Grinder Minder

Bench Grinders are everywhere. So are fingers. May the two never meet. That's where grinder minder comes in.

Similar projects worth following
After spending the last ten years running maker spaces and educational programs for K-12 students, we are keenly aware of the everyday shop risks that makers everywhere face. We all have stories to tell about shop injuries and "almosts" but it's not fair to put young students at risk and most schools don't have faculty with the breadth of experience to keep everything above board. This is the perfect opportunity for technology to take up the slack!

We started by surveying what tools are high risk and pitting that against how common a tool is in a shop. We arrived at the humble bench grinder. This ubiquitous and seemingly gentle tool that can be had at Harbor Freight for just $35. Little did we know it claims thousands of fingers every year and even causes death when the wheels shatter. Not OK!

We have set out to create a small device that provides DC Injection motor braking, access control, emergency stop, and accidental restart protection, all while just bein

Project Goal: Develop a product, in collaboration with potential customers, that can protect bench grinder operators from the most common hazards, installs easily, and is priced for "Home Depot" range consumer market.

Technical Goals: Our product will provide the following features and functions:

  • DC Injection Braking - to stop the tool quickly after each use thus preventing injuries associated with coasting. This requires the supply of up to 5kW of DC power delivered for up to 4 seconds.
  • Accidental Restart Protection - to prevent the startup of the tool in the event of power loss or accidentally leaving the power switch in the 'on' position.
  • Emergency Stop - an ANSI-compliant emergency stop button to activate braking in the event of an emergency.
  • Plug-and-Play - designed so installation requires nothing more than plugging it in in-line with the existing bench grinder power cord.


  • Size - the physical device must install unobtrusively onto a bench grinder. Initially targeting less than 40 cubic inches.
  • Cost - we're targeting a BOM cost under $30 with an all-in cost to manufacture of $50 per unit at a quantity of 1,000.

The Team

But who are the grinder minders? Let me introduce you!

  • Scott Swaaley - Educator, Engineer, Maker, Entrepreneur
  • Scott McGimpsey - Engineer, Maker, Entrepreneur
  • Phillip - Maker, Scientist, Computer Engineering Undergrad (and previously one of Scott's students).
  • Harrison - Maker, Adventurer, and College-Bound Engineer (and previously one of Scott's students).


This is two of two arduino sketches that interface with the MCP39F511A eval board and the MegunoLink interface.

ino - 5.07 kB - 05/13/2019 at 16:11



This is one of two arduino sketches that interface with the MCP39F511A eval board and the MegunoLink interface.

ino - 13.26 kB - 05/13/2019 at 16:11


Power Monitor.mlpz

This is the Megunolink project that we use to run our MCP39F511A power monitor.

mlpz - 30.20 kB - 05/13/2019 at 16:10


  • Wait...who are we doing this for again?

    scott.mcgimpsey19 hours ago 0 comments

    If there's one thing we like almost as much as making, building and braking, it's sharing what we're doing with others. For the most part, people like to hear about something different from their day to day trade, and listen in. But as anyone in this community knows, it's always more fun to share a project with someone who is -REALLY- into what you're doing. Not for the focus and interest (though that's always awesome) but more for the insight they can provide. And lately, we've heard the same message from a few of these folks. First, what we're doing is very awesome -- it seems a lot of people in the industry of heavy power tools are highly aware of the dangers, and know a friend who was injured when safety measures weren't observed, or when they weren't enough. The second message these folks shared was that our target market is off base. According to these folks, the biggest need for tool safety is not in the at-home market, but instead in the bigger, more powerful tools which require 3-phase power.

    When we started out, we made assumptions about our customer and target market. Now we're seeing real data which contradicts our assumption, and we have a decision to make: Do we trust our starting impression of the market, or do we jump in on some customer research? Given our timetable and our goals (and the spirit of the product development competition we're in) we think customer research is going lead to the best outcome. So what's next? We're going to sit down with some more people and try to answer the question "Who would see the most value in a tool braking solution?". Updates to come...

