Author Topic: modified sine wave bad for smps ?  (Read 13364 times)

0 Members and 2 Guests are viewing this topic.

Offline kalingaTopic starter

  • Contributor
  • Posts: 13
modified sine wave bad for smps ?
« on: March 19, 2016, 04:09:41 pm »
Hey Guys,

We have a backup power system as home that runs on a modified sine wave inverter, basically used for Fans and lights when the power goes out. but there are lot of people say you can't run a TV (New LED/LCD or anything with a SMPS in it), even if you do it would damage it. but all new TVs / Computers use SMPS and it basically convert the modified sine wave/AC to DC before doing anything and world runs on UPS either square waves or modified sine wave in most cases. so this basically puzzles me, can it actually damage it ?

so is this myth or a truth ? is modified sine wave is bad for smps ?
 

Offline Seekonk

  • Super Contributor
  • ***
  • Posts: 1938
  • Country: us
Re: modified sine wave bad for smps ?
« Reply #1 on: March 19, 2016, 04:17:16 pm »
I live half the year off grid.  Those SMPS supplies that don't voltage double I run on 140V DC from broken inverters, a little more efficient.  I have never had issues with MSW or heard of anyone else with SMPS.  There are reports of problems with capacitive dropping supplies.
 

Offline rx8pilot

  • Super Contributor
  • ***
  • Posts: 3634
  • Country: us
  • If you want more money, be more valuable.
Re: modified sine wave bad for smps ?
« Reply #2 on: March 19, 2016, 04:51:26 pm »
The fast rising and falling edges create noise that will be coupled to the DC busses. It is also a stress riser for capacitors and silicon components since there is a resulting current spike. That means it is shortening the life of the SMPS while generating noisier DC.

It's not a myth, but the result is not immediately fire and smoke either. I only have pure sine now that I started designing SMPS and understand what is happening.

Sent from my horrible mobile....

Factory400 - the worlds smallest factory. https://www.youtube.com/c/Factory400
 

Offline kalingaTopic starter

  • Contributor
  • Posts: 13
Re: modified sine wave bad for smps ?
« Reply #3 on: March 19, 2016, 05:07:31 pm »
I did a test with a smps, measuring voltages and spikes after the rectifiers, (the 2 smps are not power factor corrected) didn't see any spikes.  sw vs msw.

Have anyone done any extensive testing ?
 

Offline djQUAN

  • Frequent Contributor
  • **
  • Posts: 266
  • Country: ph
    • My DIY website
Re: modified sine wave bad for smps ?
« Reply #4 on: March 19, 2016, 05:12:33 pm »
I never really think of having a problem with "normal" SMPSs that rectify the AC to DC and use it directly, but was wondering what happens in a SMPS with active PFC.

I guess it's time to do some measurements.
 

Offline kalingaTopic starter

  • Contributor
  • Posts: 13
Re: modified sine wave bad for smps ?
« Reply #5 on: March 19, 2016, 05:17:03 pm »
I never really think of having a problem with "normal" SMPSs that rectify the AC to DC and use it directly, but was wondering what happens in a SMPS with active PFC.

I guess it's time to do some measurements.
Willing to help :) there is no proper answer to this (I mean factual data) can understand implications that might come with PFC type SMPS but in general, like that used in laptop bricks and TVs ?

Sent from my A0001 using Tapatalk

 

Offline rx8pilot

  • Super Contributor
  • ***
  • Posts: 3634
  • Country: us
  • If you want more money, be more valuable.
Re: modified sine wave bad for smps ?
« Reply #6 on: March 19, 2016, 05:17:47 pm »
I did a test with a smps, measuring voltages and spikes after the rectifiers, (the 2 smps are not power factor corrected) didn't see any spikes.  sw vs msw.

Have anyone done any extensive testing ?
I have not, but my customers really want me to investigate further.

How did you look for spikes? Its a difficult measurement to make. The edges are fast and will be flattened by a slow scope or poor probing.

Sent from my horrible mobile....

Factory400 - the worlds smallest factory. https://www.youtube.com/c/Factory400
 

Offline kalingaTopic starter

  • Contributor
  • Posts: 13
Re: modified sine wave bad for smps ?
« Reply #7 on: March 19, 2016, 05:21:25 pm »
I did a test with a smps, measuring voltages and spikes after the rectifiers, (the 2 smps are not power factor corrected) didn't see any spikes.  sw vs msw.

Have anyone done any extensive testing ?
I have not, but my customers really want me to investigate further.

How did you look for spikes? Its a difficult measurement to make. The edges are fast and will be flattened by a slow scope or poor probing.

Sent from my horrible mobile....
Did a basic test with out the long grounding wire, short lead. But I'm not properly equipped to do a proper measurements. My AC lines it self must be having some noise as well, fluorescent and fans. Non are PFC

Sent from my A0001 using Tapatalk

 

Offline Seekonk

  • Super Contributor
  • ***
  • Posts: 1938
  • Country: us
Re: modified sine wave bad for smps ?
« Reply #8 on: March 19, 2016, 05:50:53 pm »
Spikes, so what?  Even small diodes can take 60A.  Every switcher puts spikes into capacitors.  Let's stop coddling electronics.  It is a matter of perspective.

If the TV is plugged into the wall and dies........Crappy TV

If the TV is plugged into the inverter and dies........Crappy inverter
« Last Edit: March 19, 2016, 06:04:47 pm by Seekonk »
 

Offline rx8pilot

  • Super Contributor
  • ***
  • Posts: 3634
  • Country: us
  • If you want more money, be more valuable.
Re: modified sine wave bad for smps ?
« Reply #9 on: March 19, 2016, 07:50:09 pm »
Spikes, so what?  Even small diodes can take 60A.  Every switcher puts spikes into capacitors.  Let's stop coddling electronics.  It is a matter of perspective.

If the TV is plugged into the wall and dies........Crappy TV

If the TV is plugged into the inverter and dies........Crappy inverter

Sort of. Spikes make it through to the DC busses and cause trouble with the devices that are far more delicate than the giant diodes on the front end of the SMPS. This cause audible noise, noise for audio circuits, and digital errors that may or may not be error corrected. The MOSFETs involved are not necessarily capable of dealing with the high di/dt and dv/dt. The Capacitors may not be able to deal with the higher ripple current for extended periods.

Spikes are a genuine problem. There is no coddling of electronics. This is even more true with TV's, and other low margin consumer electronics that don't have the protections built in since they were designed to have 'ideal power' applied. Commercial and industrial electronics may be more capable in this regard, but who knows. The majority of the electronics I deal with are related to my job and I simply don't want to take a chance so I use sine wave UPS's. At home, I have modified sine because they are only for the very rare power outages to provide short term backup for cheap (relatively) electronics. I have my own balance there - cheap MSW for very occational needs and for the more critical I use pure sine only.

Sometimes I have to power a big digital system on SLA battery banks in a remote location. That system is only sine wave because I am powering it for extended periods and it is critical that it performs perfectly. With modified sine - you simply cross your fingers and hope for the best.
Factory400 - the worlds smallest factory. https://www.youtube.com/c/Factory400
 

Offline kalingaTopic starter

  • Contributor
  • Posts: 13
Re: modified sine wave bad for smps ?
« Reply #10 on: March 20, 2016, 07:00:06 am »
what is there is an isolation transformer in between the SMPS and MSW Inverter ?
and have common mode chokes such as a proper EMI filter ?
 

Offline rx8pilot

  • Super Contributor
  • ***
  • Posts: 3634
  • Country: us
  • If you want more money, be more valuable.
Re: modified sine wave bad for smps ?
« Reply #11 on: March 20, 2016, 05:15:18 pm »
My comments are based on theory and could use some bench time to understand reality. I would like to do some experiments with some common SMPS variants to see what happens.

Sent from mobile device.... Keeping it short and mis-spelled

Factory400 - the worlds smallest factory. https://www.youtube.com/c/Factory400
 

Offline kalingaTopic starter

  • Contributor
  • Posts: 13
Re: modified sine wave bad for smps ?
« Reply #12 on: March 20, 2016, 05:29:37 pm »
My comments are based on theory and could use some bench time to understand reality. I would like to do some experiments with some common SMPS variants to see what happens.

Sent from mobile device.... Keeping it short and mis-spelled
Do share your results, I'll do some reading on how to do it myself as well. I'll share the same.

Sent from my A0001 using Tapatalk

 

Offline Seekonk

  • Super Contributor
  • ***
  • Posts: 1938
  • Country: us
Re: modified sine wave bad for smps ?
« Reply #13 on: March 20, 2016, 06:08:39 pm »
MSW is not a sine wave so by definition there will he issues.

SMPS are harder on capacitors than linear supplies so failures will be more common.
A well designed SMPS should have a reasonable life cycle.

The input capacitors of a SMPS see 100,000 pulses a second. It doesn't seem like another
120 a second from the power line would have a major effect.

Attached is a scope trace of the charge current of a 9W LED lamp on grid power, only thing
I have on file.  As you can see there is a rapid rise in current once line voltage reaches
the charge of the capacitor.  That rise would be even faster if not for the internal current
limiting series resistor.  In a typical power supply this rise would be even faster since
there no resistor.  There are surges even with a pure sine wave. I just don't believe a
square wave would be an order of magnitude worse.

You could run the TV on DC like I do. Use an external bridge rectifier and capacitor.  No
pulses with DC.  It would require a good internal knowledge of the internals of the TV,
some TV have voltage doublers and that wouldn't work.  PFC controllers were supposed to be
the thing of the future but haven't seen many.

If this really bothers you, a 10 ohm series resistor might be a solution.  The SMPS would
easily handle the slightly lower voltage and the resistor would lower and extend the surge
current.  Something for the peanut gallery to discuss.  I'm not sure it would increase life
expectancy of the capacitors.  It certainly wouldn't damage anything, just waste a little
power.  If it helps you sleep at night that may be sufficient.
 

Offline kalingaTopic starter

  • Contributor
  • Posts: 13
Re: modified sine wave bad for smps ?
« Reply #14 on: March 20, 2016, 06:11:26 pm »
MSW is not a sine wave so by definition there will he issues.

SMPS are harder on capacitors than linear supplies so failures will be more common.
A well designed SMPS should have a reasonable life cycle.

The input capacitors of a SMPS see 100,000 pulses a second. It doesn't seem like another
120 a second from the power line would have a major effect.

Attached is a scope trace of the charge current of a 9W LED lamp on grid power, only thing
I have on file.  As you can see there is a rapid rise in current once line voltage reaches
the charge of the capacitor.  That rise would be even faster if not for the internal current
limiting series resistor.  In a typical power supply this rise would be even faster since
there no resistor.  There are surges even with a pure sine wave. I just don't believe a
square wave would be an order of magnitude worse.

You could run the TV on DC like I do. Use an external bridge rectifier and capacitor.  No
pulses with DC.  It would require a good internal knowledge of the internals of the TV,
some TV have voltage doublers and that wouldn't work.  PFC controllers were supposed to be
the thing of the future but haven't seen many.

If this really bothers you, a 10 ohm series resistor might be a solution.  The SMPS would
easily handle the slightly lower voltage and the resistor would lower and extend the surge
current.  Something for the peanut gallery to discuss.  I'm not sure it would increase life
expectancy of the capacitors.  It certainly wouldn't damage anything, just waste a little
power.  If it helps you sleep at night that may be sufficient.
Wow that made lot of sense! Thanks!

Sent from my A0001 using Tapatalk

 

Offline Seekonk

  • Super Contributor
  • ***
  • Posts: 1938
  • Country: us
Re: modified sine wave bad for smps ?
« Reply #15 on: March 20, 2016, 06:23:48 pm »
Many power supplies have a resistor in series to limit charge current to the capacitors when first turned on.  As this resistor heats up the resistance goes down to less than an ohm.  A fixed resistor with a square wave would just do this all the time.
 

Offline Zero999

  • Super Contributor
  • ***
  • Posts: 19345
  • Country: gb
  • 0999
Re: modified sine wave bad for smps ?
« Reply #16 on: March 20, 2016, 07:07:33 pm »
MSW is not a sine wave so by definition there will he issues.

