EEVblog Electronics Community Forum

Electronics => Projects, Designs, and Technical Stuff => Topic started by: Circlotron on November 22, 2017, 11:20:07 pm

Title: Place MOV before or after CM choke?
Post by: Circlotron on November 22, 2017, 11:20:07 pm
Several sites where we have a piece of equipment installed have suffered blowups in the 150W SMPS. We are putting it down to mains spikes. They already have a 12mm diameter round 275V MOV fitted but for these sites we are going to instead fit a 250V 26mm square MOV. Lots more meat!

As for placement, the existing MOV is across the line just after the fuse and thermistor. Following that is an X cap, CM choke, another X cap then the rectifier bridge. Would it be better to place the new MOV after the CM choke so that any resistance in the CM choke helps reduces the clamping current in the MOV and consequently reduces the max clamping voltage, or is this a totally bad idea?
Title: Re: Place MOV before or after CM choke?
Post by: Zero999 on November 22, 2017, 11:26:36 pm
I wouldn't have thought it would make any difference. The choke isn't going to limit the current that much and if the impedance of the MOV is low enough that it does, you've got worse things to worry about.
Title: Re: Place MOV before or after CM choke?
Post by: jbb on November 23, 2017, 07:08:34 pm
Several sites where we have a piece of equipment installed have suffered blowups in the 150W SMPS.  We are putting it down to mains spikes.

(Maybe you've done this already...) If you can collect and analyse the failed equipment you might get some good feedback.  Sometimes a small change can solve big problems.

Also, if you talk to your local EMC test house, they can help you test mains surge to the appropriate standards (sorry, I forget the appropriate numbers...) in a repeatable fashion.  The standards are generally designed to encompass real-world misfortunes and could be quite helpful for finding the weak points without failures in the field annoying users. (Note: sanity of standards is not guaranteed...)

fit a 250V 26mm square MOV

What's your line voltage? 'Cause it would seem like the equipment may be subject to 264V high line conditions in some parts of Australia (http://www.ewh.ieee.org/r10/nsw/subpages/history/Australian-AC-Line-Voltages.pdf) (note, this info could be out of date).  Also note that sometimes MOVs need a lot of derating of their voltage rating - I suggest you check the data sheet and look for application notes from the MOV manufacturer.

Then you have to look at the maximum clamping voltage of the MOV and make sure that voltage won't mess anything up.

The joy of the EMC standards is that they provide test waveforms (event suggested schematics) that can be simulated in SPICE, which can save you some trouble.

Title: Re: Place MOV before or after CM choke?
Post by: ocset on November 23, 2017, 07:47:05 pm
You may need a mains impedance tester...this is where you can test if the mains is inductive or capacitive...if inductive then you can get spike problems......if capacitive then less likely...underground cables are usually capacitive, overhead lines usually inductive.

Mains transients are cuased most often by fault currents in heavy equipment getting interrupted by a fuse blowing or circuit braker acting..........you can get 100's of amps flowing in the mains, then it suddenly has nowhere to flow to when the fuse blows, so an ovevoltage spike results due to the breaking of an inductive current flowing in the stray inductance of the mains......if this is bad enough, then its unlikely that anything cost effective that you could put inside a power supply would be any  use.   :horse:

You may need an external transient protection module on the line.   :phew:   

Having said that, for smaller spikes, a big electrolitic cap after the pfc stage can act to absorb smaller transients.

Movs die after so many quenches.    :horse:  :scared: 

 8)

Title: Re: Place MOV before or after CM choke?
Post by: Circlotron on November 23, 2017, 09:08:04 pm
This piece of equipment with the SMPS we don’t actually manufacture, it is made in Europe and we just use it as part of something larger that we actually do build. Seeing it is only happening at several locations and we have several thousand out there running with no problem, another approach I was thinking of is to simply put a low value inductor in series with the mains supply to the unit. I got a 100VA transformer and stripped off all the windings then threaded through the stack 20 turns of mains rated plastic insulated wire, half on one side of the bobbin for active and the other half for neutral. Phased the two windings as differential mode, not common mode. At full load the AC voltage loses about 20 volts off the peaks, so at spike frequency hopefully it should be very effective. Probably only trying it out in real life will tell.
Title: Re: Place MOV before or after CM choke?
Post by: T3sl4co1l on November 24, 2017, 02:15:27 am
After.  A choke for a 150W load will have about an ohm DCR, which improves the MOV's performance in terms of protecting the DUT.  Meanwhile, the choke is just a tangle of wire -- if anything's going to blow, the fuse will melt first (or, second after the choke fails shorted), and there you go. :)

Tim
Title: Re: Place MOV before or after CM choke?
Post by: Jay_Diddy_B on November 24, 2017, 02:28:26 am
Hi,

 I don't know which MOV I don't know which MOV you are planning on using. Here are the specifications for the Littelfuse V25S series:

(https://www.eevblog.com/forum/projects/place-mov-before-or-after-cm-choke/?action=dlattach;attach=372982;image)

The clamping voltage is 700V, way above the voltage will kill your power supply.

