My DVD player has a Class II SMPS supply with the secondary 0V directly connected to the metal chassis. The RCA sockets at the analogue output are also directly connected to the chassis. The Primary side of the SMPS is insulated and there is a 470pF Y capacitor connecting the primary input common to the secondary 0V (to reduce CM noise/EMI emission). But the chassis is not connected to the IEC earth pin.
After much thought and some reading, I intend to connect the chassis to the IEC earth. Do you see any problems?
Your advice will be much appreciated.
After much thought and some reading, I intend to connect the chassis to the IEC earth. Do you see any problems?
Your advice will be much appreciated.
In terms of safety, I am not converting a class I device into a class II device. I am treating a class II device as if it were a class I device so I am not downgrading safety but instead potentially improving safety.
I believe connecting the chassis to the IEC earth should not introduce any new problems, because with the existing DVD player the SMPS secondary 0V and the output RCA sockets are both connected to the chassis, and the interconnect cable goes to a pre-amplifier which is earthed. So, effectively, there is already a conductive path to earth, just not within the DVD player, but on the next interconnected component in the audio chain.
I believe connecting the chassis to the IEC earth should not introduce any new problems, because with the existing DVD player the SMPS secondary 0V and the output RCA sockets are both connected to the chassis, and the interconnect cable goes to a pre-amplifier which is earthed. So, effectively, there is already a conductive path to earth, just not within the DVD player, but on the next interconnected component in the audio chain.
So why would I be bothered with connecting the chassis to the IEC earth?
The SMPS is VERY NOISY! When placed next to a 65” modern LED TV the EMI affects a large section of the screen!
I have found that it is easy to get rid of the DM noise in a SMPS supply, but not the CM noise! It seems that the only easy modification to reduce the CM noise is to increase the Y capacitor linking the primary input common to the secondary 0V, as together with the primary to secondary winding capacitance Cps it forms a voltage divider for the CM noise. If Cps is about 50pF to 100pF, using a 4.7nF Y capacitor will attenuate the CM noise to about 1/100 to 1/50. The existing Y capacitor has a lower value of 470pF. In many SMPS schematics I found the Y capacitor to be 2.2nF.
If I do that, the DM noise will be increased, which is easy to deal with, while the CM noise will be reduced. But the increased CM noise current dumped to the chassis can only travel down through the interconnect shields and in that way creating more distortions, not to mention the likelihood of getting electric shock when hand touching the chassis.
Connecting the chassis to the IEC earth will hopefully provide an exit path for the CM noise current, and will also allow installation of a commercial EMI filter (Corcon, Schaffner, etc) to prevent RF contamination of the L and N lines. The DVD player has sufficient space to fit in one.
The SMPS is VERY NOISY! When placed next to a 65” modern LED TV the EMI affects a large section of the screen!
I have found that it is easy to get rid of the DM noise in a SMPS supply, but not the CM noise! It seems that the only easy modification to reduce the CM noise is to increase the Y capacitor linking the primary input common to the secondary 0V, as together with the primary to secondary winding capacitance Cps it forms a voltage divider for the CM noise. If Cps is about 50pF to 100pF, using a 4.7nF Y capacitor will attenuate the CM noise to about 1/100 to 1/50. The existing Y capacitor has a lower value of 470pF. In many SMPS schematics I found the Y capacitor to be 2.2nF.
If I do that, the DM noise will be increased, which is easy to deal with, while the CM noise will be reduced. But the increased CM noise current dumped to the chassis can only travel down through the interconnect shields and in that way creating more distortions, not to mention the likelihood of getting electric shock when hand touching the chassis.
Connecting the chassis to the IEC earth will hopefully provide an exit path for the CM noise current, and will also allow installation of a commercial EMI filter (Corcon, Schaffner, etc) to prevent RF contamination of the L and N lines. The DVD player has sufficient space to fit in one.
Hi, please help if you can.
I would like to confirm that I won't be doing something really silly and kill myself...
I read from other forums that earthing the SMPS 0V is OK. But I trust the experts in this forum more...
I would also like to know if 4.7nF Y cap across the isolation barrier is too big, as the largest value I have seen is 2.2nF.
I would like to confirm that I won't be doing something really silly and kill myself...
I read from other forums that earthing the SMPS 0V is OK. But I trust the experts in this forum more...
I would also like to know if 4.7nF Y cap across the isolation barrier is too big, as the largest value I have seen is 2.2nF.
Last edited:
Connecting the chassis to mains ground posses no risk... you could well do that anyway via the RCA leads into partnering equipment which is already mains grounded.
The cap bridging the barrier will have been chosen specifically to do the job required and I wouldn't go altering it on a whim. The value will depend partly on the operating frequency of the SMPS.
4n7 could allow very significant HF current to flow.
It is common to be able to 'draw an arc' from chassis to mains ground or an aerial lead for example while 'hot connecting' leads. If you have damp hands you will certainly know about it.
Raising the cap value therefore posses a shock hazard... don't do it
The cap bridging the barrier will have been chosen specifically to do the job required and I wouldn't go altering it on a whim. The value will depend partly on the operating frequency of the SMPS.
4n7 could allow very significant HF current to flow.
It is common to be able to 'draw an arc' from chassis to mains ground or an aerial lead for example while 'hot connecting' leads. If you have damp hands you will certainly know about it.
Raising the cap value therefore posses a shock hazard... don't do it
Many thanks, Mooly.
But if the chassis is earthed, while increasing the Y cap results in increased "shock" current, the current is dumped to the earth line. As long as it's not large enough to trigger the house safety switch, would it be fine? I mean there is no safety threat.
In any case, I have now read that this capacity can not be over 3.9nF for a 2 wire mains input.
But if the chassis is earthed, while increasing the Y cap results in increased "shock" current, the current is dumped to the earth line. As long as it's not large enough to trigger the house safety switch, would it be fine? I mean there is no safety threat.
In any case, I have now read that this capacity can not be over 3.9nF for a 2 wire mains input.
I was thinking more of you just increasing the cap value on its own tbh.
Thinking about it a little more, whatever you do has to be safe under all possible conditions of use, and while grounding the chassis is absolutely safe, increasing the cap could pose a problem if the unit were connected to a non grounded supply... one with no earth.
Thinking about it a little more, whatever you do has to be safe under all possible conditions of use, and while grounding the chassis is absolutely safe, increasing the cap could pose a problem if the unit were connected to a non grounded supply... one with no earth.
You may well make noise from the device higher by doing that however, because the added ground wire can form a ground loop that include the signal return shield of the audio cables. It goes like this -- some other component (most likely) leaks some power line noise current of some type to it's chassis, which is conducted out on its chassis ground. But now some of it will be shunted to your DVD player via your new ground wire and that current will find it's way back over the audio cable shields. That current, multiplied by the non-zero impedance of the cable shields become a noise voltage that is directly added to the audio. How much depends on the circuitry of the next device in the audio chain, there are configurations that can minimize that kind of noise path, but rather few designs seem to use them.
Hi bwaslo, if the noise source is as you described then the noise can travel by the interconnect, any way. The earth lead is in exact parallel to the interconnect shield. It can only reduce the noise effect on signal, not increase. Balanced interconnection can eliminate the ground loop and the noise you refer to. I have not chosen that path only because of the extra complexity - we need additional receiver circuit in the next audio component, and this receiver needs a very good power supply not to compromise sound quality, etc. For home audio of 1 to 2 meters interconnection, it may not be worthwhile. For professional audio use, for sure.
Safety message: Connecting the secondary side 0V directly to earth removes the isolation property of isolation transformers. If you happen to touch the, for example, secondary side 100V, current flows through your hand to your feet to earth and from earth back to secondary side 0V. And ELCB/RCD cannot detect this. If earth is not connected to secondary, there is no complete path.
Instead of direct connection, an appropriate class Y capacitor that provides high impedance at the lethal voltage's frequency (be it 60Hz or DC) should be used to limit current in a fault scenario.
Noting that a device is only class II if "no single failure can result in dangerous voltage becoming exposed". So normally you would need to somehow touch the secondary side 0V and high voltage at the same time, which means two failures. With the 0V connected to earth, you need just one more failure.
Instead of direct connection, an appropriate class Y capacitor that provides high impedance at the lethal voltage's frequency (be it 60Hz or DC) should be used to limit current in a fault scenario.
Noting that a device is only class II if "no single failure can result in dangerous voltage becoming exposed". So normally you would need to somehow touch the secondary side 0V and high voltage at the same time, which means two failures. With the 0V connected to earth, you need just one more failure.
The secondary side can legitimately connect to earth via the ancillary connected equipment. It is a valid scenario. There is no safety implication for the casual user.
Working on the equipment with the covers off is a different thing altogether, and in that case it is up to the individual to know and understand the risks.
Working on the equipment with the covers off is a different thing altogether, and in that case it is up to the individual to know and understand the risks.
It will not, if all the ancillary connected equipment follow the same rule of not connecting the secondary 0V directly to earth.
If any equipment breaks that rule, the exposed metallic 5-way binding posts will become lethal, exposing 100VDC in the case of bridged amps powered by 200VDC. Or whatever VAC voltage that is currently being sent to the speakers regardless of amp type. And I stress again, this will not be protected by ELCB/RCD.
We are talking about a very specific piece of equipment. A DVD player with an SMPS.
The chassis will already connect to the zero volt line of the PSU internally. A user of that DVD player is free to connect that player to whatever they choose, equipment that may or may not be grounded in its own right.
The chassis will already connect to the zero volt line of the PSU internally. A user of that DVD player is free to connect that player to whatever they choose, equipment that may or may not be grounded in its own right.
In that case I am wondering how a DVD player can be classified as class II (it does have the double square sign printed at the back panel), because there is a RCA connection to other equipment and the other equipment can be earthed.
Since there is no change to the class II SMPS other than my planned increase of the Y capacitor from 470pF to up to 2.2nF, to have the primary side exposed is very unlikely. The secondary side has +/-12V and +5V only. The secondary side 0V is already connected to the chassis. If the chassis is earthed, the only chance that I can touch a killing voltage is for the class II SMPS to break down and its primary voltage gets across to the secondary then onto the analogue output RCA signal line then onto the other equipment.
Here is a quote from an article:
UL1950 specifies that information technology equipment with Class I or three wire input (line, neutral, and earth ground), 240 VAC, 60 Hz input must have a leakage current no higher than 3.5 mA if earth ground is opened or one component has failed short which
restricts Y-capacitor maximum value below 0.039 μF (or 39 nF). For class II or two wire input (line, neutral, with no earth ground), leakage current must be less than 250 μA with one failed component which restricts Y-capacitor size to under 0.0028 μF (2.8 nF or 2800 pF) for 240 VAC, 60 Hz input.
So if I earth the chassis and change the two wire mains input to 3 wire mains input, I can legally try 4.7nF as the Y capacitor in order to reduce the CM noise.
UL1950 specifies that information technology equipment with Class I or three wire input (line, neutral, and earth ground), 240 VAC, 60 Hz input must have a leakage current no higher than 3.5 mA if earth ground is opened or one component has failed short which
restricts Y-capacitor maximum value below 0.039 μF (or 39 nF). For class II or two wire input (line, neutral, with no earth ground), leakage current must be less than 250 μA with one failed component which restricts Y-capacitor size to under 0.0028 μF (2.8 nF or 2800 pF) for 240 VAC, 60 Hz input.
So if I earth the chassis and change the two wire mains input to 3 wire mains input, I can legally try 4.7nF as the Y capacitor in order to reduce the CM noise.
It is because grounding it via external leads doesn't make it 'unsafe'.
The player is double insulated and safe in its own right and when not connected to any other equipment. You can hold the player in one hand and grab hold of ground with the other hand. You will not receive a dangerous shock although you might 'feel' something because of the allowable leakage.
Changing the cap would be classed as a design change and is unsafe because the player could find its way to another owner who would not be aware of all this and who could connect it to a two wire supply that has a faulty or open ground. The chassis would then have a higher leakage current than is allowable which would present a shock hazard.
Its a fine line. Technically you opening the player and adding the ground lead invalidates all the built in safety features because what you do is unverified and changes the original design. We assume you know what you are doing and can add the wire safely
Do you see
What you originally propose enters the realm of diy, and what you propose with the ground wire isn't unsafe in itself. Changing the cap is unsafe if you were ever to forget years later, or passed the player on.
I won't sell it or give it away as a gift. This is DIY, a fascinating hobby to satisfy a man’s curiosity and desire for learning.
I guess that because in many countries they use 2 wire mains for residential power supply, the DVD manufacturer makes it to use 2 wire mains input so that the same player can be sold in all countries. But 2 wire mains input limits the performance of the player. To allow for larger leakage current will reduce CM noise, which is a big deal with audio.
But I still don’t understand why the maximum allowable leakage current specified for 2 wire and 3 wire mains are different. For 3 wire mains input, in the case of a failed earth connection the leakage current will run in the device chassis, which is not different from a 2 wire device.
Fortunately, in the case of audio devices, there are multiple earthing connections. For example, if the mains cable contact becomes loose / open, the 2 interconnect cables connecting from the DVD player to the preamplifier provide the alternative earthing paths in which the CM current is discharged. So the only chance I can get a nasty shock from touching the DVD player’s chassis is to have all earth paths broken. The risk is quite small, especially when I am keenly aware of the potential hazard.
I guess that because in many countries they use 2 wire mains for residential power supply, the DVD manufacturer makes it to use 2 wire mains input so that the same player can be sold in all countries. But 2 wire mains input limits the performance of the player. To allow for larger leakage current will reduce CM noise, which is a big deal with audio.
But I still don’t understand why the maximum allowable leakage current specified for 2 wire and 3 wire mains are different. For 3 wire mains input, in the case of a failed earth connection the leakage current will run in the device chassis, which is not different from a 2 wire device.
Fortunately, in the case of audio devices, there are multiple earthing connections. For example, if the mains cable contact becomes loose / open, the 2 interconnect cables connecting from the DVD player to the preamplifier provide the alternative earthing paths in which the CM current is discharged. So the only chance I can get a nasty shock from touching the DVD player’s chassis is to have all earth paths broken. The risk is quite small, especially when I am keenly aware of the potential hazard.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.