Mountains

Mountains

Monday, November 25, 2013

Blow Hole

One of the first shocks of homeownership was the first gas bill. We left the house unoccupied while we worked on it for the first few weeks, so I was quite suprised when a 300$ bill showed up for heating an unoccupied building.

It was obvious that the many air gaps in the structure were doing nothing for it's energy efficiency.

Since then, I have been on a hunt for airleaks in the building. In a house that is designed to be windows-up in the summer, that eliminates the summer months as a rational point to find small leaks; the house breathes through the open windows and there is little movement through other routes.

Now that winter has returned, the problems of last spring have again become obvious.

The current home repair goal is to eliminate drafts and heat loss for general comfort and to further buffer the upstairs temperature, which only seems to have a Goldilocks temperature for 4 weeks out of the year. In the past few weeks I have been weather stripping the doors and windows, insulating the steam pipes (subject of an upcoming post), and draft sealing the attic.

The foundation has a small portico structure that extends under, and to a degree, supports the front porch. The porch is decked with tongue-and-groove boards, so I thought it might be air tight, especially after the brickwork in the foundation was fixed. However, while insulating the steam pipes, a strong draft was still coming in through the portico.

I solved that little problem.


I had originally designed a more elaborate, removable plug that used large pieces of polyethylene sheeting as the air seal. However, when I started cutting the board and really examining the shape of the hole and the discomfort of crouching in the crawl space, I resolved to act on the more direct cut/tape/foam permanent installation. If there is ever need to access the interior of the portico, the foam is easily enough removed, and this approach is completely air tight.

While installing the plug, the velocity of the draft through the gaps in the side increased as I plugged them. The flux of air was actually substantial, like the exhaust of the shop vac or a wimpy leaf blower, probably driven by convection of warm air out of holes in the upstairs. A few hours later, The Girly commented that it seemed a lot more comfortable. That can be taken as progress.

Sunday, November 24, 2013

ChevOldsmoBuiac: Front Half-Shafts, Wheel Bearings, and Halfshaft Seals

Some photos and notes for anyone looking to replace the halfshafts and transmission seals in their A-body Oldsmobile Cutlass Ciera.

Some bygone aphorism stated that every journey starts with a step. In the case of half shaft replacement on a 24 year old car, that step requires a 36" long wrench and a 24" cheater bar. The axle nut was originally put on with Loctite and 190 ft-lbs. Then it had all of time and every weaving frost-heaved winter road in New Hampshire to rust over.


Marco Polo, eat your heart out.

The main inspiration for replacing the seals and and halfshafts is that they were all leaking. The shafts had been leaking for years, and though they still seemed good and tight, were in the way of replacing the seals that were leaking transmission fluid more quickly than I was comfortable with. I replaced the bearings too, since they are easiest to remove while the shafts are out, and the were getting a little chatting in hard curves.



The wheel bearings are simple to remove on the Cutlass Ciera: They're held in by 3 star-drive bolts to the steering knuckle. No hydraulic press needed. The half-shafts came out of the bearings really easy. A half-hearted rap with the deadblow mallet broke them free.


Leaky tie rod ends: a project for another day.

The knuckle needs to be separated from the strut in order to pull the halfshaft out. I painted guidelines on the assembly so I could reinstall without throwing the alignment. The bolts holding the strut to the knuckle were fairly corroded, and a huge amount of work was required to break them free and unthread them.


The halfshaft is held onto a splined haft inside the transmission by a spring circlip. It takes a bit of force to pop it of. I ended up jacking the car up another few inches so I could put the breaker bar between the frame and the shaft housing, then hit it hard with the deadblow hammer. The housing looks like plastic, but is really cast aluminium, and it suffered no ill effect from the whack.


The exposed shaft in the transmission.


Getting the axle seal out of the transmission was not an easy job. Haynes called for a hammer and a cold chisel, but there is not enough room to get a good angle on the seal, or a good swing with the hammer, so the process of deforming the old seal required patience, several flatblade screwdrivers, and a needle nose vice grip. To get the seal out, I pulled little bits off until I could get between the seal and the transmission to really deform the seal casing. It took over an hour to finally dig it out.

