Author Topic: Problems with triac switching a toroidal tranformer  (Read 3898 times)

0 Members and 1 Guest are viewing this topic.

Offline SebiTNTTopic starter

  • Newbie
  • Posts: 8
  • Country: de
Problems with triac switching a toroidal tranformer
« on: December 25, 2018, 05:34:21 pm »
Merry Christmas everyone!

I have a BTA100-800 triac (actually it is a Chinese spot welding controller) which is used to switch a toroidal transformer on for a few cycles. My input voltage is 230 V. If everything is well, I get a nice sine wave output from the transformer with a loaded (almost shorted) secondary winding. My problem is, that (with nothing changed) for most of the time the output looks like crap (like on the second picture) and I have no clue what is causing it.
There is a snubber circuit already connected over the triac consisting of 10 nF cap and about 47 Ohm resistor.
Additionally I have connected one 1 uF capacitor in parallel to the transformer, because at first I had problems with the triac not shutting off the current under load (probably because of a too high di/dt). This problem has gone after connecting the 1 uF cap but the problem with often getting crap on the transformer's output is persistent.
It would be great if someone could give me a hint. Thank you all in advance!

Sebi

Edit: (hopefully :-) removed typos
« Last Edit: December 25, 2018, 09:08:48 pm by SebiTNT »
 

Offline Benta

  • Super Contributor
  • ***
  • Posts: 6420
  • Country: de
Re: Problems with triac switching a toroidal tranformer
« Reply #1 on: December 25, 2018, 07:20:36 pm »
What you probably have is a toroid in deep saturation. This gives unpredictable output voltages when loaded.
The trick is to use the Triac to switch on the primary at maximum mains voltage, or just after. It would be interesting to see your input current, not just the voltage, and also the unloaded output voltage.
.
 

Offline Gyro

  • Super Contributor
  • ***
  • Posts: 10172
  • Country: gb
Re: Problems with triac switching a toroidal tranformer
« Reply #2 on: December 25, 2018, 08:36:47 pm »
Yes, that looks like saturation from switching on at the wrong part of the cycle. Switch on at mains peak as Benta says... and also switch an equal number of positive and negative half cycles. You're going to need to add intelligent mains cycle / zero crossing detection to make it reliable I'm afraid.
Best Regards, Chris
 

Offline not1xor1

  • Frequent Contributor
  • **
  • Posts: 716
  • Country: it
Re: Problems with triac switching a toroidal tranformer
« Reply #3 on: December 26, 2018, 07:56:54 am »
Yes, that looks like saturation from switching on at the wrong part of the cycle. Switch on at mains peak as Benta says... and also switch an equal number of positive and negative half cycles. You're going to need to add intelligent mains cycle / zero crossing detection to make it reliable I'm afraid.

the OP could use this nice trick to switch the triac on close to the peak of the AC voltage

(image originally posted by BravoV)

 

Offline SebiTNTTopic starter

  • Newbie
  • Posts: 8
  • Country: de
Re: Problems with triac switching a toroidal tranformer
« Reply #4 on: January 08, 2019, 04:29:22 pm »
I finally had the time to make the requested input current measurements. Please find these attached.
Is this really a saturation related issue? Sometimes I'm getting a "normal" sinewave output and sometimes not. Wont't the transformer be saturated every time?
Edit: For me it looks like something is rectifying the input current.
« Last Edit: January 08, 2019, 04:31:10 pm by SebiTNT »
 

Offline Benta

  • Super Contributor
  • ***
  • Posts: 6420
  • Country: de
Re: Problems with triac switching a toroidal tranformer
« Reply #5 on: January 08, 2019, 04:52:41 pm »
Your traces show uncontrolled switch-on behaviour of a transformer.
The first trace is perfect, you've hit the right spot, which is appr. peak AC voltage.
The second trace shows core saturation, because the switch-on point is at appr. zero AC voltage. This brings a current excursion to only one side of the AC waveform, which saturates the core, and it just goes on. Recovery takes quite some time and will not occur in just a few cycles.

Very educational traces, can I use them for other (non-commercial) purposes? Thanks.
 

