dummy parts

Den tirsdag den 27. marts 2018 kl. 02.27.15 UTC+2 skrev M Philbrook:

Back in the early days of transistor radios I recall hearing of some overseas manufacturer(s) claiming that their portable radio was a seven (or nine or whatever) transistor radio when the circuit actually only used 5 active transistors. I can't find any evidence of this practice at this time though.

It may be that these dummies are for similar purposes - so the circuit looks more complex than it actually is. This bit of misdirection could be aimed at either the consumer or someone who wants to copy the design...

John :-#)#

The other day I came across a site that were selling dummy SMT components..

Diodes, SOT23 etc...

I thouht that was a great idea for confusing those that think they have a fully implemented board and just got some bad components! ;)

Jamie

Some parts are apparently for solder training, or test runs on pick-and-place lines. I guess someone might want to do a trial run of soldering a $15,000 FPGA or something.

I actually have one board now that I'm obfuscating a bit.

Intentionally?

But real parts can't easily be used as part of a circuit without having some effect.

Cheers

Some parts can in some places. I'm not convinced how much obfuscation they'd get you though.

NT

In the 486/pre-Pentium era, some motherboards were proudly distributed with fake DIL cache chips.

The pressures of marketing :-|

Hey, this discussion just gave me a GREAT idea. So, you add a component that has a readily visible part number, maybe a 74HC00 or something. The firmware does something that exercises this part but DEPENDS on it NOT working! You put each one in a fixture and burn out the output before mounting to the board. It would take a REALLY sharp clone maker to figure out that part HAS to be defective for the instrument to work.

Jon

OTOH your circuit now relies on the behaviour of a faulty part. Probably ok if you can fuse it completely oc, not otherwise.

NT

If you recall 5-1/4" floppy discs, one company's security measure against copying was to intentionally drill a tiny hole in the disc which was smaller than the head. If the hole wasn't there then the program wouldn't run...

Some discussion on that:

John :-#)#

afaiu the copy protection for sony playstation CDROMs was that they were pressed with a side-to-side wobble in the data "groove" that could be detected via the laser tracking

a copy would have the correct data but not the correct "wobble"

I think it may have been the other way around.

CD-R (and similar) disc blanks are manufactured with a track wobble. The CD-R burner detects it (by sampling the voltage in the laser system's side-to-side servo) and uses it to phase-lock the speed of the rotation, so that the disc is rotating under the burner at the speed needed to give the correct linear velocity along the track.

CD-ROM-based videogame systems can use a drive which can detect the wobble in the same way, and say "Aha! This is a copy, burned onto CD-R, and not a mass-manufactured/pressed CD-ROM. I reject it."

At least, that's how CD-R copy protection was done in the 3DO videogame system (I wrote the CD-ROM driver and filesystem code for its O/S kernel). We had to have a special development-mode override in the O/S to allow development systems to boot and run games from the CD-R test discs we burned (and man, were those blank discs expensive back then!!)

Am 28.03.2018 um 02:13 schrieb John Robertson:

Racal Redboard had this. We wanted it also on our home PCs, but it requested all 640 KB RAM in the PC and could not be loaded with any RAM spent for the debugger.

But we had a PC with RAM between EGA/VGA and the boot proms, so one could load Codeview into the high ram, and then the crack was a piece of cake, it took a little bit more than a day.

Nevertheless that was not worth the time. I must retract my statement of last week that Windows Orcad was the worst layout tool. Redboard was.

Not a bad idea. You can pretty reliably fuse a logic device so that it has increased supply current and a stuck input or output pin of some resistance. The test fixture can apply the fusing current, and verify the "write".

It's not very informative though: a few bits of "key", for as many pins very obviously doing some sort of analog sensing. Trace the circuit, patch out the resistors and it's repaired-for-good, as in bypassed.

A picosecond domain circuit could use its code to verify its PCB, using a funny shaped trace on an inner layer. One would have to determine how much error a typical copy would accumulate, though. Might not be very reliable against FR4 already.

Tim

Den onsdag den 28. marts 2018 kl. 03.17.32 UTC+2 skrev Dave Platt:

that doesn't sound right, afaik a blank CD-R is just that, blank. so there wouldn't be any track to servo on. The writer know the rotational speed and how far in the head is already

its optical qualities demonstrate that it's not literally blank

NT

you are right, I had to look it up and it does say that there is a pre-groove etched on CD-R

anyway, it still means that you can make a CDROM virtually impossible to copy by encoding data into the groove "wobble"

tabbypurr is correct. A "blank" CD-R is not literally blank - that is, it's not just a flat surface.

The clear substrate has one flat surface (the one the laser shines through), and one surface that's been very carefully molded. On this lower surface there's something known as the "pre-groove", which guides the "burning" laser as the disc is recorded. The burning laser "follows" the pre-groove, in much the same way that the playback laser follows it (and the line of "pits" recorded in the dye) - there's an active focus-and-tracking system which measures the reflected light from the disc, and drives the horizontal and vertical servos on the laser head so that the beam stays properly focused on the groove. It's an active feedback process. Without this, there's would be no way for the burning laser to record the tracks as close to one another as necessary - the laser-positioning system would be operating "open loop".

As I mentioned, the pre-groove is molded into the disc with a side-to-side "wobble", which provides direct feedback to the recording drive as to the actual linear speed at which the track is passing under the laser. This, also, is an active feedback process.

The groove "wobble" is actually modulated a bit, and this modulation contains a small amount of information which describes the type of CD-R disc (e.g. whether it's an "unrestricted" disc, or "restricted" to data use only), its dye type, manufacturer, and the correct strategy for modulating the laser during burning (some dyes give a better "burn" with longer or shorter laser pulses).

The Wikipedia article on CD-R has some details, and a photo of the groove on a blank disc. The article at CMC (one of the manufacturers of such CD-R blanks) has some technical details (see section 7).