Actually, the heading would be more accurate as "Mis-reading" rather than mis-leading.

As I have mentioned in previous posts, I run a small club for electronics enthusiasts. Members experience varies from novice to expert and we endeavour to assist and educate when any problems are encountered.
One of the on-going projects being run by several members, is the constuction of 3D pinters and CNC machine control from parts salvaged from old printers and other parts that have been donated to the club.
One donation was a large box of various bridge rectifiers that had been picked up at a computer fair. The part numbers that were printed on some KBPC35-00's was a little blurred, but otherwise seemed ok. 35 Volts and max average current seemed ideal for the projects power supply that required 10volts at 4 amps after regulation.
Secondary of the Torroidal mains transformer used was rated 12-0-12 volts rms @ 7amps. Off load it actually produced 13.6 Volts rms, the peak being just over 19 Volts. They had connected only half of the bridge to provide a positive output, the negative terminal was not used yet. (negative was the transformer center tap at this point in testing)
To all intents and purposes, this should have been ok, but after two rectifiers failed, (not caused by transients, as the mains is isolated and has transient suppressors) I did a couple of tests on the remainders.
Using an isolated supply and a variac I carried out a reverse voltage leakage test and found that they were way off from the specified levels from the data sheet.
When studying the spec sheet, some of the parameters are shown as "Per element" meaning each individual diode, so that if each diode has a maximum recommended reverse voltage of 17.5 Volts, then when used as a bridge rectifier, two diodes are in series at any one time. Not quite the configuration that they were being used in. By using higher voltage devices, the problem was solved.
It seems as though the manufacturer/supplier had rated the device at the breakdown voltage of two diodes in series for those that failed as each individual diode was breaking down at around the peak output of the transformer. Sold as Bridge rectifiers, and expected to be used as such.
Moral being, Check spec sheets carefully, select parts wth a sufficient overating to handle expected peaks, and buy from known reliable sources.

 

Good to know and commit to my memory bank.

 

I have used that bridge, I used the KBPC3506 of the KBPC15, 25, 35/W SERIES. Looking at the linked data sheet I see the Forward Voltage Drop listed as 1.2 volt and that is the only per element I see. Then too,there is likely a dozen data sheets out there for that bridge rectifier. Over the years I never had an issue with any I used mostly in linear power supply applications.

Ron

 

I have used that bridge, I used the KBPC3506 of the KBPC15, 25, 35/W SERIES. Looking at the linked data sheet I see the Forward Voltage Drop listed as 1.2 volt and that is the only per element I see. Then too,there is likely a dozen data sheets out there for that bridge rectifier. Over the years I never had an issue with any I used mostly in linear power supply applications.

Ron

Yes as you say, many different data sheets!
I think the ones we had were possibly re-branded rejects as they came from an unknown source.

 

I recall this topic from my college days (more than 40 years ago).

Consider a bridge rectifier rectifying 240Vac mains. The rectified dc voltage will be 340Vdc, but the Live input will be varying from +340V to –340V with reference to the Neutral potential.

As shown in the attached diagram, the peak reverse voltage across diode D1 of the bridge rectifier will be 680V (based on 240Vac input).
Therefore the reverse diode voltage will be twice root 2 of the rms ac voltage.

 

Here are useful explanations, but for more information, there are dozens of articles describing rectifier configurations if you search the internet.
Also take a look at "alternating current" in the education tutorials on this site.

 

I recall this topic from my college days (more than 40 years ago).

Consider a bridge rectifier rectifying 240Vac mains. The rectified dc voltage will be 340Vdc, but the Live input will be varying from +340V to –340V with reference to the Neutral potential.

As shown in the attached diagram, the peak reverse voltage across diode D1 of the bridge rectifier will be 680V (based on 240Vac input).
Therefore the reverse diode voltage will be twice root 2 of the rms ac voltage.View attachment 168094

Your bridge rectifier looks a little peculiar unless I am missing something. Should the anodes of the line marked live be tied?

Ron

 

Yes as you say, many different data sheets!
I think the ones we had were possibly re-branded rejects as they came from an unknown source.

That would explain quite a bit. By the way I think what you are doing with your group is really cool. Reminds me of older days with "radio clubs".

Ron

 

Your bridge rectifier looks a little peculiar unless I am missing something. Should the anodes of the line marked live be tied?

Ron

one of the diodes is the wrong way round (doh!)

 

13.6Vrms is 19.2V peak. That is already above the rated reverse voltage of the recitifier, being two diodes in series or not. Then consider what Hymie said, which you must never forget when dealing with rectifiers in gerneral, specifying at least double the peak working voltage is almost allways necessary.
A 40V diode would be awfully close to death all the time, 50V might be an option but I would go with at least 60V.

If you want reliability, derating to 2/3 of the specs is a good way to achieve it. For critical components where it would have a really bad effect on the overall circuit (much higher Rdson in a mosfet, much higher price...) you can go closer to the limit, but you need to verify or ensure that in worst case it doesn´t get too close to it, for instance line voltage in europe is specified as 230V +15/-10% and that should be one of the main concerns.

 

one of the diodes is the wrong way round (doh!)

Been there and done that!

Ron

 

13.6Vrms is 19.2V peak. That is already above the rated reverse voltage of the recitifier, being two diodes in series or not. Then consider what Hymie said, which you must never forget when dealing with rectifiers in gerneral, specifying at least double the peak working voltage is almost allways necessary.
A 40V diode would be awfully close to death all the time, 50V might be an option but I would go with at least 60V.

If you want reliability, derating to 2/3 of the specs is a good way to achieve it. For critical components where it would have a really bad effect on the overall circuit (much higher Rdson in a mosfet, much higher price...) you can go closer to the limit, but you need to verify or ensure that in worst case it doesn´t get too close to it, for instance line voltage in europe is specified as 230V +15/-10% and that should be one of the main concerns.

Yes I agree, when they selected the rectifier based on the data sheet, they had assumed that as the reverse voltage was stated as 35Volts rms, meant that the peak would be 49.9 Volts and would therefore be ok, but as you say, a bit too close to the edge. Possibly an in spec quality device would have survived, but these on test showed to be far below spec.
The one I gave as a replacement was the 100Volt version so plenty of headroom with that.

 

I have always gone much higher with the PIV ratings because of the need for reliability. The bridges mentioned would be OK for a five or six volt supply at the most. So now you know.
"Specsmanship" is still prevalent in some areas, it appears.

 

I have always gone much higher with the PIV ratings because of the need for reliability. The bridges mentioned would be OK for a five or six volt supply at the most. So now you know.
"Specsmanship" is still prevalent in some areas, it appears.

It can be difficult to know from some of the spec sheets whether the rating is "per element" or is the rating when used as designed as a bridge rectifier where 2 diodes are in series at any one time.
Using the minimum rule of thumb guide of 3:1 rating, would mean that 13.6V rms will be 19.2V peak. which gives 57.7 Volts when x 3 rule applied. So even though the info from the data that shows the rating being 35 Volts rms, which is 49.5 Volts peak, the selection of the rectifier was underated. Even more so, as it was being used in a bi-phase configuration with only one diode at a time in reverse condition.
I prefer to use 3:1 as an absolute minimum and would normally go for 4:1 as a safegaurd. Hence the replacement that I had on hand, being rated at 100 Volts , provided a nearly 5:1 margin.
I think it is important to look at the actual manufactures spec sheet for the device, rather than for a generic one that may or may not be the same. I uploaded the spec sheet for that series of rectifiers, but it is still "buyer beware" when using parts from unknown sources.

 

Here is a useful video that may help when choosing bridge rectifiers.

Also see the attached file from Hammond mfg.