Harbor Freight #63585 100 Watt Solar Kit

Seeking more power than what a 1.5 watt solar panel could provide, it’s time to step up to the 100 watt solar kit, Harbor Freight item #63585. The manual, posted online as a PDF, fails to describe a few useful details which we’ll cover here.

Every product picture showed the four panels lined up in a row. But in fact the four panels are capable of standing separately as each panel is in their own frame and has their own folding stand. Bolting them together is optional. If the panels are to be deployed and stowed frequently, leaving them separate might make sense as the panels are much easier to handle individually.

The package content lists wires but not their length. Each panel has a 3 meter long wire permanently attached. This wire terminates in a connector common to Harbor Freight solar products but its exact type specification is unknown. It is definitely not the MC4 connector common in rooftop solar installations.

(UPDATE: Thanks to a tip in the comments, we now know this is a connector commonly used in the automotive world and can be purchased from auto parts stores. For example it is commonly used to make electric connections to trailers. While this connector follows the pattern of SAE J928 and J1239, it is not explicitly covered by either specification.)

The four panels connect into a 4-to-1 module. The four wire side are half a meter long, and the unified side has a 3 meter long wire towards the controller. A final half-meter long adapter has the unknown HF solar automotive connector on one end and a barrel connector on the other. (~5.5mm OD, ~1.5mm ID, 12mm length) The barrel connector fits into a corresponding jack on the controller.

HF 63585 100W power adapter

Adding it all up: Each of the panels can be up to 3.5 meters from the central 4-to-1 hub, and that hub can be up to 3.5 meters from the controller. The package includes a 1 meter cable to connect controller to battery.

The kit included two LED light bulbs, each of which have a 5 meter long wire. Curiously, the long wire ends in a standard light bulb socket. But instead of the 120V AC household voltage we would expect from such a socket, it carries the battery DC voltage. This is a decidedly nonstandard and confusing way to do things. (UPDATE: An earlier version of this paragraph incorrectly stated 120V AC conversion took place, a bad assumption based on the standard light bulb socket. Voltage meter told the truth and paragraph has been rewritten.)

HF 63585 100W LED bulbs

The simple charge controller covers the basics, guarding against battery overcharging and over-discharging at adjustable voltage thresholds. The manual claims there is over-current protection as well, but there appears to be no way to adjust the current limit, either for charging or for discharging.

Hunt For Larger Solar Panels

Now that the project ambitions have grown beyond a little 1.5 W solar panel from Harbor Freight (Item #62449) the hunt is on for something larger. The 1.5 watt panel is intended to trickle-charge automotive batteries and is sized well for the job. We’re aware of much larger multi-kilowatt installations for household rooftop solar. What kind of market would support solar equipment in between that range?

One answer is the outdoor activities market, where some people desire a bit of electric power while away from civilization. Cell phones won’t work in the wilderness but there are still other reasons to have a power source: LED lanterns, GPS equipment, and cameras to document the adventure. The products designed for this market place focus on size and weight, important for carrying in a backpack. But since those values aren’t important for the current experiments, there’s no reason to pay the corresponding price premium.

Another answer is the market of people who want a less rugged experience away from home: boats and RVs. While these leisure vessels have power generators, supplementing them with solar panels reduce fuel consumption and associated noise and fumes. For this market there isn’t much desire to make trade-offs for size or weight, and so we can get more watts for the dollar.

Which brings us back to Harbor Freight who offers two products for this market. A small single 15 watt panel (item #96418) or a larger package featuring an array of four 25 watt panels plus a controller module (item #63585). The constantly varying world of Harbor Freight coupons means the exact dollar-per-watt changes for any given day. But the general trend is clear: between the 15 watt kit and the 100 watt kit, we pay roughly double the money for over six times the power plus a control module to treat the battery properly. The choice was easy to make.

HF 63585 100W

Old SLA Battery for a 1.5 Watt Solar Panel

The little solar panel (Harbor Freight #62449) has proven itself to be capable of sending out 1.2 W, within reasonable reach of the 1.5 W announced on the box. However, we’ve also learned its actual power output varies tremendously depending on its orientation relative to the sun and the weather. As a result it’s not terribly useful on its own. We’ll need to add a battery in to the mix.

Enercell SLA

An old sealed lead-acid (SLA) battery from the parts pile is thereby enlisted in the project. We can start the experiment by hooking up our solar panel directly to the battery terminals. It’s not ideal, but a big lead acid can tolerate this abuse, at least in the short-term. (Never do this with lithium-ion batteries of any size.)

