xox, I found that I had to lower the output level of the USAMO plugin -7dBFS in the DAW to get it to work, even with the trim all the way down the signal was too hot with the DAW fader at unity. What also threw me off at first is that the clip indicator on the USAMO doesn't seem to be affected by the trim knob - if I set the DAW level so the clip LED is just starting to come on like the manual says, I can turn the trim control across its whole range and it'll stop working right at the extreme settings but the clip LED will react exactly the same across the entire range. I ended up just leaving it at the halfway point, setting the level in the DAW, and then saving a channel template for it. I also noticed that it seems to need different levels in different software, although that might be that some of the programs I tried were running at 96kHz and others were at 48, but whatever reason I ended up having to set the plugin output 2 or 3 dB hotter to get it to work right in ReNoise than I did in Reaper. Another reason to set the levels in software and save a preset rather than trying to use the front panel trim control.
So if you were having trouble getting it to work reliably, try using your software to set the output level of the USAMO plugin rather than using the trim knob, that might work better. I'm sure it depends on your specific interface and software.
I was completely OTB with an MPC2000xl as the center of my setup for years and since I went back to a sort of hybrid setup with hardware sound sources/ITB mixing (but with some hardware effects), I've mostly stopped sequencing because whether I was sequencing in the DAW or sequencing in hardware and slaving that to the DAW clock, the timing was loose enough that everything always felt kind of off and it was impossible to get really tight interlocking stuff happening like I did when I was completely OTB. The USAMO completely fixed that. If anything it's probably even tighter than the MPC or Octatrack, but both of those were good enough that I didn't notice any jitter except if I tried to trigger the same sample twice on the same sequencer step in the MPC2000xl - the timing in that thing is just bad enough that you get phase cancellation issues when you layer identical samples in it (that was a common trick to get bigger drum sounds in the MPC3000 because its internal sequencer has sample accurate timing; according to a friend of mine who used to be at Akai tech support there were a lot of MPC3000 beat producers who were pretty upset when they got newer machines and the timing wasn't good enough to do that anymore; for regular MIDI sequencing the MPC2000xl is more than tight enough for me to notice, though - the main reason I rarely use it these days is a bunch of the buttons need to be replaced and it's a big hassle to pull it apart and get access to them).