If your power supply has a universal input, 90 - 260 Vac, I would suggest using a 250VA auto transformer in front of the power supply to reduce the input to 120V. You now have the impedance of the auto-transformer and 120V of headroom.

Regards,
Jay_Diddy_B
Title: Re: Place MOV before or after CM choke?
Post by: Circlotron on November 24, 2017, 04:16:04 am
Yep, the one I have is a V25S250P.
Here we have nominally 240V mains so I put the MOV across the output of an isolation transformer and wound the variac up to it's max of 265V. The MOV didn't get the slightest bit warm and there was no evidence of the AC wave being clipped on the scope.The SMPS has a minimum input of 230V -15%, so about 195VAC.
Title: Re: Place MOV before or after CM choke?
Post by: jbb on November 24, 2017, 05:01:38 am
Yep, the one I have is a V25S250P.
Here we have nominally 240V mains so I put the MOV across the output of an isolation transformer and wound the variac up to it's max of 265V. The MOV didn't get the slightest bit warm and there was no evidence of the AC wave being clipped on the scope.The SMPS has a minimum input of 230V -15%, so about 195VAC.

Unfortunately it's not quite that simple.  You would need to test tens or hundreds of MOVs at elevated temperatures for quite a long time.  They age and are very temperature sensitive. Someone at Littelfuse has already done all that hard work for you to make the data sheet.

For difficult cases you sometimes get a chain of series fuse, shunt MOV, series impedance (e.g. inductor) and than a shunt TVS. MOVs can take a lot of energy but are a bit slow and have widely varying clamp voltages. TVSs are faster and have better clamping voltages but can't take much energy.  A combined circuit can take a lot of energy and achieve fast clamping.

For industrial-strength cases you can even add a Gas Discharge Tube to the chain (it's a shunt type device which can handle heaps of energy).

Also, if you modify the SMPS module (e.g. swap out the MOV), you'll quite possibly void any certifications (e.g. safety) that it comes with.
Title: Re: Place MOV before or after CM choke?
Post by: ocset on November 25, 2017, 04:43:48 pm
MOVs are pretty well just as quick as TVS.
MOVs can take more grief than TVS , generally.
TVS have more accurate  clamping voltage than mov, but its still poor.

What you shoudl do is have two fuses.....have the mov just downstream of a very high current fuse, so the mov  shunting high current doesnt blow it, -unless its really really high current....then have a lower value fuse downstream of the mov...otherwise you will keep having your fuse blow when the mov shunts high currents.

Feel free to have a diode_blocked  high capacitance  just after the  mains rectifier bridge (diode blocked so it doesnt interfere with the pfc stage).....if you like, have your low value fuse downstream of this...so that transient overcurrents that  flow into the diode-blocked capacitance dont blow  the light current fuse.

Some people rely on the post pfc stage capacitance to shunt the transient overcurent, and stop it from developing a big overvoltage,  but if you do it like that, then  your light current fuse might nuisance trip when a transient overcurrent occurs.

Title: Re: Place MOV before or after CM choke?
Post by: Seekonk on November 25, 2017, 06:04:18 pm
Ever notice that MOV testing is done with 1/8 inch leads.  Every little bit of inductance or resistance helps (or hinders).  This guy in communications would tie several knots in power cords.  Yea, choke off those electrons.  He did say the power cord would blow out at the first knot in a lightning strike.
Title: Re: Place MOV before or after CM choke?
Post by: intabits on November 25, 2017, 06:46:58 pm
I don't recall ever seeing the term "diode-blocked capacitance"
Could you please explain what that means?
Title: Re: Place MOV before or after CM choke?
Post by: Seekonk on November 25, 2017, 07:47:14 pm
That is a tracking noise clamper that is quite effextive  A diode bridge is placed across the line.  The rectified voltage goes to a large capacitor bank which has a little bleed off resistor.  So the capacitor voltage always rests at peak sine wave voltage.  Any spike above that, even just a few volts, is shunted by the capacitor.
Title: Re: Place MOV before or after CM choke?
Post by: floobydust on November 26, 2017, 03:20:16 am
Consumer electronics use spark gaps across the CM choke; i.e. US Patent US8345400 (http://www.google.com/patents/US8345400) "... surge events can cause resonant ringing in the components of the RFI filter, which may greatly amplify the magnitude of the voltages applied to the electronic components..." Seems to be around 500V breakdown.

If this is why your PSU's are failing, then I don't see a bigger MOV (CM choke input side) helping.

Title: Re: Place MOV before or after CM choke?
Post by: T3sl4co1l on November 26, 2017, 05:02:52 am
That's relevant for ESD and EFT.  Indeed, I've seen ESD sparks jumping between turns of the CMC on an open frame PS module.

Surge is slower, and lower voltage, 2.5kV typically.  It won't jump those gaps, nor will it have enough dV/dt to induce that much voltage across a filter component.

Tim