I’d say one thing that will help you in the long run is to stop using this term ‘throwing money at the problem’.
You’ve got 3 options here. One is tons of design, testing and verification work involving probably expensive test equipment to find the cheapest solution. One is to put what may not be the cheapest solution. One is to do nothing and live with having boards come back or the e-mails from people asking for help.
It’s not about the cheapest board or the lowest eBOM or the fewest parts count. It’s about overall price vs performance. No sense in saving yourself 10c/board and spending an hour e-mailing every 50th person who happened to go full Palpatine on the board because they were wearing polyester underwear while they were using their CNC or whatever. How much per-board is worth not sending out replacements? Etc. etc. etc.
Anyway, on a less preachy note.
Definitely wait to see what is actually dead and how bad it is. That’ll give a rough idea of next steps. Under the assumption that it is ESD on the probe and not something else, I suspect you’d see dead diode arrays on the input (either completely open circuit or short circuit) and/or a dead/unresponsive micro input.
My personal approach would be that if the probe circuit is dying to ESD then all the inputs probably need beefing up, but that’s likely overly conservative for what you’re after. If it is just the probe then are you ok with having a separate header for the probe or a specific input that the probe must be on? If so then we can add ESD protection to that one specifically.
Adding something to the probe could work. ESD straps build a 1MR resistor into the line to limit peak currents to prevent ESD damage. Adding that to the probe could work, but you’d want the ‘unprotected’ side insulated enough that it couldn’t take a direct strike. Alternatively you could have the probe connected through an adapter board that had some form of ESD protection and added impedance to both sides of the circuit.
Ideally, you’d probably want the earth strap for the probe to either be permanently connected to the router or to have its own limiting , under the assumption that a potential source of ESD is connecting the ground clip. That may be mitigatable by ensuring that the circuit itself has at least some impedance to the installation earth, even if it’s high impedance.
My logic here is that you’ve got 2 open connections to your circuit that are being handled by a charged human and then being connected to something that could be charged or that your circuit may be getting charged via the long lines running everywhere and then be discharging into an earthed/uncharged person or machine. ESD strikes via people tend to be higher voltage but less damaging. ESD strikes via machines/other circuits tend to be lower voltages but higher energy so can be more damaging. An ESD strike into the ground clip won’t be as likely to hurt things, but it could still cause damage. It will depend on how the rest of the circuit is wired, whether the power supply 0V is earthed, what the inductances/impedances everywhere are, what the impedances to other connections are, etc. A big pulse of current down the ground strap and out to installation earth can cause the 0V reference (GND) on the Jackpot to change voltage briefly vs other things its connected to which may be enough to hurt stuff. An ESD strike into the probe itself will take one of two forms depending on polarity. It will either be a high-impedance strike to the diode array which will cause a high voltage to appear on the input, potentially tracking to other adjacent inputs, or it’ll be a low-impedance strike through the diode array, potentially damaging it. All of those situations can also occur as no damage, partial damage or catastrophic damage, depending on the scale of the strike. Even repeated small strikes may cause long term drift in specifications at best or outright damage at worst.
