BR101 Bridge rectifer wierd fault

[dicky96]

Hi guys I have this BR101 bridge here. It is definitely faulty, with about 9V going in to the rectifier I only get about 6.8V DC coming out of it (and the output voltage is unstable with load) and the bridge gets quite warm with less than 1 amp load.  If I increase the input voltage, the output doesn't increase proportionally.

I was a bit puzzled how a bridge rectifier could fail in this sort of way and I was seeing some odd readings around the bridge in circuit with my dmm on diode mode so I pulled the device and tested it on the bench

Here is what I found - these are Forward Biased voltages@

AC1 to negative 0.552V
AC2 to negative 0.501V

AC1 to positive 0.511V
AC2 to positive 0.503V

All above reverse biased = OC

AC1 to AC2 are OC in both directions

All this looks OK apart from some misbalanced voltages on the AC1 and AC2

But here's the really odd one....  testing from bridge negative to bridge positive reads OC with black meter lead to bridge negative and 0.960V with red lead to bridge negative!

OK so it's faulty. No worries. 

But I can't for the life of me figure out what is going on in there to produce those readings.    After all it's just four diodes in a bridge circuit.    I've seen faulty Bridges before - Short, Open, Burnt and Exploded. But never 'wierd'

Someone please enlighten me.

best regards

[Armadillo]

Hi, Your AC 9V should be connected to the 2 AC Leads, the + lead and the -ve lead is the output.

You sounded like you are connecting DC 9V into the bridge. What do you mean by increase the voltage?

Maybe you draw how you connect up so that we can understand better. 

[dicky96]

Hi Armadillo.

OK some more info, the device I am repairing an scrolling LED sign.  It had a 12V 2A DC PSU with it but that isn't the original one.  I powered it from my bench PSU and tried it at 12V DC as that was the same voltage as the PSU that I was given with it.   It draws less than an amp mostly but sometimes more depending on how many LEDs are lit on the scrolling text at the time.  I never saw it go near to 2A though.  I removed one side panel and there was a 7805 with unstable 6.8V going in and 4.8V coming out.  The regulator doesn't get hot

The scrolling sign works but at random times it make a beep from the built in buzzer and resets.  This seems to happen more after it has been on for some time.  I upped the voltage from my bench PSU to 12.5V and that seemed to make it work stable but after 20-30 mins it started beeping/resetting again.  I noticed if I reduced the supply voltage it would start to reset/beep continuously.

I then did a google search and found this device should have a 9V PSU.   

I tried lowering the input voltage to 9V but it was continuously beeping so I let it cool down.  After 15 mins I tried at 9V but it continuously beeped/reset.  Upping the supply to 9.5V it worked for some minutes then beeps/resets.

I then decided to dismantle it and found it has a bridge rectifier and a lot of smoothing electrolytics in parallel on the output of the bridge so I figured the original PSU is 9V AC, which probably gives about 12V DC on the smoothing caps at a guess.   Feeding DC instead of AC into the bridge shouldn't cause a problem, apart from only 2 diodes are conducting and effectively taking 100% of the load but seeing as this is a BR101 which is a 10A bridge I can't see that being a problem 

This is when I found out that with 9V DC going into the bridge I only get an unstable 6.8V DC coming out.  I then reversed the polarity of my bench PSU to use the other diode pair in the bridge (its a bridge after all so reversing the DC input polarity isn't gonna affect the output polarity) but that made no difference to the output voltage 6.8V.   I then tried varying the DC input voltage to the bridge and found the output voltage didn't vary proportionally to the input (2V more input didn't give 2V more output)  also the bridge was getting quite warm with only less that 1A load.

From there I tested the reading around the bridge with the DMM on diode mode, found some odd readings, unsoldered the bridge from the device, did some more readings on the DMM diode mode and posted the above question.

Still can't figure out how a bridge rectifier can fail in this manner.  Hence the post.

