Posted January 10, 20232 yr Heya, so I was busy testing a UPS and figured a nice and easy load for testing would be a couple 200W floodlights. So the UPS seems to be working fine but the MOV on the floodlight input was completely fried... I am going to be getting an Oscilloscope back into the lab on Thursday, hoping to check the output of the UPS to check why it would do this. In the meantime, I was hoping to pick the communities collective minds as to why this would happen or how I can test the UPS to make sure it's not faulty?
January 10, 20232 yr 1 hour ago, Psy said: Heya, so I was busy testing a UPS and figured a nice and easy load for testing would be a couple 200W floodlights. So the UPS seems to be working fine but the MOV on the floodlight input was completely fried... I am going to be getting an Oscilloscope back into the lab on Thursday, hoping to check the output of the UPS to check why it would do this. In the meantime, I was hoping to pick the communities collective minds as to why this would happen or how I can test the UPS to make sure it's not faulty? Over voltage as regulation can be very bad. One also has to content with the voltage measured. AC vs True RMS. Edited January 10, 20232 yr by Scorp007
January 10, 20232 yr 56 minutes ago, Psy said: Heya, so I was busy testing a UPS and figured a nice and easy load for testing would be a couple 200W floodlights. So the UPS seems to be working fine but the MOV on the floodlight input was completely fried... I am going to be getting an Oscilloscope back into the lab on Thursday, hoping to check the output of the UPS to check why it would do this. In the meantime, I was hoping to pick the communities collective minds as to why this would happen or how I can test the UPS to make sure it's not faulty? Is the N-E correctly bonded on the output of the UPS? What's the voltage between: L-E - should be ~230V L-N - should be ~230V N-E - should be ~0V
January 10, 20232 yr Author 46 minutes ago, system32 said: Is the N-E correctly bonded on the output of the UPS? What's the voltage between: L-E - should be ~230V L-N - should be ~230V N-E - should be ~0V Ok, so I think you're on to something. I was busy measuring this to check and I found something interesting... So I have these PVC boxes that have a Voltmeter, Ammeter and Hertzmeter built into it. I plugged this into the UPS and when mains is connected it reads 230V at 50Hz, however, when running off battery the meters read 170V at 150Hz. When checking these measurements with a proper multimeter and its still reading 230V at 50Hz, so I assume that the UPS creates a dirty wave form with some harmonics or its a modified sine wave. I will be able to confirm this on Thursday but the only time I have seen my measuring box offer these odd kind of readings is when there is some kind of harmonics on the line...
January 17, 20232 yr Author So, just updating this thread in case some runs into something similar... So it looks like it has to do with the way the inverter generates it's waveform. So looks like this unit is "simulated" sine wave, so not pure and found out the hard way. With the way the modified waveform, it picks up like 40% or more harmonics on the 3rd and 17% on the 7th. This combined with a light that doesn't have a power supply or transformer in it leads to instant burning... So when it comes to you lights, you can connect lights to a pure sine wave but connecting an LED light to a modified sine wave can end badly. It's not like every LED light on earth will pop on modified, but a lot of them don't have the power supply needed to handle being connected to modified sine waves.
January 17, 20232 yr 3 minutes ago, Psy said: So, just updating this thread in case some runs into something similar... So it looks like it has to do with the way the inverter generates it's waveform. So looks like this unit is "simulated" sine wave, so not pure and found out the hard way. With the way the modified waveform, it picks up like 40% or more harmonics on the 3rd and 17% on the 7th. This combined with a light that doesn't have a power supply or transformer in it leads to instant burning... So when it comes to you lights, you can connect lights to a pure sine wave but connecting an LED light to a modified sine wave can end badly. It's not like every LED light on earth will pop on modified, but a lot of them don't have the power supply needed to handle being connected to modified sine waves. I don't see how that would pop the MOV? Please post some pictures. Edited January 17, 20232 yr by P1000
January 17, 20232 yr Author Well it's one of a couple high power floodlights I use as load, adding more lights as needed to see how inverts etc act under load. If you look at the power supply of the light, its basically a rectifier with voltage regulators... however, the MOV is before any of that stuff and burnt right through. Even the input resistors are burnt completely, reading as open circuit.
February 5, 20232 yr Live and neutral are ~120v in respect to earth. This is done to simplify the electronics. If the MOV was connected in a way that assumes neutral is truly neutral in respect to earth it would blow it up.
February 5, 20232 yr 26 minutes ago, Gnome said: Live and neutral are ~120v in respect to earth. This is done to simplify the electronics. If the MOV was connected in a way that assumes neutral is truly neutral in respect to earth it would blow it up. Ney broe, nou verstaan ek nie? if Mov is connected E-L then it gets 170V or 120V, should be ok, if its N-L, then it should get 240V and should be fine, if its E-N, it should see 170 or 120V and still should be ok... hoe nou? I can't see how it can be connected that would cause it to eject the magic smoke in this scenario... it probably has something to do with harmonics and I assume some higher Voltage spikes appearing out of the UPS, you'd have to have a storage scope and store 5 seconds worth of trace from when you switch on the light and thus should be able to see the HT spikes beyond the MOV's rating, I'd think... normally a MOV is expected to go short circuit and blow a fuse between it and the power source, to try and protect what's downstream of it from the excessive Voltage...
