Date   : Sat, 27 Nov 2010 10:00:39 +0000
From   : pete@... (Pete Turnbull)
Subject: Electron Ferranti ULA reverse engineering progress

Alan Williams wrote:
> oopse I meant 6850
> 
> 
> -----Original Message----- From:
> bbc-micro-bounces+awilliams=linkme.com.au@...
> [mailto:bbc-micro-bounces+awilliams=linkme.com.au@...]
> On Behalf Of Alan Williams Sent: Saturday, 27 November 2010 2:29 PM 
> To: bbc-micro@... Subject: Re: [BBC-Micro] Electron
> Ferranti ULA reverse engineering progress
> 
> As the next  chip is the 6550, which is asynchronous I speculate that
> the cassette interface is as well.  So data bytes on the tape are
> framed with start & stop bits.  No clock recovery necessary.

Indeed, and modems typically don't have zero-crossing detectors either. 
  James may have been thinking of the NRZ/NRZI schemes often found on 
1/2" magtape.

The way most UARTS work, and this includes the 6850, is based on use of 
a clock that's 16 times the bit rate.  When the leading edge of a start 
bit is received,, the UART counts 8 clocks to find the notional centre 
of the start bit.  Thereafter it samples the data every 16 clocks, which 
ought to be the centre of each data bit, and then the stop bit.  That 
allows for quite a large variation on speed over 8 bits.  At the point 
where it should be sampling the centre of the first stop bit, if it 
finds the wrong level, it generates a framing error, and you know the 
speed was too far off.  Thus it gets re-synchronised every ten bits or so.

-- 
Pete                                           Peter Turnbull
                                               Network Manager
                                               University of York