Bridge rectifier/SMPS question

[yuka42]

Hello all. I could use some help understanding bridge rectifiers. I’m repairing an Allen and Heath ZedR16 PSU and am slowly understanding most of the circuit. I’ve never worked on an SMPS before. Can I expect the same DC voltage (around 80VDC from 120VAC) to be present with different rectifiers? Specifically, the circuit has a 2KBP206, would a KBP208 have the same DC output and can I safely replace it? The data sheet seems to have their max input voltage as the main difference. Forward voltage is 1.1V on both rectifiers.

I actually don’t think it needs replacing as it tests fine but the question is just more for my learning. Awake and pondering bridge rectifiers… haha. I guess if both rectifier’s forward voltage is the same then they should put out the same DC voltage correct? Is there other factors that need to be considered other than max voltage?

Another question for anyone familiar with SMPS power supplies, but can I safely power up the unit without the mosfet (f8nk100z or stp5nb100fp) and controller chip (UC3842an) installed to check voltages? I would like to make sure the controller isn’t getting too much voltage at VCC before I install it.

Not sure if I’m allowed to upload the schematic to this site but I can if it’s permitted. I’m pretty wary of uploading things I don’t have copyright to.

Thanks

Ryan

 

[radardoug]

Just meter the diodes, if they all measure good dont change the bridge. You can leave out fet and controller, but make sure after you check voltage on caps to discharge them. Also, 120 volts ac will give you 1.4 times that voltage as d.c.

 

[Khron]

Another question for anyone familiar with SMPS power supplies, but can I safely power up the unit without the mosfet (f8nk100z or stp5nb100fp) and controller chip (UC3842an) installed to check voltages? I would like to make sure the controller isn’t getting too much voltage at VCC before I install it.

That won't do much of anything. The UC3842 will, at best, be in some "hiccup mode", since it supplies itself from the main switching transformer, once it starts up. And if it's "getting too much voltage at VCC", it WILL let you know about its own discomfort by going "poof", long before you have time to probe anything.

Speaking of which, you'll wanna be VERY careful poking around the mains-input / primary area. AC isn't all that bad, but i still wouldn't recommend it. And high-voltage DC with a fat(ish) reservoir electrolytic won't be fun either.

If anything, using a series bulb tester (with an old-school ~100W tungsten filament bulb; nothing else will work) would do a lot more good.

https://www.bristolwatch.com/load_lamp.htm

 

[user 130681]

These power supplies are very reliable and the only faults I have found with them is the secondary diodes can fail putting the 3842 into protect mode.
It is not advisable to take the FET or 3842 out of circuit!!!

 

[ccaudle]

some help understanding bridge rectifiers

A rectifier bridge is just a convenient package for using four diodes together in a common configuration. By placing either multiple diodes on a single silicon die, or multiple die into a single package, you have a compact arrangement for the footprint, and you have a device which can be attached to a heatsink if needed. Individual diodes are typically difficult to attach to a heat sink until you get into the large TO style packages.

Can I expect the same DC voltage (around 80VDC from 120VAC) to be present with different rectifiers?

Yes, the primary difference in different bridge rectifier devices will be maximum allowed reverse voltage, maximum current, and if the package style is different the thermal differences associated with that package.

Although an SMPS typically has a rectifier connected directly to the AC line, so for 120V input you would have around 170V on the filter capacitor at the output of the rectifier. I guess I am assuming you mean an AC to DC SMPS, but mention of bridge rectifier would normally mean offline AC-to-DC design.

That is for lower power designs without power factor correction. Modern high power designs are required to use power factor correction in most jurisdictions, and if those designs are made to work with worldwide 100V-240V input ranges, they end up with around 370V on the output of the power factor correction capacitor.

the circuit has a 2KBP206, would a KBP208 have the same DC output

Do you have any links to datasheets? I am not able to find any data on a "2KBP206" device.

if both rectifier’s forward voltage is the same then they should put out the same DC voltage correct?

