Group Actions
Engagement
  • Author Author: jw0752
  • Date Created: Date Created
  • Views 2659 views
  • Likes 12 likes
  • Comments 23 comments
Related
Recommended

Incorporating a PWM Control Module into a New Test Device

jw0752

Several weeks ago I purchased a couple of Modules that were advertised as able to provide 1 Hz to 150 kHz signal with a 0% to 100% duty cycle control.

 

image

 

The modules were able to be powered with 3 volts to 30 volts and would output a signal of this same magnitude. The output was rated at 5mA to 30mA depending on the supply voltage. The accuracy of the device at 2% for the Frequency and 1% for the Duty Cycle is acceptable for the type of experiments that I do in my lab.

 

I played with the modules for a while and then I didn’t think too much more of them until a couple days ago when I needed to run some experiments on a few servo motors. I found the module easy to work with and helpful in my experiment so I decided to see if I could design it into a new piece of lab equipment for the shop.

 

After some thought I decided to build a variable DC power supply very similar to the ones I have put together before. The power supply would provide the foundation for the new piece of equipment and the PWM module would provide new PWM features for special applications and the convenience of having the module hard mounted and powered.

 

I wanted to have the following outputs on the device:

 

  • DC 0 to 27 volts from the power supply

 

  • 5 volt signal output from the PWM module

 

  • A MOSFET controlled output that would link to the signal from the PWM module and allow a Load attached to the DC output of the power supply to be pulled to ground for a low impedance inverted output.

 

  • An earth ground / chassis ground terminal which is useful in some applications.

 

A block diagram of the proposed build was drawn out.

 

image

 

A schematic was also roughed out but rather than post the preliminary schematic, here is the finished schematic. I have put the power supply board, the volt amp meter, and the PWM module in as black boxes.

 

image

 

Fortunately I already had a suitable enclosure for the unit with a front panel that had been cut for a dual bench power supply. The meter hole in the front panel for the right side would need to be enlarged for the PWM module but otherwise only minor modifications would be needed to the front panel. Here is what the inside of the chassis looked like at the beginning of the build.

 

image

 

And here is the preliminary front panel from the back side:

 

image

 

When the unit was ready for testing I could not get the power supply to work properly. It seems I had switched the plugs for the voltage control and the current control. It was also noted that higher power output using the MOSFET caused the V/A meter’s screen to flash and not display the voltage of the DC output. A couple of 1000uF capacitors solved this problem. One was placed across the DC output and the other was put on the power supply to the meter module. When I wire a build I always plan for routing and cabling of the wire after the unit is tested and debugged.

 

image

 

Here are pictures of the finished unit:

 

 

image

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

image

 

image

 

image

 

Since my scope probe is between ground and the drain of the MOSFET the duty cycle is inverted. You can see the unit is showing 12.3 Volts at 50 Hz and 16.6% Duty cycle while the scope is telling us that we have 12.8 volts at 49.9 Hz and a calculated duty cycle of 16.9%. These numbers are in good agreement for the level of precision that I need. Sorry about the reflections on the oscilloscope screen. I am still too lazy to get a screen print and transfer it. Fortunately I usually work with my shirt on.

 

The simplicity of the overall project, the modules, and the fact that the case was all prepped made this build go very quickly with a start to finish time of two days.

 

I will be getting in some more servo motors for testing later this week and I am looking forward to giving the unit a proper test.

 

If you go looking for the PWM Modules on line be aware that they come in two versions. One version has a push button that allows the output to be turned on and off. The other version is a dollar or so cheaper but lacks this feature.

 

Thanks for reading my blog.

 

John

  • Since I have been having quite a bit of fun with this PWM Generator I decided to make a simpler version with just a 5 volt power supply and the module as well as output binding posts. This will give me the ability to test small servos and to have a second more portable unit to test with.

     

    Here are some pictures of the build:

     

    image

     

    The housing is a plastic box that I got at Walmart for a couple dollars.

     

    image

     

    Here the unit is set to 50Hz and 7.2% Duty cycle which puts the servo at the 90 degree position.

     

    image

     

    Side View. I did as good a job as possible on the strain relief considering the thin plastic case. The metal washer inside is glued to the box and spreads out the strain. always pointed it out when my strain reliefs were not good enough. Thanks to his prodding I am much more conscious of this safety concern now.

     

    image

     

    The power supply for the unit is just a board from a 1 amp USB power supply. I have removed it from its wall wart case and mounted it to the bottom with a standoff.  The white post in the middle of the case is mounted to the bottom and provides extra support for the control module. It is necessary to push the control knob in order to cycle through the functions of the module and without the support pillar the top of the case would be flexed too much.

     

    image

     

    Most of the modules that are available like this PWM module come with plastic fingers on either side that ramp into place and hold the modules into the front panel. These have never worked well for me as it is often difficult to get them to compress enough to fit through the panels. Recently I broke a front panel pushing one into place and subsequently I have begun to remove the fingers and to just glue the modules into the panels.

     

    This unit is simple to build and does not require any expensive parts. Between the box, the USB power supply, the PWM module and the 3 binding post I believe that I have around $15 in this build.

     

    John

  • in reply to jw0752

    Hi John,

     

    Given that the devices you are familiar with, and are currently testing, have been driven with a single Darlington BJT, then it sounds like your instrument is designed correctly for this application. (Well done!).  Magnetostriction seems to be described as expansion and contraction in the presence and absence of a magnetic field, so this seems to reinforce the on/off nature of the driver (like a single transistor would provide).

     

    Good work!

     

    Gene

  • in reply to genebren

    Hi Gene,

     

    I do not pretend to know the more detailed engineering for the cavitron devices that I work on. Of the 20 or so different machines that are currently in use by the dentists perhaps 20% use a piezo element to drive the tip and the rest are magnetostrictive in nature. Of the magnetostrictive devices I know of none that use an H bridge. All are driven with a single Darlington BJT. I can't speak to the piezo devices as I see them infrequently and the cause of their failure is usually the cord between the machine and the handpiece. These machine's design technology goes back to Tube (Valve) electronics technology and it is entirely possible that no one has ever considered the advantage of H-Bridge technology but rather have just blindly followed the designs of previous units. In any case the idea of an H Bridge output on my next iteration of this piece of the PWM Power Supply is worth looking into even if I do not have an application immediately in mind.

     

    John

  • in reply to jc2048

    Jon,

     

    This is  a very good point.  Depending on the element that is being driven it may be necessary to have a true push-pull driver.  I was working on building a driver for a piezo device, when it became clear that the device needed current in both direction to cause a full cycle.  I had initially designed a single ended driver (open drain), which supplied power through the device while energized (similar to John's setup).  Even though the voltage and current specifications of the device were met, the device failed to generate the desired forces.  I then redesigned the driver circuit to use an H-bridge (I switched over to transistors to limit the switching rate a bit), which did cause the piezo device to operate correctly.  My knowledge of piezo devices is limited, but it appears to do with whether the device has single or multiple rest states, and/or whether there is sufficient forces applied to the device to return it to a given rest state.

     

    It sure seem that the cavitron device was working (sounds like the same scary noises that I hear while in the dentist chair), so maybe the single ended drive is correct for this device.

     

    Gene

  • in reply to jc2048

    Hi Jon,

     

    No I hadn't considered it but it is an excellent idea. I will investigate it.

     

    John