No schematic, but here is what I am experiencing:

I built a simple PWM control to replace the battery on a portable saw:
It's rather simple, having a 120v power cord>fuse>switch>large-transformer>diode-bridge>electrolytic-filter-capacitors>pwm circuit>DC power saw.

My former battery-operated circular saw now resides upside-down, doing service in my shop as a mini-table saw. It is plugged into the same circuit as my shop computer tower. Sometimes, I notice that my computer kicks off after using the saw, but the circuit-breaker doesn't trip. At first, I thought I had a bad neutral, as once I lost power completely. I was able to hunt down a fiddly wire nut, and I thought I had found the trouble, but the situation is still occurring. I suspect that I am getting some kind of back emf/emp from my DIY table saw that whacks my computer/power-supply.

Does this seem a worth-while direction to go, or should I look for something else?

 

You just need to look into how much Electrical-Noise you are broadcasting.
The wiring between your PWM and Saw-Motor, must be as SHORT as possible.
Also be sure to tightly twist your Power-Wires to lessen their propensity for broadcasting Noise.

Brushed-Motors automatically generate RF-Hash, (RFI), from the sparking of the Brushes,
this Hash can easily exceed twice the Voltage that you are driving the Motor with,
and smoke components in your PWM Circuitry.

Small Ferrite-Beads slipped over the Power Wires to the Motor "may" absorb most of it.
Use lots of them, more is better.
They are super Cheap and come in a huge variety of sizes.

Also, various different sizes of Ceramic Capacitors installed between both
Motor-Terminals and the Metal-Housing of the Motor.
6-Capacitors, 2-100nf, 2-10nf, 2-1nf.
Multiples are needed because Capacitors do not have a
"flat" Frequency-Response at extremely high Frequencies,
and you probably don't have a high-end Oscilloscope that can "see" this Hash.

This RFI Hash is "not necessarily" created by your PWM Circuitry,
but the 2 sources of Noise could be interacting, and/or, creating severe Oscillations.

You need to stop the RFI from "escaping-out-of" the Motor-Case
and into other circuitry where it can be broadcasted out into the room.

You also need to make sure that your PWM Circuitry is not getting into
any unwanted Oscillations.
Inadequate MOSFET Drive Current will quite often cause severe Oscillations,
even up to the Ghz Range.
This is a classic problem when trying to drive a Big-MOSFET with a Micro-Controller.

 

Thank you for that excellent rundown of RFI, it's causes and solutions.

Can it leak into the mains with enough current to knock-out, or reset
computers located along the same mains circuit? (I'm thinking of a situation
where perhaps tens or hundreds of amps, coming or going, disrupts the mains voltage
or current supply.)

 

This is not a problem with AC-Power delivery.
This is a problem with super "High-Frequency-Noise" being transmitted through the Air.

This noise "could possibly" be directly connected via the Mains Wiring,
but it's extremely unlikely.

It is only slightly more likely that this Noise is using the Mains Wiring as an "Antenna",
but generally, Frequencies this high are filtered-out of the Mains Wiring,
either on purpose, or by default.

If I were to hazard a guess,
I would say that the Wiring to one of your Computer-Peripherals is acting
as a receiving Antenna,
and is injecting this RFI "Hash" directly into your Mother-Board,
and that,
installing Ferrite-Beads and Ceramic-Capacitors on your Motor-Leads
will completely eliminate the problem.

But of course, Your PWM Circuit could be part of the problem if You
have severe "Ringing" coming from Your Switching-Device(s).
The Motor doesn't care much about Noise from the Power-Supply,
so it may be performing just fine,
even with heavy Ringing / Trash / Noise possibly being generated.
It's good practice to "simulate" your PWM Circuitry in Software to
(hopefully) reveal any "Ringing" problems with the design.