... SPI for example uses 4 lines,...
I am currently working on an SPI (=serial) problem, and it's more parallel than you think.
I try to connect 3 different modules (ETH, GLCD, MCP) on the same SPI "port" of a PIC processor, the way it's theoretically described for ages in every book. Mosi, Miso, Clock. Finished.
in practice:
-every SPI module needs his own CS_EN pin.
-to rule out problems and easy time-based debug, I chose to give every module temporarily it's own RST pin.
-for some reason, the GLCD has an AO pin: data or command. You can drive the backlight led with pwm.
-the ETH and MCP module can give feedback on an interrupt pin.
-to control start/stop sampling of the analyser, I control an output bit in my program on suspicious places
This gives:
3-clk, miso, mosi
3-cs0 cs1 cs2
3-rst0 rst1 rst2
2-AO, led_pwm
2-int0 int1
1-bit for start/stop analyser
"only" 14 channels, to debug a "serial" protocol, witch I consider easy and basic. (and shitty)
I have a 8-channel analyser witch is enough for this, but less is really not workable.
You may also consider Saleae Logic-products in addition to the oscilloscope. The Saleae Logic-products are very handy and can really save your day. There might be some other similar products in the market, but I have used the Saleae Logic-devices in my professional life with great success.
+1, due to their "sold out" at the time I bought the cheapest $10 replica, that device seems to replace everything in the -y2010 -500$ market.
After this experiences, I would never choose a scope with integrated analyser, except maybe a $10K scope with a $4K analyser, for a $50BN project, witch is far away from my abilities and needs at the moment.
The only need to get it integrated I see at the moment is the ability to see synchronisation between analog and digital, but in my accuracy and frequency range this can easily be done by triggering one device with the other, or record 1 channel with both devices the same time.