Decoupling on ac input

[SkrillBill]

While I was taking apart an old psu, I came across this on the ac input. It looks like capacitors, so assume it'd decoupling (tho it's hard to tell what the blue ones are). Am I safe in assuming thats what they are or am I way off the mark(probably am)?

[Benta]

The blue ones are MOVs.

[SkrillBill]

Ahh. So this is protection from transient voltage/spikes?

[Dave]

The blue ones are MOVs.

Those are Y-rated capacitors.

[David Hess]

The blue parts are Y-class capacitors between hot and neutral to ground.  The yellow part is an X-class capacitor between hot and neutral.  The two resistors discharge the capacitors to prevent shock after the unit is unplugged.

These capacitors are all for interference suppression.

[Gyro]

Just for completeness... There are two discharge resistors in series to avoid exceeding their individual breakdown voltage rating (normally around 250V for a 1/4W Carbon film).

[Benta]

My apologies. Y capacitors it is.

[SkrillBill]

Ah, ok. I was on the right track just the wrong train. EMI/RFI suppression. Sounds like i would want to leave them connected when i re-use that plug in my device. I will do some more research into x and y class but it seems pretty straight forward so far.

Thanks!

[David Hess]

Ah, ok. I was on the right track just the wrong train. EMI/RFI suppression. Sounds like i would want to leave them connected when i re-use that plug in my device. I will do some more research into x and y class but it seems pretty straight forward so far.

The significance of X and Y class parts are the certifications for capacitors installed on the AC line side.  Both must fail without causing dangerous conditions and Y class capacitors must fail open to prevent elevating ground.

If the load does not produce and is not susceptible to conducted EMI like an incandescent lamp or resistive heater, then the capacitors are not required.

[SkrillBill]

Ah, ok. I was on the right track just the wrong train. EMI/RFI suppression. Sounds like i would want to leave them connected when i re-use that plug in my device. I will do some more research into x and y class but it seems pretty straight forward so far.

The significance of X and Y class parts are the certifications for capacitors installed on the AC line side.  Both must fail without causing dangerous conditions and Y class capacitors must fail open to prevent elevating ground.

If the load does not produce and is not susceptible to conducted EMI like an incandescent lamp or resistive heater, then the capacitors are not required.

I'm just building a pretty basic power supply with some LM78xx regulators, a transformer and a full-wave rectifier.

[David Hess]

If the load does not produce and is not susceptible to conducted EMI like an incandescent lamp or resistive heater, then the capacitors are not required.

I'm just building a pretty basic power supply with some LM78xx regulators, a transformer and a full-wave rectifier.

The rectifiers can produce noise during reverse recovery which gets coupled back into the AC power line although if that happens, it is better to fix the problem at the rectifiers.

I would leave the capacitors there unless you have some reason to remove them.

[SkrillBill]

If the load does not produce and is not susceptible to conducted EMI like an incandescent lamp or resistive heater, then the capacitors are not required.

I'm just building a pretty basic power supply with some LM78xx regulators, a transformer and a full-wave rectifier.

The rectifiers can produce noise during reverse recovery which gets coupled back into the AC power line although if that happens, it is better to fix the problem at the rectifiers.

I would leave the capacitors there unless you have some reason to remove them.

Other than leaving the caps, how would i fix that at the rectifiers? And, i don't really have a way to detect if that is happening. My only 'bench tool' is a Multimeter :/ . Should i just build in the protection from the start?

[paulca]

The two resistors discharge the capacitors to prevent shock after the unit is unplugged.

I have met more than one device that omitted these.  One was a cheap 12V DC wall wart the other a cheap PC PSU.  Both gave you an impressive little zap when you touched the plug pins after disconnect.

[Zero999]

The two resistors discharge the capacitors to prevent shock after the unit is unplugged.

I have met more than one device that omitted these.  One was a cheap 12V DC wall wart the other a cheap PC PSU.  Both gave you an impressive little zap when you touched the plug pins after disconnect.

Discharge resistors aren't always needed. If the device doesn't have a power switch before the SMPS and it discharges the capacitor to a low enough voltage, before shutting off, then a discharge resistor is unnecessary.