February 6, 20232 yr On 2023/02/05 at 11:16 AM, Kalahari Meerkat said: Ney broe, nou verstaan ek nie? if Mov is connected E-L then it gets 170V or 120V, should be ok, if its N-L, then it should get 240V and should be fine, if its E-N, it should see 170 or 120V and still should be ok... hoe nou? I can't see how it can be connected that would cause it to eject the magic smoke in this scenario... it probably has something to do with harmonics and I assume some higher Voltage spikes appearing out of the UPS, you'd have to have a storage scope and store 5 seconds worth of trace from when you switch on the light and thus should be able to see the HT spikes beyond the MOV's rating, I'd think... normally a MOV is expected to go short circuit and blow a fuse between it and the power source, to try and protect what's downstream of it from the excessive Voltage... Earth and neutral should be close to 0v. You wouldn't size a clamping device between earth and neutral for your peak to peak voltage. That would be idiotic and pointless. "Harmonics" whatever that means (fairly broad term) is not a factor for an MOV. The expected peak to peak voltage is the only thing that matters. If by harmonics you mean that the wave form is not sinusoidal, that is irrelevant because of peak to peak sizing for MOVs. If you mean high frequency noise, MOVs only clamp up to a certain frequency given the response time of the device (PTCs are used for shorter time spans). Either way neither are of concern. Edited February 6, 20232 yr by Gnome
February 6, 20232 yr On 2023/02/05 at 10:41 AM, Gnome said: If the MOV was connected in a way that assumes neutral is truly neutral in respect to earth it would blow it up. please explain how, in the LEDs above you would connect the MOV so that it will let out its smoke as it did, since we can assume its rated at 250V or more and since the multimeter reading provided 230V, how can we magically increase this to get the MOV to pop? 26 minutes ago, Gnome said: Earth and neutral should be close to 0v. You wouldn't size a clamping device between earth and neutral for your peak to peak voltage. That would be idiotic and pointless. obviously, I was just trying to figure out how you can make the MOV in this case perform sepuku, since the voltages shouldn't allow this... except if the multimeter is oblivious to higher frequencies 3d harmonic maybe 150Hz and this may carry a higher Voltage, but the multimeter may be unable to register this, doubt that this is the case, but hey, you wrote about the MOV being connected in a certain way, how/which way if it were connected to earth and live, maybe if live and neutral is floating way above earth, not very likely, but @Psy didn't mention measuring between earth and neutral or earth and live and I suspect that the LED has no actual earth connection either....
February 14, 20232 yr On 1/10/2023 at 3:51 PM, Psy said: Heya, so I was busy testing a UPS and figured a nice and easy load for testing would be a couple 200W floodlights. So the UPS seems to be working fine but the MOV on the floodlight input was completely fried... I am going to be getting an Oscilloscope back into the lab on Thursday, hoping to check the output of the UPS to check why it would do this. In the meantime, I was hoping to pick the communities collective minds as to why this would happen or how I can test the UPS to make sure it's not faulty? The MOV (Metal-Oxide Varistor) on the floodlight input is a component designed to protect against transient voltage spikes and surges. It is not uncommon for MOVs to degrade or fail over time, especially when exposed to frequent or high-energy transients. However, it's possible that the UPS may have produced a voltage spike or surge that exceeded the MOV's maximum rating, causing it to fail. This could be due to a fault in the UPS itself, such as a failed component or an incorrect setting. To test the UPS, you could try connecting different loads of varying sizes and types to it and monitoring the output voltage with a multimeter or oscilloscope. You could also check the manufacturer's documentation and support resources to see if there are any recommended tests or procedures for checking the UPS's performance. It's also a good idea to check the input voltage and waveform of the UPS to ensure that it is within the UPS's specified operating range. This can help ensure that the UPS is not being subjected to excessive or unstable input power that could cause it to produce undesirable output waveforms or voltage spikes.
February 14, 20232 yr On 2023/02/06 at 5:25 PM, Gnome said: Earth and neutral should be close to 0v. You wouldn't size a clamping device between earth and neutral for your peak to peak voltage. That would be idiotic and pointless. So you say we are idiotic What if your N is interrupted between point of supply and say a gate motor. The distance between DB and gate can and is over 50m on many gates and you get a lightning strike on this cable? This again is very common and does take out gates and electric fences.
February 16, 20232 yr On 2023/02/14 at 9:56 AM, Scorp007 said: So you say we are idiotic I don't say you were idiotic. I said putting a neutral to earth MOV that has a line to neutral voltage would be idiotic. In other words. If you put a MOV with a clamping voltage of 400v between neutral and earth, that would be stupid. On 2023/02/14 at 9:56 AM, Scorp007 said: What if your N is interrupted between point of supply and say a gate motor. The distance between DB and gate can and is over 50m on many gates and you get a lightning strike on this cable? This again is very common and does take out gates and electric fences. Yes, that is why you would size the neutral to earth as close to 0v, in terms of a clamping device. Meaning you'd put a clamping device there with something low like 35v, 50v, etc.
February 16, 20232 yr The neutral-earth mov is usually the same as the live-earth mov. Swapping live and neutral is a common enough fault (especially on things like extensions, or at the installation point of a light fixture) that you don't want to blow all your arresters/equipment with movs everytime that happens, it would cause too many warranty returns.
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.