Offline Zero999

  • Super Contributor
  • ***
  • Posts: 20357
  • Country: gb
  • 0999
Re: Problems with triac switching a toroidal tranformer
« Reply #6 on: January 08, 2019, 05:08:30 pm »
Your traces show uncontrolled switch-on behaviour of a transformer.
The first trace is perfect, you've hit the right spot, which is appr. peak AC voltage.
The second trace shows core saturation, because the switch-on point is at appr. zero AC voltage. This brings a current excursion to only one side of the AC waveform, which saturates the core, and it just goes on. Recovery takes quite some time and will not occur in just a few cycles.

Very educational traces, can I use them for other (non-commercial) purposes? Thanks.

Yes, nice traces. I expect it's fine to use them anywhere in this forum, as long as you give credit and link back to the post. I'd just use [img] tags and a link. If someone doesn't want anything they post to be shared, then they shouldn't post it.
 

Offline rstofer

  • Super Contributor
  • ***
  • Posts: 9963
  • Country: us
Re: Problems with triac switching a toroidal tranformer
« Reply #7 on: January 08, 2019, 05:15:21 pm »
Notice the drift downward in the last photo.  The base of the waveform starts at 0V and drifts down with time.  Interesting.

If I were doing this, I would consider the diac approach and discard it in favor of using a uC to get an interrupt at zero crossing and then, 16.7 ms later I would trigger the triac.  20 ms in the other parts of the world.

I suppose this could be done with various multivibrators but I think I would still use the zero crossing + delay approach.

The reason for switching at the peak is that the rate of change of voltage is at a minimum.  The worst case is to switch at zero crossing where the rate of change is at a maximum.

The math reason:  The rate of change of sin(t) is cos(t) and cos(t) is a maximum where sin(t) goes through 0 volts.

I would think that de-energizing the transformer should also occur at a voltage peak because this, again, is the low point in current.  More important, this is the zero flux point and that's what we want.

https://www.allaboutcircuits.com/textbook/alternating-current/chpt-9/practical-considerations-transformers/

Since I would be using a digital <something> in the control system, adding zero crossing + delay is inexpensive.

Here is one approach to detecting zero crossing with a uC, there are many others:

http://microcontrollerslab.com/zero-crossing-detector-circuit-using-pic-microcontroller/

 

Offline Benta

  • Super Contributor
  • ***
  • Posts: 6420
  • Country: de
Re: Problems with triac switching a toroidal tranformer
« Reply #8 on: January 08, 2019, 05:34:06 pm »
Notice the drift downward in the last photo.  The base of the waveform starts at 0V and drifts down with time.  Interesting.

Yes, this is the transformer starting to recover from saturation. It'll probably take several hundred cycles before it's back on track.
 

Offline SebiTNTTopic starter

  • Newbie
  • Posts: 8
  • Country: de
Re: Problems with triac switching a toroidal tranformer
« Reply #9 on: January 08, 2019, 05:49:52 pm »
Everyone can use my pictures under the CC BY-NC license. So you are of course free to use them. Thanks for asking!

Thank you all for your answers! So I think there is some inconsistent fault in my circuit. I will investigate this.
 

Offline Benta

  • Super Contributor
  • ***
  • Posts: 6420
  • Country: de
Re: Problems with triac switching a toroidal tranformer
« Reply #10 on: January 08, 2019, 05:59:00 pm »
As we don't know your circuit, there's not a lot we can say.
But the target is to fire the Triac at UMAX, take it from there.
 

Offline rstofer

  • Super Contributor
  • ***
  • Posts: 9963
  • Country: us
Re: Problems with triac switching a toroidal tranformer
« Reply #11 on: January 08, 2019, 06:09:30 pm »
Everyone can use my pictures under the CC BY-NC license. So you are of course free to use them. Thanks for asking!

Thank you all for your answers! So I think there is some inconsistent fault in my circuit. I will investigate this.

It's the residual magnetic flux, not the circuit.  Well, except for the part where the transformer isn't firing at a particularly advantageous point in the voltage waveform.
 

Offline BravoV

  • Super Contributor
  • ***
  • Posts: 7549
  • Country: 00
  • +++ ATH1
Re: Problems with triac switching a toroidal tranformer
« Reply #12 on: January 08, 2019, 06:19:25 pm »
So I think there is some inconsistent fault in my circuit. I will investigate this.

