Though I used to record to cassette tape via a small 6-channel Yamaha mixer, it is so much easier to go all digital and save using software via stereo line input that is connected to your mixer.
To start, I find it important that you start thinking what final product you want in terms of a stereo image. What instrument you want where in that spacial image as a listener. To this end, it helps a lot if you don't want ultimate control over every section individually for a big final mixdown at the end by setting up the initial stereo tracks at the mixer and then adding enhancement tracks later by playing the initial tracks through your headphones and recording accompaniment tracks later. This saves cutting down the number of times you play/record the sections.
Instead of recording bass, then treble, then drums, then accompaniment, then main track, let me explain how I did things.
I would configure the mixer settings in terms of my arranger volume and image (pan) settings, then do the same for the bass, and then the treble settings, giving emphasis of volume to the bass and treble settings. Basically, I would set it up as if I was playing a live gig. For me that meant that of the 16 channels on my mixer, 8-9 would be configured how I wanted things to sound. After that I configured effects like echo/reverb to taste.
This was then dumped into 1 left/right track to my computer, saved using Adobe Audition CS6. Now any similar program like Cubase, Audity and other free applications would do the same thing.
Done right, it saved me a TON of time by not needing to record the song 10 times over and separate each section, because the balancing and mastering of all these parts was already done via the mixer on the initial setup.
Where you are going wrong is losing sync between tracks, and here is the trick on how to avoid this... after Track 1 is recorded, you are going to listen to this track through your headphones from the recording and simultaneously play/record an accompaniment track that records only to track 2. Perfect sync every time! And you just keep adding tracks and other accompaniment tracks until you are finished.
The second last step is to balance your accompaniment tracks to your #1 track. Once done, output it all to a nice MP3 and make it available for us to all hear!
Yes, I know that this is not how "real studios" do it, they put everything on it's own track and then spend weeks balancing each track to the overall project, but seriously, unless you are mastering for professional mass production and are paying an expert to do this all for you, it is not always necessary.
Oh, here is a small hint... Unless you are recording from an analog source (a microphone or pickup mics in your accordion), you don't need to use headphones, you can listen to the music through your computer's speakers and record the next track, but make sure all analog mics are disabled or you introduce these background sounds into the track you are recording on when you do NOT want it there.
Need a small video tutorial to show it? Let me know and I would be willing to try to make something for the group. I am not an expert, but do have some experience doing this and would be willing to share if someone wants.
The difference between this setup and a digital "recording" is that you don't play back midi sounds that can be changed and then mixed down, but rather is a kind of more old fashioned way where everything is saved in a digital sound file and the only changes one can make is changes to the track, but that said, you have the choice of adding in effects while recording the track, or if your software supports it, you can do all other kinds of effects additions via software. The one big thing that I like is that when recording audio analog tracks, one can re-EQ and/or de-noise or de-hiss them before mixing down to the final MP3.
In a more complex setup, one could save output to a midi-file that could change the tone/sound/instrument on the fly. I cannot do that in my setup and honestly, don't really miss it, because if you look, one of the first things I do is "design" the sound and final product before I start recording. Most of the time, I would have played that song so many times that I would know in advance how I wanted it to sound and no major changes would be needed after the recording was made. I can see the advantage, though... but its no biggie for me.
