Author Topic: PWM a TRIAC  (Read 5465 times)

0 Members and 1 Guest are viewing this topic.

OperaHouse

  • Hero Member
  • *****
  • Posts: 1308
  • Country: us
PWM a TRIAC
« on: January 03, 2017, 12:58:06 PM »
Skipping pulses to save energy is a similar to an idea that NASA had about 40 years ago and
licensed the patent to something called the Green Machine. This was phase triggered and
designed to be used on refrigerators, fans and anything that had a motor. It would sense when
the load dropped it would delay turn on of each cycle.  I have been meaning to try that with
my refrigerator at camp that is powered by a MSW inverter. It has just a simple TL494 creating
the 60Hz and duty cycle. This model even has a pot to adjust both of these. A simple switch to
reduce duty cycle after a minute would probably save a little power and reduce motor heating.

Not sure of how I feel about missing just a half cycle.  Seems innocuous enough unless it is
going into a decent power toroid. When I did my PWM test well over a year ago I grabbed a cheap
5W 277 to 12V transformer and put it across the test lamp so the scope would be isolated. Wasn't
concerned about the purity of the signal, just wanted to see duty cycle.  The strangest thing
happened. Some AA batteries started rolling about 3 inches on the desk and then back. It was like
haunted. On the scope you could see numerous half cycles of one polarity in a row and the transformer
was becoming a magnet. This transformer had too much resistance for it to be troubled in any way.
A high powered toroid would create some noise, a motor go into braking. This is a serious problem
facing many utilities in their large distribution transformers.

"This residual short term DC can cause significant problems with equipment using large toroidal
transformers in their power supplies. Toroidal power transformers have very low DC resistance on
their inputs and any input DC present will cause very high currents to occur. This high current
can cause saturation in the core during peaks in the AC power cycle.  This saturation causes the
input AC power to see a greatly reduced inductance and as it is going into a virtual dead short
on these peaks, there is very little to impede this peak current flow and it can reach hundreds
of amps."

"DC can traverse the ac power system and add unwanted current to devices already operating at their
rated level.  Overheating and saturation of transformers can be the result of circulating DC currents. 
When a transformer saturates, it not only gets hot, but also is unable to deliver full power to the
load, and the subsequent waveform distortion can create further instability in electronic load
equipment."

OperaHouse

  • Hero Member
  • *****
  • Posts: 1308
  • Country: us
Re: PWM a TRIAC
« Reply #1 on: January 03, 2017, 01:03:15 PM »
I repeated my tests again using using a SSR. This is a triac with a zero crossing MOC3041 opto
isolator. These are cheap and it is easy to make your own SSR with triacs or dual back to back SCR
for even higher power. Check the data sheet for schematics and coponent values. This is similar to
what you would find inside a power block. I used UNO at the the 30 and a fraction Hertz PWM to create
the test images. This gave the best proportional results of all the frequencies.



The first scope image has a string of well ordered single cycle bursts. The bottom green image is
the PWM signal that turns the triac on. You can clearly see that the triac waits for the zero crossing
of the sine wave before turning on.  Notice that trailing edge of the PWM keeps getting closer to
the zero crossing. If the next cycle was visible, only a half cycle would be seen. It tool a lot of
tries to capture this and the other images. The lamp is visibly pulsing at about a one second rate
in all the images.

The second scope image at a different duty cycle has groups of positive and negative half cycles
separated by a long off period. The triggering signal on the bottom is the PWM. It is trying to
trigger the triac on but misses the zero crossing quite often. There doesn't appear to be any
correlation between the PWM duty cycle and the triggering of the triac.  Again the lamp is pulsing.
If this is powering a resistive element with a long thermal time, there is a rough proportionality
of the duty cycle to the average power that could be used.

Next is a higher PWM duty cycle representing more power. The trigger pulses have been removed since
these would just be a blur. Notice again, there are groups of just plosive and negative half cycles.
Around the zero voltage there is an imposed sine wave.  Two light vertical blue lines on the left
record this as being 1.2Hz. The transformer serves to filter this saturation into a more visible
sine wave. This AC voltage could not be used on anything inductive with low resistance.