Transmission without any seals. Drivers side, I believe.

Getting the new seal in was as hard as getting the old seal out. Finding a round object suitable to evenly push the seal in was not easy. I tried for another hour or two to use various scraps of PVC and the socket for the halfshaft nut, but the seal refused to go in flush. I eventually went to the hardware store and purchased a 2 foot piece of pipe that fit over the seal exactly. I also removed the strut to give myself a straight shot at the seal.

In the end, 10$ at the hardware store did in 10 minutes what I had spent several hours failing to accomplish.

Say hello to my little friend.

New seal, installed.

New and old halfshafts, compared. The new shaft slide right in and clicked into place without a lot of force.


Wheel bearing removed from knuckle. There is a dust seal on the back of the knuckle that matche a seal on the halfshaft. However, there is also a seal on the bearing itself. I still stuffed this with grease to keep anything from rusting together.

New wheel bearing.

Passenger side leaks, for posterity.

The passenger side halfshaft has a splined in the plugs into the transmission. I suppose it's impossible to put it together backwards, even if the different shaft lengths didn't give it away.

I pulled the bearing off the passenger side before removing the halfshaft to give myself more room to work. It only helps a bit, still a tight squeeze.

Installing the new shaft revealed that it came with a ABS sense ring. This oldsmobile doesn't have ABS, and, strangely, had a large iron ridge in the steering knuckle that almost seemed to exist to keep the ABS shaft from being installed correctly. The method of attachement for the abs ring wasn't readily apparent (glue? pressed? spotweld?) so I opted to grind of the ridge on the knuckle rather than try to remove the ring on the halfshaft.

After grinding, I had to clean up all the shavings to keep them from getting into the assembly. In retrospect, it took so much time to grind the knuckle that putting a notch in the ABS ring to see if it would easily come off would take a lot less time.


General notes:
-Every fastener seemed to be rusted together. This job could probably be done in less than 8 hours if nuts and bolts would spin freely.
-Paint guidelines on the interface between the struts and the knuckles to preserve the alignment.
-Remove the struts. It's just 3 nuts and it makes a lot more space in the wheel well.
-Jack up both sides of the car so it's easy to push/pull the rack without having to twist the steering wheel. Getting the end of the halfshaft out of the wheelbearing required a lot of experimentation to get the knuckle into just the right spot.
-Come prepared with lots of grease and threadloosening fluid
-Be patient and have lots of pointy tools for removing the axle seals in the transmission.

Blinkenlichten II

Once taken, twice removed, I suppose.


By which, I mean to say that it was fairly certain that fixing the broken christmas light timer was possible if all the electrical components were intact. We bought two last winter, and thereby instantly removed the bedtime "Who's going to unplug the lights?" discussion and also the equally awkward "Wow look at the december electric bill!" associated with running the lights 24/7 during visits with family. On at 4:50, off at 11:00 every night.

Then, sometime this summer, one of the timers stopped working. Screen blank. No clicking relay. Zilch. Reset button did nothing. Plug/unplug did nothing. Appeared dead. But of course, I had already opened it and determined that it was almost too bloody simple to die. As long as the ballast capacitor stays in spec, it should work about forever.

 I pried the timer open to see this again:

For reference, the left side PCB with all the components is an AC/DC converter, and an SSR driven by the transistor, which is in turn driven from the blue wire that leads to the right hand PCB, which is the timer microcontroller, the screen, and buttons (not shown). The major components on the timer board are few. On the back, there's a oscillator, a photoresitor ("Dusk"), and a battery (the green shrinkwrap thing). The timer charges from a 1.45 volt line through the red and yellow wires. I think the AC/DC conversion is actually 12 volts (for the relay), and then a seprate resistor drops it further to supply the controller. Though, that leaves the purpose of the big red capacitor unanswered. Regardless, I wasn't there to diagram circuits.

