Speaking from experience, I have found:
1 - Ultrasonic
When I first started out (back in 2012) one of the first bits of kit I bought was an ultrasonic cleaner from Maplins (remember them!!). It cost about £70 and had a heater (ON/OFF) and a timer (max 480 seconds).
I used this for 100's of hours and it NEVER failed me.
Last year I upgraded to an Elma ultrasonic (S10 H).
I must say I felt a bit guilty at the time, because this effectively consigned the old one to the bench (in sporting terms).
When I first used the Elma, my guilt subsided immediately. Temperature control from cold to 80 degrees or so. Time control from 1 minute to run forever and a sweep mode that means the 'sweet' spot moves around in the bath.
It is an order of magnitude better. Yes - its expensive, but this is a professional bit of kit that does (and I hope will do) the job for many years.
2 - Cleaning machine
For the first couple of years or so, I used to clean all parts (case, bracelet and most movement parts) in the ultrasonic.
But, I found that the water based cleaning agent I was using could attack certain parts (on the mainplate in particular) - causing tarnishing and sometime serious corrosion.
I also had on a few occasions pinions separating from their respective wheels.
From this point on I stopped cleaning movement parts in the ultrasonic and moved to hand cleaning them.
I found this worked for me - it got me really close to each part, and in doing so I could appreciate each part and ensure that it really was clean. I did this cleaning under a 2nd hand stereo zoom microcope.
I'd wanted to try out a 'proper' cleaning machine for quite some time. A) to see what the difference it would make and B) so I could start to use cleaning solvents.
Last year I bought a
VERY old cleaning machine (from the '60's or earlier) and restored it to life. This allowed me to start using L&R cleaning solvents and also see the difference in this cleaning process over what I had been doing.
Being non water based meant that all parts (bar the balance/hairspring and the pallet forks) could now be cleaned without fear of corrosion.
The baskets meant that I could separate the respective parts, grouping them so screws don't get mixed up etc.
And the heater element meant that after rinsing, the parts could be hot-air dried.
Given how meticulous I was (and still am) with hand cleaning everything, I can't say that the cleaner has dramatically improved things.
However, I'm pleased that I'm now using the L&R cleaning solvents and being able to clean an entire movement in one go is a real bonus. It is quicker, though I still give most parts a wash in Renata before, and a micro clean / inpsect afterwards (Renata again followed by rodico).
Shortly after restoring the vintage cleaning machine, I purchased an Elma cleaning machine (the ElmaSolvex SE).
Visually this is like going from the Cutty Sark to the Starship Enterpise. But, to all intents & purposes it still does the same job.
The are four cleaning jars, I have two with cleaning solvent and two with rinse. It has a timer (with cutoff) and variable speed control. And the hot air drier also has a fan and cutoff.
Its not fully automatic, which means it needs to be attended to cycle through the various stages (clean, spin, clean, spin, rinse, spin, rinse, spin, dry) which all told takes about 25 minutes.
But while this is taking place I'm ususally cleaning the case conponents in the ultrasonic, so its not dead time.
3 - Magnification
I started out with two loupes (1.5 and 3). I surprised myself by being able to keep the loupe in my eye without much trouble - I felt like a 'real' watchmaker!!
But, I found there was always a risk that the loupe might drop out, onto what I was working on.
I found myself to be quite proficient with the loupe, and this allowed me to progress through a variety of procedures from cleaning, hand fitting, movement stripdowns & rebuilds and even onto luming.
At one of the watch fairs, I saw a stereo zoom microscope that had come off an electronics assembly line. It was old and the eye piece optics were scratched, but as soon as I started to use it I realised what a game changer this was going to be.
I was able to see everything (upto x40 magnification) and in particular this allowed me to see the pinions and where they were in relation to the bridge that I was trying to fit.
Up until this point I had been unable to re-fit a 7Axx main bridge, despite numerous attempts (over a nearly 2 year period). A week after getting the microscope and within a matter of minutes I had the bridge back on. The reason I'd been failing was that quite simply I'd been unable to see exactly what was going on (under the bridge) and had been effectively poking around hoping I could tease each pinion into place.
I can't recommend highly enough a stereo zoom microscope for this hobby. For dissasembly, cleaning and inpection, assembly and in particular oiling.
Last year I upgraded my microscope to a Leica S6.
The optical clarity is superb and the amount of light, lack of spherical aberations (at high zoom) makes looking through it a pleasure.
Given how much time I spend with this device I felt it well worth the investment.
4 - Timing & Regulation
I started out working on quartz watches (the 7T34, then the 7Axx, then the 754x, then the H55x and several others).
I liked these because they tended to be binary: working or not working. If they worked then they kept time. If they didn't then I would strip, clean, rebuild and usually end up with it working.
I switched to mechanical watches after buying a non working 7005 based watch (again at one of the watch fairs).
It had a snapped hairspring, which I fixed by swapping in a donor balance. I serviced the whole movement, learning about left handed screws in the process, and was proud when the watch burst into life.
Over the course of the next 9 months my focus switched from quartz to mechanicals and I ended up investing in a Timegrapher (the Weishi 1900).
And I immediately found out the difference between a watch that ticks and one that runs properly!!
I found this an invaluable tool to measure my abilities against. The objective to take a watch that was either running (badly) or not at all, and then improve this - looking for stability of trace and amplitude.
The 1900 comes with a pickup microphone stand that can be orientated in one of 6 positions (by hand). This to measure the isochronism of the movement (how it behaves in each position - to mimic how it will be positioned as it worn).
Last year I invested (long term investment) in a MTG-5000 Timegrapher that can automatically test in the 6 positions meaning I can run a full test automatically without having to keep manually changing the pickup head orientation. It also has a printer so I can print (and keep) the results.
More recently I purchased the Horotec FlashTest for quartz movements.
This is a great bit of kit especially for quartz movements whose rate can be adjusted (such as the 7Axx series). It can also measure coil resistance, battery charge, plus a few other things I've not yet tried.
5 - Staking & Jewelling
Like many hobbyists I'm always searching ebay for 2nd hand watchmakers tools (becase the new ones are so very expensive).
A couple of years ago I picked up a Seitz Jewelling kit.
At the time I didn't have a specific need for it, but I 'knew' that it would prove useful. Within a matter of months I had a need to replace a cracked jewel in a chronograph bridge - after researching how I should use the kit I'd bought I found this very easy.
Since then I've used it numerous times, and use it most frequently to up-jewel the 6139 movements I work on. Reaming out the mainplate and fitting the jewels is a simple process with the equipment designed specifically for this job.
More recently (and yes, at one of the watch fairs) I bought a Boley staking set - the full set including the extra jewelling press (which I don't need as I have the Seitz kit, but its still nice to have).
You might think this wouldn't be necessary owning the Seitz kit, but while there are things they have in common, there are significant differences.
I have used the Boley kit to re-stake a number of parts that without it would have been impossible.
Has it paid for itself yet? No, but it will.
So, there you have it. My first hand experiences gained over 7 years and probably 1000's of hours.