The volt meter indicated this battery was overly neglected in storage, because its voltage had self-discharged down to 6 volts. This is far below the recommended range for lead-acid batteries and may have caused some damage. Fortunately it was able to handle a charging cycle and held an open-circuit voltage of 12.5 volt. Good enough to continue the experiments.

Once the battery is in place to cache power delivered by the little solar panel, we can now power a 12 volt USB charger and charge a cell phone on solar power. But the small panel does not track the sun throughout the day, so it could deliver only a fraction of its maximum power. As a practical matter this means the panel need to charge the lead-acid battery over several days before enough power is collected to charge a cell phone for a single day of use.

Based on the latest findings, we can take the solar investigation in one of two directions:

  1. Wring more power from the little panel: build a sun tracker so it can face the sun throughout the day.
  2. Throw money at the problem: buy bigger solar panels.

The sun tracker can be a fun project, but it’ll have to wait. The vote was decided by the arrival of a Harbor Freight coupon for their solar kit. So: option #2 it is!

Observing Behavior of 1.5 Watt Solar Panel

Now that we have a power meter for the small solar panel (Harbor Freight #62449) it’s time to take it into the sun and see what it does. The first lesson learned is that a solar panel’s voltage range is far wider than a battery. A charged and rested “12 volt” lead-acid battery’s open-circuit voltage is around 12.5 volts, and the charging voltage should never exceed 14.4 volts. In contrast, the open-circuit voltage of this “12 volt” solar panel sitting in the sun is more than double that nominal rating. Yikes!

Open 27V

While we expect the voltage to drop as soon as a load is put on the circuit, there’s still that momentary spike of voltage which might cause problems until we better understand how to handle it. Digging through the parts pile for a test load found a 24V cooling fan that was retired due to a bad bearing. Since it was designed for 24V operation, a quick spike of 27V (or possibly higher) should not be immediately fatal. The maximum amperage listed on the label is 0.1A, which translates to a maximum power ceiling of 2.4W so it should be able to handle the power of a 1.5W solar panel.

Upon connection to the voltage output, the fan twitched but did not start turning. A tap of the fan started it turning and we can see the solar panel delivering 9.3V * 0.0328A = 0.3W. This is only 20% of the advertised power while the panel is sitting flat on the ground.

Flat 0.3W

This was in the mid winter afternoon, when the sun is already at a fairly shallow angle relative to the ground the panel was sitting on. Now we have this baseline, the next experiment is to prop up the panel so it faces the sun directly. We expect the power output to increase, and the meter will tell us by how much.

Facing 1.2W

The answer: 19.3V * 0.0625A = 1.2W, or roughly quadruple the output, just by finding a better angle into the sun. This reinforces why solar installations prefer to face into the sun and some photo-voltaic solar systems even track the sun’s movement across the sky. Since this is not a rigorous test, there may be other factors involved that may overstate (or understate) the effects. But after this experiment it should be fair to state:

  • The advertised power rating of 1.5W probably represents the most optimistic value under ideal conditions, but we can get reasonably close.
  • Orienting solar panels to face directly into the sun makes a huge difference.

Measure Output of 1.5 Watt Solar Panel for “Free”

Now that we have a small cheap solar panel (Harbor Freight #62449) to play with, we can start exploring solar panel behavior. The initial quick and dirty test was to connect it to a car USB charger and while we were able to light a little power LED, the panel didn’t do much beyond that.

Before we repeat the experiment (and tackle new ones) we’ll need a way to monitor the voltage and amperage output. The multi-meter in the standard tinkerer toolkit can perform these measurements, but not both at the same time. Aside from the obvious problem of having only a single numeric display, there’s also the fact voltage measurement has the meter wired in parallel with the circuit while the amperage measurement requires it to be in series. Constantly changing wires around would get old very quickly.

We can buy instruments that are designed to monitor power output and simultaneously track voltage and amperage. But for the sake of a small side project, we’re going back to the Harbor Freight catalog. They sell some basic digital multi meter as item numbers #69096, #90899, and #92020. These multi-meters are frequently part of the Harbor Freight “free with purchase” coupon offering, so their prices are “free” subsidized by other sales.

Two of these were wired together so one monitors voltage and the other amperage, displaying their readings side-by-side. A few zip-ties to hold the contraption together and now we have a cheap clunky power meter. Here it is, showing that the solar panel has almost 18 volts of open-circuit voltage just sitting under the light of the photo booth.

