Tonight I rebuilt a leaking hydraulic cylinder on the mower. The leak was coming from the scraper/seal that seals the rod to the end of the cylinder. Unfortunately since the end of the rod is larger than the seal, the cylinder has to be disassembled to replace it. The process went as follows:
#1 – Remove cylinder from mower.
#2 – Remove the internal cir-clip that holds the cylinder end inside the cylinder bore. This was extremely fiddly and took 20min or so with dental picks and small screwdrivers to get the end of the clip pulled out.
#3 – Remove the cylinder end and piston/rod from the cylinder bore. The cylinder end’s O-ring catches on the cylinder’s cir-clip groove, so while it moves freely to a position that’s ‘almost’ out of the bore the last bit of movement requires clamping the rod in place and persuading the cylinder downwards with a hammer.
#4 – With the piston/rod/end removed from the cylinder, the nut can be loosened and the piston removed. The piston is a press fit and requires a gear puller to remove and a hydraulic press to reinstall.
#5 – With the piston removed the cylinder end slides off of the rod and the scraper/seal can be replaced. The scraper has it’s own small cir-clip that’s much easier to remove.
#6 – Re-assembly is the opposite of assembly, while I was at it I also replaced the cylinder end O-Ring.
#7 – Reinstall
After the rebuild the mower no longer has any hydraulic leaks; the last remaining leak is an engine oil leak that I’ll be tracking down next…
Christina has had an eye out for a piano for a while. Interestingly the standard price for used pianos is ‘free, must be able to move’. With this in mind we were shopping on ease of moving – looking at the surroundings in the listing photos to determine where it likely was (i.e. garage vs inside) was almost as important as the piano itself. The right one popped up this week, so this weekend we took the trip to pick it up. I had a particular interest in this one also since it’s a player piano. Player pianos are one of the earliest (the earliest?) wide-spread programmable automated ‘machines’; it’s fascinating to see how these were designed to be built using mostly hand-made parts and to operate without electricity.
Moving: Pianos are the archetype of things that are difficult to move. The trick, as usual, is letting the equipment/physics do the work. At the pickup side it was already in a garage; after backing up near it I jacked the tongue of the trailer up so the ramp/bed formed an even & shallow slope. I was ready with the winch at this point, but because their driveway sloped away it actually rolled onto the trailer without needing it. To unload, I backed the trailer around the back of the house to the back door and then raised it up on jack stands until it was about even with the threshold. We then put a 4×4 timber across the bottom of door frame of a nearby interior door and used this as an anchor point to winch the piano inside with ratchet straps. Once it landed on the tile floor it was easy to push around.
Rebuild: As near as we can tell from the serial number the piano seems to have been built in 1920. It was also signed inside with a 1996 date, likely associated with a rebuild. The good news is that everything seems to be in place and after sealing a few vacuum leaks it was able to play automatically; It definitely had/has room for improvement though. I was able to make it sound noticeably cleaner by dialing in the correct ‘capstan’ height on the back of the keys. The next priority is getting the keys working more consistently; several of the keys (~40) were missing felt from their front hole. The front hole felt prevents the key from wobbling side-to-side. Several keys (~20) also had cracks where the back hole goes through the key. The back hole is where the key pivots, so these cracks allowed the keys to tilt side-to-side. Felt was added to the front holes and the cracked back hole parts were glued back in place; this resolved the loose key problems. For just a few keys (3) the cracked part is missing and I’ll need to create and glue in a repair piece. Once I’m finished rebuilding the keys I can then go through and set the key level and adjust the hammer action. Along the way I’ll also be coming up with an electric vacuum pump – pumping with the pedals is a work out!
This week’s challenge was a flat tire; I switched to the spare relatively quickly and uneventfully except that when I was tightening the security lug bolt the security key slipped out a bit. When it slipped it broke a little chunk off of the security key and warped the security pattern on the lug bolt. I ordered a new key and lug right away but since I didn’t want to risk the warped lug bolt breaking the new key, I had to come up with some other way to remove it.
First I tried a broken lug extractor – these are cheap and readily available at the local car parts place. They’re basically a regular impact socket with a reverse thread on the inside rather than the usual hex pattern. The theory is that as you loosen the lug the socket’s threads bite into and turn the lug bolt. This actually would have worked great if it weren’t for a collar that’s integrated into the security bolt – the socket only bit into this collar and it spun on the lug bolt without turning the bolt body. It’s almost like they thought of this scenario when the security lug was designed…
Next I tried to cut some small grooves in the bolt head and deform the collar into these grooves so that the collar would be able to drive the bolt out. I think this was an OK idea, but since the collar was glass-hard steel it chipped instead of deforming.
Running out of options, I decided just to weld a nut to the security lug bolt and unbolt it. I had held off on this ‘nuclear option’ because there is a level of risk – the ground current could actually weld the bolt into the hub somewhat or damage bearings/electronics – I chose the ground clamp location carefully and also cleaned it carefully to avoid this. It was a messy weld (prioritizing not damaging the spare) but worked great. I think the heat from welding may also have helped by expanding the bolt and relieving some of the pressure against the spare.
