Easy locking Ariens snow blower axle

I've got a big Ariens 1028 with an axle lock pin (also called "lynch pin") on one side. (It doesn't have remote wheel lock control.). Looking forward over the handlebars, the left wheel is always powered. It's easy to maneuver the machine when the right wheel is unlocked. The lynch pin normally sits in the outer hole to hold the wheel on the axle.

The problem is how to re-lock that wheel to the axle for extra traction. You need to pull the pin from the outer hole and pass it through the holes in the wheel hub and axle. But the freely turning wheel has two degrees of freedom: in and out along the axle, and rotating on the axle. So it can be hard to align the holes to pass the pin through, especially while standing in lumpy dense snow!

My solution is to keep the wheel in a fixed position along the axle with all the holes in the same plane. I experimented to add the right washer thickness, and added a short pin to hold it all in place. Now I can easily push and pull the handlebar, in an arc centered on the other wheel, until the holes align. Then the lock pin drops easily into place. Flip the ring over the hub to hold it in place and you're ready to go. For storage, I clip the ring to the handlebar.

This particular axle is 7/8 inch diameter and has an existing thrust washer. A 14 gauge "machine bushing" washer was the right thickness to add on this machine. Then I cut down a 1/4 inch clevis pin to be shorter than the diameter of the hub and drilled a new hole for a cotter pin. (I since learned of universal pins pre-drilled in several places; just cut to length.)

The axle with two washers.

Cut and drilled clevis pin

Assembled with a cotter pin. The lynch pin drops easily into place.

Ready to go

View fullsize

Storing the pin

Modulator fixed

The Channel Plus 5445 four-channel analog modulator with H838 distribution amplifier (DA) includes infrared (IR) remote control of our cable DVR from various locations around the house. I've had the system, running continuously, for about 15 years. The modulator powers the DA over the coax and the same cable returns IR from devices at the TVs, through "IR blaster" ports on the modulator, to an IR emitter at the DVR. Over the last couple of years, the DA has occasionally shut off spontaneously. After I unplugged the system for a while, it would work okay. Finally, in Fall 2015, it stopped altogether. I powered the DA with a separate supply, but lost IR control of the DVR. Since this system's IR targets run on 5 volts, and nothing made today works with them, I was looking at a complete system replacement for more than $600. I considered upgrading to high definition distribution, but it just didn't seem worth the substantial cost for a kitchen or basement TV.

I couldn't find any schematics online, and the Channel Plus company has been through many name changes and owners. Their customer service number rings busy all the time, and currently, they are owned by a distribution company that has no customer support line at all. I was going to have to fix it by myself, or buy it all over again.

I traced through it and made hand-drawn sketches of the whole thing, except for the delicate cable input filter on the H838. That probably works fine, but it's not in use.

The circuit I've drawn here is the over-current limit circuit, in the modulator power supply going to the DA, to protect the supply from shorts. The function is needed because if you forget to put a DC blocker on a DA output that doesn't have an IR target in-line, the TV can draw a lot of current on that cable.

The specific problem is the 10 µF electrolytic capacitor on pin 9 of an inverter.

Current limiter components

The large blue cylinder is the 5.1 ohm sensing resistor. Below it are the contact pads for the pass transistor. The first time I tried to diagnose this, I broke that transistor while trying to unsolder it. It was an On Semi J31C and after this photo, I installed a TIP29. The circled parts are the ones in my schematic. X marks the bad 10 µF capacitor. (Plus I've labeled the transistor that drives the IR outputs from the modulator.)

The circuit measures the voltage across the sense resistor. The MC14069 inverter chip is supplied with the same 15 volts as the rest of the circuit, so its thresholds are related to its supply. As long as the sense voltage drop is minimal, the 11/10 inverter holds pin 9 low, so pin 8 stays high and the pass transistor stays turned on. The 10 µF capacitor holds pin 9 low during initial startup, when a brief current surge through the system pulls the sense voltage down, pushing pin 10 high. This 10 µF capacitor was partially shorted, so it was acting as a resistor in parallel with 510KΩ, rather than as a current sink to keep pin 9 from rising. Instead, at startup, pin 9 jumped up and shut down the pass transistor, and that was that. A replacement capacitor, worth about 30 cents, solved it completely and it's been running fine for over a month.

