Programming the 1200

Go To Last Post
13 posts / 0 new
Author
Message
#1
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I bought a tube of the 1200's and can't program them.
I've built about 5 different versions of programmers from the net and none seem to program the 1200. With the very same programmers, I can program the 8535 and the Tiny15 but the 1200 is the same on all programmers, the programmers say that either they can't program it or it's locked. Now because I can program the 8535 and the Tiny15, I am beginning to suspect the chips. The atmel representative in South Australia said that they have had no complaints or emails about a bad batch or 1200's.
Can anybody help.

admin's test signature
 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I went round and round with emails to Atmel (wasted a week of time) and could not get them to say if they were having troubles with them programming. I only have 8 of them, and finaly got 5 to take a load using a program named PonyProg (www.lancos.com) I have no problems with 2313's but the 1200's just wont erease.And if I do a read then I get a false count if 00 01 02 03...ff. If you do a search here on "cant erease" you'll find my other posts.

admin's test signature
 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

It's good that you mentioned the 00 01 02 03 because that's what I get as well. It looks even more toward a crook batch.

admin's test signature
 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

It's good that you mentioned the 00 01 02 03 because that's what I get as well. It looks even more toward a crook batch.

admin's test signature
 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Hi all,

It's a bit difficult for me to comment on 3rd party programmers like Pony Prog, but I'm investigating this matter a bit at the moment. I'm quite sure this is not a "crook batch", but maybe some cute quirks with the 1200's ISP protocol, which differs some from the '8535 and indeed all other devices. Here's my first comment:

The 1200 has to go out of programming mode between erase and write. Your programmer must:

- Enter programming mode
- Erase the chip
- Go out of programming mode
- Enter programming mode
- Program the chip
- Go out of programming mode

if your programmer doesn't do this, it's a classic! If you have a way of excluding the erase cycle from the programming sequence, do this. Then erase the device separately, and program the device without erase.

I'll look into it further, but please give me some feedback on what I just wrote here.

Best regards,

Morten, AVR tech. support, Atmel FAE

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Hi again,

I've checked the matter out a bit, and have found the following which might help you: There are two reasons for the chip read to return "00 01 02 03... ":

- The lock bits are programmed ('0'). Then only the last bytes of the address are returned.
- The programmer has lost synchronization with the device, and the address is returned instead of the data.

Note for people who use the STK500/ AVR Studio package: To program the 1200, uncheck the "Erase Device Before Programming" box. Then "Erase Device", then program.

Morten, AVR tech. support, Atmel FAE

admin's test signature
 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Try this: http://jaichi.virtualave.net/avr...

I couldn't get anything to program the 1200, but this does program and verify properly every time. I still can't clear the fuse bits with this, but it will program just fine. The cap value from SCK to ground is critical, so try to keep it close to 1 nF, 1000 pF. You have to set up external power for the atmel. The parallel port just loads down. I used a simple 9V wall transformer going through a 7805. Also, make sure to manually configure for 1K EEPROM. The auto-select feaure can cause problems.

admin's test signature
 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I had the same problem (0x00, 0x01 ,0x02) and foujnd out that the main clock to the AT90S1200 was not running(due to a missing component in my application). As soon as I got the clock running to the AT90S1200 all worked just fine :-)

So check that the XTAL1 pin on the AT90S1200 is oscillationg. (When using STK500 wrong OSCSEL jumpersettings may cause no clock).

Bjorn

admin's test signature
 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Hi,

The reason why your programmer couldn't program/erase the fuse bits, is most likely that the programmer doesn't support the parallel programming mode required to do this. The STK500 features this programming mode.

Best regards,

Morten, AVR tech. support, Atmel FAE

admin's test signature
 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Yep, I built this programmer to do some 1200s I had, and they are already set to externally clock, so I didn't need to access the fuse bits. That was just a side note. I have about 1000 1200As I need to de-program (set )the RCEN bit on. I just got confirmation my STK500 is shipping today, so I'm sure I'll be back for help on setting those fuses. :) The STK500 looks like quite a versatile machine.

admin's test signature
 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

For those of you that are having problems programing the 1200's. it is most likally software related. I've been using a program called Fbprg to program my 2313's and it has worked fine. I received and tried some 1200's and they kept comming back with 01 02 03 bytes. and when I tried to read the device it came back with an orderly ascii dump. don't know were it got that from. By the way this was version Fbprg1.6 that I was using. there was an older version #1.3 which seems to work but only with an external 5 volts applied. to the chip being programed. The point here is that if your getting errounious readings check out the vcc level on the pin. my expereance with the 1200 and this make shift programer which works great for all the other atmel that I've used. has a problem with the 1200's. while monitoring the vcc during programing the level was at 3.5 volts which is below the threshold for the 1200. Once I put an external power supply of 5volts on the chip. the erase/programing/verify and reading of the device bits went smothly. Hope this helps.

admin's test signature
 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Hi all,

I have programmed the AVR1200 as long as there are available with my own serial programmer. You can download the tool for free at www.roesink.com. (The windows software and a scheme for building the serial programmer)

I must say that I haven't tested the latest software on the 1200 because I'm using the 2313 instead. (because this part has sram available.)

I have tested the latest software on the AVR, AVRmega and Tiny family.

Because all you need is a MAX232 (or ST232), a transisitor and some passive components, it's worth trying.

Another thing; ATmel had some problems with the fuse bits. It was not possible to enable programming again after locking it when the VCC is above the 4.5V

Greetings,
Robert

admin's test signature
 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I use a serial programmer that was designed for the 1200,
it even comes with a 1200 in the socket and a serial
9pin cable for your PC to get going.
Since it's serial not parallel programmer you can't blow
the RCen fuse using this programmer but there
is a 12MHz crystal on the board to run the chip once
programmed. Every I/O line is brought out to a single row
header of pins.
You can plug the whole thing into a solderless breadboard
for prototyping and just leave the 1200 in its socket
on the programmer. Very rapid prototyping.
Available from a guy in Canada, and it's cheap,
it uses the avrasm and avrprog right off the
atmel site. This should be on the tools page... but its not.

check out www.avrboard.com

admin's test signature