Harbor Freight Router Speed Control Mod





Introduction: Harbor Freight Router Speed Control Mod

About: Engineer/Scientist Interests include: * Physics based toys and demonstrations * Things that spin, like Levitron, Fidget Spinners, Origami spinners * Electronics * Cooking authentic ethnic food * Magne...


The Harbor Freight Router Speed Control is a bargain-priced (on sale for $19.99, with a 20% OFF coupon, just $17.40 with the tax) general purpose electrical controller, despite its very limited name. It is widely used for controlling heating elements, motors (be careful here, it only works with certain types of motor), and high-power stage lighting, just to name a few uses. I bought one to control a popcorn popper to roast coffee. The popper was getting too hot so I needed to be able to control its temperature and had read about using this device for the purpose. You just plug in the power cord and you can use the 3-way switch for Var(iable), Off, and Full. It is spec'ed to be able to handle 15 amps (1500W).

The need to mod…

After using it several times, I decided that the chip, which is a BTA26-600B A320 Triac, was getting too hot. Touching the backplate, with the unit not even on full power, I could only stand the heat for about 2-3 seconds. This was designed to provide intermittent power to a router, not continuous power to a heating element. Even though I can replace this chip for around $2 on eBay, you never want to overdrive a component like that, so I decided to add a heatsink, like the guys who use this unit for controlling heating elements for beer brewing and controlling industrial lighting. The heatsink cost me more than a replacement chip and some time and labor but I like to 'do it right the first time' (old IBM slogan), or in this case, the second or third time ;), whenever I have the luxury to do so.

Step 1: The Mod

The heatsink

So I ordered this nice 60x150x25mm High Quality Aluminum Heat Sink for LED and Power IC Transistor H148 from eBay seller goldpart.

It took weeks to get here from China but it is a quality product and only cost me $4.49 shipped. The backplate on the controller, which serves as a poor heatsink, measures 110mm x 54mm. You can get a heatsink on eBay that closely fits those dimensions but, for a dollar more, I opted to get this oversized one.

Detach components from the backplate

I decided that the easiest way to do this was to simply detach the components from the backplate and mount them directly to the heatsink.

Upon removing the backplate's 4 screws, the first thing to do is to detach the mounted components from the backplate.

Drill out ground wire rivet

Drill out the rivet holding the 2 green ground wires. I used an electric hand drill with a 1/4" bit.

Go slowly to the point where the flange of the rivet just releases.

There was a steel or aluminum ball inside that I poked out with an awl.

Not done yet. The inside of the rivet still has the crimped sleeve attached to it. I carefully drilled that out with a drill bit that was small enough to avoid damaging the rings of the terminals on the ground wires. Probably around 1/8" but YMMV.

One terminal freed but the other one is still attached.

Easy does it.


Unmount the triac

Remove the screw attaching the triac, with a screwdriver.

Drill out the rivets of the hanging clip

I also removed the hanging clip from the backplate so I could use the plate as a template for drilling the holes in the heatsink.

Terrible heatsink…

I was horrified to see what the manufacturer thought was a heatsink. A pebble-grained rectangle of 1/8" aluminum and NO heatsink compound! There are so many things wrong with that. First of all, a heatsink has, you know…, fins, to dissipate the heat quickly, not act as a thermal mass, building up heat to the point where it burns your fingers. Second, a heatsink is nice and smooth on the side where you mount your heat-generating component. I've even seen where over-clocking fanatics will actually lap their heat sinks with a progression of grits up to 6,000 to get mirror-finish smoothness. Over-clocking overkill to be sure but it does make the point about the requirement for smoothness. Reserve pebble-grain for leather hand bags or something else… Third, heatsink compound is absolutely mandatory to insure intimate contact between both surfaces, to promote maximum heat conduction. It fills in any micro-gaps, that surely exist on both mating surfaces, with a heat-conductive paste.

Lousy soldering job…

I was also horrified by the lousy soldering job I saw on the neutral wire direct from the wall socket.

Had to re-solder…

So the first thing I did was to re-solder it.

Not perhaps an expert job, but a lot better than it was.

Prepare the heatsink

Lay out the drill hole template

Lay out the positions of the drill holes for the mount holes and the components on the back of the heatsink. Note that the side of the backplate that was OUTSIDE the box is placed against the back of the heatsink to assure the greatest accuracy.