  • Two steps forward, one step burnt.

    scott.mcgimpsey06/02/2019 at 04:45 0 comments

    Today was all focused on testing out the braking effects of a pulsed rectified sine wave on a motor. But to do such a test, we needed a circuit which would be able to switch fully rectified mains at a frequency several times higher than 60Hz. Our first circuit was composed of a power MOSFET for switching the power on and off, a bridge rectifier to give us a full wave rectification of AC power, an Arduino to give us a cheap and easily configured square wave, and an optocoupler to protect the Arduino from accidents. We used a bar-style 50 ohm resistor for the load to just start testing, to be replaced with the grinder once the circuit was doing the basics correctly.

    We configured the Arduino to produce a roughly 500Hz square wave with 50% duty cycle, and kicked on the power. Our first test went off without issue, producing a very nicely chopped up sine wave, and a little heat from the resistor.

    We wanted our later motor tests to have a wider envelope in terms of tested frequencies (upwards of 5kHz), so we configured the arduino to send out a faster signal. At this point, we just started seeing the fully rectified wave from the rectifier. We tested to ensure the optocoupler hadn't failed closed, and found no fault. Then we double checked it's timing limitations. Unfortunately, the time to turn on was ~10μs, while the time to turn off was ~50μs, which gives us a frequency limitation of 1.6kHz. So, we thought hard about the actual need of the optocoupler, and decided to do without it.

    Having bypassed the optocoupler, we ran the test again, and sure enough, we landed a very nice higher frequency chopping of the sine wave.

    With our basic test completed, it was time to hook up the Grinder and see what happened.

    It actually worked quite nicely, altogether. Sure, we had a lot of inductive kickback whenever we turned the MOSFET back off, and the flywheel diode we had got a little toasty from dropping that current on it's return trip, but altogether, it was at step in the right direction. This is where we started having a few issues.

    To start, our flywheel diode broke down from heat on the second test. We were treated to some arcing and charred breadboard before we shut it back down. We found that the diode, the mosfet, and the Arduino were shot (Blew the cap right off the FTDI chip). We replaced the smoked parts, went to a larger protection diode, added a optocoupler with a faster response time, and directly soldered the connections to the pins of the MOSFET.

    Again ready to proceed, we kicked on the power, and immediately blew the MOSFET and charred the heck out of the pulldown resistor on the MOSFET gate. Whoops.

  • Motor Characterization

    Scott Swaaley05/11/2019 at 19:26 0 comments

    Today we spent some time characterizing the motor for a mid-size grinder. Specifically, a 5/8 HP DEWALT DW756 6-Inch Bench Grinder. I've been working with the eval board for the Microchip MCP39F511A Energy Monitor IC and finally tweaked my code enough to run it via an Arduino at an overall read-rate of about 50hz. I then use MegunoLink (which I highly recommend) for basic GUI and visualization. This is the test setup:

    I was able to plot the startup and run characteristics of the motor and shared the plots below. Of note is the impact of mechanical load on the electrical characteristics. That's just one of the many things we need to characterize. Pretty cool plots. One thing doesn't look right though - can anybody spot it?

  • Specifying a DC Supply

    Scott Swaaley05/10/2019 at 19:27 0 comments

    Now that you understand the DC power requirements from our previous post, it's time to get into the details. Let's start by looking at the typical specs for an  AC/DC supply (listed below). Now, we get to start crossing off ones we don't care about.

    • Input Voltage (AC): 120VAC
    • Max Output Voltage (DC): 90V
    • Output Power: 5kW
    • Output Current
    • Efficiency
    • Line Regulation
    • Load Regulation
    • Ripple & Noise

    So basically, we don't give two shits about most of the qualities that make a DC supply a "good" supply.  This is a lucky break because finding affordable output caps in the milliFarad range rated for 100V is next to impossible. We're just pumping juice into a huge inductor (the motor) so we get to do it dirty. So if we remove all those unnecessary features of a DC supply ... what's left? In the simplest form, you have a half-wave rectified AC signal like the one shown below with an average "DC" voltage of 54V, a rms voltage of 85V, and a ripple voltage of 170V. That's pretty darn close to the 90V output we need. Let's try it.