Don't think its a fault, as that is normal for inductive load, like transformer, motor, vacuum cleaner and etc.

I know it sounds funny, its matter of luck, whether you switched on at the moment the mains line was at peak AC voltage, or at near zero, that will make huge difference, its like tossing a coin, literally.

Watch my thread as pointed by not1xor1 above, soon I will update my little project there, with details measurements (voltage and current) with/without softstarter, and also what happened when I switched a inductance load at peak or zero voltage, hell, this is also matter of luck if my scope can single shot capture at the right exact moment as yours, thru switching on and off like crazy perhaps.  :-DD

Suggesting to read this -> Beware of Zero-Crossover Switching of Transformers

Your scope shots below is very similar to Figure 4.A at above document.


Offline Benta

  • Super Contributor
  • ***
  • Posts: 6420
  • Country: de
Re: Problems with triac switching a toroidal tranformer
« Reply #13 on: January 08, 2019, 10:07:28 pm »
I hereby retract all my previous comments!
A closer study of the AC current traces provided do NOT look like saturation.
It looks much more like an asymmetric current draw on the secondary of the transformer. What exactly it is, is unknown at this time, and why it's intermittent is also a mystery.

More info is needed.
 

Offline rstofer

  • Super Contributor
  • ***
  • Posts: 9963
  • Country: us
Re: Problems with triac switching a toroidal tranformer
« Reply #14 on: January 08, 2019, 10:48:24 pm »
The triac will turn off at current zero crossing which will be very close to the maximum voltage.  For 60 Hz, I would turn the triac on at voltage zero crossing + 16 ms but to turn it off, I would probably do that at voltage zero crossing knowing the triac would turn off a few ms later at current zero crossing.

As to scoping the signals, if I were doing this digitally, I could bring out any trigger I wanted including zero crossing.

 

Offline Sylvi

  • Regular Contributor
  • *
  • Posts: 85
  • Country: eg
Re: Problems with triac switching a toroidal tranformer
« Reply #15 on: January 08, 2019, 11:02:25 pm »
Hi

I did not read the whole thread but I think the issue might be fixed easily witha DC blocker. large PTs especially toroidals can become satruated with no load because of asymmetric noise on the mains. Big PTs will growl.

The DC blocker is simply antiparallel high-C electros in parallel with a high-current bridge wired as a bilateral cvoltage clamp. The latter is a cheap way to have antiparallel diodes of high current rating in an easy to deal with package. Do not worry about the electros having up to 1V reverse across them in this application. The higher the PT VA the higher total C required.

There is no need to wire the caps in reverse series in this application. That is unneccesary due to the voltage present, and it cuts the net capacitance in half.

Have fun
 

Offline SebiTNTTopic starter

  • Newbie
  • Posts: 8
  • Country: de
Re: Problems with triac switching a toroidal tranformer
« Reply #16 on: January 09, 2019, 11:13:26 am »
Thank you again for your answers!

I also really don't think that this is a saturation related issue, because it appears really random and not only when I fire it up with short delays. Please see the video I just made under https://youtu.be/gJdt6XoCtFI
The blue curve is the input voltage and the yellow one is the input current (as before). Please note that the issue only occurs if the triac switches on at the falling edge of the input voltage. Might this be the problem? Of course (as you told me before) one should switch on at Vmax but is this really causing my problem? I am using a circuit for a spot welder (please see the attached picture) and I really don't know how to modify the time when it is switching on.
 

Offline Circlotron

  • Super Contributor
  • ***
  • Posts: 3362
  • Country: au
Re: Problems with triac switching a toroidal tranformer
« Reply #17 on: January 09, 2019, 11:56:01 am »
I'm thinking maybe the triac gate gets a short pulse every zero crossing of the voltage and seeing the current lags the voltage, the gate pulse should extend from the voltage zero crossing to a little after the current zero crossing. Or better yet, plain continuous DC applied to the gate for as long as the transformer needs to be on.

Interesting also that the fault is only one polarity, so maybe the triac lacks gate sensitivity in one quadrant?

Or maybe the MOC3021 opto is sus?
« Last Edit: January 09, 2019, 11:57:58 am by Circlotron »
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf