been working on lately is a set of circuits for the beetle's warning buzzer. The way the stock system is wire, if either door opens, the dome light goes on (if the dome light switch is set that way). Additionally, if the key is in the ignition and the and driver's door is opened the buzzer sounds. I think that is imminently sensible...however, I also very much want to have the buzzer go off if I open the driver's door and the headlights are on. And there is the additional requirement (solved before) that the driver's door switch that I got in the car doens't work and I can't find a replacement.

So I'm building a set of circuits that will solve these problems. I sketched this up in a waiting room a few weeks ago:

The dotted box on the left is a double-pole, double-throw (DPDT) relay. The two dotted boxes in the right are single-pole, single-throw (SPST) relays. This design sets up a circuit that duplicates the original functionality of the dome light and buzzer and adds the buzzer sounding with the headlights on.

Here's a small "perf board" from Radio Shack upon which I've mounted the relays. The black box is the DPDT relay, and the three tubes are reed relays, which are SPST relays. There's three so that there's an extra.

And I've begun soldering the connections together:

So I guess all this is saying I hope that there are people like me, who have gone through school and have good technical knowledge about their subject, can be inspired to become someone who makes things, and doesn't settle for what they can buy, but that what you create yourself can be infinitely more satisfying and leads them on an eternal quest to find out more about the world and make it a better place. I guess part of that this the scientist in me talking. I really do believe that you can understand anything if you really put your mind to it, and that in doing so your knowledge and wisdom about the world will be profoundly impacted for the better.


it's only a flesh wound

2008 June 14 21:56

I thought I was doing well yesterday. I had voltmeter and current meter set up in the beetle. I finished what I though were the final connections, and then I reconnected the battery ground strap. Big spark (meaning something's drawing a LOT of current). Uh-oh.

So it seems that the current meter that I have requires an isolated power supply--basically the "-" on the power input for the current meter is connected to the "-" on the sense input. Fortunately, the only thing I lost was the current meter.

Lesson learned. Before starting to hook a piece of electronics into any system, particularly your car, doubly particularly your vintage car, ask yourself "what happens if I hook something up ass-backwards and connect any random bit to ground?". If there answer isn't "a fuse blows" then you need to go back and re-think what you're doing. If I'd put fuses in the wires that brought the shunt voltage to the current meter, I wouldn't have smoked the meter.

So...I need an isolated power supply. This isn't such a terribly exotic device...however, its pratical availability in rural southeastern Kentucky is limited. I went to Radio Shack looking for something to use. Nothing, so I left the store degectedly. If I was a more patient person, I would have ordered an isolated-output power DC to DC converter and waited for it to arrive before shipping. However, I'm not a patient person in a certain mood, so I walked back into radio shack and got the parts for a simple, chemical-based isolated power supply:

It's a nine-volt battery wired to a 7805 regulator that puts out exactly 5 volts, which is what powers the meter. I brought the wires out through the front of the panel so that I can plug it in or not when I go to drive. This isn't the elegant solution I was envisioning, but it allowed me to get it going today. Pure stubbornness, in other words. :-)

Here's the luggage bay with everything installed.

The black box in the lower right contains the shunt. The bright red cable carries the main bus current. On the left side of the photo, you can see the two white in-line fuses in the shunt voltage wires to prevent any more mis-haps from mis-wiring. At the top of the photo you can see the backs of the meters through what was the speaker grill in the original design of the car.

What I've been working on. Huzzah! (Note the box at the very upper right of the photo; that's powering the current meter.)

The top meter is the voltage (in volts) at the main wire from the battery as it enters the front luggage compartment. The bottom is the current in amperes coming through that same wire. So effectively it's the voltage and current of the entire electrical system.

Now I can start using the meters to characterize the electrical system. A few preliminary values: key on: 5.5A key on + low beams: 14.3A key on + high beams: 17.7A key on + high beams + brake lights: 20.9A (at which point the votage has dropped to 11.2V)

The red lights to the right of the meters are the warning lights at the bottom of the speedometer. The generator light is mostly covered up by the steering wheel.

I also tried plugging in one of the halogen bulb sport headlights from Mid-America Motorworks. The headlight connector fit. On each low beam and high beam, the sport headlight drew about one amp more on one side. Not too bad.

By the way, I'd like to give a shout out to the outfit that I bought the panel mount meters from. Discrete panel mount meters seem to be going the way of the Dodo. They're very hard to find in even catalog stores. How I found an outfit called colfusionx who sells these and other electronics on ebay. The meters are cheap and they deliver the product as described. They don't send instructions, I guess they assume that you know what you're doing. (If I'd first measured the resistance between the terminals on the current meter OR had proper fuses in the shunt wires, it would not have been a problem). One thing is they have been VERY responsive to my e-mails. Twice I've written to them with a question, and someone got back to me within 12 hours. The most recent time was to tell me that "no, I can't use the current meter to measure its own poser supply, and yes that will fry the meter, so you have to use an isolated power supply" and that was what I needed to hear. So anyone reading this who's interested in panel-mount electronic meters, I would recommend these guys.

Incidentally, they also sell output-isolated power supplies, so I think I'll pick one up.


burble burble

2008 June 12 23:36

My first real upgrade project for the beetle is to create a panel to the left side of the speedometer that has a voltmeter and ammeter for the electrical system. This will help bootstrap up to doing other electrical stuff. I started working on this panel the day we had people over to watch the Indianapolis 500 race.

Here's the back of the panel; I finished soldering tonight:

The power wires are on the lower and right side of the photo. The wiring convention in the beetle is that ground wires are brown, so I followed that. The wire with the in-line fuse brings in power. The other red wire is to power something else with the same switch that turns on and off the meters (GPS, maybe?). The wires in the the upper left of the photo are the sense wires that connect from the tap connections from the shunt. The wires with one connector each are the current sense wires. The one with the double connector is the voltage sense wire; I will connect it to whichever end of the shunt has the higher voltage.

After soldering everything together, here's a final hot test of the meters before I start installing things permanently in the car:

This is taken inside the luggage compartment where the wiring is. Notice the hole in the instrument panel to the right of the photo where the panel has been removed.

I made an enclosure for the shunt from a "project box" from Radio Shack. It should keep the shunt isolated from the car body. The heavy wires carry the current and the skinny wires are the tap that runs to the panel where the meter is.

By the way, after a second round of adjustments to the brakes, they're working very very well. On the first out-of-walking-distance drive, I noticed the steering wandering a bit. Not necessarily new, I just wanted to check it out. Well, there's a bushing in the steering system in the front end that I've ordered a replacement for. So I'm doing less driving until that comes, probably the middle of next week, and thus I'm working on system-level stuff.