Thumbs up for ultrasonic record cleaning

I have about 25,000 LPs.  I've noticed that the LPs I cleaned 20 to 30 years ago using a stiff brush and Torumat fluid on a VPI 16.5 sound as good and clean as when I first cleaned them.  I'm playing them on a VPI VI/SME IV (mod)/Benz Ruby 3.  No increase in surface noise even after 100 playings of several of them.  Clean stylus and clean records make both last a long time.

I've been experimenting with USC in the past few month, having cleaned about 30-40 2nd hand classical Lp's so far. I've added USC as a second stage after my regular Okki Nokki vacuum cleaning.

I must say it has not been so far such an epiphany to me as it seems to others, rather incremental. The biggest disappointment is that it does not seem to reduce clicks&pops, sometimes making them even more pronounced, probably due to quieting the background. I'm quite sensitive to clicks as I'm listening on high efficient Stax headphones, so all is in my ears.

Whats it does seem to do, provided I use right chemistry (see below), is  increase of micro and macro dynamics, and detail retrieval (so perhaps all the clicks I hear are due to permanent vinyl damage). On some LP's the sound becomes more direct which is a big plus for me.

Here is the hardware and procedures I use:

GT Sonic 6L 150W 40kHz machine + 1um filter+ 1rpm motor for rotation. I clean at most 3 Lp's at the time as evenly spaced as possible. I can clearly see the standing waves between the vinyls.
Currently I'm using 5% IPA + 0.05% Tergitol S7, which is available in Poland albeit v expensive. This Tergitol concentration is recommended by the archivists and the solution is already foaming. I was afraid to  use the higher 0.13% concentration as recommended elsewhere after my failure with PhotoFlo. I used it instead of the Tergitol and it was giving me unpleasantly "plasticky" sound despite a mandatory 2 step rinse on the vacuum machine. Had to US rinse all vinyls I cleaned with it and clean all that had contact, incl. the tank, the filter, brushes, etc.

The procedure is:
*1 step vacuum: Apply the above solution, let it soak for 5mins, biderectional scrub, vacuum.
*USC: 60% power 40kHz, 32-33C, 15mins (first degass for 10mins, filter before the batches)
*2 step vac rinse: First high purity DI water then 3% ethanol

I've once heard opinion that alcohol dries vinyl surface and one can never get rid of pops&clicks. I tried just DI water USC too but no real change and Tergitol softens the cavitation action at 40kHz, lowering the surface tension. I'm thinking of investing in a second 80kHz tank to see if the more refined cavitation would improve. Perhaps my initial vac + scrub is quite efficient in removing what's easy to remove.

Cheers,
b

As a follow up, directions I'm considering:

*80kHz tank since I use it as the 2nd stage and 80kHz process is more penetrating but gentler
*slower motor; anyone knows of a cheap 1/3rpm or so motors in EU? 230V synchro would be the best
*submicron filter - are carbon filters good? would they not filter out also the chemical solution out of the water?

I’ve been following this thread for some time now, very interesting reading and thanks to all the contributors.

Some observations:
I owned one of the original AD US cleaners and was not impressed with the results or the build quality. I replaced it with a KLAudio KD-CLN-LP200 with (IMO) much better build quality, but still unimpressive results. I’ve looked at the DIY machines and have been reading everything I can get my hands on wrt to US technology. There seems to be a considerable difference between low frequency US cleaners (28kHz) and higher frequency (80-100kHz), the former used primarily as industrial cleaners (car parts, large metal components etc), the latter used for more delicate structures, especially fine jewelry. The cleaning action is created by the collapse of the cavitation bubbles which can create extremely high pressure (10’s of Mpa) and high temps (5000°C), but on a very small scale, determined by the size of the bubbles: The lower the frequency, the larger the bubbles and the more concentrated the released energy. On high power, low frequency US machines, this process can be so aggressive the sides of the stainless steel tank are subject to etching.

I was curious about this, so I did some measurements on my KLAudio RCM which is assumed to be 40kHz. It actually runs at 34.72kHz which would seem to be more aggressive, almost midway between 28 and 40kHz. There was a frequency adjustment pot, and I monitored the AC power being consumed as I adjusted the pot. Increasing the frequency lowered the power consumption and lowering it increased the power; at 33kHz, it went from 200W to almost 350W. The power consumption also changed drastically between running the cleaner with and without a record (200W with vs 150W without) indicating that the load seen by the transducers affect power consumption and the load changes with mass, density and distance from transducers.

So my question is this: With so many variables affecting the performance and operating parameters of the US process (temp, surface tension, power, frequency, cavitation efficiency), how does one know what they are really getting and what they are doing to the record surface (not only whether it is effectively being cleaned, but whether it is being damaged in the process)?