Solar power monitor

The upside of a super cheap power meter is that we would shed no tears if an experiment accidentally destroys it. The downside is that we have to be realistic about the (in)accuracy of cheap Harbor Freight instruments. For one data point, we can connect a good Fluke multi-meter to compare readings. This difference is not great but acceptable for exploration.

HB and Fluke




Solar Experiments Begin with Small Panel

Solar power can be a part of everyday life in many different ways, from tiny solar-powered calculators to a home rooftop solar power system. There is great potential for interesting solar power projects. But before that: some investigation to get orientated.

The low-power capability of the 8-bit PIC micro controller might make an interesting pairing with calculator-sized solar panels, but let’s not overly constrain ourselves on power budget until we are comfortable dealing with it. Similarly, a home rooftop solar system is well into the realm of power that can kill, and thus a bad idea for beginner experimentation.

Let’s learn with cheap things first by starting with a small solar panel from Harbor Freight. Item #62449 is designed to be placed in a parked car to keep its battery topped off with solar power. With the coupons typical of Harbor Freight, it should be obtainable for less than $10.

The panel is advertised to supply power to a car battery. So our first quick-and-easy experiment is to wire it up to an accessory meant for car power: an USB charger designed to plug into the lighter socket.

Small solar

This particular USB charger has a blue LED to indicate power. When the solar panel is placed in direct sunlight, the blue LED illuminates. Unfortunately it doesn’t do much beyond that – if a USB peripheral is plugged in to charge, the LED goes dark and there is no sign of charging taking place.

Well, we’ve tried the easy thing first. Now we start poking around to better understand what we’re dealing with.

One Month of Living With Moto X4

The Motorola X4 “Android One Edition” is the mid-range offering for Project Fi subscribers as a more affordable ($399) alternative to the flagship Google Pixel phone. ($649) Here are some person notes on how they compared after using the Pixel for over a month followed by using the Moto X4 for over a month.

Moto X4

Exterior: The Pixel is a much more distinctive design compared to the relatively generic Moto X4. But the fragile aluminum & glass construction meant they both ended up encased in the cheapest TPU case from Amazon, muting the design distinction. Result: Tie.

Display: The Pixel’s OLED screen promised brighter colors, longer battery life, and a more responsive display. The responsiveness part was a requirement for Daydream VR. Outside of VR, the day-to-day user experience is equivalent. Result: Pixel wins in VR, but otherwise a tie.

Fingerprint Sensor: The Pixel sensor is far more reliable at reading and unlocking the device. The X4 sensor can be frustrating at times, occasionally sending the user to the backup screen pattern unlock. Advantage: Pixel

Camera: The Pixel camera is pretty fantastic, but the X4 camera isn’t very far behind. The X4 adds a second camera with a wide-angle lens that was occasionally very useful. People who need a high quality camera usually have a dedicated camera, phone cameras are useful as backup in a pinch and needs to offer flexibility. Advantage: Moto X4.

Storage: They both start at 32GB of storage but the Moto X4 augments that with a microSD slot for storage expansion. Apps need to stay in internal storage but movies, music, pictures, and some other data (Google offline maps) could be shifted to the memory card. If a Pixel owner needs more storage they need to buy a higher capacity device up front. If they find they need more storage later… they are out of luck. Advantage: Moto X4

Convenience: The Google Assistant on the Pixel hasn’t turned out to be terribly useful so far, neither has the Pixel-specific launcher. In contrast, Motorola’s customization to the X4 have proven to be more useful. A surprisingly handy feature is to turn the flashlight on/off by shaking the phone twice in a karate-chop motion. This was initially dismissed as a gimmick but it ended up being used multiple times a week. Advantage: Moto X4

Summary: There are clearly advantages to the flagship Pixel device, but if none of them are important to a user, they can save a lot of money with the competent Moto X4 and also get some nice features missing from the Pixel.

Disassemble Smoke Detector

One of the smoke detectors in the house has started raising a lot of false alarms and was replaced after a 3:30AM episode. Naturally, we’re going to take it apart. Today’s home smoke detectors come in two types: ionization vs. photo-electric. This particular detector, a First Alert P1210, is a photo-electric detector.

The cover was held on by four screws, only one of which was immediately accessible. The remaining three were blocked by the paper label which was easily punctured for access with a screwdriver.

Smoke Detector Lidless

Inside the plastic housing is a surprisingly large circuit board, its battery, and piezoelectric alarm buzzer. A cursory examination of the circuit board revealed some pads for absent components, which likely meant this board is shared with higher-end models with more features. Beyond those empty pads, almost a third of the circuit board real estate serves no apparent purpose.