I first drew lengthwise guidelines by measuring the widths of the backplate and the heatsink, subtracting them and halving the remainder, and then using that value for the offset from the side edges of the heatsink. I then carefully positioned the backplate, aligning its sides with the guidelines, so that the inner edges of the mount holes just snugged up against the edge of a fin opposite them, tracing each hole with a fine permanent marker. Note that the holes for the 'bottom' (component side) are opposite the left side of the 6th fin from the left, and the holes for the 'top' side are opposite the right side of the 7th fin from the right. This was the best I could do to avoid drilling into those thick extrusions next to the fins. I drew lines under the 'inner' edges of the 5th and 6th fins and measured and marked the 'horizontal' and 'vertical' center between them for the drill hole for the 2 ground wires. I somehow messed up when drilling, and the hole ended up against the 6th fin. No biggie since I was able to get a rivet through it anyway. The hole for the triac just happened to line up with the left side of the 8th fin when the mount holes were lined up. I carefully selected a drill bit that was small enough to allow me to tap the hole for the screw.

Drill the holes

I clamped the backplate to the heatsink, being careful to line up the mount holes with the traces on the heatsink. I placed a strip of wood under the fins to avoid damaging them and rested the whole thing on 2 pieces of wood. I then drilled the 4 mount holes and the triac mount hole, using the backplate as a guide. Choose a drill bit that fits the mount holes of the backplate. Then choose a smaller bit to drill the triac hole--one that is small enough to allow tapped threads. Be sure to tap the triac hole with an appropriate size and pitch tap. You'll have to determine this on your own. I drilled the 1/8" rivet hole for the ground wires by center-punching the cross-hairs and then drilling, but my bit slipped a little as mentioned before. I couldn't use the backplate as a template because the hole would've gone right through the 6th fin.

Smooth all rough edges

Sand the cut sides and file the edges of the heatsink

I didn't take any photos of this tedious step. I first sanded the 2 saw-cut sides of the heatsink with 400 grit wet-or-dry to remove any rough spots. I removed any rough edges from the holes I drilled, both on the bottom and between the fins, with a small fine needle file. I then filed every rough edge of every fin with a small fine flat file. The tops are smooth already so didn't require filing. I beveled the sharp-pointed corners at the tips. I beveled the 4 bottom edges and the 4 corners. I went over it several times until I was satisfied it was smooth with no sharp edges or corners anywhere. Then it was ready to mount the components to.

Mount the components

Ground wires

Place a 1/8" pop rivet through the 2 ground wire terminals and into the 1/8" hole on the left. Arrange the wires so they are both down far enough to allow plenty of room for the triac.

This was the original placement:

Pop the rivet, and the wires should be firmly in place with no movement.

You could tap the hole and use a machine screw instead, if you don't have a pop rivet tool, but the Harbor Freight Hand Riveter Set is cheap ($4.99 sale price) and is a great addition to your tool collection.


Apply a generous quantity of heatsink compound to the back of the triac package.

Carefully position the triac so the holes line up.

Push down on it to start spreading the paste, and then tighten the screw firmly, but not too tight, or you could crack the plastic triac case. The paste will spread out very thinly, which is what you want--just enough to fill any gaps while maintaining as thin a layer as possible. I tapped the hole with the closest tap I had. It was off by about 1 TPI but I was able to get a good firm connection. You will have to determine the best tap to use from what you have on hand.

Put it all together

Finally, screw the heatsink in place and you are done. I had to replace the original sheet metal screws with ones that had smaller heads because the original heads were too large to fit between the fins.

Finished at last…

Here is the modified controller with the popcorn popper coffee roaster plugged in and ready to go. Even with the pot set at full power (1250W), the heatsink gets barely warm. Now I have much better control over the profile of the roast.



    • Clocks Contest

      Clocks Contest
    • Water Contest

      Water Contest
    • Creative Misuse Contest

      Creative Misuse Contest

    40 Discussions

    Excellent Instructable. After reading some of the comments, some folks want higher current capacities. I am reminded of my first electronics design classes where the instructor impressed upon my class the need to design in a 100% safety factor. Too often manufacturers design a circuit with components that just meet the circuit need and nothing more just to save a penny or two granted a good quality product does not generate repeat purchases due to product failure. When that is done you have a design that will last. I like the over sized heat sink. Since this circuit is fused at 15A my personal preference is to parallel two of the BTA26's giving greater total current capacity of 50A thus reducing the strain on an individual triac. I like the fact you re-soldered the wiring in the unit. One thing I like to do is use silver bearing solder to give better electrical conductivity also a little higher melting temp to counter the heat of the current draw of the 15A. Keep up the good work. ;)

    3 replies

    Thanks for your comments. I agree about 'just meeting' the design requirements, or in this case, falling short I suspect. That's a good idea, paralleling 2 triacs but still keeping within the 15A requirement. I hope others will read this. For my requirements this would be overkill though.