With the lug bolt out the spare was swapped to the (now repaired) wheel and tire. Not sure how a dealer/mechanic would handle this scenario, but I’m guessing it means the welder has now paid for itself for at least the 3rd or 4th time.
Last fall prior to Hurricane Florence I took down some good-sized white and red oaks. These were great trees but both had large limbs overhanging/threatening the garage, so they had to go. Rather than cut them up for firewood as usual I let them stay where they fell through the winter. Today was unseasonably warm and I took a stab at milling one of the logs. Drying takes a long time, so the point of this was to get at least something started – this way if I don’t finish milling the rest of it for a while I’ll at least have something in the queue.
To make the milling cuts I just free-handed with the chainsaw and there’s a good bit of variance in thickness as a result; to get straight boards I’ll have to deal with this variation at the joining/planing stage. For the next attempt I plan to build a metal guide frame over the saw to allow it to rest level on top of either a ladder laid on the log (first cut) or the level surface made by previous cuts. Also, my saw is a middle/low power model (3HP) and bogged down occasionally. Normal chainsaw chain is meant for cross-cutting and takes too big of a bite for ripping, for the next attempt I plan to modify an old chain into a ripping chain by grinding back some of the teeth – these converted teeth will help clear chips out of the cut rather then cutting themselves and it should mean less bogging down.
Once milled, the log was reassembled with some ‘sticker’ pieces I cut from scrap 2×4’s to create airflow gaps. It’s under an overhang that should provide enough rain protection. The drying happens from the inside out over a long period of time, getting wet from rain only temporarily increases the moisture level on the outside; this dissipates quickly and doesn’t hurt the overall dry time. Drying should take about a year per inch, the slabs are about an inch and a half on average so I may be able to use these as soon as next summer.
We have some ideas for the wood but no immediate plans; this is just a long-term thinking/prep. The white oak is good outdoors and may become some much-needed patio furniture. The red oak may be used to upgrade the fireplace mantel and be used for a headboard and side tables.
This weekend I installed the shop dust collection system. The system consists of several parts:
Fan/Motor: I repurposed a portable dust collector fan I’ve had for a while that’s been underutilized (collecting dust, but not as intended). Space is limited in the mechanical room so since the fan won’t need easy access I mounted it high up above the air compressor near where the dust collector pipe enters the mechanical room.
Pipes: 4″ PVC DWV pipes; there are a few branches leading to the different tools. I tried to keep the overall length as short as possible and the bend radius’s large.
Blast Gates: The blast gates control the air flow though the system by blocking off unused branches. I made these with 1/2″ plywood and 1/4″ hardboard. Circle cutouts were made on the lathe to match the pipe outside diameter exactly.
Filter and Collection Bin: The portable dust collector came with a light canvas bag that restricted the air flow massively while still allowing fine particles to escape. To improve this I replaced the bag with a semi truck air filter mounted to a trash can. The theory is that air will exit the filter and larger dust/chips should fall into the trash can below. There are purpose-built dust collection filters available, but the costs are much higher for these and the semi truck filter has the same specs; different economies of scale. To mount the filter to the bin I made a plywood ring, for now they’re just taped together but I may add latches at some point. The design may need some tweaking; I’ll know more after it gets further use, but for now the airflow is excellent.
Return Vent: Having the collection bin in the mechanical room created a problem; the mechanical room is well sealed for noise reduction, so there was nowhere for the air exhausted from the filter to go. For heat/air to be retained in the shop, the exhaust air needed to return to the shop via a vent. Since I also wanted to keep the mechanical room noise level as low as possible this meant the vent needed to be sound proof. I built a sound proof vent by creating a 3ft long box and offsetting baffle plates inside of it. The sound has to reflect a dozen or more times off of the baffle plates; at each reflection it gets absorbed some by a fiberglass lining. The air, however, is able to snake around the baffles and find its way out. The inlet to this vent also points directly at the floor away from the sound sources. Somehow after adding this vent the mechanical room noise is actually noticeably quieter than when it was completely sealed. I think this may have had to do with the air pressure changes resonating in the previously sealed room, whereas now any fluctuations are equalized through the vent.
Control: For now control of the system is via a remote control outlet (repurposed from controlling the vacuum at the old shop), at some point I may integrate some low voltage switches with the blast gates so the motor will turn on as soon as any gate is opened.
There are a few things that need to be solidly mounted to the workbench in use but that otherwise make sense to be easily moveable either for storing out of the way (bench grinder, polisher) or to make way for big projects (vise).
To allow this, I installed threaded inserts into the workbench in key locations and then used bicycle style quick release bolts to attach/detach tools when needed. I also cut the bolts down so they have 4 or 5 turns of engagement to make threading quick.