The brains of the operation is a single chip printed with the Microchip logo which made it easy to look up its datasheet. A search on its numbers identified it as a Microchip RE46C190 which is apparently a turnkey solution for anybody to build a smoke detector around the chip.

Smoke Detector Sensor and Baffle

The nose of the operation is an infrared LED paired with a detector, kept in a housing that keeps the detector out of direct line of sight to the LED. When smoke particles enter the detection chamber, it will be illuminated by the LED and reflect light into the detector.

There is no obvious indication of why this smoke detector started sounding false alarms. Perhaps some dust entered the detection chamber? A smoke detector chip shouldn’t sound the alarm for just any reflection, it should look for specific characteristics of smoke particles. But there’s a chance we are expecting too much of this little detector.

And even if it does panic at any reflection regardless of source, the knowledge is not particularly helpful. The detection chamber and the baffles surrounding it is not accessible for cleaning without taking the detector apart.

For the immediate future, these parts will sit in a plastic bag. Added to the stockpile of electronic parts for potential future projects.

Disassemble Old Cordless Drill

While working on my NEXTEC Dustbuster project, I took the work-in-progress to various local maker meets to as a show-and-tell subject. This inspired another local tinkerer to bring a really old cordless drill for a compare-and-contrast session. It hasn’t run in years so nothing was risked by taking it apart. Which we did.

Old Cordless Drill

We see a motor that’s roughly in the same class as the motor in my Dustbuster. Instead of an air-moving fan directly attached to the motor output shaft, we have a simple reduction gearbox instead of the planetary gear popular with modern counterparts. Other missing convenience features common in current generation products include a torque-sensitive clutch and key-less chuck.

The most surprising part of this old design is how they implemented the two-speed mechanism. The trigger moves a switch into one of three positions corresponding to “off”, “low”, and “high” speed. Obviously “off” is an open-circuit and “high” speed connects all five (six?) battery cells to the motor. It’s the “low” that was a surprise – as far as we can determine, it connects three of the battery cells to the motor and bypasses the rest. This will certainly send less power into the motor, but it also results in uneven discharge pattern for the battery cells. Such behavior is considered absolute no-no with modern lithium-ion battery cells, and it couldn’t have been very healthy for these old NiCad cells, either.

In theory this drill could be revived with a battery transplant, or maybe upgrade to two-cell Lithium-Ion power. Whether it’ll happen depends on the owner, who should definitely find an alternative implementation for variable speed.

I Should Have Bought a Real Wire Stripping Tool a Long Time Ago

A lot of the talks at Hackaday Superconference 2017 were inspiring, informative, entertaining, or a combination of the above. But one of them is the first to have a significant impact on my hands-on projects and that honor goes to the Wiring Bootcamp talk by Bradley Gawthrop.

Your first reaction is probably the same as mine: “wiring? really?” Yes, really. At first glance a boring subject, Bradley turned it into an engaging presentation. One portion of the talk preached the wonders of having an actual wire-stripping tool. After the talk I felt motivated enough to try the Knipex tool he recommended.


After using it in a few projects, I found myself really enjoying the luxury of stripping wire insulation with a single motion. This purchase has thus been categorized under “Where have you been all my life?

Knipex Jaw.jpg

Key to the magic is the relationship between the handle, the front jaw (black plastic) and the cutting blade (shiny metal.) When the handle is first pulled, the motion goes towards closing this assembly. When jaw closes on the wire insulation, the blade closes a little bit further to cut into the insulation. Beyond this point, motion on the handle is translated into horizontal movement so the blade pulls the insulation away from the conductor.

There’s no obvious way to adjust the distance between the jaw and the blade. It is either fixed or inferred from some spring tension. This works fairly well, the only problem surfaces when cutting wires with very thin insulation. In these cases the blade bites too deeply and nicks the conductor.

But that is a minor nitpick. I certainly nick conductors at a far higher rate when using my previous wire strippers. Which have been assigned the job of collecting dust while waiting as backup in case the Knipex breaks.


I got myself a real wire stripping tool and loved it. You should do it, too.

Here’s the wiring talk posted on the Hackaday YouTube channel:

With Great NEXTEC Power Comes Dustbuster Responsibility

An old and tired Dustbuster BHD9600CHV is now upgraded with the power module of a NEXTEC LED work light. The eight-cell nickel-cadmium rechargeable battery pack built into the Dustbuster is gone since it could no longer deliver anywhere near its 9.6 nominal voltage rating. Now cordless power comes courtesy of the NEXTEC three-cell lithium-ion battery sending nominal 11.1 volts to the motor.

