ultrasonic cleaner, help needed

[jimathan]

Hi all.
New here. Hope some one can help me.
I’ve just rebuilt the eight transducers on my ageing lucas dawe ultrasonic tank due to the ingress of moisture. The amp is made by soniclean, both units date back to the mid/late eighties. All the transducers are now resonating and pulling around 20ma when fed with a low voltage signal from a signal monitor/ generator. (about 1volt. All the transducers had their original labels on showing their khz, (25.8/26.6..They are all resonating after the rebuild within a 450hz spread. The system is working reasonably well after playing with the tuning coils in the amp.
My problem is there are two adjustable tuners in the amp and I don’t understand why. One I guess is to tune the frequency, but whats the other for? Tweaking them both one way or the other increases the power drawn and output in the tank. Going to far saturates them and the power drops of. Backing them off a little restores operation; however I would like to know exactly what I am playing with to be able to get the best out of it. I have not been able to find a drawing for the amp. The only label visible states 25k, and the number on the board is type 6221a. I know it’s a long shot but any help would be appreciated.
Regards Jim

 

[shrtrnd]

I don't know, but I'm interested now too.
Are you sure they're both resistor pots?
Could one be a resistor and one a capacitor?
Could one be an inductor, and the other a capacitor? (for tuning the circuit)
Might be current control through the transducers?
Hope somebody that knows answers your question (so I'll know too)

 

[jimathan]

From what little I know I would say they are both inductors. Apart from a bridge and a handful of diodes, everything else is big and robust. The two tuning devises have 45mm approx steel wheels on top with notches and locking pins to set them in place. From what bit of reading I have done, I recon one of them is to set the feedback behaviour but im not sure. Fortunately, there are no ic’s, the frequency appears to set/generated/controlled by caps.
I really would like to understand a little more about the setup, as I spent four full days dismantling, cleaning refacing and reassembling the transducers. I’m happy that I have it working, and I don’t want to fry it through setting it up incorrectly.
Jim

 

[shrtrnd]

If you can identify how the two trimmers are wired, that would help figure-out their
purpose. (If you can get at the wiring).
I'm wondering if they're wired in parallel, kind of a 'tank' circuit to balance frequency/current flow.
Or if they're wired differently to two separate points in the control circuit.
Can you see any identifying information on the two variable devices? Usually imprinted
on the sides.

 

[jimathan]

Ok thanks. I have the bored out and will try and work out the circuit but im not the best at this. Here is a pic of the board. The feed to the transducers is on the bottom left, just to the right of the mechanical timer. I have done some more reading and it appears that a trimmer should be in the base of one of the transistors to control feedback. the numbers on the transistors are long gone so i need to work out whats what on those.
jim

 

[shrtrnd]

Wish I could get my mits on that board for a closer look.
Can't see everything as well as I would like.
I assume the two devices you're talking about are the (bright in the picture) cogged wheels.
Do the cogs mesh with a micro-switch lever? I was wondering if maybe the devices are
small motors?
Some of the old ultrasonic cleaners had timed two-frequency outputs.
A low freq for a while, switching to a higher freq, back and forth; for more effective cleaning. Cogs to a lever-actuated micro-switch could accomplish this. As the cogs
slowly turned, the lower & higher freqs would be fed to the transducers.
You'll have to excuse me if that's obviously not the case, I'm just trying to get a handle
on what you can see.
I'm just wondering about the cogs. Do they mesh with anything, or they're just there?

 

[jimathan]

Hi The cogs are the rotatable handles that you can turn, after removing the locking peg. Beneath them is the ferrite core. The way I have them set it pulls around 490watts. I don’t know if this is correct or not, but if I tweak them anymore I think they saturate. The power drops to 120watts and the tank goes quiet. There are four distinct test points on the board but I don’t know what to test for on them. The transistors are running similar temperatures’, about 32c. the tinfoil test is quite good, so I guess I should be grateful it works at all after rebuilding the transducers.
regards Jim

 

[shrtrnd]

Sounds like tunable inductors for your transducers.
With no documentation, you're doing what I'd do:
Tune the transducers for what you'd expect to be optimum performance while
ensuring your current flow isn't excessive (excessive heating of the transducers).
They were probably tuned at the factory, and stop pins installed because they wouldn't
have to be adjusted again.
Installing new transducers causes the need to retune the circuit.
Maybe somebody here will give you a better explanation, but my theory seems to
be pretty close to yours.
Good luck with the project.

 

[daddles]

I can't help with the electronics, but in lieu of any other information, I'd assume that the goal would be to turn the controls to get the maximum power output. If you have a way of measuring the power of interest, you can use a designed experiment to help you find the optimal point. Basically, you assume a response of the form

power = b0 + b1*c1 + b2*c2 + b3*c1*c2

where c1 and c2 are the control inputs and the b's are constants. This is a basic linear equation with the important term being b3*c1*c2, which accounts for an interaction of the controls. It is a model often used in experimentation; you can consult design of experiment books for the details in both the designs to use to sample the independent variable space and how to estimate the variances in the predictions (important if you want to make decisions with the model).

If you find the model isn't accurate enough, then the next step can be to fit a quadratic model to the data:

power = b0 + b1*c1 + b2*c2 + b3*c1*c2 + b4*c1^2 + b5*c2^2

You'll need more data, of course, because now you're estimating six parameters. You can use least squares linear regression with these models; it's easy to apply, as the matrix equations are simple. I use some simple python scripts with numpy to do the heavy lifting.

 

[shrtrnd]

Another thing to consider, is the possiblility that the two variable inductors are each
running 1/2 of your transducers. I can't think of a reason why two would be needed
in a single circuit to run all the transducers.
When they built your cleaner, the engineers had already determined the current for
their design. All the assemblers needed to do, was install the transducers and dial-
up the correct current reading to the specs given to them.
Your job, with new transducers, is a little tougher.
You can trace the wiring, just thought I'd mention you may be dealing with two separate
transducer circuits there.

 

[seby]

Please can you tell me which type of transistor use in this amplifier ?
I have model 6223a and stop to work

many thanks