Vacuum Record Cleaner Rreplacement

@ljgerens

To compare these static exposure times to a rotating spindle, the time has to be multiplied by 3 because 1/3 of the record is in the bath solution at any given time

Thanks for your great experimental work!

I do query your rule-of-thumb factor of three.  I set my ultrasonic bath level so the bath is a bit below the level of the label.  Records are vertical in the bath, so the inner part of the grooves get less than a third of a turn in the bath, depending on where the actual recording stops.

Do you have any observations showing the influence of cleaning temperature?

I use both a VPI and a Humminguru. Humminguru 1st so all my drying is done on the VPI. Since the Humminguru uses so little fluid I rerouted the discharge to a tube and dump it into the VPI's waste jug and use a new US fluid batch every time. I found a marked measuring cup a better fill vessel than the Humminguru's container.

@thom_at_galibier_design 

Thom said "This has always been a concern of mine and I never bothered to perform experiments as you have done.  I’m not sure what criteria I’d use to assess the damage.  Is visual (microscope) inspection sufficient?  Can you determine whether you’ve inflicted minor damage (not visible) but which will produce damage over multiple cleanings?"

For these experiments two techniques, Optical Microscopy and Photoelectron Spectroscopy, were the primary techniques used to analyze samples for every cleaning condition before and after cleaning. Optical microscopy was primarily useful for gross contamination analysis such as large particles. No evidence of surface damage was ever observed with optical microscopy. Optical microscopy does not have the resolution to detect surface damage at the sub-micron level. Any evidence of such damage would require a more extensive use of Scanning Electron Microscopy (SEM). The SEM was not in my laboratory and does require extra sample preparation time so was only used on select samples. 

Photoelectron spectroscopy provided a quantitative analysis of the contaminants on the record surface. A number of additional supporting techniques were also used to identify specific contaminants on the record test samples.

Thom said "This raises a follow-up experiment:  multiple cleanings which are below the threshold of visible damage.  Would multiple 3-minute (static in your example, above) cleanings produce damage?  Can you trust your inspection with 100% confidence?"

From experiments on other polymeric materials, multiple short duration exposure to ultrasonic is less destructive compared to one long exposure. There are localized thermal effects that increase over time that can induce chain-scission and damage to a polymeric material. So short exposures would minimize such effects, This is one reason that a rotating record would lessen these thermal effects because part of the record is out of the bath at any given time. 

Yes, I trust our analysis 100%

@richardbrand 

I do query your rule-of-thumb factor of three.  I set my ultrasonic bath level so the bath is a bit below the level of the label.  Records are vertical in the bath, so the inner part of the grooves get less than a third of a turn in the bath, depending on where the actual recording stops.

You are correct. The factor of three is just an approximation.

Do you have any observations showing the influence of cleaning temperature?

Higher temperatures do improve the ultrasonic cleaning efficiency for some contaminates. However because vinyl records can begin to warp at ~45 degrees C, it is recommended to keep the bath temperature below 40 degrees C. I typically operated my ultrasonic bath at ~35 degrees C. 

Instead of relying upon microscopic or submicroscopic examination to determine whether or what amount of US can damage the information in the grooves, I would look at the frequency response before and after.  To do that, one would require a test LP with a series of frequencies encoded in separate bands, say 100Hz, 500Hz, 1000Hz, 5kHz, and 10kHz. Take two copies of such a test LP and test them for response at the indicated frequencies, using an oscilloscope or other sensitive instrument.  You could normalize to 1000Hz; in other words, set the 1000Hz response to 0db.  Do this for both LPs in advance, so you know about any inherent differences in amplitude of the responses between the two LPs in advance of any treatment. Then take one LP and clean it with US.  Then test both LPs for frequency response again, using the untreated LP as a control.  This could get very complex, as one would want to test several different US conditions, including an extreme case that you know in advance is likely to do damage.  If such a test LP exists, that would be an interesting study. (In practice, one would probably think of more internal controls, but I think I have the main ones here.)