Bang and Olufsen 3i headphone repair

These B&O headphones stopped working with the iPhone 6 Plus (no audio out), and were intermittently opening Voice Control or playing music on their own on an iPhone 5s. Here's what I learned while fixing them.

1) The iPhone 6 Plus seems to shut down the headphone audio if there's a problem with the mic/control circuit. The iPhone 5s copes, at least in the short term.

2) Others have reported similar odd behavior:

iPhone 6 iOS 8 Voice Control comes on randomly while using headphones, how do I disable??
 
I wonder if small interruptions of the common circuit can cause static that's basically inaudible to mobile users but the phone is recognizing them as command pulses. Or if the mic side opens, or shorts to common due to a bad cord, there's no audible signature unless the user is on a call.

3) The B&O 3i cord has five conductors that go to the four-circuit plug. They are:

Tip    -  Green - Left speaker
Ring  -  Red - Right speaker
Ring  -  Red/Green - spiral-wrapped shield for microphone, twisted together with
         -  Copper - common for the speakers.
Sleeve - White - Microphone center conductor

I used four layers of shrink tubing to reconstruct the plug housing. Here's photos of the wiring and rebuilt plug.

Canon i80 Bubble Jet printer gear fixed

We hadn't used this portable printer much, but then I came up with an application. On testing, it started to work, then made terrible noises. Others have seen the same cracking failure of the two-part metal-nylon paper-feed gear (e.g. http://youtu.be/HvR47ZOuYNA). I believe the manufacturer forgot to allow for shrinkage of the nylon. The core is metal, and the plastic became stressed as it shrank down around the core. It was okay for a few years, but eventually the stress pulled the gear apart.

There is some mechanical repair information available online, but no schematics or other electronic repair guidance. Since it was electrically functioning, I just had to try a mechanical repair. I am very pleased with the results.

How I fixed the gear

1) Stop-drilled stress relief holes at the ends of three cracks: the main one and two others I found under a microscope.
2) Extracted the metal core (the stop holes allowed me to manipulate the plastic without growing the cracks). I measured the core and estimated how much to reduce its diameter to allow the crack gap to close, starting with 2% from the shrinkage tables for nylon.
3) Ground the core down from 12.7 mm to 12.3 mm. Ground the taper on both ends. Found that I needed to deepen three vertical grooves as well. Reamed out the plastic part to give more relief. This took 4-5 fittings; the cracks didn't get much worse and the part stayed together. I was satisfied when the crack closed under light pressure and there was still enough friction in the core (due to the three grooves) to drive the gear.
4) Drilled holes across the gap and closed them with dental ligature wire, with the twists in a groove on the metal gear side.

How I got the gear off

Getting to the gear isn't easy. I found some helpful info from the Parts Catalog (photos show some part relationships) and Service Manual (some advice on assembly).

http://elektrotanya.com/canon_i_80.zip/download.html

I suggest you take photos as you work to help identify what goes back where. I turned the printer over once, and parts fell out. My photos showed me where they belonged.

1) Remove three hinged lids by bending them in the middle slightly.
2) Remove Bluetooth adapter if present.
3) Remove screws at back.
4) Remove screws at front edge.
5) Press lock tabs and remove dark plastic corners at back.
6) Press lock tabs and pull up on top cover from the back.
7) With top cover partially open, use a tool to press on the lock tab under the paper feed button in the center section and lift the cover toward the front.

Getting to the gear means taking the chassis out only partially. Stop dismantling as soon as you have access to the gear. That way, you can avoid dealing with all the many connectors and cables.

8) Untape the Bluetooth socket's cable where it's attached to the bottom case.
9) Remove the paper lifting plate. Looking from the back, on the left side there is a latching tab. Press the pin toward the outside and swing the lever to the back. Release two springs and pull the plate out. (Service manual advises replacing it with small tabs above paper return tabs. I didn't see them. Maybe it depends on the state of the machine at shutdown.)
10) Under the battery, there is a metal shield. Untape the wires passing over it, then remove four screws at plastic tank, center back edge of plate, forward and down near circuit board, and on the right above the USB port. Pull out the shield enough to remove another screw on the right at bottom that anchors a bracket to the base.
11) There are two release tabs. From the back, one is on the left, inside next to the carriage motor (revealed under the removed lifting plate), and the other is on the right, outside between the USB port and the brass motor gear. Also pry the shell away from the power connector, and now you can lift the back of the chassis out of the case, and pull back slightly to free the front edge. Just lift it enough to get the gears clear of the case and prop it up.
12) Pry the right gear off its shaft.
13) Use fine-tipped tools to open and release the plastic split ring holding the cracked gear on its shaft. (I use dental tools.) Maybe add some tape over the work area to reduce its fly-away range. I did find mine eventually, but Canon considers it throw-away.
14) You can now slide off the gear, helping the cogged belt slip off the metal core on the inner side. Take a photo of the area; to put the gear back, you may have to remove and reassemble the cogged belt's idler pulley.

