Ultrasonic cleaner

[MotorBikeLes]

I am still waiting for my pal to do the welding on the tank which vibrates, but decided to look again at the other "faulty" tank. (It is about 5 years since I first acquired this equipment, when we tested all the tanks and found two non-workers). The one shown in the photo attached to the first posting is clearly faulty, as mentioned in previous posts. The third tank however I had "left on the shelf" until this weekend. I removed the bottom cover, but everything appeared physically perfect. However, further close examination shows where the linking braid is in fact open circuit or nearly so in a number of places.
The braiding. A small piece was cut off, and found to be constructed from 50 strands of 0.11mm diameter copper, which MAY have been silver plated. (That makes it 0.48mm in modern terminology) It seems to have been resin coated presumably to prevent vibration damage as Ed suggested, but this has probably set hard with time and become brittle. The likely repair solution is to replace ALL the braid on the free (live) end of the transducer blocks. It seems unlikely that the braid on the earthy end will have become open circuit in view of the total resin encapsulation.
I have some ariel coax, and if I pull out the inner, I can slip off the outer. This in fact is 48 strands of 0.14mm copper. Would this do to repair it, or does somebody know of a source of braiding of 50 or more strands of 0.11mm diameter wire.
I do have some PTFE insulated wire (22 awg 19X 0.15mm silver plated I think) and wondered if this might be OK. I would need to double up to get the same cross sectional area.
Two possible routes there, any more advice?
Les.

[Ed_Dinning]

Hi Les, coax braid should be fine, but may be a bit "light"; Farnell and others sell braids in various thicknesses. If well cleaned a dribble of a silicon sealer (not and acetic type) may well act as a damper.

PTFE may be ok, but could be a bit stiff for this application..

Ed

[FrankB]

Be sure to check the cleaning solution to see if it is aluminum compatible also.
The local Harbor Freight Tools sells a few sizes of cleaners for about 100.00 or less (USD).
The instructions on the "special" cleaning medium says it is NOT aluminum compatible.

I just use Dawn brand dishwashing soap, followed by a good rinse in the hottest water I can stand- unless the part is plastic and can heat deform.
Works great on VCR selector switches- add a drop of silicone oil afterwards to re-lube the switch; and on the switch decks in stereos with the multi contact push button switches.

[Radio Wrangler]

You have to be very careful using silicone lubricants near switches in case any creeps onto the contacts.

When I joined HP, one of the things I noticed was that all the components we used had "Must withstand a dishwasher cycle using Calgonite detergent" in their purchase specifications. This was to ensure that flow-soldered PCBs could be hot washed to remove acid flux. We'd never heard of Calgonite on this side of the Atlantic and it sounded like something superman ought to be wary of. Most dishwasher stuff etches aluminium, is "Dawn" something aluminium safe?

David

[Ed_Dinning]

Hi Gents, never mind the solution being aluminium safe, the standard test at Kerry's was to fold a piece of cooking foil in a vertical "V" shape and give it 30 secs in a tank. It should come out full of holes !!

Ed

[MotorBikeLes]

The contemporary thread on Nitrogol capacitors made me think about my ultrasonic cleaner, mentioned here a couple of years ago in a different thread. The comments in the "Viscogol Capacitor" thread suggests both that and the Visconol should be pensioned off. Another similar thread on another list suggests using ex-microwave capacitors. I am appending both a view of the capacitors (and other components, taken from a second unit with S/C mains transformer) and my attempt at its circuit diagram.
Anybody care to suggest replacements for the included capacitors?
Les.

[Radio Wrangler]

That mains transformer might be an interesting design. It seems to be hot with lots of 40 kHz on both secondaries.

The nitrogol capacitor has a voltage rating specified at 50Hz, yet a current rating specified at 40kHz. I had to redesign a 40kHz inverter which had to have a good sinusoidal output. The original was a resonant tank with Philips high-speed SCRs (Actually the type for the Syclops PSU/line output design). When finished the rated output was increased from 240W to 320W and it used IGBTs. But the choice of resonating capacitors was critical. I remember the good ones where made by TRW. The others had a tendency to suddenly glow bright red followed by PCB tracks doing the same thing. Throughput may have been 200-300W but the resonant circulating current required components more suited to a couple of kW.

Your circuit is higher impedance, the voltages are more threatening than the currents, but you may wind up with a bank of parts to replace that capacitor.

David

[MotorBikeLes]

David, the transformers were different between the working and non-working machine. It may have been an early poor design which failed, but then older scope transformers with the heater winding (at CRT cathode voltage) and EHT winding were very common for shorted turn failures. I recall that with all secondaries removed it quickly got hot with big current, so I scrapped it.
I will start looking for old microwaves, and if I find anything (or combination of anythings) that look OK to me, I will report back for further guidance.
Les.