    So something weird happened. I did the first test with a phase-chopped waveform to limit voltage (more later on this). Let's start with looking at the output voltage waveform of our half-wave rectifier.

    The strange thing is that when the output voltage should be zero (between positive ac waveforms), it goes negative as the magicnetics of the motor feed back into the supply. While this is happening, the motor continues to run at full speed (though sounding funny). Definitely not an effective braking means, even though it is "DC". What I think is happening is that the inductance of the motor wants the current to continue (as inductors like to do) so it creates a reverse voltage (the initial spike). This initial spike then discharges over the half cycle until it gets voltage again on the subsequent half wave. Without an actively-supplied current to force that missing half-wave into positive (and thus brake the motor) - it doesn't work. Lesson learned. You might be thinking that we should just use smoothing capacitors. We did the math and simulations and the capacitance (and voltage tolerances) would be HUGE to source energy for the 8ms between half-wave pulses!

    Next up is trying full wave rectification. The numbers for full-wave are a little scary, with an rms voltage of (not surprisingly) 120VDC. Added to that is the fact that when the motor sees "DC", it's overall impedance is MUCH lower and it draws much more current. We want to stop the motor slowly, not rip the shaft in half. We have some ideas to deal with this - more soon.

  • Our Subtle Ambition - Big Power, Little Footprint

    Scott Swaaley05/10/2019 at 18:25 0 comments

    We've talked about how DC Injection will play a large role in this product but we haven't yet talked about how we'll actually make that DC. We're all used to thinking about a DC Power Supply as a trivial purchase or design, but this is something altogether different. The driving factor that makes this different is our gargantuan power requirements. We're designing our product to work with grinders motors up to 1HP so let's start by looking at the specs for a 1HP induction motor at 120V (based on NEMA MG-1):

    • Full Load Current: 11 A (1.3 kW)
    • Locked Rotor Current: up to 80 Amps (9.6 kW)

    When a motor starts, the initial current hits locked-rotor-current for an instant then settles into a startup current roughly four times it's full load current. That means a 1HP grinder coming up to synchronous speed will draw 44A (5.2 kW) for a few seconds. Just as that 5.2 kW decelerated the motor from stopped to synchronous speed, we'll need roughly the same amount of power to decelerate it from synchronous speed to stopped. That's a ton of juice. For some perspective, a commercially available 5kW DC power supply costs nearly a thousand dollars and measures 17"x7"x6" and weighs over 30 lbs. Our entire product (of which DC supply is only a part) should measure less than 5"x5"x2", weight less than 1 lb, and cost less than $100.

    And we haven't even begun to talk about DC output voltage. Most commercially available supplies max out around 24V but we need up to 90V DC to source enough power into the relatively high impedance windings of an induction motor. Wowzers. Now you see why this project is so ambitious. But don't worry, we have some tricks up our sleeve. More soon!

  • Ouch - Those Are My Nurtles!

    Scott Swaaley04/26/2019 at 08:00 0 comments

    As we are getting ready to move into prototype mode, we've identified another set of requirements that we need to keep in mind as we move forward. Nationally Recognized Test Labs (NRTLs, sometimes pronounced "Nurtles") have explicit requirements for product safety and we'll need to find what category we fit in. The fact that we are an AC mains powered device makes NRTLs especially restrictive in order to limit the potential for electrocution or fire.

    At first glance, it looks like we fit best under UL 508 - Motor Controllers. Unfortunately, it seems that the standard we need to meet is nearly a thousand dollars to purchase and hundreds of pages, not to mention the myriad of other standards it references. Ugh.

    Excerpts I've found online list lots of conductor spacing and thermal testing requirements. We'll also want to make use of UL recognized components whenever we can. More soon!

  • Friction is so 19th Century ...