    You are correct it is definitely overkill in the end however your equipment is more reliable. The only other consideration is the price of the triac for the most part not a real budget breaker. The other is construction technique which is as important as the components used. Amazing that it takes no more time to be neat and accurate than it does to be sloppy and inaccurate.

    I don't know about your second point. It took me the better part of a day to do this mod right. Careful measurement, careful drilling and alignment, lots of filing, getting everything to fit back in tight quarters, all ads up to a lot of time.

    Very informative instructable; thanks for posting.

    Due to carelessness, I've created a problem for myself. While drilling the back plate, I accidentally broke two of the tabs on the BTA26-660B. Now I need to find a replacement (or buy a new unit from Harbor Freight). When I search BTA26-600B on the internet, a bunch of stuff comes up. But what's the A320? Do I have to find a part that also has the A320?

    13 replies

    The 320A is a mystery. I could not find it anywhere either. The spec sheet I posted does not show that suffix so I would assume it is irrelevant. I just bought a backup chip on eBay for $2.82:


    eBay search:


    Followed your link and the Triac is $1.79 as of today! Glanced at my back plate and the Triac is riveted in not screwed. Ordered a heat sink.

    Nice mod for me as I plan on using the unit to control the speed on a HVLP turbine compressor where duty cycles "might" be longer.

    That's a great price. Curious that the triac is riveted. That is usually discouraged in the data sheet because it can break the case due to uncontrolled pressure.

    I did order one just in case. I should open it up to see how it looks on the inside.

    How did you know how/if the four screws for the back plate would land between the fins on the heat sink? If I don't get lucky on the one I ordered, I'm able to machine it to whatever is needed. Aluminum is easy to machine.

    Read how I did it. I just barely got it right.

    Ah, I see now as I didn't read all the details closely. I notice now that I made a mistake on my heat sink as fins are going the wrong way so machining will be required for sure. I may cut it to fit the back plate pocket exactly now.

    Wrong way?! Which way? Down or vertically? LOL!

    Fins run parallel to the long side of the box! I was paying too much attention to the overall size and missed that. Oh well........

    The pic is a bit overexposed but the bad soldering point isn't that bad but look at the fuse, a sure current reducer there. I'll have to loosen the nut to check the other side.


    That definitely needs a touch-up.

    Ya think? The other side looks okay. The fuse holder is marked as 10A and the fuse is bit hard to read but looks like a 20A. One metal cap moves too, a real cheapo!

    Think of it more as a Speed Controller KIT. 'Some assembly required'. ;-) Still, for the money, it is quite a useful little unit.

    I got this heat sink from Amazon but didn't know that they had Amazon stores from China; http://www.amazon.com/gp/product/B00BGAO2W0?psc=1&... so it took 17 days to get here but now showing it "to be" sold by Amazon when in stock.

    Like I mentioned I forgot about the direction of the fins so I had to machine slots, not too bad with a stout drill press and the proper end mills. I taped a 8-32 hole for the ground wires and had to use a smaller 6-32 for the smaller mounting hole on the triac. I also machined the pocket that the old cover sat in which proved to be a PITA since the plastic was rubbery and harder to cut without a wreck.

    I packed the case with paper towels to keep the chips out. At one point the end mill grabbed the paper towels which caught the dangling wires. Luckily I stopped the drill press quickly with no damage and removed what paper toweling that was still left.

    I ran my HVLP turbine which draws 13A at just over the low/medium mark for five minutes and it barely got warm.

    Thanks Laral for idea! :-) Fuji Spray Systems now has a 5-stage turbine with a variable speed control but I wonder how they dissipate the heat as these turbines already generate lots of heat. I've burned my knuckles removing the hose at the unit right after using it!!

    Heat Sink_1.jpg

    Wow, you managed to make even more of a project of it than I did. :) But it looks really cool and it works. That's the main thing. Thank you for sharing.

    I'm not sure where you get the triac that size for $2, but, god bless you for finding one so cheap. I had to revert to the Chinese to get them at a bargain for $5 each, maybe $50 or $55 for a 10 lot, plus shipping, come to find out the chip wasn't even bad after using it to operate my Bob Boyce hydrogen system that didn't work as described because of the lack of credible schematics, instructions and parts. (I had to be very creative and utilize Stan Meyer's designs to come up with a more logical approach).

    To make a long story short, the triac I thought was bad. I ordered them, replaced it and the speed controller still did not work. I bought a $50 speed controller to find it worked differently, and that one blew too. I thought maybe it was the switch, I replaced it and it didn't work either. Then, replaced the trimmer. Still nothing. I finally gave up and bought a new one good for 240 volts and turned it way down.