[MotorBikeLes]

Only another two years, and today I got my one working tank back with the three holes welded up.
I connected up, filled with water added a small amount of detergent and hung in a "test" carb. and switched on. Well, there seemed to be a slight fuzzyness in the tank, but nothing dramatic. Adjusting the tuning/coupling control gave a max, but nothing special. I have looked at a few Utube videos, and with some of them, it seems "all hell let loose" when switched on. Typical Utube about 100W, this nominal 500W! It seems a bit of tuning, diagnosis and even repair may be needed. I attached the circuit back in post No.1, and have noted that the output, presently taken from tapping No.10, could also be on 8 or 9. Similarly the coil's primary connections are on (No.0 and) No.4, but it could be No.3. (this is the conclusion from examining this and the other generator which had the S/C mains transformer. Of course I have no idea of the valve's emission. Any in circuit tests for emission I could safely make? Do I just play about with the alternative transformer tappings (Primary 3 or 4, secondary 8,9 or 10)? Do I simply asses on the basis of liquid activity, or do I get out a scope and monitor? I did think about attaching an ultrasonic transducer ( Grundig remote control receiver type) to my little audio signal tracer which has a L/S output, as well as a scoping point and a meter display. It is based on the output of an old mono tape recorder, not intended for U/S, but who knows. With that I could just "listen".
All suggestions welcome, I have two Del'orto carbs which urgently need cleaning. Bike off the road!!
Les.

[MotorBikeLes]

OK, I have been looking at the circuit (not my strong point) and it looks like this:
The 2kV is grounded (via the jumper) at its cold end, the hot end going to the anode of the valve, The cathode is on tag 3 or 4 of the coil, with the other end of the coil (tag 16) being at ground. Does this mean that choosing tag 3 or 4 will give some measure of primary frequency selection (??). The grid circuit is from tag 0 to tag 3/4.
The output is taken from tag 8/9/10 to ground. Presumably this choice is to match the primary frequency into the resonators in the tank?
I assume I monitor anode current by sticking a milli-ammeter in place of the jumper without any especial risk (connected with unit OFF of course). This tells me something about valve condition, but do I play with tag 3/4 to optimise anode current?
I checked all passive devices a couple of years ago, and don't expect any problems from C1 or C3, C4 is trivial, only C2, the visconal being a failure risk and an expensive or unobtanium.
I want to get this thing shifting dirt quickly, but don't want to kill it (or myself) in the process, hence wise advice is sought.
Les.

[MotorBikeLes]

I had a look at this today, armed with my Tek 212 portable scope (200kHz BW, Xs 1 direct probes), and a NOS Bulgin 200mA meter (still boxed, cost 10/6d). I have redrawn the circuit to clarify operation.
I connected the scope between Tag 0 (0v) and Tag 1, the lowest tapping on the output section.
At switch on, I noted a valve current of 78mA, with the 'scope showing the pulsed (mains frequency) ultrasonic waveform of 32vpp, and period of 25+ usecs, corresponding to about 39.9kHz.
If detuned, the current falls by about 10mA, with the U/S frequency falling to around 36kHz (27.5usecs period).
I then moved the cathode connection from Tag 4 to Tag 3 and repeated the measurements. The valve current fell slightly to 73mA and the waveform at Tag 1 to about 30vpp. I returned to connection from Tag 3 to Tag 4.
I had a spare valve, so the Mullard TY-125 was replaced by a clean looking Zaerix valve.
At switch on, an increased current was noted at about 89mA, and the scope now showed 100vpp, a three fold increase, still at 39.9kHz.
Obviously, this is an improvement, and maybe it can be further improved by moving the output pickup point from Tag 10 to either Tag 8 or 9. I will only do this if I monitor the full output voltage, which will require me to take a better scope (HP1703A) and suitable probe (Xs 100 and 5kV rated) to the motor 'bike workshop.
The observed improved output does appear to give more disturbance to the liquid, but still nothing like the violence seen on some Utube videos of 150W commercial cleaners.
Les.

[Ed_Dinning]

Hi Les, one of the standard tests at Kerry's was to use a GDO to check the frequency.
This had 2 uses, if it was not on frequency there was no reading on the GDO. If it was on frequency the tuning C variation would easily peak the GDO.
It sometimes took a while to stabilise the action in the tank if the water was gassy, straight from the tap.

The circuit you posted is very similar to the Kerry early units, in fact the company was bought from MULLARD IN THE LATE 50's.
I believe they still have a service department at their Hitchin offices

Check the transducers individually to see if the bolts have released at all, or if there is a low resistance across the transducers (the PTFE sleaving sometimes wore through depending on the particulate transducer design that was used.

Ed

[MotorBikeLes]

Ed, I will try Kerry later on. I will be fitting the milliameter permanently today, and I think max current equals optimum tuning.
Gassiness. I have not bothered boiling the water, so that will be first job today. I do have a previously not used kettle in the shed. It will be good if that improves things, and a lesson learned.
The info on the front says not to add hot water to a cold tank, and not to pour cold water into a hot tank. Paradox.
If anybody looks at my latest circuit, I just spotted the connection between junc. C2/C3 across to Tag 0 is missing. I will correct that later.
The circuit (transformer marking) suggest 200mA HV current, so what sort of valve current do I expect? Can I increase it (assuming too low) by playing about in the grid circuit?
C2 is the "oily" capacitor. If I suspect any component, it would be this. It is rated at at 4kV DC wkg, could I safely replace with say 5 off 0.05uFd in series?
Cheers, Les.