The controller battery was a) charging and b) charged. That narrowed the problem down to the controller and the screen. I removed the control board and noticed that one of the fingers that holds the screen to the timer PCB was loose. LCDs often use a silicon surface contact. I removed the screen and found the controller CPU underneath, along with test, reset, ISP and ground pads. I grounded the reset pad for 5 seconds cleaned the LCD contacts,and pressed it back together until it snapped. The screen started working, blinking time like a VCR.

Rock on green box, rock on.

Friday, November 22, 2013

Pop Goes the Breaker

The breaker blew.

Not for the first time.

The lab was silent as all the pumps, computers, heaters, and devices on the north wall went dark.

My frustration drove me to reset the breaker to get the ball rolling again. The pile of shit to do was high and the tolerance for failure low. Then, I realized something: every time the breaker had blown previously, I moved things to different outlets to distribute the load between other breakers. The bench had outlets labeled #2, #4, #6 and even a few #7 and #8s. Why were they now dead too? Shouldn't it only be outlets of a common number?

I turned off some pumps and went to the electrical box. Breaker #2 was warm and popped. #4 and #6 were stone cold and in the on position.

The outlets in the lab were mislabelled.

A thermograph showed that only a few breakers were carrying any load at all, with #2 getting the brunt of the use, probably overloaded.

 

Time to call an electrician.

Wednesday, November 20, 2013

Just After a Successful Alarm Clock Hunt

The girly, with the look of satisfaction after giving an alarm clock justice.

Li'l Tommy Knocker: The Joy that a New Knock Sensor Can Bring

Our Volvo 740 (1990 B230f with 368000+ miles) has not been a smooth operator, at least not in the three years that have passed since I saved it from the scrapper. Things have incrementally gotten better as I have moved through various ignored and differed repair and maintenance tasks, but it is still some distance from original performance, whereas the ChevOldsmoBuiac is usually ~90% of new.

Part of the Volvo's problem is engine related. Things got better when I replaced the long worn plugs, and stopped the oil leak into the distributor that was causing misfires. It still bothered me that the engine tended to lug in high gear, moderate speed, and moderate RPM. Pinging 2k rpm seemed unusual. I had ignored it, but this summer, it seemed to get a bit worse.

As a wild guess (that I later justified using a note on swedishbricks), I replaced the knock sensor. It is a true one-bolt job and it's conveniently located on the side of block where it's in easy reach. You can even fit a normal ratchet with a normal socket into position with clearance! It's like it was designed to be replaced! I practically threw a party when I realized the swap would be done before the coffee got cold.

I didn't really expect the replacement to improve anything; usually I'm wrong when I think a sensor is bad. However, in this case, I was immediately rewarded with an engine with smooth power delivery at low RPM.


Given that the symptoms were not super-obvious, the old sensor must have had decayed sensitivity. Inspecting the old and new did not reveal any signs of failure, and, frankly, a little cleaning would make them almost indistinguishable.

Tuesday, November 19, 2013

Poof.

It has been three years since NMT decided to a) eliminate alumni TCC accounts b) go back on the agreements with those who had prepaid for long periods or purchased lifetime accounts, and c) alienate me as an alumni and insure I would never donate money to the school.

In the interceding time, my original site soldiered on in zombie form. I was unable to login, but the page was still there, and worked as recently as a few weeks ago. However, it now appears that www.nmt.edu/~holstien/ is now truly expunged from the TCC servers. For several weeks, that URL has returned a 404.

You can, of course, still see it on the Wayback Machine. I don't have vital statistics, but those of you still interested in Power Macintosh G4 fan noise might still find that information useful.

I intend to continue to maintain a presence here for the immediate future. While I had originally planned on a quick move to a different host I directly control, Blogger works well enough and the amount of space available for storing photos is increasing faster than I care to create and post new entries. At some point, it is likely that google will neuter blogger and ultimately attempt to merge it with google+ into a single product no one thinks is cool or useful. Then I will once again consider the options for my broadcasts into the nothing.

I may resort to writing letters and mailing them.

Until then, keep drinking the formaldehyde.


