OK, I pulled the processor out of the project board and set it up on a proto board with nothing other than the minimum circuitry to power it up ( Vdd, Vss, reset, serialin, serialout) and communicate with it serially. I took out the RS232 driver (yes, I discovered the inversion after I did it). Still have the same results. I am using a desktop computer with a "real" serial port. I am using the internal resonator on the PICAXE. There is no question that when the serialin pin is pulled to +5v I get output on serial out, I don't know if it is valid data or garbage. I am using a 5v power supply with plenty of current capacity. When I do a firmware check, I get the hardware not found message. What actually does the "firmware" button do? What is the message that the bottloader returns to the computer? Is it printable characters? Is it a bit pattern I can detect with a scope?
28x1 bootloader problem
- Thread starter HowieD231
- Start date
It is not printable - it is a handshaking sequence that will not mean anything useful in ASCII. The only download circuit should be the 10k/22k resistor - do not use a RS232 driver. Make sure 10k pull-down is on computer side of 22k.
However the fact that the serial signal is there proves the chip is working fine. Which narrows the problem down to
1) a faulty serial cable / serial connection
2) wrong COM port selected
3) COM port not correctly operational. There are no settings to make under Windows, this is all done automatically via the ProgEdit software.
However the fact that the serial signal is there proves the chip is working fine. Which narrows the problem down to
1) a faulty serial cable / serial connection
2) wrong COM port selected
3) COM port not correctly operational. There are no settings to make under Windows, this is all done automatically via the ProgEdit software.
Last edited:
That's exactly how I have it connected. I can do the com port test and watch the serial in line toggle as it should. If I short the tip and sleeve on the programming cable, I can send characters to myself with Hyperterminal. I have the control lines on the DB-9 jumpered as follows: RTS-CTS and DTR-DSR-CD. I didn't know if the program or operating system was looking for these signals or not. The pattern I see on the scope looks like alternating 1's and 0's on serial out in response to a 1 on serial in. I'm about ready to launch this thing into low earth orbit...
hippy
Ex-Staff (retired)
That could be your problem ...
Tip should be 0V
Ring should be PC TX into Serial In via 10K/22K
Sleeve should be PC RX from PICAXE Serial Out
Shorting sleeve and ring should create the loopback for HyperTerm.
BeanieBots
Moderator
Which project board. I don't know of one which has a RS232 driver?
Does it still work when replaced in the "project board"?
The download circuit is three wires and two resistors. Nothing more.
Reset must be high. 0v should applied to BOTH 0v pins.
Resonator pins left 'floating' if using the internal one.
Maybe post a photo.
Last edited:
The "project board" is a point to point wired perf board with a circuit for an amature radio project. I an effort to eliminate all outside conditions, I put the chip onto a proto board and wired pins 8 and 19 to ground, pin 20 to +5, reset pin 1 is pulled to +5 through 4.7K resistor, serial in has 22K resistor with 10K to ground on computer side, serial out has 180 ohm in series, Com port ground is tied to PICAXE ground....I am able to do all this from memory because I have been over the wiring many many times now. Running the com port test indicates the correct conditions on serial in, so that along with other cable tests make me confident the cable is good. I can see data generated by the PICAXE when the serial in line goes high. Does anybody know what the computer expects to see from the PICAXE? There doesn't seem to be any way to determine conclusively if the problem is in the computer or the PICAXE without determining if the PICAXE is sending the correct information to the computer.
Monitoring the serial port both ways when a Firmware? check is done shows:
- apparently no outbound ascii characters from the PC - maybe it's done by manipulating the serial port in some other way (Technical grins to himself as people stumble around in the dark trying to work out what's going on)
- a string of ascii charcters coming back from the PICAXE (about 100) where on a sample of two different models of PICAXE , the last two bytes returned are identical and equal to the firmware version in a Binary Coded Hex (?) notation e.g. last byte is 0x76, firmware is reported as version 7.6, last byte is 0xA5, firmware is reported as version A.5 (the latter being a 28-X1)
- apparently no outbound ascii characters from the PC - maybe it's done by manipulating the serial port in some other way (Technical grins to himself as people stumble around in the dark trying to work out what's going on)
- a string of ascii charcters coming back from the PICAXE (about 100) where on a sample of two different models of PICAXE , the last two bytes returned are identical and equal to the firmware version in a Binary Coded Hex (?) notation e.g. last byte is 0x76, firmware is reported as version 7.6, last byte is 0xA5, firmware is reported as version A.5 (the latter being a 28-X1)
Last edited:
Your issues are very unusual. Everything still points to a wiring mistake in your hardware/cable - particularaly as you are using a non-standard arrangement. Have you tried the recommended 4.5V from 3xAA cells instead of your power supply? Are you sure the chip has not been damaged by the previous incorrect (e.g. rs232) wiring attampts?
The format of the serial data is completely irrelevant, the PICAXE chip will either send the right data or nothing at all!
You also make no mention of testing PICAXE to PC connection, only the other way.
Try this to check the serial out pin to computer connection:
1) Select correct COM Port (View>Options>port)
2) Open PICAXE>Terminal menu to get the serial terminal open. Select baud rate 4800.
3) Connect the serial in pin (the actual leg of the chip) to V+ momentarily.
You should then see a stream of garbage data appear in Terminal. What it is irrelevant, but if there it shows the TX pin is working correctly.
The format of the serial data is completely irrelevant, the PICAXE chip will either send the right data or nothing at all!
You also make no mention of testing PICAXE to PC connection, only the other way.
Try this to check the serial out pin to computer connection:
1) Select correct COM Port (View>Options>port)
2) Open PICAXE>Terminal menu to get the serial terminal open. Select baud rate 4800.
3) Connect the serial in pin (the actual leg of the chip) to V+ momentarily.
You should then see a stream of garbage data appear in Terminal. What it is irrelevant, but if there it shows the TX pin is working correctly.
No secret, it's called an RS232 'break signal' and you can do it yourself by uing the Port Test LED (View>Options>Serial Port>Test)
Problem solved!
I received the new 28x1, plugged it in and everything worked fine. I was able to download my program without problems and everything worked. I guess these chips really don't like have serin and serout reversed... in any event, thanks again for all the help...case closed
I received the new 28x1, plugged it in and everything worked fine. I was able to download my program without problems and everything worked. I guess these chips really don't like have serin and serout reversed... in any event, thanks again for all the help...case closed