[RTTY] About N4II's experiment...

[Kok Chen]

On Mar 15, 2014, at 2:37 PM, Joe Subich, W4TV wrote:

> Manufacturers who skip the chipmaker's recommended regulator for the
> sound ADC, who fail to properly decouple (filter) the USB power, who
> fail to regulate the V/2 reference for their op-amps and who fail to
> adequately decouple their power distribution bus certainly can't be
> trusted to know what they're doing when it comes to instructions
> about setting AFSK levels.

This reminds me of the time I bought the original digiKeyer (not the digiKeyer II, which is a different animal).  It would randomly refuse to connect to Macintosh computer.   

The designer (OM7ZZ) was kind enough to send me the schematics.  I compared it against TI's Application Notes.

Sure enough, TI had decoupled the USB *data* lines with small caps (10 or 20 pF, if I recall) and those were missing in the digiKeyer.  You would think the caps would cause the data line in the USB cable to ring, but when I added the caps to the digiKeyer, the unit started to connect consistently.  I reported it back to the designer and sending the TI app note with my report.

When the above happened to me, MicroHAM was just in the process of switching from thru-hole components in the digiKeyer to SMD components.  The designer added the caps to the new board layout, and was gracious enough to send me a completely new free digiKeyer that has the new board.  That new digiKeyer kept working until I stopped using it.

The moral of the story is that things like this can happen to any manufacturer.  And, when the application note of a device from a reputable chip manufacturer includes a component that does not make sense, include it in your design anyway! :-).

The difference between Mac OS X and Windows appears to be Windows keeps trying to make a USB connection even when an initial connection is not solid.  Mac OS X disconnects the USB interface if it notices that the initial connection is not perfectly solid, with either few or no retries. 

73
Chen, W7AY