Tuesday, November 12, 2013

Weepy: The Brief Saga of Replacing the Thermostat in the ChevOldsmoBuiac

Last spring, the Girly started to complain that the 1990 Cutlass Ciera's heater wasn't keeping her warm, requiring her to arrive at work more frozen than she started, and more annoyingly, leaving defrosting as an exercise for the driver to undertake. Longer trips would eventually yield hot air, but it required a lengthy sit in traffic rather than a 4 mile lull to the office.

That seemed like a classic lazy or stuck thermostat symptom that could possibly precipitate an disasterous overheating situation. It's easy to fix by swapping the thermostat, a ~$5 part that takes about 10 minutes to replace, yet could ruin the vehicle if left unattended.

So, of course, I let it glide all summer until I had enough of a wish list to include a new one in an autoparts order. Shipping costs totally ruin the fun of buying things online, you know?

Oops.

For the record, I did fix the problem before anything bad happened, at least, beyond mild frost bite. That heals, right? Right.

The thermostat comes out by removing two bolts from the thermostat cover and pulling it back, which releases a torrent of green antifreeze from either side of the union.

I got a stant superstat instead of a standard OEM replacement. It seemed like it would get the heater rolling a little sooner and worker harder to regulate the car's temperature.


The old thermostat looked somewhat better than i imagined. The radiator has a considerable amount of white corrosion inside, but the thermostat and the housing where clean. There was some wear on the body of the old one, and a small gap between the gate and the thermostat housing, but otherwise, it looked functional.


While the parts store stocked a gasket that claimed to fit the car, the thermostat was put in with an RTV silicone gasket. I scraped off the RTV and put the loose fibre gasket on instead, since I had no RTV handy.


I refilled the antifreeze, started the car and let it got warm and the heater worked.

To check my sanity, I dropped the old thermostat in boiling water. The OEM spec is to turn at 195 C. The thermostat waited until 200-210 to turn, then waited until 150 or so to close again. Motion wasn't always smooth. Indeed, the old thermostat was the issue.

After a week of driving, a puddle of antifreeze appeared. Green goo everywhere! The fibre gasket was leaking. I first tried tightening the thermostat cover bolts. A careful exercise since the bolts are iron and the cover and housing are aluminium.

That didn't stop the leak, so I got a tube of RTV and pulled the cover again.


The fibre gasket was a hair small for the bolt pattern on the thermostat housing. When I put the bolts in, it tore the gasket just inside of the crimp line, causing the leak. RTV, obviously, doesn't suffer from these kinds of problems.


I am pleased to report the heater starts producing warm air within minutes.

Friday, November 8, 2013

Flowers


Flower photos: the cheap generics of art photography. Pretty. Universal.

Thursday, November 7, 2013

Yet Another Derby Photo


In my photography, I work very hard to get the camera settings nailed so I have a minimal amount (by which, I mean, no work!) to do in post processing, beyond eliminating the shots that aren't interesting and attractive. The creative dimension of editing is mostly cast aside. Every now and then, I remember that the elements that make an image interesting can be amplified and distorted to change and enhance what the image actually says. There are knobs and buttons that don't get pressed, but could.

Right now, rollerderby is the most interesting subject I work with. I've been playing with capturing the raw banality of the DC suburbs, but, by definition, that's a very stale subject to work with, and requires an extremely careful editorial eye, otherwise, the photos are just pictures of 'burbs. There's a lot of depth in derby that i'm leaving untapped, either by vision or technique. I should think about how to get a little closer to the nerves, and see what digging will yield.

Monday, November 4, 2013

Photographic Reconstruction

Repetition:

Depth:

Offset composition, rule of thirds:

Depth:

Repetition, composition, organic detail:

Rule of thirds, organic detail:

Construction sites are intriguing places to take pictures. The raw components and geometry of the building are open to examination, and many of the components of visually stimulating art are represented. There is also room to think about the work. There's no rush to manage a living subject, and the different components can be abstracted from each other. The material itself is almost inconsequential: the photographer, you, are giving it a voice through the medium.

When I lived in New Hampshire, there were many construction projects around, and I had a lot of opportunities to explore the potential. One of my favorite prints is printed on a poster and hung on a wall over my desk. Here, such things are not quite so common.