[David Hess]

The rectifiers can produce noise during reverse recovery which gets coupled back into the AC power line although if that happens, it is better to fix the problem at the rectifiers.

I would leave the capacitors there unless you have some reason to remove them.

Other than leaving the caps, how would i fix that at the rectifiers? And, i don't really have a way to detect if that is happening. My only 'bench tool' is a Multimeter :/ . Should i just build in the protection from the start?

I usually detect it when the 120 Hz buzz comes through my audio amplifier.

Using fast recovery or soft recovery rectifiers which store less charge and recover faster helps.  It is a problem with rectifiers from some manufacturers and not others and may vary between part numbers and lots.

The usual solution is to add a snubber across each rectifier.  10s to 100s of picofarads across each diode usually works well.  Alternatively a larger capacitor, like 0.01 microfarads, across the transformer secondary can work also.

[SkrillBill]

The rectifiers can produce noise during reverse recovery which gets coupled back into the AC power line although if that happens, it is better to fix the problem at the rectifiers.

I would leave the capacitors there unless you have some reason to remove them.

Other than leaving the caps, how would i fix that at the rectifiers? And, i don't really have a way to detect if that is happening. My only 'bench tool' is a Multimeter :/ . Should i just build in the protection from the start?

I usually detect it when the 120 Hz buzz comes through my audio amplifier.

Using fast recovery or soft recovery rectifiers which store less charge and recover faster helps.  It is a problem with rectifiers from some manufacturers and not others and may vary between part numbers and lots.

The usual solution is to add a snubber across each rectifier.  10s to 100s of picofarads across each diode usually works well.  Alternatively a larger capacitor, like 0.01 microfarads, across the transformer secondary can work also.

Google has shown me that are specific devices called snubbers, and also that capacitors can act like a snubber. Would i just a cap on the AC side of the rectifier?

[David Hess]

Google has shown me that are specific devices called snubbers, and also that capacitors can act like a snubber. Would i just a cap on the AC side of the rectifier?

A snubber limits dV/dT.  Usually it includes a capacitor and low valued resistor in series but in this case the resistor can be left out.

A low value film or ceramic capacitor, 0.01uF, 0.1uF, whatever, on the AC side of the rectifier but very close to it may be sufficient.  Alternatively, smaller capacitors across each diode will work; 10s to 100s of picofarads typically work well.  I have also seen EMI suppression beads added to each diode to accomplish the same thing; the added inductance of the beads limits the rate of change in current as the diode recovers.

[SkrillBill]

Google has shown me that are specific devices called snubbers, and also that capacitors can act like a snubber. Would i just a cap on the AC side of the rectifier?

A snubber limits dV/dT.  Usually it includes a capacitor and low valued resistor in series but in this case the resistor can be left out.

A low value film or ceramic capacitor, 0.01uF, 0.1uF, whatever, on the AC side of the rectifier but very close to it may be sufficient.  Alternatively, smaller capacitors across each diode will work; 10s to 100s of picofarads typically work well.  I have also seen EMI suppression beads added to each diode to accomplish the same thing; the added inductance of the beads limits the rate of change in current as the diode recovers.

will probably use a low-value capacitor on before the rectifier. I'm using a package rectifier instead of 4 physical diodes. Could this cap be connected directly across the AC leads of the rectifier?

[David Hess]

will probably use a low-value capacitor on before the rectifier. I'm using a package rectifier instead of 4 physical diodes. Could this cap be connected directly across the AC leads of the rectifier?

Yes, it should be mounted as close to the diodes as possible.

[SkrillBill]

will probably use a low-value capacitor on before the rectifier. I'm using a package rectifier instead of 4 physical diodes. Could this cap be connected directly across the AC leads of the rectifier?

Yes, it should be mounted as close to the diodes as possible.

https://www.digikey.com/schemeit/project/powersupply-v1-1-GDI647G303I0/

C11 - 100pF ceramic cap placement. It's directly after the transformer, so it should only see a max of about 36V.