Saturday, 28 April 2012

The Turntable Motor Controller Project - Part 5 - The Project Gets a Name

I was really happy with the result of the prototype PCBs when they arrived, nicely made, awesome silkscreen colour.  So it was time to populate the board and see if it was going to work, fingers crossed.  Here is a block diagram of the system, pretty self evident from the diagram of how it is meant to work (you can click on the images to see them larger).  When I was designing the PCB, I wanted to label the board so I realized I did not have a name other than turntable motor controller, so giving it some thought I decided to call it the Bi-Onda  TT Motor Controller.  Since this is a dual sine wave generator and 'wave' in Italian is onda, I named it Bi-Onda, which if you know Italian, bionda means blonde, as is my darling wife.  Sounds ok too.

Test Bed for the Bi-Onda
In order to hold the whole prototype system together I decided to make a wooden base to hold all the bits and facilitate connections.  You can see in the picture the brackets for the LCD and the button panel I wired up.  While all this took extra time, the advantage of keeping everything neatly together, and easy to modify, is really worth it and keeps the frustration factor down when testing.  

However, it did not keep the frustration totally at bay.  I first populated the PCB with all the parts up to, but not including, the output op-amp stages.  This way I would be able to test the power supply, the signal generator and low pass filters.  This worked like a charm, while the supply rail voltages were a bit higher than I liked, it worked fine and I could get two very nice sine waves at the output of the filters.  The LCD looked good and the debounce components worked well for the tactile buttons I chose.  So I thought, ok, lets put in the power op-amps.

Back to the drawing proto board !
Problem city, the waveforms looked horrible and the op-amps were overheating badly!  Now, I know that the heatsink I made was a bit small, but the over heating seemed way too high.   I thought it was a problem with the supply rails being higher than spec so I ordered a lower voltage transformer and put it in, but that did not resolve the problem.  The issue was getting serious, I had to go back to more basic prototyping to find the problem.  This meant removing the op-amps from the PCB and wiring them on the breadboard to see what I could find.  You can see the picture here of what it looked like when I re-wired the whole thing. The potentiometer flying off the right side of the PCB is for LCD contrast.

So after a bunch of trial and error, here is what I found to be the problems with the system fully put together as per the initial design:
  1. Power supply capacitors need to be of much larger value.
  2. Additional bypass capacitors have to be installed on both inverting and non-inverting inputs of the op-amps to stabilize the signals.
  3. Some minor component value changes and circuit improvements.
So, overall not too many changes, but the whole thing really threw me considering everything looked pretty good as per design on the breadboard version.  Just goes to show you can not be confident until the whole thing is really put together, completely.  So I will look to use a better heat sink in the final unit and I will probably have to design a new PCB to accommodate the changes, but I should be able to make these PCBs work with a few 'flying' components.

Next, hook up both output transformers and try driving the turntable motor!
That will do it for now, 'til next time.