On Mon, Feb 22, 2021 at 07:39:42AM -0500, Michael G. Katzmann wrote:

On 2/22/21 3:52 AM, Johan Hovold wrote:

quoted

Does your updated algorithm also result in 110 baud (8n1) being encoded
as:

	a8 a6 01 80 00 02 07

And are you using some official Prolific Windows driver or something
that came with the device?

Johan,

  On Windows I did not install a new driver. It was recognized by the
system and uses the Microsoft provided Prolific driver Ver 3.8.38.2.

Thanks for confirming. Then this sounds like something which Prolific
should be very well aware of.

On windows everything looks fine (no sign of distress (i.e. no yellow
caution triangle)).

Where should I look for the encoding (a8 a6 01 80 00 02 07) ? (110bd
encodes as 80 00 C3 54 using the algorithm I described))

That was the encoding used by Joe's device (same driver I think) for
110n81.

It may not work with your device, but 0x10000 could just be a
2-prescaler bit:

	32 * 12*10^6 / (2 * 0x6a8 * 4^5) = ~110.03521126760563380282

where as you would have:

	32 * 12*10^6 / (0x354 * 4^6) = ~110.03521126760563380282

Could you try hardcoding 0x1a6a8 and and see if you end up with 110 bd?

quoted

I tried asking Prolific about this but I'm still not sure whether these
are official chips or counterfeit. 0x0300 is supposed to be a PL2303TA
and Prolific claims that the current driver is working fine with these
so we'd need to key off something more than just bcdDevice.

Johan