And what an upgrade it is! It now generates enough vacuum to pick up anything we can expect a small handheld vacuum to handle. It is now genuinely useful again and the handle blended well enough that ergonomics didn’t take a hit as I had feared. But all is not well; this upgraded super powerful Dustbuster now hints that there may be such a thing as too much power.

Intake Exhaust

The massively increased airflow is great for picking up messes, but all that air pulled into the vacuum has to exit through the exhaust. This massively increased exhaust has become a new problem that didn’t exist before. Vacuuming small objects off a surface now requires a bit of planning in terms of approach angle and the distance at which we turn on the power to the motor. Approach too shallow or turn the motor on too early, and the exhaust airflow will reach the object before the vacuum side can pick it up.

When this happens to small objects, say picking up a few peanuts on the ground, it means the peanuts run away and we have to chase it down. But when it happens to a loose clump of light objects, say an area filled with dust bunnies, it means the exhaust blows everything up into the air and we end up with an even bigger mess on our hands.

So yes, this is now a massively more powerful handheld vacuum, but that new power has consequences.

(Cross-posted to Hackaday.io)

Behold The NEXTEC Dustbuster

After waiting 24 hours, the adhesive should be fully cured. First the clamp holding the trimmed Dustbuster body to the NEXTEC battery compartment was removed. Then the screws holding the two halves together were removed. Finally the big moment: try to pull the two halves apart. If they were accidentally glued shut, it’ll be difficult to put all the components back inside the case.

With a soft pop, they opened up. The pieces that we’d want to stay together stayed together, and the parts that we want to separate separated. Whew!

Successful SPlit

From there it was straightforward to reinstall all the necessary components back in. Then the soldering iron was warmed up to make the electrical connections. The new wires are thicker gauge than the original wires, but it makes sense to upgrade the wires due to the increased battery voltage.

Assemble internals.jpg

As a cleanup task, the former charging port’s connection to the motor frame were severed. This avoids any accidental charging of the new lithium battery pack at nickel cadmium battery voltages. We’re keeping the charging connector plate solely for cosmetic reason of not leaving an empty hole in the bottom of the enclosure.

Before putting the two halves back together, it was wise to perform a quick test with the battery to verify the switch functioned and the motor is turning in the correct direction. The final step is to put the two halves back together and reinstall all the screws.

NEXTEC Dustbuster

This old Black & Decker Dustbuster is now powered by the Sears Craftsman NEXTEC battery system. It is definitely far more powerful than a week ago on weakened Ni-Cad battery cells. And likely far more powerful than when it was new. It may even be more powerful than its modern successors, as a search indicated they are running on two-cell lithium-ion batteries instead of our three-cell upgrade.

This concludes a fun upgrade project with a practical end result.

(Cross-posted to Hackaday.io)

Cut & Paste For NEXTEC Dustbuster

All the investigation and planning has led to this point: cutting apart the old Dustbuster and the unused NEXTEC LED work light so we can put them together into a NEXTEC-powered Dustbuster. Time to deploy the Dremel and install the cutting wheel!

First the components are set out. The original plan, outlined in red, was to cut off the Dustbuster handle up to the flat internal brace, and cut the work light hinge so it will clear the Dustbuster body.

Cut 1

The handle cutoff was straightforward, and once the handles were cut off the pieces could be put up against each other. Once in place the Dustbuster body curvature was found to be more complex than originally expected, making it tough to cut the LED work light body to fit. However, the work light hinge mechanism is a simple cylinder that is almost the same diameter as the Dremel cutting wheel. It would be much easier to cut a scallop into the Dustbuster body to accommodate the cylinder so that became the new plan. (Red arrows.)

Cut 2

While the now-exposed flat brace inside the Dustbuster body is indeed flat, the front face of the work light body turned out not to be perfectly flat. As a result they didn’t mate flush so we’ll needed adhesive that can bridge small gaps. A search of adhesives already in the workshop nominated the Loctite construction adhesive for the job. While applying the glue, the center had to be avoided since we don’t want to glue the two halves together.

The instruction for the adhesive said to brace the joints with physical force while it cures over 24 hours. So the two parts will be held together with a small clamp while we wait.


(Cross-posted to Hackaday.io)


Fitting NEXTEC Compartment to Dustbuster Body

We have a Black & Decker BHD9600CHV Dustbuster. We have a useless LED work light that’s a part of the modular NEXTEC lithium-ion battery system. The Dustbuster needs new batteries, so let’s remove the LED and attach its battery compartment to the Dustbuster. This will give the Dustbuster 12V lithium-ion power, a healthy boost over its original 9.6V Ni-Cad power.