SMPS are harder on capacitors than linear supplies so failures will be more common.
A well designed SMPS should have a reasonable life cycle.

The input capacitors of a SMPS see 100,000 pulses a second. It doesn't seem like another
120 a second from the power line would have a major effect.

Attached is a scope trace of the charge current of a 9W LED lamp on grid power, only thing
I have on file.  As you can see there is a rapid rise in current once line voltage reaches
the charge of the capacitor.  That rise would be even faster if not for the internal current
limiting series resistor.  In a typical power supply this rise would be even faster since
there no resistor.  There are surges even with a pure sine wave. I just don't believe a
square wave would be an order of magnitude worse.

You could run the TV on DC like I do. Use an external bridge rectifier and capacitor.  No
pulses with DC.  It would require a good internal knowledge of the internals of the TV,
some TV have voltage doublers and that wouldn't work.  PFC controllers were supposed to be
the thing of the future but haven't seen many.

If this really bothers you, a 10 ohm series resistor might be a solution.  The SMPS would
easily handle the slightly lower voltage and the resistor would lower and extend the surge
current.  Something for the peanut gallery to discuss.  I'm not sure it would increase life
expectancy of the capacitors.  It certainly wouldn't damage anything, just waste a little
power.  If it helps you sleep at night that may be sufficient.


That makes sense. Perhaps some SPICE simulation would provide more answers.

And with regards to power factor corrected power supplies; you obviously haven't encountered any modern equipment designed for the European market, which mandates power factor correction of switched mode power supplies, over a certain power rating.

Many power supplies have a resistor in series to limit charge current to the capacitors when first turned on.  As this resistor heats up the resistance goes down to less than an ohm.  A fixed resistor with a square wave would just do this all the time.
Why would it do that all the time with a squarewave? An NTC resistor limits the turn on surge current because its thermal time constant is much longer than half a cycle of the mains frequency so it shouldn't make any difference whether it's a sine or square wave as long as the fundamental is significantly higher than the thermal time constant.

 

Offline Seekonk

  • Super Contributor
  • ***
  • Posts: 1938
  • Country: us
Re: modified sine wave bad for smps ?
« Reply #17 on: March 20, 2016, 07:13:43 pm »
Thought you were worried about high pulse currents with a square wave vs nice ramp up with sine wave.  A resistor would limit maximum pulse currents.

I'm sure you are trying to make a point.  I just can't figure out what it is.
« Last Edit: March 20, 2016, 07:21:29 pm by Seekonk »
 

Offline Stonent

  • Super Contributor
  • ***
  • Posts: 3824
  • Country: us
Re: modified sine wave bad for smps ?
« Reply #18 on: March 20, 2016, 07:44:26 pm »
I would like to think there was a way to smooth the rise and fall of the wave. Some sort of very low pass filter to put some angle on the rising and falling edges of the modified sine wave and maybe catch the overshoot and undershoot.

That makes me wonder, why wouldn't a triangle wave be used instead of a modified sine wave?
The larger the government, the smaller the citizen.
 

Offline Zero999

  • Super Contributor
  • ***
  • Posts: 19345
  • Country: gb
  • 0999
Re: modified sine wave bad for smps ?
« Reply #19 on: March 20, 2016, 08:08:49 pm »
I would like to think there was a way to smooth the rise and fall of the wave. Some sort of very low pass filter to put some angle on the rising and falling edges of the modified sine wave and maybe catch the overshoot and undershoot.

That makes me wonder, why wouldn't a triangle wave be used instead of a modified sine wave?
A square wave with a step in it, is used in cheap modified sine wave inverters, because it reduces power/switching losses. Reducing the rise/fall time would increase the power dissipation and reduce the efficiency drastically.

A filter could be added to produce a sine wave but it would be bulky and expensive.

A modern, pure sine wave inverter use PWM and a small filter to generate a perfect sine wave.
 