Also BTW I don't have the original PSU or another 9V AC one or similar.  I was thinking of using the 12V 2A PSU that came with it, and just soldering 2 Diodes in place of the bridge to protect against a reverse polarity PSU being attached, and still giving the normal voltage drop of a bridge rectifier.  I would have to check the 7805 is not overheating with 12V supply, it's bolted to the Chassis.   The 12V or whatever the electrolytics on the output of the bridge would charge up to with a 9V AC attached seems to be purely +ve supply to the LEDs

Rich

[Armadillo]

Hi, Firstly the 7805 will not be supplying 2A to the running LEDs load because it's running current is 1.5A about max.

Depending on the number of LEDs, the current could be quite high, so I suspect the 7805 is for control purpose only since LEDs need constant current source.

You should attached picture of the scrolling LEDs and the PCB so that we can understand better to advice you.

A picture is better than millions words.

[dicky96]

Sorry Armadillo I thought I explained clearly.  It's at my workshop so I can't get pics until tomorrow.  However.....

No the 7805 does not supply the +ve to the LEDs, it powers the logic board.   With the fault I only had about 6.8V (unstable) input to the 7805 and 4.8V output.  Obviously it wasn't dissipating any significant power then so was running cold.  This problem is almost certainly causing the reset/beeps as the 5V is only 4.8v and probably unstable as well.

The LED + comes directly from the Bridge +ve.

It generally draws less than 1A from my bench PSU with the current program running, though occasionally up to 1.5A if lots of LEDs are lit at the same time.  I suspect 2A would be sufficient to run it.  However at the moment the output voltage from the Bridge is lower than it should be so most likely the LEDs will draw more power once that is fixed, and be a bit brighter too.  Having said that, I didn't go so far as to see if this uses PWM to limit the LED current in which case a higher higher LED supply voltage should actually give less current draw.

I guess my question boils down to:

If this originally had a 9V AC supply, what voltage would have been on the smoothing caps after the bridge rectifier.  I imagine it would have been nearer to 12V as the caps tend to charge towards the peak voltage and there is quite a few of them in parallel so a decent amount of capacitance.?  I want to use a DC supply as I don't have a suitable AC one.   The LED sign works from my bench supply now (apart from the occasional resets as the output from the Bridge Rectifier is unstable)

And - Out of interest - how can a Bridge rectifier, which is basically just 4 encapsulated diodes in one block, fail in this odd way so as to give the readings on my DMM in diode test mode that I observed?

Rich

[wraper]

But here's the really odd one....  testing from bridge negative to bridge positive reads OC with black meter lead to bridge negative and 0.960V with red lead to bridge negative!

That's completely normal. You are measuring 2 diodes in series (x2 in parallel). You just get twice of forward voltage.

[Armadillo]

Hi;
The Linear voltage regulator like the 7805 requires a +3V to regulate properly. Meaning if the regulator is 5V, you must supply 5V + 3V = 8V to regulate properly.

Example 7812 is a 12 V regulator. Therefore supply voltage = 12 + 3 = 15Volts.

Hope it clarifies.

Unless it is a LD0 type linear voltage regulator.

EDIT: [THE VOLTAGE SUPPLY IS AT THE INPUT PIN OF THE REGULATOR. NOT YOUR POWER SUPPLY VOLTAGE]

[ArthurDent]

I also have a LED sign and one supply I used would cause it to reset when all the LEDs lit up in test mode because of the drop in output caused by the current drawn but that has little to do with your problem, just an interesting problem I had with a sign.

I'm guessing that if you remove the bridge and connect each of the diodes in the bridge forward biased across your bench power supply set to 2-4 amps in CC mode that you'll see that the forward drop isn't around .6-.7 volts but is much higher.

[trevatxtal]

Ref""OK some more info, the device I am repairing an scrolling LED sign.  It had a 12V 2A DC PSU with it but that isn't the original one.  I powered it from my bench PSU and tried it at 9V DC as that was the same voltage as the PSU that I was given with it.""

Is it possible that you are testing this devise with a DC input!
A Bridge rec requires AC input 9v ac will give 12X1.414v (12.726v) out peak less the voltage drop over a diode normally .7 volt (12.26v). This is under no load condition.
If you supply DC in then only 2 diodes of the four will work and the voltage out will be input minus diode drop.(9-.7=8.3v)

[Zero999]

Hi;
The Linear voltage regulator like the 7805 requires a +3V to regulate properly. Meaning if the regulator is 5V, you must supply 5V + 3V = 8V to regulate properly.