Loose fit

We’ve disassembled both devices and can start with some loose test fits for spacing. Here’s an arrangement that removes as little plastic as possible. It keeps most of the Dustbuster handle and attaches the work light body to the end. Physically this design would be very long and has a lot of wasted space. Let’s tighten things up.

Tight Fit

We want to keep the Dustbuster switch, so we’re not willing to cut any forward than that. We don’t care about the work light switch, so we can cut that off. In fact everything is fair game up until the battery connector which we want to keep. Pushing the battery connector plate all the way up against the Dustbuster switch gives us the most compact layout feasible.

This will work from a component layout perspective, but it will be tricky to precisely cut all the pieces to fit. The industrial design of both devices have a lot of curves for us to deal with. However, they both have some flat surfaces we can try to use to make our work easier.

Flats Fit

The Dustbuster has a flat brace just behind the switch. The work light has a flat surface where the battery compartment ends and the swivel LED begins. If we line them up, we should be able to attach the flat surfaces together. There will be more wasted space this way but the reduced fitting headaches should be worth the trade-off.

(Cross-posted to Hackaday.io)

Disassemble NEXTEC LED Work Light

Today’s project is to disassemble the NEXTEC LED work light and see if we can use it to adapt an old Black & Decker BHD9600CHV Dustbuster to lithium-ion power.

The wish list of the disassembly operation are:

  • Battery compartment: If we could use the battery compartment of the work light, don’t have to reverse engineer the dimensions of the battery pack and the slots needed to clip the battery in place.
  • Battery connector: If we could use the battery connector of the work light, we don’t have to reverse engineer the precise location and dimension of the metal contacts for drawing power from the battery.
  • Battery over-discharge protection: Unlike Ni-Cad batteries’ tolerance of discharge, over-discharging lithium-ion cells could cause permanent damage. As a result, most lithium-ion devices have a protection circuit and I’d like to pull that in.

The things we don’t care about:

  • LED array: There are plenty of LEDs of all color and brightness on every electronic tinkerer’s workbench. One fewer array would not be missed.
  • Switch: The work light’s switch is the type where a press closes the circuit, then another press opens the circuit. This is the right behavior for a light but not for a vacuum. Also: this switch was designed for a low-amperage LED and while it looks sufficiently beefy, it might not tolerate the amperage draw of a Dustbuster motor.

With these goals in mind we start with the obvious task of removing the four visible screws. After the screws were removed there was one more fastener: a small metal C-clip holding the two halves together near where the pinkie finger rests while holding the light. The clip was designed to require a specific tool for a clean removal. For people who are unconcerned about cosmetic damage, it could be persuaded to abandon its post with pliers.

Work Light Disassembled

Looking at the disassembled light, we see we can easily re-purpose the battery compartment and associated battery connector for the project. The third item on the wish list – the over-discharge protection circuit – is unfortunately incorporated onto the LED circuit board and not an easily separated part. We’ll just have to be careful when using the upgraded Dustbuster and not discharge it too much.

Having the battery compartment and electrical contacts is great. It bypasses a lot of iterative CAD work and 3D printing to pin down proper dimensions to fit the battery. The next step is to join the two devices together.

(Cross-posted to Hackaday.io)

Lithium Batteries for an Old Dustbuster

An old Black & Decker BHD9600CHV Dustbuster has been crippled by its degraded battery. The prime suspect for the degradation is the simple charger unaware it was overcharging the battery pack. We’ve decided to leave behind both the degraded Ni-Cad battery and suspect simple charger. We will replace them with rechargeable lithium-ion batteries.

We have many lithium-ion battery designs to choose from. The different physical form factors are visually obvious, but there are also less visible internal differences. The motivation for these internal differences come from optimization for different usage scenarios. A battery with the wrong usage pattern could result in poor performance, poor battery life, or maybe even fire.

The first candidates are the battery cells extracted from dead laptop battery packs and occasionally used for small electronics projects. These cells were designed to hold a lot of charge then dispense power slowly and evenly over several hours of a laptop’s battery run time. A Dustbuster has a different discharge profile, pulling a lot of power in short bursts. They might not be happy together.

The next candidates are the batteries for remote-control hobbies. RC aircraft and cars are very hard on their battery packs, drawing high amount of power almost constantly over a period of time usually measured in minutes. There were some batteries on hand for small remote-control aircraft, but they could not sustain the amperage draw of the Dustbuster motor. While it is an option to buy some higher-amperage rated battery packs, and they would happily power a Dustbuster, but they are an expensive overkill.

