Envelope penetration detection 4

[joaomartins]

Hi,

I have one PCB with electronics that I want to protect from ouside world probing! I use on Flat for the interface.

I am considering the use of a sheet to enclose the PCB.
It would be great to measure the tin sheet and if some hole is made on it this resistance measurent should be different in order to trigger some action.

Any Idea on how to measure correctly this!?

Please note that this must be working on all circustances so probably whatever design idea is brought it must work properly powered by a lithium battery.

Thanks 2 All

 

[nbucska]

There are many microprocessors which have internal, unaccessible RAM. If part of the circuit is on FPGA,
someone may copy it but without understanding, couldn't manufacture and maintain it ...

<nbucska@pcperipherals.com>

 

[lcsjk]

I like the epoxy as a slow down deterrent, but I think you should use a box inside a box approach where the second box must be in place before data is stored. Then, if the outside box (with epoxy) is disturbed or removed, the destruct circuit will fire, destroying the data.
Some small wiring embedded in the epoxy would also make entry by epoxy removal difficult.
A magnet and hall sensor are one approach, with circuit that senses the hall current or other and etc.

 

[Wedwin]

How about randomly chop up chunks of encrypted data and store them in different physical RAM's?

It would add to the complexity of the PCB.

It would slow down puzzling together the shreds of data to something interpretable.

If a pro really wants the data, he will stop at nothing.

And if someone wanted to get the data for the sport of it, he won't stop at anything either ...

 

[joaomartins]

&quot;How about randomly chop up chunks of encrypted data and store them in different physical RAM's?
&quot;
I believe that the uP does this! By encrypting the addresses you do not know where is the vector table for instance.

&quot;but I think you should use a box inside a box approach where the second box must be in place before data is stored.&quot;
why? I haven´t understand yet what is best! Place the EPOXY covering the components or covering another internal box where the secBoard is?!

 

[melone]

Why is this worth all of this trouble?

 

[Heydave]

No single scheme will provide the protection you are looking for. FET probes can be made microscopic and guided to the desired circuits by Xray with extremely fine drills. As for raw die, I have personally probed working raw die and have even modified them by cutting metal traces. I do not doubt that some partical beam could be employed to probe a working circuit as well. What might be one of the best schemes is a soldered metal cover where alternating conductive and non conductive layers are deposited on the cover after the soldered cover is attached. Virtually any effort to probe the inside would result in a short and/or change in capacitance erasing the data. The test circuit would also monitor the capacitance of the cover coating from inside the cover. All the traces to and from inside the cover would be inner traces of the PCB allowing a complete seal all around. Similar monitoring of traces in the PCB would check for any effort to delaminate the PCB. A very low power CMOS processor with a lithium battery would be running the testing. Using a few resistors it could constantly test the layers from the time the unit is assembled. Any time the parameters went out of a predetermined range, the SRAM power would be cut until the parameters were back in range. The AC signals from the micro, used to test the capacitance/resistance, would vary with time according to a randomization scheme in the micro. Such algorithms could easily be made to not repete for many decades making it difficult to syncronize some system to defete it. Cost of this scheme would be fairly low as such microcontrollers are cheap. How long that would delay determined engineers is still a gamble. Many would simply see it as a challenge.

 

[joaomartins]

Hi Heydave,

I have two questions:
1) The C is not temperature and pressure dependent?
2) &quot;The AC signals from the micro, used to test the capacitance/resistance, would vary with time according to a randomization...&quot; why is random necessary?

 

[Heydave]

The C is variable but we are not talking precision, just a range. The value of the capacitance is small making it difficult the physically switch out and switch in a dummy. Multiple contact points to the detection coating can make the switching difficult and allow for an added continuity test.
A switched state from an output pin through a resistor to an input pin and the capacitance being tested will produce time delay to reach a given level. This would be capable of detecting a short or change in capacitance by having a time window around this delay. Using a randomized PWM scheme can complicate this measurement but make it difficult to defete.
A device driven by a fet probe could record the waveform virtually without detection by the intrusion detection circuit. Later, this device would force the waveform during removal of the barrier. Randomization of the test signal defetes this intrusion scheme. A possible scheme to defete this could include a focused radiation on the control circuits for the power switch making it permanently active. Perhaps this could be bolstered by redundant power switches that cycle between each other. If a switch does not quit, then a crowbar circuit steps in......................
It does not matter what scheme you come up with, it can be broken given enough effort. The point is to make that effort unreasonable without making the protection unreasonable. The more diciplines needed to defete a protection, the more unlikely it will be that such skills can be assembled.