Avoid copy Eprom.

No. If the CPU can read from it, by manipulating its buses, then so can anything else.

You could, though, scramble the address and data lines, so that the PCB wiring automatically corrects things. This won't work if the potential copier has access to the circuit board.

You want to use a more modern secure microcontroler, which avoids having open buses between the CPU and EPROM.

Richard [in PE12]

I don't see how. If you prevented anyone from reading the data, you'd also prevent your processor from using the chip.

You could possibly try encryption or other some other obfusciation method but that wouldn't stop a serious attacker. You might be better off using a uC with onboard flash and enabling the code protection. That could also be bypassed but it takes alot more effort.

How valuble is your data?

Al

I have made some changes to a program for a Z80 processor, and I sell a machine with this modification.

There are some person that read my Eprom and sell the same machine with my modification.

So the only thing is to avoid the reading.

How hard is this modification to reverse engineer?

There is no point in stopping someone from stealing your copyright, if they can just implement the same change themself in a few days.

tim

That's been my thought for a while. There does seem to be an inordinate amount of paranoia regarding protecting code that's hardly worth protecting, since it can be reimplemented with less effort than copying. I know, I know, there are exceptions.

That said, there are some people in this world (the other side of the world from me) that will spend more effort copying than reimplementing - and that may be the point. Make it very difficult to copy, and your knockoffs will be later and later to market.

And while you're at it, post full schematics with some errors inserted, and post full source with some errors included, and the knockoffs may never make it to market.

I have seen this done before (not with electronics or code), and the results were hilarious.

... snip ...

Electrolytic capacitors in the recent past, and the results were not exactly hilarious.

--
Some useful references about C:
  
  
  
   (C99)
             (C-library}
                    (GNU docs)

Yes. If you break the pins off the package, it will be much more difficult for someone to copy it.

Are you implying that the party(ies) who inserted bogus specs into whatever documents were involved have some culpability? I don't know any specfics about this episode.

That seems to be a general principle in academia and the arts -- the people with the most fear of having their ideas stolen are those who don't have a lot of really good ideas.

*chuckle*

Can you remove the Z80 and perhaps replace it with one with on-board secure flash, and put the program in that, instead of in EPROM.

You could try using potting compound to prevent either the processor or EPROM being removed, and/or things attached to its pins.

Richard [in PE12]

>

The only way to protect it would be to add some other hard-to-copy hardware like a PLD or microcontroller to the system - either to some available port, or by putting the eprom on a small PCB with additional hardware that the hardware uses for authentication. This is common practice in the automobile ECU re-tuning industry.

Another trick you can do is include some random data in the eprom, which has no purpose. That way you can easily prove if someone has copied your device. If the product has a suitable user interface, a good trick is to have a secret 'hidden' function, e.g. a combination of keys, that displays your name etc., from text that is encrypted in your eprom so it's not obvious to the casual observer looking at the contents. This makes it even easier to prove an anauthorised copy.

In article , Mike Harrison writes: |> The only way to protect it would be to add some other hard-to-copy |> hardware like a PLD or microcontroller to the system - either to some |> available port, or by putting the eprom on a small PCB with additional |> hardware that the hardware uses for authentication.

Why copy the dongle if you can remove the dongle routines?

That would require, of course, some significant amount of reverse engineering (depending on the dongle quality), but still it doesn't keep people from copying if they really want to.

Rainer

Nothing is 100% secure. You just need to decide how much effort a potential copier is able/willing to put in. Adding a properly-implemented dongle solution increases the skill level required to copy by a substantial amount. In some cases it is also possible to hide important algorithms, keys etc. inside a dongle to make it very hard to remove..

LOL! The only answer so far that has gotten close to answering the OP's question directly. No amount of encryption or obfuscation can protect an Eprom from being copied, as every bit of every byte is accessible. Just as photos cannot be copy protected as anyone can run them through a scanner etc. But I think that the OP should really of asked "how do I make it harder to for someone to make a working copy of my system" or suchlike.

The answer lies in adding an extra bit of hardware and having the program verify it. What comes to mind is a suitably programmed secure micro in an small package that can easily be added to the system, perhaps through a single I/O pin. It could even be powered from the same pin! The program needs to verify that the device is attached and that it's response meets the challenge given.

So yes, the Eprom can still be copied but it won't work unless they crack the secure device or the Eprom code and overcome the basic security measure. But that's work they mightn't be up to hopefully.

my2cents

*Peter*

It's not that hard to disassemble the EPROM code to see how it accesses the "secure device" and see what it expects back from it. It would be better to encrypt everything in the EPROM and have the "secure device" hold the key for decryption. But presumably all code runs from the EPROM, the boot code will have to be 'clear' and decrypt other areas (or the whole EPROM) into RAM and then execute the code segment from RAM. This becomes a significant effort for only a mild amount of security. A determined individual or group WILL disassemble all the code if they can get to the object code.

It would be much better to have a processor that has its code memory completely internal on the chip, and has a 'security' bit that keeps the internal memory from being (easily) read. This was already posted:

Are there Z80-compatible microcontrollers? I've heard of the Rabbit, with a superset of the Z-80, but I'm not really familiar with it.

-----

To disassemble an Eprom requires knowledge, time, and effort. Using a secure device just makes it that much harder then COPY,BURN. You or I plus many others would be able to hack the system, but then again, why would we when we can just design a new system anyway. This is only a basic security mechanism and any determined hacker will eventually overcome it no matter how secure it is made.

The OP is working with an existing design so I guess he doesn't have an option of using some other cpu with on-board flash. Anyway, he may be more of a software guy and not all that keen to delve into the hw.

*Peter*

Ben Bradley wrote:

Try Zilog?

They have Flash models ( but not likely to plug into an existing Z80 design, unless you make an adaptor ) - maybe Zilog could do a eZ80_BGA => StdZ80 module ?

IIRC std Z80's do not have Adr latch, but another scheme to give medium security is to replace the Latch with a PLD, that scrambles the adr lines, and adds some state logic.

-jg

The machine is build by a company, I only have made the change to Eprom, and I sell this Eprom to the person that have this kind of machine.

This machine have 3x8 DIP swicth, I can make a function to validate a combination of an unused swicht. This can stop the machine if I want.

Thanks, S.L.