We would prefer something between the above two extremes of the spectrum. Looking around at other lithium battery packs already available, attention settled on a set of cordless power tools powered by 12V lithium battery packs. (Sears Craftsman NEXTEC series.) A handheld vacuum has a power profile very similar to most power tools: short bursts of high power draw. The 12V nominal voltage is higher than the 9.6V of the old battery pack but close enough the motor should be OK as long as used in short bursts. And hopefully the higher voltage gives the vacuum a boost making it even more capable than new.

These cordless tools were purchased in a combination pack, all using the same NEXTEC battery. The LED work light is the least useful member of the pack in this household already full of LEDs. Let’s take it apart and see how we can use it to adapt a Dustbuster to run on the NEXTEC lithium-ion rechargeable battery pack.

Nextec Work Light

(Cross-posted to Hackaday.io)

Dustbuster Battery: Next Steps

The battery of my old Black & Decker Dustbuster BDH9600CHV is too weak to do its job. Now that we’ve found enough evidence to suggest the battery was degraded by constant over-charging, there are a few options forward: repair, replace, or upgrade.

Cells Removed


The first option is to try to repair the existing battery cells. The information on Wikipedia indicates the cells could be repaired by performing several deep discharge+charge cycles. This must be done on an individual cell basis, because deep discharge of a pack risks irreparable damage to the weaker cells from cell reversal. It would be time-consuming even if the equipment to automate this process is on hand.


The second option is to purchase new Ni-Cad battery cells and replace the degraded ones in the Dustbuster. Ni-Cad batteries are cheap, but assembling them into packs are usually done with the help of a small spot welder. In theory the battery tabs could be attached with a soldering iron, but it’s very difficult to solder a battery cell because the metal can acts as a heat sink drawing heat away from the solder joint. If too much heat goes into the battery cell chemicals, it will damage the battery.

Whether repaired or replaced, the existing power adapter that constantly overcharges the battery pack will need to be retired. What we would need is a good charger with a controller that knows how to properly charge a NiCad battery. Without this knowledge, we’ll quickly return to the same predicament with a battery pack ruined by overcharging.

Usually, when working on a project, it’s fun to buy new tools necessary for the job. Our candidates are:

  • Individual Ni-Cad cell deep discharge+charge cycler.
  • Ni-Cad battery tab spot welder.
  • Ni-Cad aware smart battery charger.

Unfortunately, they’re all dealing with Ni-Cad battery cells, which has faded to niche applications and their use is not expected for future projects. The present (and foreseeable future) solution to portable battery are lithium-ion chemistry cells, and that’s the motivation for the next option:


Since the battery charger would need to be replaced anyway, that removes the main motivation to stay with an electrically compatible chemistry. Freed from that constraint, the most interesting path forward is to find a way to power this old Dustbuster with an entirely different type of battery.

Let’s bring this Dustbuster into the 21-st century with a lithium battery upgrade.

(Cross-posted to Hackaday.io)

Investigating Dustbuster Battery Degradation

When embarking on a project to repair something, it’s always helpful to understand and articulate what went wrong so we have confidence we’re fixing the right thing. The starting point for this project is seeing this old Black & Decker BDH9600CHV Dustbuster trying to do its job: It couldn’t spin its motors fast enough to generate vacuum to pick up household debris. The most obvious suspect is the battery pack, so let’s examine the batteries.

Dustbuster screws.jpg

Five small screws held the two halves of the vacuum enclosure together. Once the screws were removed, the two halves separated easily without any additional glue or plastic clips to worry about. The internals were as expected – a battery pack, hooked up to a switch, and wired to the motor driving the vacuum vanes.

The battery pack is built from eight nickel-cadmium (Ni-Cad) battery cells. Six arrayed around the motor, and two more tucked in the handle. The first hypothesis is that some of the cells have died. The cell voltage levels were probed as the motor spun, looking for any cells that has sunk to zero volts or possibly a cell-reversal situation. All eight cells delivered under 1 volt but well above zero, disproving the initial “dead cell” hypothesis.

The next hypothesis is battery memory effect. Technically the term applies to a very specific issue with Ni-Cad battery, but in popular use the term has become an umbrella for several different conditions that afflict Ni-Cad batteries.