Example 7812 is a 12 V regulator. Therefore supply voltage = 12 + 3 = 15Volts.

Hope it clarifies.

Unless it is a LD0 type linear voltage regulator.

EDIT: [THE VOLTAGE SUPPLY IS AT THE INPUT PIN OF THE REGULATOR. NOT YOUR POWER SUPPLY VOLTAGE]

The LM7805 only needs 7V to regulate properly, but will will work at slightly lower voltages without dropping out, if the current draw is significantly lower than 1A.
http://ee-classes.usc.edu/ee459/library/datasheets/LM7805.pdf

I think you're right though. He said:

Sorry Armadillo I thought I explained clearly.  It's at my workshop so I can't get pics until tomorrow.  However.....

No the 7805 does not supply the +ve to the LEDs, it powers the logic board.   With the fault I only had about 6.8V (unstable) input to the 7805 and 4.8V output.  Obviously it wasn't dissipating any significant power then so was running cold.  This problem is almost certainly causing the reset/beeps as the 5V is only 4.8v and probably unstable as well.

The fact that the 6.8V is unstable implies you've got a lot of ripple. How stable is the 9V supply?

4.8V is within tolerance of the LM7805, so the circuit should work at that voltage but if there's a lot of ripple on the output, it could be dropping way below 4.8V, which would cause a problem.

2.2V of voltage drop is on the high end of what you should expect from a bridge but doesn't seem unreasonable. Are you sure the original power supply was 9VAC? I think you should try a 12VDC power supply. It shouldn't do any harm, as long as the smoothing capacitors are appropriately rated.

[Armadillo]

The LM7805 only needs 7V to regulate properly, but will will work at slightly lower voltages without dropping out, if the current draw is significantly lower than 1A.
http://ee-classes.usc.edu/ee459/library/datasheets/LM7805.pdf

The dropout voltage of LM7805 is 2V. So if you plainly think purely mathematically that 5 + 2 = 7V, then you are right.

But practically, please couple in the dynamic ripples of the input voltage and applies logic of tolerance allowance for a stable and worry free design considering that the design was originally meant for AC power input to bridge. I think "murphy law" does applies in Electronic.   I mean, how will you want to explain to variants of members here? Will you agree?

[dicky96]

To clarify again - this originally had a 9V AC PSU but that has gone AWOL so I was using my bench PSU to supply DC via the bridge. I was getting the resetting/beeping problem with supply voltages from 9.5V to 12.5V.  Below 9.5V it wouldn't boot properly

Today I put a couple of 3A diodes in place of the Bridge Rectifier and powered it up again from my bench PSU at 12V.  It worked reliably and the current draw peaked at 2A at occasional points when there was lots of yellow scrolling text (red+green LEDs illuminated) but 95% of the time it was around 1A

I then tried with the 12V 2A PSU I got with the LED sign and it worked reliably.  I'm getting about 10.5V across the smoothing caps now but noted I have about .4V drop on the wires from my Bench PSU, plus the voltage drop of the two diodes so that sounds about right.  The 10.5V is stable now, not unstable like it was with the original Bridge Rectifier in circuit.  So it's working nicely using a couple of diodes instead of the bridge and with the available 12V DC PSU.  The 7805 does not get hot.

best regards
Rich

[Armadillo]

Great, nice to hear that.

But I have a question.

If you now intend to use your DC power supply, Why need the Diode?

[dicky96]

I would have to solder something across the AC to DC terminals where the bridge was previously located, be it wire links or something else

2. Protection against someone plugging in a reverse polarised DC PSU in the future (customers tend to do that sort of thing)

3. I wasn't 100% sure what voltage the original 9V AC PSU would have been giving on the smoothing capacitors when under load - so  I decided to put a couple of diodes instead of the bridge to drop the 12V down a volt or so

Possibly not necessary but the sign is working happily enough so it didn't do any harm

Rich