[vcf-midatlantic] Zilog System 8000

[Oliver Lehmann]

By the way,

the DEI drive has a Controller and Codec board as shown on page 19 in
http://bitsavers.informatik.uni-stuttgart.de/pdf/dei/CMTD-3400S2_4tk6400bpi_1979.pdf

So you already get NRZ Data out of it. You could just "select" the tape
and then let the tape read the whole tape at once and record the NRZdata
incl. clock.....

Oliver


Jonathan Gevaryahu via vcf-midatlantic  
<vcf-midatlantic at lists.vintagecomputerfederation.org> wrote:

> On 11/16/2015 4:16 PM, Evan Koblentz via vcf-midatlantic wrote:
>>>> I hope that it is ok that I signed on to your list. I am from germany
>> Welcome!
>>
>>>> and I'm looking for information about the Zilog System 8000. I  
>>>> know that you owned at least one in the past. I don't know if you  
>>>> still have it, but I hope so :)
>> We have it, but we haven't done much with it. Perhaps at our next  
>> repair event. We have workshops here every few months.
>>
> Note I've been told QIC tapes of this age most likely (95% chance)  
> have to be "baked" to prevent them from turning to mush against the  
> drive heads. (I'm not 100% sure how the baking process is gone  
> about, but I believe it requires disassembling the cartridge and  
> removing the tape spool, and wrapping the tape around a special  
> metal spindle, then heating it to a certain temperature for a few  
> hours. I could be way off here.)
>
> Also, to recover the data from these tapes, given the drive capstans  
> (and cartridge capstans?) always seem to turn to goo and we're not  
> sure if the z80-based interface card works or not...
> Would it make more sense to get any old QIC drive that is  
> sufficiently old enough to be hackable but has intact rollers, and  
> (after baking the tapes) use an FPGA development board and some  
> breadboarded analog amplifier magic to directly control the QIC  
> motor and tap the QIC drive heads for the 4 tracks, and either  
> record them as analog waveforms at a high samplerate, or log  
> time-deltas (in 50mhz clocks) between flux transitions? Then decode  
> the resulting flux stream entirely in software. 50mhz may be extreme  
> overkill, its possible digital sampling at 192khz or twice that may  
> be fast enough, I don't really know...
>
> I could have sworn someone on this list, or in another similar list  
> (classiccmp?) had already worked out a setup for decoding tapes just  
> like this...
>
> -- 
> Jonathan Gevaryahu
> jgevaryahu at gmail.com
> jgevaryahu at hotmail.com