I couldn't find a random triggered SSR at the time of this test.  Those images would be even more
interesting.

DamonHD

  • Administrator
  • Super Hero Member Plus
  • *****
  • Posts: 4125
  • Country: gb
    • Earth Notes
Re: PWM a TRIAC
« Reply #2 on: January 03, 2017, 01:20:34 PM »
I found that some of my devices/appliances were very tolerant of half-cycle dropping, and some so intolerant that I suspect they may not even meet EU standards for resistance to flicker problems.

In my case I was trying to avoid any obvious bias (for example) by randomising the dropped half-cycles, to a degree.

Rgds

Damon
Podcast: https://www.earth.org.uk/SECTION_podcast.html

@DamonHD@mastodon.social

Ungrounded Lightning Rod

  • SuperHero Member
  • ******
  • Posts: 2865
Re: PWM a TRIAC
« Reply #3 on: January 07, 2017, 05:46:06 PM »
Half-cycling on the grid can be a problem for other devices than the one being half-cycled.  The parasitic resistance in the pole-pig transformer and the wiring from it means you can end up applying a nontrivial DC component to the input power of other loads.

This DC component can result in a substantial current in the input winding of any transformer-powered device on the same circuit - or even the same drop.  That, in turn, can result in the transformer saturating.  Then its inductance drops a couple orders of magnitude and it pulls massive current until a fuse/circuit breaker blows (or it burns up).

If it's fused in the device, with a non-socketed protective fuse (as was typical for such devices, such as older hi-fi amps), it's "dead" until it's taken to a shop and the fuse replaced, at substantial expense.  Then, when it's plugged back into the circuit near the half-phasing appliance it goes out again.

(You're unlikely to saturate the Pole Pig and explode it.  Devices that could do that, like big motor controllers, are carefully designed to avoid exactly these issues.)

DamonHD

  • Administrator
  • Super Hero Member Plus
  • *****
  • Posts: 4125
  • Country: gb
    • Earth Notes
Re: PWM a TRIAC
« Reply #4 on: January 08, 2017, 02:23:09 AM »
@ULR: thanks for that observation: I've added it to my notes as a caution.
Podcast: https://www.earth.org.uk/SECTION_podcast.html

@DamonHD@mastodon.social

OperaHouse

  • Hero Member
  • *****
  • Posts: 1308
  • Country: us
Re: PWM a TRIAC
« Reply #5 on: January 08, 2017, 09:25:26 AM »
An interesting example demonstrating this core saturation is the magnetic amplifier.  You can easily control more than a 100W of AC with just two transformers and and a 9V radio battery.

Rainwulf

  • Newbie
  • *
  • Posts: 20
  • Country: au
Re: PWM a TRIAC
« Reply #6 on: February 05, 2018, 03:38:52 AM »
Be extremely careful with chopping of waveforms/dimming what's basically a very large inductor.  Its vital to have the cuts symmetrical, and even then, you run the risk of nasty high voltage/current artefacts. Its nearly worth investigating your own home made VFD for the slowing of induction motors.  Its something im looking into myself, i want to use a STM32 in high speed PWM mode to synthesise a variable frequency waveform for the purposes of making my own inverter/vfd for driving a pool pump with my excess solar power. With my 3kw, (soon to be 3.75kw) group of panels, i will have excess power i wish to use to either run a fridge or the pool pump.

The other important issue is with the slowing down of non locked induction motors that have slip, so that pretty well includes any single phase induction motor. In the the case of fridge/freezers, the RPM of the compressor is vital to lubrication. They use a centrifugal oiling system, with a little bent pickup pipe sitting into the oil pool at the bottom. As the rotor spins, it throws the oil up inside the pipe which forces it outwards and then upwards throughout the motor. Slowing down the RPM which will happen if you reduce the input power as the inductive slip will increase will potentionally drop the RPM below which the oil system works.

This kills the motor.