Offline Seekonk

  • Super Contributor
  • ***
  • Posts: 1938
  • Country: us
Re: modified sine wave bad for smps ?
« Reply #20 on: March 20, 2016, 10:12:02 pm »
This is a MSW waveform from my 2000W inverter at camp, the spikes are from an uncompensated probe.  I just wanted to see how close the the frequency was after five years of operation.
 

Offline djococaud

  • Regular Contributor
  • *
  • Posts: 78
  • Country: fr
  • General rule : Simpler is Better
    • Djo Labs
Re: modified sine wave bad for smps ?
« Reply #21 on: March 25, 2016, 08:07:19 am »
Modified sine wave should be ok for simple SMPS appliances, but bear in mind that a lot of consumer appliances (like oven display, timers, coffee machines etc...) now uses simple capacitive droppers, relying on the mains frequency so the harmonics can litterally kill them !
HVDC could be safer than any non-sine voltage (capactitive dropper won't work on DC but should not take fire ! :phew:)...
With DC, the half of the rectifying of a SMPS must be able to withstand all the power though...
« Last Edit: March 25, 2016, 08:09:52 am by djococaud »
 

Offline kakureru

  • Contributor
  • Posts: 38
  • Country: us
Re: modified sine wave bad for smps ?
« Reply #22 on: March 27, 2016, 09:11:46 pm »
I like the idea of lowering the voltage to the device. Ive had things with a SMPS happily run on voltages as low as 85vac.
 

Offline Zero999

  • Super Contributor
  • ***
  • Posts: 19345
  • Country: gb
  • 0999
Re: modified sine wave bad for smps ?
« Reply #23 on: March 27, 2016, 10:40:07 pm »
I like the idea of lowering the voltage to the device. Ive had things with a SMPS happily run on voltages as low as 85vac.
The trouble with that is it increases the current draw.

Switched mode power supplies can often run at much lower voltages than they're designed from, even DC. My Rigol DS1054s oscilloscope runs fine off  36VDC but it's not specified to work below 100VAC. I did consider building a battery pack for it but quickly lost interest as I already have an oscilloscope with a built-in battery.
 

Offline kakureru

  • Contributor
  • Posts: 38
  • Country: us
Re: modified sine wave bad for smps ?
« Reply #24 on: March 28, 2016, 01:36:06 am »
The trouble with that is it increases the current draw.

Switched mode power supplies can often run at much lower voltages than they're designed from, even DC. My Rigol DS1054s oscilloscope runs fine off  36VDC but it's not specified to work below 100VAC. I did consider building a battery pack for it but quickly lost interest as I already have an oscilloscope with a built-in battery.

Nice to know since ide like my 1054z to be portable.
 

Offline uncle_bob

  • Supporter
  • ****
  • Posts: 2441
  • Country: us
Re: modified sine wave bad for smps ?
« Reply #25 on: March 28, 2016, 01:48:39 am »
Hi

Ok so going back to the top of the thread:

The real question is: Are modified sine waves an issue with *modern* TV power supplies?

The issue isn't that they are switchers. The issue is the power factor correction. If you have a PFC supply (as most modern gear is), it will work to level out the mess that a rectifier based load puts on  the mains. That circuit *expects* to see a sine wave. When you give it something weird it does something weird. In this case weird is not good. The current leveling pulses happen at the wrong time and both the inverter and the load get a bit bothered. The PFC heats up and the inverter is not as efficient as it should be (= it gets hotter as well). No, it will not burst into flames. Yes, it will have a shorter life that it otherwise would have had.

Bob
 

Offline Zero999

  • Super Contributor
  • ***
  • Posts: 19345
  • Country: gb
  • 0999
Re: modified sine wave bad for smps ?
« Reply #26 on: March 28, 2016, 08:49:57 am »
The trouble with that is it increases the current draw.

Switched mode power supplies can often run at much lower voltages than they're designed from, even DC. My Rigol DS1054s oscilloscope runs fine off  36VDC but it's not specified to work below 100VAC. I did consider building a battery pack for it but quickly lost interest as I already have an oscilloscope with a built-in battery.

Nice to know since ide like my 1054z to be portable.
Seed the following thread: https://www.eevblog.com/forum/testgear/battery-pack-for-rigol-ds1054z/
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf