Hello Everybody,
I have a EPROM programmer at my college. I wanted to burn code in my DS89C450 uC. the programmer had only DS89C420. i checked the datasheet it says that the 2 IC's are pin compatible and applications using DS89C420 can be replaced with DS89C450. When i tried burning my code the programmer says " MANUFACTURER'S CODE NOT MATCHED". what does it mean and what should i do?
It means exactly what it says!
Programming EPROMs involves applying high voltages - obviously, if it's done wrong, this will damage the chip.
Each chip has its own specific programming requirements (voltages, timings, which pins to use, etc) - sometimes known as the "Algorithm".
Therefore, before an EPROM programmer actually starts programming, it reads some identifying codes from the device to ensure that it is the correct device - so that it doesn't apply the wrong "Algorithm" and (potentially) destroy the chip.
A DS89C450 will almost certainly have different identifying codes from a DS89C420.
The fact that they are (claimed to be) "pin compatible", and that "applications using DS89C420 can be replaced with DS89C450" has nothing to do with programming them!
Can't these chips be programmed through the UART, anyhow?
Can't these chips be programmed through the UART, anyhow? if I recall, when I, a long time ago, looked at the DS ISP I said to myself "what a crazy scheme". With that recollection, I doubt it is UART programmeble.
also
Programming EPROMs involves applying high voltages - obviously, if it's done wrong, this will damage the chip. if I recall, the DS does not require any externally applied high voltage (that would not make sense for an ISP chip)
Erik
Quite possibly not.
But the OP seemed to be talking about a general-purpose EPROM programm - and that would have to have the capability to apply high voltages for those chips that do need it.
Another problem with parallell programming is that wrong programming algorithm can result in chips that "looks" like they are programmed, but the chip may start getting bit errors after months, weeks or maybe even hours.
That is a big advantage with chips with IAP - the chip itself have all the hardware or at least optimized firmware to get the best result. But even then, chip manufacturers now and then have to send out bootloader updates because the factory code didn't perform as expected over the full voltage and/or temperature range.
Another problem with parallell programming is that can result in chips that "looks" like they are programmed, but the chip may start getting bit errors after months, weeks or maybe even hours. also, the "wrong programming algorithm" can even blow the chip (e.g. applying 12V "programming voltage" to a pin)
but the chip may start getting bit errors after months, weeks or maybe even hours. which is precisely why most manufacturers have stopped supporting kitchen table programmer makers.
one related note: when shopping for a parallel programmer be sure to ask "how long will you support this model with updated software to cover newly introduced chips". I got so badly burnt by Xeltek, 9 months after the purchase I got "to program that chip you need to buy our new model". Of course, for the sake of full disclusure, I must say they 'generously' offered me $50 for the old one.
Yes - that was my original point!
The OP didn't seem to understand this...
Not too much problems with chips that lets out the magic smoke from incorrect programming parameters.
The big problem is chips that just got a little bit damaged, or chips that was "almost" programmed. I.e. chips that gets sent out to a customer and then fails.
When the chip blows up directly while being programmed, it's just a fixed learning cost, not to program with wrong algorithms.
When the chip blows up directly while being programmed, it's just a fixed learning cost, not to program with wrong algorithms. the pro will agree with you, the amateur not