semi-OT: do you routinely attempt projects out of your comfort

[Tony Duell]

> 
> Tony Duell wrote:
> > Notice I said 'provide it as a kit', not 'provided it as a kit or ready=
> =20
> > made'. The point being that if _only_ the kit is available, people who=20
> > want it have ot buy the kit, and they have to be clueful enough to put =
> it=20
> > together.
> 
> Like I said before, though -- the DiscFerret is an almost completely SMD=20
> design. In fact, the only parts of the board that aren't SMD are the USB=20
> connector (a Lumberg thru-hole B-type "device" connector), the power=20
> connector (Switchcraft RAPC722 2.1mm DC barrel socket), and the=20
> interface connector (40-pin right-angled IDC with latch ears).
> 
> As a bare minimum, if I made a kit, I'd want to supply it with the power=20
> controller chip soldered down (the aforementioned QFN chip -- TI=20
> TPS75003). That's simply because the leads don't extend up the side of=20

Is thatere any particular reason that you're using this chip? I haven't 
got the data sheet, but what does it do that can't be done with 
easier-to-solder components?

> 
> Would this be more to your liking?

Err, yes.. But it doesn't need to be 2 separae manuals. Nothing wrong 
with a comminder 'operating/sercice manual'

> 
>    User's Guide / Programming Guide
>      Software installation / basic usage
>      Hardware pinouts -- disc interface / power connectors
>      API documentation and USB control protocol
>      Firmware theory of operation
>        (possibly useful if you want to play with it at a lower level=20
> than DiscAPI allows)

Firmware source listing? 

> 
>    Service Manual
>      Schematics
>      PCB layout diagram (x2 magnified)
>      PCB top and bottom layer (x2 magnified)
>        (useful if a trace gets burned away!)
>      Parts manifest
>      Assembly instructions (for kits)
>      Hardware theory of operation (for each major section)
>        (no sense in building a kit if you don't know how it works!)

FPGA souce? PAL/CPLD equations?

> 
> > Given the scheamtic and the knowledge that I put it together in the fir=
> st=20
> > place, it's likely I can fix it.
> 
> Unless a failed part is NLA...

Not so. If I genuinely know what the defective part is, there;'s a good 
cahnce I can kludge up some replacemnt from bits I can get. Of course if 
the unit is am FPGA + PSU + not much else ad the FPGA is what's failed 
and NLA, it might be easier to make up a new PCB to take the replacement 
FPGA and move the PSU parts onto it...

I can assure you that if parts being NLA stopped me from repairing 
something then many of my classic computers wouldn't be operational.

-tony