In general yes. The forward voltage is slightly dependent on current, so you would need to see if the forward voltage was for the same rated forward current for both devices. As a general principle if the replacement rectifier has same or higher maximum reverse voltage and forward current rating it should be OK. If the diode is being used for other than power line rectification (e.g. in a clipping circuit, as signal line protection, other signal processing use) there are probably more subtle effects to consider, but for 50Hz-60Hz rectification there are not a lot of issues to worry about. Sometimes cut-off behavior can cause changes in noise, but that is usually more a system performance issue than a safety or reliability issue.

can I safely power up the unit without the mosfet (f8nk100z or stp5nb100fp) and controller chip (UC3842an) installed to check voltages? I would like to make sure the controller isn’t getting too much voltage at VCC before I install it.

Make sure you understand the circuit, often there is a bootstrap circuit to get the controller powered up which has some small capacitor voltage divider, or resistive voltage divider, and the voltage may not necessarily be correct without the controller current load. I don't know without looking at the schematic, but if that is the case you may not necessarily get the correct voltage at that pin if the device is not in place.

 

[yuka42]

Okay thanks for the advice everyone. That makes a lot of sense. I actually built a dim bulb tester specifically for this repair. Right now there’s a 40w bulb in it. And I try to be extremely cautious. One hand in your pocket kinda deal. I’ll just make triple sure everything is in order before I power it up. I’ve already replace a few resistors that were open and one on the gate of the mosfet that was 75kohm when it should’ve been 22r according to the schematic and the Color bands. I replaced all the leaky caps in it a few years back after it failed the first time. It worked for a good three years but gave up the ghost again.

I’m sure I’ll have more questions. thanks everyone.

 

[yuka42]

So in this circuit the ac mains comes directly into the bridge rectifier and then branch’s out 170VDC to the primary on the transformer and the mosfet drain (with two diodes), and the controller vcc through parallel 150k resistors (to limit current I think?). The secondary windings on the transformer have varying voltages for the mixer and phantom power with diodes and regulators etc.

So just to clarify in my brain, the mosfet is used as a fast on/off switch which is controlled by the gate signal coming from the controller? Essentially making a high frequency square wave for the transformer? Is that correct?

Would this type of circuit need a load on the secondary to work? Can I power it up without installing it back into the chassis?

 

[yuka42]

Elektrotanya for electronics experts

Link to view the circuit.

 

[gswan]

This is a fairly rudimentary SMPS that appears to use the controller VCC as the regulation feedback. Regulation won't be all that good as the feedback is quite slow and hence they put linear regulators on the output.
Don't take out the MOSFET or controller chip. Instead check the secondary diodes and replace them if defective.
If they are all good then remove all the secondary diodes to leave the transformer secondary floating and check if it powers up and the VCC on the controller (which is generated from the feedback winding). If that looks OK then it's probably one of the regulators that may have failed. You can easily check these.

 

[Somkereki]

Just meter the diodes, if they all measure good dont change the bridge. You can leave out fet and controller, but make sure after you check voltage on caps to discharge them. Also, 120 volts ac will give you 1.4 times that voltage as d.c.

I agree ! The voltage should be 1.4 times higher if the capacitors are good! The weakest links! Few people know that their guaranteed lifetime is 1000 - 5000 hours (nowadays). I measure all electrolytic capacitors, that's the first! And they even depend on the temperature. The series loss resistance (E SR) can be high ! It can only be measured with a reliable measuring bridge! The hand tools are deceiving!

 

[yuka42]

I agree ! The voltage should be 1.4 times higher if the capacitors are good! The weakest links! Few people know that their guaranteed lifetime is 1000 - 5000 hours (nowadays). I measure all electrolytic capacitors, that's the first! And they even depend on the temperature. The series loss resistance (E SR) can be high ! It can only be measured with a reliable measuring bridge! The hand tools are deceiving!

Do you remove the caps to measure? Is there an easy way to tell if a cap is bad due to ESR?

 

[user 130681]

This is a sumary of my practical experience over the last 50 years of servicing electronics.
The easiest way is to look at the capacitor and check if it is bulging or not. Having said that, when they bulge, it is a long time after the ESR is too high to 'smooth' ripple from the supply line.
If capacitors are faulty, the power supply will produce output momentarilly, go into overvolt mode and shut down. If a secondary diode has gone short circuit, it will go into protect mode and 'chirp' at a very high frequency, easily observed with a good oscilloscope.
However, the capacitors can dry out and the only real way to test them is to remove them from the board and use a dedicated capacitance meter.
DVMs only give an idea, not a true reading.