While laying out the third version of the jGPS board, it suddenly
occurred to me: we chose the MAX2769 because we were goaded into
thinking that the serial output was actually... well, serial. And not
some weird strange bizarre useless thing, as it really is. Since it's
not usefully serial, we're switching the MAX2769 to parallel mode,
throwing the output at a CPLD, and then putting SPI'ing it into the STMF407.

So: if we're doing the MAX in parallel mode, why not look at *other* RF
front end ICs, and use them? I found two:

The ST STA5630. This one looks great, but is useless because there's no
useful datasheets on this chip. So, no.

The Skyworks SE4150L [1] is just the old SiGe SE4150L made now by
Skyworks. It's a dumb and easy chip to use: no SPI, no configuration to
speak of, and it only outputs 2 bit Mag/Sign data. So we lose a couple
of bits, but we now have no weird configuration problems. But it has a
serious problem for us: the IF is 4.092 MHz, and the sample rate is
16.384 MHz. So that's 32.768 Mbps, which is a whole lot of data, and is
> 25 Mbps of the usual SPI limits. On the other hand, at those rates we
get real bandwidth.

Currently on the MAX2769 [2], our IF is 2.048 MHz, the sample rate is
4.096 MHz at 4 bits, so our bitrate is 16.384 Mbps.

In a nutshell:

- SE4150L = less bits, more data, more bandwidth, simple
- MAX2769 = more bits, less data, less bandwidth, complicated

Anyone have anything to say about this? I'd appreciate thoughts on this!



[1] http://www.skyworksinc.com/Products.aspx?CategoryID=562

Andrew Greenberg

Electrical and Computer Engineering
Portland State University
a...@ece.pdx.edu  C: 503.708.7711

psas-avionics mailing list

Reply via email to