Thanks for your help. I've fixed it now.

There were several problems:
1. As you suggested, I was not clearing the CH (clock halt) bit. I think 
I also need to clear the WP (write protect) bit before doing anything 
else, since the data sheet says "The initial power-on state is not 
defined." for both CH and WP.

2. There was a typo in the read portion of the GetTime function. I wrote 
"response >> 1;" and I of course needed "response >>= 1;".

3. I wasn't addressing locations properly. I saw A0-A4 in the command 
byte, and saw addresses 80h-FFh in the address table, and assumed that I 
somehow had to fit 80h into 5 bits. I realised that I really have to 
just transmit the hex number given in the table.

In Hi-Tech C, if one writes "RA5 = 1;" then pin RA5 will be set. 
Therefore, if one does "#define CE RA5", "CE = 1" will be translated by 
the preprocessor into "RA5 = 1;" (or that's my understanding). It does work.

Thanks again for your help.

Robert Rolf wrote:
> I am not familiar with the Hi-Tech compilier,
> but I use a DS1302Z chip in many projects, and bit bang it using
> ASM, without problems.
> 
> Have you checked that your CE line is in fact going high before
> you clock out your command? I see where you assign CE to 1,
> but does that actually CHANGE the port bit with Hi-Tech?
> As far as I can see you have just defined CE as RA5 which in
> normal C would give you an initial value of 0x20
> and do nothing to the port.
> 
> I would suggest triggering the scope using the rising edge
> of CE if you don't have a 4 channel scope. Then you'll see
> each command and if the setup is sufficient (which is generally
> is when bit banging, as long as you don't enable and clock
> with the same port instruction).
> 
> You should be able to do a 'read status' command, and see
> all zeros except for the 'clock stopped' bit.
> You do have to CLEAR that bit explicitly to start the oscillator.
> I don't see where you are doing that so even if you were able
> to load the time, it wouldn't increment.
> 
> I would hang a 1k pullup on the data line and see if the
> DS1302 is even driving the line. If it isn't, (0xFF response)
> it didn't see the command.
> 
> Robert
> 
> Trevyn Watson wrote:
> 
> 
>>Hello,
>>
>>I'm trying to bitbang communication from a 16F877A (running at 4 MHz) to 
>>a DS1302, and I haven't managed to get anything out of the DS1302. I'm 
>>using Microchip MPLAB with Hi-Tech PICC-lite 9.50. I'm using delay 
>>functions by Shane Tolmie from http://microchipc.com/sourcecode/#delay. 
>>My program is included.
>>
>>So far, I have:
>>
>>Read and applied the "gotchas" page
>>(http://www.piclist.com/techref/microchip/gotchas.htm)
>>Checked the connections (at least 10 times).
>>Replaced the 32.786 kHz crystal with one that is known to have the
>>correct capacitance (new from Digikey).
>>Replaced the DS1302.
>>Read the data sheet multiple times.
>>Searched the PicList archive (I found very few mentions of the DS1302).
>>
>>I have put a scope on the clock and data lines, and as far as I can
>>tell, the proper command byte is being sent. You can see it at 
>>http://trevyn.ca/sclkio.jpg
>>
>>The command byte is being transmitted in the way that I intend (but that
>>may not be the way the DS1302 expects it). There are 8 falling edges
>>after the last bit of the command byte.
>>
>>My only ideas are:
>>1. The command byte I am sending is malformed/incorrect. I can't figure 
>>out how to translate a data address (range 81h-FFh) into a 5 bit number 
>>that goes in the middle of the command byte.
>>
>>2. I'm connecting the DS1302 wrong. I still haven't had any confirmation
>>that the chip actually works. I've seen nothing out of it.
>>
>>3. I am assuming that the counter starts when power is applied, and that 
>>reading it every half second will give some sort of output after one 
>>second. If this is not the case, then what I am seeing is perfectly normal.
>>
>>Basically, my questions are:
>>1. Should I use a different chip? (I need one that is available in 8-SOIC)
>>2. Is there a better way to talk to the DS1302?
>>3. If not, what am I doing wrong?
>>
>>Thanks for reading.
>>I appreciate any help or ideas you have.
>>Trevyn Watson
> 
> 

-- 
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist