Hello all.

I'm working on my SBC using the NXP SC28L92. I've been reading over the great "Interfacing the NXP SC28L92 DUART" by BDD. It's a wonderful document but I'm having trouble understanding the transceiver part. BDD uses a MAX248 but in my design, I decided to go with a MAX238 as he recommended it as a secondary ("simpler") interface.

As best as I can understand, I **THINK** I have this wired correctly but I sure could use some constructive criticism if not. My goal would be to have at least one external serial port using full RS-232 levels and hardware handshaking. This would be my interface to the outside world.

For general communications with my main PC, I decided to put headers on and expose all pins of both channels. I thought using Channel B and a simple RS-232 to USB adapter would be good enough for debugging purposes. The SBC will eventually go in a project enclosure box and the "debug channel B" headers will not be brought out (to protect from the higher voltages of RS-232).

Attached is a piece of my schematic for the serial portion (converted to B/W).

I realize my understanding of RS-232 in general needs a lot of work.

Thanks for any help!

_________________
Cat; the other white meat.

I attached "Interfacing the NXP SC28L91," which illustrates the use of the MAX-238 on page 17. Note how the MAX-238 is drawn in the schematic—it conforms to Maxim's schematic symbol, which naturally places the UART connections on one side and the DTE or DCE connections on the other side.


It appears your hookup to the MAX-238 is correct. Don't forget the 1 µF bypass capacitor between VCC and GND for the MAX-238. All capacitors should be as physically close to the device as possible. Not shown but recommended is a large electrolytic at the MAX-238's VCC and GND, alone with the 1 µF bypass.

Despite all the Sturm und Drang in my life in the last 11 months, I've tried to maintain some momentum on my 28L9x documentation. Attached is a yet-to-be-completed paper on writing a driver for the 28L91. Any typos or layout boo-boos within are due it being a work in progress. The 28L92 is essentially the dual channel version of the 28L91, so the information should be of some value.

_________________
x86?  We ain't got no x86.  We don't NEED no stinking x86!

Wow. Once again, you've out done yourself. That is exactly what I was looking for!



This would be in addition to the normal bypass cap for the MAX-238 right? Something to allow a large bank of power for the charge pumps of the chip?

I was thinking something like a 47uF cap. I admit, I have no idea why I chose 47 other than it seems to be quite common in designs. What do you suggest?

Thanks!!!

_________________
Cat; the other white meat.

Correct.


The electrolytic I use is 100 µF at 10 volts.

_________________
x86?  We ain't got no x86.  We don't NEED no stinking x86!