I’ll also be cutting the grinder and polisher cords to the exact length to the nearest outlet from this position, this makes it easier/neater to store; if I ever need to use these somewhere else temporarily I can always use an extension cord.
(I’ve had this built for a while now, but I’m catching up on documenting shop progress tonight)
A while back I got an old Delta ‘Portable’ power planer at a garage sale; it’s helped out in a few projects already so I thought it deserved it’s own spot in the shop design. The planer requires a fair amount of space for it’s infeed and outfeed to keep the material flat, however the planer isn’t used enough to justify keeping it out on the workbench permanently. I also wanted to avoid the need to lift it out of a cabinet, so with this in mind I came up with the cart below that integrates with the cabinets. The cart rolls on double-locking casters and has folding infeed/outfeed tables that align with the planer’s table when upright. I used ‘drop leaf’ style hinges/supports to allow the tables to lock in place.
The ‘dead’ space inside is currently holding the router and router bits; I may convert this to a drawer at some point. When the cart is stored it just looks like any ordinary cabinet door.
Over the last few nights I’ve been working on handles for the cabinets. The handles are 4″ sections cut from 1.5″ aluminum angle. Before cutting to length I ripped ~1/2″ off of one side of the aluminum so the handles weren’t too wide. This was my first time cutting aluminum on the table saw, it was very quick and effective but also shot scalding hot aluminum chips in every direction and sounded like a Pterodactyl being fed through a wood chipper.
Since I may add/change cabinets in the future I wanted to make sure that I wasn’t locked into a particular brand/model of handle that could be discontinued, by making my own with standard materials this isn’t a problem and the cost was also kept to a bare minimum.
Once all the sections were cut to length I sanded the rough edges and drilled/countersunk holes on the drill press. I made a quick fixture with clamps and wood blocks to get the holes consistently located. To attach the handles to the cabinets I chiseled out a 1/8″ recess and then attached the handles flush with the door edges.
I’m undecided on whether these will keep their current ‘brushed’ looking finish or whether I’ll polish or paint them. I’ll decide that when the time comes to finish/paint the cabinets.
This weekend I resumed work on the shop and made major progress on the cabinets:
– Upper cabinet built
– Another cabinet built for small parts storage
– Doors for existing base cabinets built
The doors are ‘shaker’ style for simplicity; building these is as easy as cutting a piece of 1/2″ plywood to the right size and then wrapping it in the same 1×2 poplar used for the cabinet face frames. All the door frame pieces got a 1/2″ dado to fit around the plywood edge and then the top/bottom frame pieces had 1/8″ removed from either side of their face on the ends to make a tenon. I found that the majority of the build time is in changing between the various setups needed (cut to length vs dado cutting vs tenon cutting vs cutting plywood, etc) so after the first test door I tried to build as many doors at a time in parallel as possible.
I went with ‘inset’ mounting of the doors because i like the clean/simple look. This gives a lot less room for error compared to overlay mounting since the gap is visible and it and needs to be small and consistent. It’s critical that cuts are within 1/16″ and that everything stays perfectly square or things go downhill fast. For the dados I used an old ‘wobble’ style dado blade (these are so sketchy they’ve been banned in europe); the geometry of this contraption is such that it can’t leave a perfectly square bottom. It’s not really a problem since the dado bottom is internal/unseen, but it made measuring and setting the correct dado depth difficult.
I definitely had a few mistakes to correct along the way, but overall I think it was just enough of a challenge to help improve my woodworking. Next steps are to make handles, make/install drawers for the last open base cabinet, and (eventually) do the finish work of filling/sanding/painting.
I’ve recently been playing with whole-house energy monitoring and came across a way to log and retain data using Google Sheets. This monitor (long before I got it logging automatically) also helped size the generator project from earlier posts.
#1 – Aeon Labs Gen5 Home Energy Meter. The meter installs inside the breaker panel and reports current, voltage, power, and energy back to the SmartThings hub via Zwave. The meter monitors both ‘legs’ of the standard residential 120/120/240 feed.
#2 – Samsung SmartThings Hub. The hub connects the Energy Meter to the internet where the SmartThings server allows for custom applications to be developed and run in their cloud.
#3 – SimpleEventLogger. This is an application that’s been developed by the SmartThings community that posts data to google sheets via the google API anytime a particular dataset changes on the SmartThings side. Setup is only slightly more involved than adding any other ‘custom’ apps in the SmartThings IDE.
In addition to the energy data I also have temperature data from the various HVAC zones; the thermostats that came with the house were “RedLink” compatible, so we only needed to add a RedLink hub for data and control to be enabled in the ‘cloud-to-cloud’ connection between SmartThings and Honeywell.
Having this energy and temperature data available will be key in optimizing energy usage and identifying trends/problems over time. I’ve had it running now for a few weeks without issue; eventually/hypothetically the log entries could run the sheet file size up against the maximum storage size for the google account, but the entries are so tiny this can’t really happen in practice.