    I found that some on the market only adjust from maybe 99 volts to 120, and nothing less. This one sent much less voltage when variable, literally 0 volts all the way up to 120 volt, but, at only a 9 amp rms capability before destroying the device. It stood up to 12 amps instantaneously, and, I found for a mere few seconds, and, 8 amps rms with getting so hot I couldn't handle it.

    I went and took the next one I bought apart and added heat sink with a fan to it and it could now handle maybe 10 amps rms without destruction, but, the variability is limited because there are other limiting factors about the design of this unit. The variable resistor is very weak and only capable of 25 to 50 watts if you are lucky, so, if the device is pushing the power back to the variable resistor, it's only a matter of time before it is destroyed as happened in mine. The triac went bad, but tested good, so, the switch popped, the triac was bad, and the variable resistor was also bad. 3 parts that basically make up the entire unit.

    They build them so they are hard to replace the parts (on the higher powered units), so, this 15 amp unit is a good choice just as long as you don't overdo it.

    This is an ideal unit to use for a variably powered hairpin system, which I use it to power the hairpin instead of a variac and I use the hairpin's output to push power into the high side of an ungrounded/ground isolated utility transformer to use the cold electricity to power my barn.

    I have found that the system is capable of driving several thousand watts of power, including sensitive components such as computer chargers, phone chargers, all of my lighting, flourescent and incandescent, and, machinery/motors which use inductive loads.

    I get a total of 12 times the input power which I am able to regulate by feeding the system through a killawatt meter. The input uses 250 watts, 70 to power the transformer, 180 to power the hairpin, and, the output doesn't affect the input levels at all. It is at maximum power the entire time, I just have to be extremely careful not to exceed the transformer's primary voltage, so, I have to use primary metering that we use for overhead line systems to detect the voltage levels (with an analog meter ) This has to be used because cold electricity is extremely variable for the voltage and amperage on the high side of the hairpin and transformer coils. However, once the transformer is stepped back down, the 120 side if easily detectable again with the killawatt meter.

    No, I am not anywhere near delusional and this works perfectly. I have come up with several instances that show it's full potential without affecting the input power used, just as long as I can feed my neon transformer with 180 watts, all is well.

    Cold electricity starts to react with devices to release it's potential when a high resistance is introduced to it. I found that this is completely opposite of conventional systems and it is highly recommended to generate the power needed to drive the system with a permanent magnet asynchronous generator that does not utilize a field excitiation coil using it's own power. This must rely on a magnet rotor to feed the system because it draws in an additional spike of voltage from the vacuum at a very large level which will literally place much of it's used charge back into the system to replenish what was used in it's original power up plus much more energy that shouldn't exist, but, for some reason it does. This is created by introducing the back emf pulse after the magnet passes the coil, a massive spike of energy is created from the field differences from the south and north magnet in the "elimination zone" where the energy manifests from the difference in potential between the high and low magnet excitation, as well as the phase differences which draw in additional energy in 3 ways. First is by induction, second, one spike per pass is from back emf, and, the third is from the vacuum, or vector potential from the out of phase introduction of the energizing and de-energizing coils (between them in the elimination zone) This zone will no longer eliminate the stray energies and cause harmonics or noise that used to be a nuisance to one's car stereo in the old days, but, since we are finding that it causes an overvoltage even under load since the system can not be grounded, we can introduce grounded noise isolation units to clean up the additional noises at the source of the systems powered up that need to be noise free.

    We could always transform this energy to isolate the power lines from the grounds we introduce to it, but, care is needed to be taken because the energy wants to escape to ground and it must be blanketed with rubber goods where it comes close to ground. By adding capacitors in series between the high voltage in and out, a simple Marx generator will be built and can introduce a deadly arc if the voltage is higher than the transformers can handle. Remember, high voltages can jump, and, since this mixes conventional and unconventional energies together, the output can be lethal. Please do this at your own risk, it is very dangerous. i do not recommend anybody who does not work with high powered lethal systems such as myself on a daily basis to even touch these devices unless you know exactly what you are doing and the dangers behind it.

    All personal protective rubber goods need to be used, also, remember, high voltage utility transformers being ungrounded will produce a static energy from their capacitance on the outside of the can that can't be touched or a deadly result may occur.

    1 reply

    It is not recommended to exceed the specs of the unmodified unit. There are too many factors that could cause components on the board to fail, many of which you mentioned. The heat sink simply allows the unit to run more safely in its intended range.