I finally had a chance to try out the Saleae unit today, and I have to say I didn't love it. Maybe it's a case of wrong expectations rather than anything fundamentally wrong with it, but I'd say this almost isn't a logic analyzer at all. It's more like an 8-channel, 1-bit oscilloscope. The software has no state analysis features at all, and only the most rudimentary support for parallel busses. Given that, I guess I understand why the hardware doesn't have more channels, because the software just isn't designed for working with that kind of system. If you want to do something like trigger when the value on the address bus is between 0x400000 and 0x410000, or when signal state ABC is followed by DEF, or just look at the state of the system from one clock cycle to the next in a list view, you can't.
You also can't connect an external clock source to determine when samples are captured. That means you have to run the Saleae's internal sampling clock at a few times faster than the clock on the system under test, oversampling the signals. Then you have to deal with the fact that there's an unknown phase relationship between the internal and external clocks, and the Saleae's 10 MHz clock probably isn't precisely the same speed as the system's 10 MHz clock. All of this may cause occasional weird or wrong-looking samples, if the Saleae takes a sample in the middle of your system's clock cycle, when a signal value is right in the middle of a transition.
Maybe I'm being too hard on this thing, or my harping about state analysis and external clocks reflects a "legacy mindset" that's only relevant for bus-based systems. If you were using the Saleae unit solely to troubleshoot an SPI, I2C, or other serial connection, it would probably be fine, though even there some basic features seem to be missing. For example, you can't trigger on a particular SPI value, and the options for searching for particular values or states in the captured data are fairly weak.
I'm not really sure what kind of audience Saleae is aiming for. It seems like it may be the casual hobbyist or Arduino user, but $400 - $500 for a tool in that marketplace is pretty expensive. I don't necessarily expect Saleae's hardware to have the same features as professional equipment from Agilent that costs thousands of dollars… except that most of what's missing seems like it would only require a software change, not anything on the hardware end. The Logicport LA1034 that trag mentioned is in the same price range, and from what I can tell it has most of the features the Saleae is missing. Even the open source hardware "Open Workbench Logic Sniffer" has many of those features, and it also has 32 channels and only costs $50.
You also can't connect an external clock source to determine when samples are captured. That means you have to run the Saleae's internal sampling clock at a few times faster than the clock on the system under test, oversampling the signals. Then you have to deal with the fact that there's an unknown phase relationship between the internal and external clocks, and the Saleae's 10 MHz clock probably isn't precisely the same speed as the system's 10 MHz clock. All of this may cause occasional weird or wrong-looking samples, if the Saleae takes a sample in the middle of your system's clock cycle, when a signal value is right in the middle of a transition.
Maybe I'm being too hard on this thing, or my harping about state analysis and external clocks reflects a "legacy mindset" that's only relevant for bus-based systems. If you were using the Saleae unit solely to troubleshoot an SPI, I2C, or other serial connection, it would probably be fine, though even there some basic features seem to be missing. For example, you can't trigger on a particular SPI value, and the options for searching for particular values or states in the captured data are fairly weak.
I'm not really sure what kind of audience Saleae is aiming for. It seems like it may be the casual hobbyist or Arduino user, but $400 - $500 for a tool in that marketplace is pretty expensive. I don't necessarily expect Saleae's hardware to have the same features as professional equipment from Agilent that costs thousands of dollars… except that most of what's missing seems like it would only require a software change, not anything on the hardware end. The Logicport LA1034 that trag mentioned is in the same price range, and from what I can tell it has most of the features the Saleae is missing. Even the open source hardware "Open Workbench Logic Sniffer" has many of those features, and it also has 32 channels and only costs $50.