[David G4EBT]

Obviously there are two aspects which affect the performance of any ultrasonic bath - the frequency, which you've ascertained is near enough 40 kHz, and the amplitude. In your case, given that it's a commercial/industrial cleaner and that there are four transducers, it should have considerable amplitude and whilst there should be ripples on the surface of the water, it won't be anything like boiling water - more like simmering, but there should be an audible fizzing sound too.

Don't know if you've tested the bath to see if it's effective - the standard test is a piece of aluminium cooking foil. You don't need to use an additive to the water to do the test, but I'd be inclined to add a few drops of fairly liquid to break down the surface tension of the water and help cavitation, and to switch the bath of for a few minutes to 'degas' the liquid. If you then suspend a piece of foil - say 7.5cms square in the liquid, with your bath, if it's working correctly, the foil should have holes in it in a minute, and should start to disintegrate within two or three minutes.

I use a small (2L) commercial bath in which the ultrasonic power is 70 Watts like the one at the link below and within three minutes, the foil is full of holes, so yours should easily do likewise. It ripples the water and there's quite a fizz, which rattles the lid if I put it on.

http://www.allendale-ultrasonics.co....allendale.html

Before I bought that bath I'd tried one of the little jobbies from Maplin for cleaning jewellery and the like, which shut off after 3 mins. It had no impact on the foil after several 3 mins cycles, so that went back. Then I was tempted buy one from Lidl - a bit larger, and with an 8 min cycle. It took about four cycles before there was any sign of pinholes in the foil, so it was really no better. That went back too, and I decided to buy a small commercial unit.

Don't know what your bath size is Les, but here's an example of the foil test in a medium sized commercial bath:

https://www.bing.com/videos/search?q...=0&FORM=VDQVAP

This bath is larger:

https://www.bing.com/videos/search?q...=0&FORM=VDFSRV

Incidentally, using foil isn't just a 'DIY' test - it's a requirement for the regular testing of baths used in dental hygiene and other medical applications:

https://www.bing.com/videos/search?q...=0&FORM=VDQVAP

Hope that helps a bit.

Good luck with it Les.

[MotorBikeLes]

Well, an interesting day today. I made the milliameter a permanent fixture, boiled out some water and left it to cool. Early afternoon, I poured the water in and tried the kitchen film test. After one minute, there were holes. Not as pronounced as those David linked to, but promising. Meanwhile, I had boiled a second kettle full of water which I added slowly to the tank, putting a third on to boil. I hung an old carb. body in and set about stripping the first REAL one destined for deep cleaning, adding some more water to get the test carb. fully immersed. I then realised the valve current was down to only 50mA. Due to higher temperature perhaps?
Yesterday I had been thinking about that C2 capacitor in the grid circuit, and decided to check the spare one, getting a value of 108nF. It was only today it clicked that it should be 10nF if marked 0.01uFd, had I read it wrongly? I checked with the better of the testmeters in the shed, and it was 108nF. I wonder what the one in the generator is then? So, cover off and struggle to eventually get it in my hand. A close examination shows the marking says 0.03uFd, not 0.01uFd. Looking again at the spare, I see that probably also is 0.03uFd, the print worn away because of the fixing clip and oil. The unit in the generator measures 48nF. Something wrong. I assume that means it is leaking, what the true capacitance is I don't know. I will get the Wayne Kerr out maybe tomorrow, and see what that says. Meanwhile however, it obviously needs replacement, but I don't have the correct value. I do have a 0.22uFd at 6kV, and three new 0.033caps rated 1500v and marked 21kHz. I could probably series those three in parallel with the 0.022uFd. Not sure where the 6kV rated cap came from, but the others are NOS stuff intended for line output/scan for colour TVs. Comments?
Or does anybody have a decent 0.03uFd cap rated at 4kV DC wkg.?
Will the valve current be higher tomorrow when the tank and water are cold?
I am attaching to corrected circuit with a couple of minor additions.
Les

[MotorBikeLes]

I went out this morning and switched on to see if the now cold tank would behave differently. valve current rose to about 160mA, but when tuned in, just in excess of 200mA. That looks more like it.
Came back this afternoon and left it with the first of the priority carbs hanging in the tank. The same high tuned current, but which fairly quickly fell off to about 160mA, needing retuning for maximum. There was a continued slow reduction in current, and after one hour it was down to 80mA. The lower the current, the less sharp the tuning. Both carbs cleaned, new gasket sets installed and fitted to bike. Raining heavily, so shed doors wedged open and bike started. No time for gauges to correctly balance, but it looks promising. Rain or not, I will be out on it tomorrow, then set correctly later.
I will be replacing the grid circuit capacitor once I locate one, but it seems as if I now have a working ultrasonic generator / cleaner of about 1 gallon capacity.
Les.