The most promising item under the umbrella was “Voltage depression due to long-term over-charging”. The voltage has already been verified to be low but not zero. There should be a charging control circuit to prevent over-charging, perhaps that failed? A search came up empty: there didn’t seem to be a charge controller at all. The batteries seemed to be connected directly to the output of the charging stand AC power adapter.

The nominal voltage of this battery pack is 8 * 1.2V = 9.6V. The maximum output of this Dustbuster’s AC to DC adapter? 24V. Ouch! That’s significantly over nominal and the battery pack has been held at that level for years.

These circumstances imply this battery pack has indeed suffered under long-term over-charging. Explaining why it now deliver depressed voltage levels.

24V DC

(Cross-posted to Hackaday.io)

New Project: Handheld Vacuum Upgrade

Taking a break from reviving old computers, the next project is to revive a small household appliance. The subject of the upgrade is a handheld vacuum. Specifically a Black & Decker BDH9600CHV, a member of the “Dustbuster” line whose success defined a whole new product category.


A major factor of the success is their easy of use. Whenever there’s a cleanup task, it’s easy to pull the vacuum off its charging stand and clean up the mess. No need to pull a big heavy vacuum out of the closet, no need to look for the nearest plug. A small cordless handheld vacuum is very convenient and people are willing to pay for that convenience.

The basic design of a Dustbuster is straightforward: a battery pack hooked up to a motor controlled by a switch. As a result, the majority of the vacuum are durable and reliable thanks to their simplicity. With the notable exception of the battery pack. The battery pack is what makes the cordless vacuum possible and easy to use, but the battery is also the weakest link.

This particular Dustbuster had been sitting in the standby charging base and the battery power capacity gradually dwindled over the past few years. Now the battery pack, even when freshly charged off its charging stand, could only offer a little bit of power before the motor slowed down and couldn’t generate enough vacuum to be useful.

In today’s disposable society, it’s easy to just throw away such a simple and inexpensive appliance and buy a new one. But where’s the fun in that? Since the rest of the vacuum seems to be OK, the goal of the new project is to give this vacuum new life by some combination of repairing, restoring, and/or revamping the battery pack.

Let’s open it up and see what we find…

(Cross-posted to Hackaday.io)

Time-of-Use (TOU) Electric Bill: Good Concept, Poor Execution.

Household electric power is a great convenience of modern living and a triumph of engineering. Most people living in first-world countries are so used to having power whenever we want it, we only think about the work behind the scenes when there’s a power failure. Most of the electricity generated is immediately consumed. So when consumption changes throughout the day, generation has to be kept in sync. This balance act is ongoing at all hours of the day, every day, and most of us don’t have to think about it.

One detail of this balancing act is the increasing cost of power as demand rises. Obviously the power company would like to keep their cost as low as possible, so the power plants that are the cheapest to run (base load power stations) run constantly. As demand outstrips the ability of the base load stations, they begin generating power via increasingly more expensive means. These peaking power plants are usually cheaper to build but more expensive to run so they are only used to handle peaks in demand.

Residential electric bills are typically insulated from this change. The standard home electric meters simply record the running tally so the homeowner is billed on the total amount of electricity consumed. With the introduction of more sophisticated meters, it becomes feasible to track energy usage in more detail which makes it possible to bill based on time of use. (TOU)

TOU rates correlate cost of consumption to cost of generation. This gives the consumer a financial incentive to be more energy-efficient. If enough energy usage is shifted around, the consumer can save money. Over a year ago, the local electric utility sent out an invitation to join a TOU pilot study. It would be an interesting experience to see that theory put into practice so the invitation was accepted.

The TOU rates were listed on the bill, where the peak hour rates are indeed appropriately expensive, roughly triple the non-TOU rate. But the off-peak rates were tremendously disappointing: only a 20% discount off the non-TOU rate. A 300% penalty for on-peak vs. 20% discount off-peak means it’ll be difficult to actually save overall money under this plan.

But the study was on and it’s time to put in the best effort. The most significant changes came from running the laundry machines and the dishwasher during off-peak hours as much as possible. On some days this was a severe inconvenience. The effort continued but the consumer was not always happy about it.

At the end of the one-year study, they mailed out a TOU cost summary. They took the year’s electric use and computed it two ways: Once through the TOU pilot study rate, and again on the non-TOU rate.

The reward for ecological awareness? The windfall for severe inconvenience?

SCE TOU Unimpressed

$1.93 over the entire year.

From the perspective of encouraging people to save, this was a complete failure. The utility needs to discount the off-peak rate much more significantly than they did during this study before people would see enough savings to be worth the inconvenience. The kind of time and effort expended during this year was not remotely worth saving $1.93.