If you have questions, please contact me.

Intermittent Sonos ZonePlayer ZP90 fixed

One of my 6 year old ZP90 players (now called "Connect") has had an intermittent problem for several months. Sonos Customer Support were very patient and thoughtful but didn't see any errors in my diagnostic dumps, and didn't have a solution. Finally, it seemed to stay dead.

The problem was sometimes no audio would play from the house speakers. From the controller's view, everything was fine. I connected sensitive headphones directly to the jacks and could hear that the sound was actually there, just very faint. Also, sometimes while making a small change on a controller, like changing a queue entry, the sound would fade down, very briefly, and come back. I knew it was specific to this unit because two other devices continued playing elsewhere in the house, suddenly audible in the local silence.

Opening the ZP90 involves going into territory with very few DIY-serviceable parts. Here's how I opened mine: Lift the rubbery bottom corners away from the double-sided tape and remove four screws from the corner wells (after passing through the now-expired warranty sticker). Slightly lift the rubber along the sides and push small flat screwdrivers about 2 cm into the two snap slots on each side. Pull the shell off carefully because the volume/mute button panel is cabled to a circuit board. You can pull off the panel's connector at the board and set the shell aside. Undo three antenna cable snap connections, and one screw at the back of the top board. Lift the board off by rotating it slightly up from the front toward the back so the multipin connector to the lower board gently separates. Remove the power supply connector from the top board, then pull the board forward and up to let the ethernet ports clear their back panel opening. Reassembly is just the reversed steps.

A bulging, oozing capacitor on the power supply board caught my eye. (Others had black marks, but no leaks.) It was next to one lead of the high power switching diode. So, I expect it was heat damage, not capacitor plague.

I worked with the power supply board in place because removing it involves damaging the back panel to get to the screws. The hardest task was cleaning lead-free solder from the through-holes. I twisted the diode lead to the side to give a tiny bit more clearance from the new capacitor. The player now works fine. I guess the problem was board-wide voltage fluctuations as various circuits drew power, like the WLAN when controller state changes needed to be propagated.

Bulging capacitor. High power diode lead is on the right.

LED replaces T1.5 lamp

A station needed a simple switcher. In a storage room, I found a 3M Mincom 101 vertical interval switcher collecting dust. In the lab, it appeared to be fine inside and out, flat to 4.8 MHz. Maybe it was retired because half the selection indicator lamps had burned out, making it impractical to use. The lamps are T1 1/2 (Tubular, 1.5 x 1/8 inch diameter, or 4.8 mm). They're labelled S12V 30 mA, and appear to be available from Europe (R286 dashboard bulbs). But that would be very inconvenient.

I salvaged some white LEDs from broken cheap flashlights, chose a series resistor for about 6.5 mA current (your value depends on the voltage), and built a set of 10 replacements on pieces of Vero stripboard in an evening. Note that the stripboard should be trimmed to the edges of the copper traces to stay under 5 mm wide. And the leads go along the outer edges of the strips because the sockets are H-shaped. The new lamps are a bit brighter than the originals, so maybe I should have used more resistance, but I hope they'll last substantially longer than the 3000 hour spec on the old lamps. Another refinement would be to sand the LED ends to illuminate the buttons more diffusely.

Lamps, buttons.

iPhone 6 Plus case

My wife just got an iPhone 6 Plus and wanted a case for it. She really liked the Case-Mate Wallet Folio but the sleep button didn't work. We tried two samples; both had the problem.  Others online have reported hard operation. I could see why.

The sleep button on the right side of the iPhone 6 Plus is the same shape as the volume buttons on the left side, but it's flush with the body of the phone.  The case designers knew where the buttons are located, but they missed that profile. I surprised the clerk by buying the case anyway. With three layers of aluminum foil duct tape added to the case button (Nashua 324A trimmed to about 1.5 x 7 mm),  it presses the sleep button correctly. Works great.

Foil tape, 3 layers