1
\$\begingroup\$

I have an overmolded PCBA on a cable with passive components only. Some of the components are hard to identify, but I'm still concerned about the possibility of competitors trying to reverse engineer it. What are the best materials I can put on the components to make it extremely hard (expensive, time-consuming, etc.) for anyone to get to those components in-tact? I'm concerned that standard potting materials, conformal coating materials, and epoxies would be too easy to remove with heat, chemicals, sanding, or milling. Essentially I'd like to have something that bonds so strongly to the components that trying to remove the material also destroys the components. Of course if the material is too hard on the hardness scale, or has a high thermal expansion co-efficient, then the board components might get damaged during manufacturing or during use in the field.

\$\endgroup\$
1
  • \$\begingroup\$ This topic is too broad, there are a few things that could work, but it would be difficult for an online community to find anything because your requirements are also broad. I'd contact an epoxy supplier and see if they have something. \$\endgroup\$
    – Voltage Spike
    Jun 10 '16 at 18:45
4
\$\begingroup\$

One version of the history of modern electrical engineering states that it arose out of the need to systematically study circuits without disassembly, because it had anti-tamper measures inside (explosives) or you didn't want to have evidence of tampering (all during war time). This is were electric network and circuitry theory arises from.

If as you say there are nothing but passives inside your circuit then there is no need to disassemble it to determine what is happening inside. Thus there is no need for tamper prevention.

\$\endgroup\$
2
  • \$\begingroup\$ While I'm not an expert in electrical engineering history, that explanation sounds strange to me. Do you have a source for that? \$\endgroup\$
    – user39382
    Jun 9 '16 at 17:26
  • 1
    \$\begingroup\$ Yeah, I think the first paragraph has no basis, but the second has a point - if it's just passives, then it can be completely characterised by applying signals and measuring the response. \$\endgroup\$
    – pjc50
    Jun 9 '16 at 20:19
1
\$\begingroup\$

Well, similar to cryptography, with "infinite" time (& resources) someone will always find a way to reverse engineer what you are building. So the idea is to get as close to "practically infinite" as possible while still balancing cost/time etc. on your end.

I'll make a big assumption that the additional cost and engineering involved in securing this is worth it if the attacker can't reverse engineer your product.

I would use some type of mix of epoxy and an aircraft-quality adhesive on components. This way if they start removing epoxy or potting they will start ripping apart the PCB also.

\$\endgroup\$
2
  • 1
    \$\begingroup\$ Marine epoxy, with fine sand as a filler material. It's a sadistic thing to do to cutting tools though. \$\endgroup\$ Jun 9 '16 at 16:55
  • \$\begingroup\$ Nice @BrianDrummond, you have a twisted mind. I love it. \$\endgroup\$
    – Jim
    Jun 9 '16 at 20:18

Not the answer you're looking for? Browse other questions tagged or ask your own question.