    Scott Swaaley04/20/2019 at 20:45 0 comments

    Today we want to share one of the core ideas of this project and how we believe an optimal solution would work. As we shared in our first post, the continuous spinning of a grinder after the power is turned off poses a safety problem. Fundamentally, when the power is off, the device should be safe to those around it. Our goal then, is to stop the motor quickly and safely after the power is removed.

    When we searched around for how others are stopping their shop tools, we mostly came back with mechanical solutions -- some industrious souls even weld bicycle brakes to their devices. Fabricating and welding custom-made hardware to a tool isn't exactly an accessible solution for most users. Even if it was, the maintenance and impact to the surface on the machine where the brake pad comes into contact with a tool are all elements of concern. With this, we began thinking about the magnetic forces within the motor, and how we could manipulate them to get our desired result.

    Before I dive into the diagrams and theory, if you’re not familiar with induction motors and how they operate, I highly recommend reading Hackaday’s article on the history of how the induction motor came to its modern incarnation and/or watching this video by Learn Engineering:

    Read on below ...

    Read more »

  • Pain Points & Pain Data

    Scott Swaaley04/18/2019 at 07:38 0 comments

    In the product development and design thinking world, people often reference "pain points" when referring to a user need that is sufficiently "painful" to warrant them buying something to alleviate it. With #Grinder Minder, our investigation of pain points is a tad more nuanced. It was put best by Socrates in 381 BCE.

    "In our case it all boils down to one question - how dangerous is a bench grinder and how much do people actually care?" ~Socrates, in his treatise on pre-industrial power tool safety.

    Today's project log is about the first part of this timeless question - how dangerous is a bench grinder - I mean, really!? The first step in that research was learning what not do in this kind of research. For example, don't turn off image safe search when searching "bench grinder injury". Just don't.  It's gross.

    Based on our informal conversations with hobbyists, students, and professionals alike, people have WIDELY varying views. Some folks are scared to death of the spinning death trap that screams at them for two minutes every time they grind their tungsten for another tig weld. Others yarn on about the inherent risks of using power tools and believe that "anyone stupid enough to injure themselves on a grinder shouldn't be using one".  Neither of those perspectives are particularly helpful so in advance of our more formal customer research, we spent some time looking for some data. Here are our initial findings:

    • On the consumer side, the National Electronic Injury Surveillance System (NEISS) reports over 2,800 bench grinder injuries in the last five years, with over 200 being injuries to minors.
    • The U.S. Bureau of Labor Statistics Injuries, Illnesses, and Fatalities (IIF) Database reports over 4,900 occupational grinder injuries and 3 deaths over the last five years.
    • And both of these quantities are just reported figures and don't include the myriad of small injuries that never get reported.

    Ok, now go do that image search and multiply those images times the numbers you see above. Yikes!

  • First log entry!

    Scott Swaaley04/18/2019 at 07:18 0 comments

    We are extremely excited to start in on this competition alongside so many other talented designers/engineers/thinkers/makers. Here’s wishing everyone a fun, educational, and innovative experience!

    In this log, we’re hoping to share and document our journey from identifying the pain-point to customer research to design considerations to how we’ll approach manufacturing. And to kick it all off, we thought we’d share what problem we’re trying to solve, and the kind of customer we’re designing for.

    Read more »

View all 9 project logs

Enjoy this project?



Scott Swaaley wrote 05/10/2019 at 19:35 point

Hi Mike! We're still finalizing the feature set (trying to keep costs low so consumers can afford it) but yes - access control is high on the list. We'll already need a relay in there so might as well add a key-switch in series with the relay coil. In my previous life running shops, fancy fob/rfid/card systems never worked well as the back-end was never maintained.

P.S. I need to mess with my notifications settings - I never saw this comment until just now. Sorry!

  Are you sure? yes | no

Mike Szczys wrote 04/22/2019 at 16:22 point

This sounds great! Are you focusing on a lockout-type system that ensures only users trained on this tool are able to turn it on, or is there a more preventative technology like sawstop?

  Are you sure? yes | no

Similar Projects

Does this project spark your interest?

Become a member to follow this project and never miss any updates