Dc pwm through full bridge failure

Thread Starter

Notebook64

Joined Jul 13, 2015
33
Hi all and thanks for reading
will provide more info tomorrow
using existing equipment, a 28 vdc pwm to 12 vdc through a bridge rectifier supplying an led. I know, no reason, but making use of existing kit.
the rectifier is failing after a few months use. It Blows and trips the ecu.
also, it appears to happen more frequently to an identical circuit / supply that has a much longer lead to the circuit from an ecu supplying the pwm.
any obvious reasons that this would occur?
will update tomorrow
thanks
 

Tonyr1084

Joined Sep 24, 2015
7,853
OK, wondering why you're using a full wave bridge on a DC circuit. Just using a 28 volt to 12 volt buck converter should do just fine. No need for rectifiers unless you meant to say your 28V is AC and not DC. If that's the case are you using any other components like a filter cap maybe? If so, 28VAC when rectified and filtered will be 39.6VDC (filtered). Are your calculations based on 28V or on 39.6V? You could be over stressing the rectifiers in your bridge. At the higher voltage your LED circuit will draw a higher current. If the diodes are rated for 1/2 watt of power and you're pushing 3/4 watt through them they will fail. I'm just guessing but I think the error might be in the numbers - of course, assuming you're starting voltage is 28V AC.

[edit] Oh, wait - you said 28V PWM. Different story. We're dealing with DC, so there should be no need for a bridge rectifier. Your 28V should be on for 42.9% of the time and off the balance to give you an average 12 volts. Depending on the frequency of the PWM (not the period), the diode might not be fast enough to switch on and off. But it's not blocking any reverse current, so I'm still confused by the use of the bridge. Well, I'll just have to wait for further clarification.
 

Thread Starter

Notebook64

Joined Jul 13, 2015
33
OK, wondering why you're using a full wave bridge on a DC circuit. Just using a 28 volt to 12 volt buck converter should do just fine. No need for rectifiers unless you meant to say your 28V is AC and not DC. If that's the case are you using any other components like a filter cap maybe? If so, 28VAC when rectified and filtered will be 39.6VDC (filtered). Are your calculations based on 28V or on 39.6V? You could be over stressing the rectifiers in your bridge. At the higher voltage your LED circuit will draw a higher current. If the diodes are rated for 1/2 watt of power and you're pushing 3/4 watt through them they will fail. I'm just guessing but I think the error might be in the numbers - of course, assuming you're starting voltage is 28V AC.

[edit] Oh, wait - you said 28V PWM. Different story. We're dealing with DC, so there should be no need for a bridge rectifier. Your 28V should be on for 42.9% of the time and off the balance to give you an average 12 volts. Depending on the frequency of the PWM (not the period), the diode might not be fast enough to switch on and off. But it's not blocking any reverse current, so I'm still confused by the use of the bridge. Well, I'll just have to wait for further clarification.
Thanks. Wasn’t my choice. It made use of existing equipment as I said. The unit was already built. These are failing and I needed to know why.
thanks
 

ronsimpson

Joined Oct 7, 2019
2,988
You could add a 20k resistor in the wire and forget the PWM.
the rectifier is failing after a few months use. It Blows and trips the ecu.
The diodes are good to 50 volts, 140mA. That seems good. Maybe some one in China switched to a lower voltage part.
Maybe a wire touches the case or another wire.
 

Tonyr1084

Joined Sep 24, 2015
7,853
Hmmm.

12V (lets assume DC).
BRBGX50A (common forward voltage drop 1.4V)
R1 = 16KΩ
LED1 (common forward voltage drop 2.5 typical)

12V - 1.4Vf - 2.5V = 8.1V
8.1V ÷ 16KΩ = 0.00050625A (506 micro-amps (not milliamps))
506µA x 12V = 6mW

I don't see where anything should be burning up. Or even working for that fact.

If this circuit isn't working properly then start over and build your own. No need for PWM or a BR (Bridge Rectifier). The board as you've drawn it (or someone else drew) can work on AC or DC. And if with DC it can be connected with either polarity. In my opinion it's a lot of wasted circuitry for a simple LED. If there's something you want to build then develop a plan and a schematic and then post that. We'll help you understand the circuit and make it work the way you want it to. If you want dimming capability we can assist with that as well.
 

Tonyr1084

Joined Sep 24, 2015
7,853
To be perfectly honest, 28VDC without the BR should be, assuming the numbers in post #12:
28V - 2.5Vf = 25.5V
25.5V ÷ 16KΩ = 0.00159375A (1.6mA)
1.6mA x 28V = 44.6mW.

Even that shouldn't work.

The common LED (green in this case) should operate from 2mA (not micro) to 30mA. And at 2mA it should be barely glowing. Good target range for current would be 15 to 20mA. Using a 16KΩ resistor, to achieve 20mA you'd need a starting voltage of 320VDC. So we can't see how this functions - OR fails with 28V AC OR DC.
 
Last edited:

Thread Starter

Notebook64

Joined Jul 13, 2015
33
The output to the bridge is 28v dc pwm to 12v
If I can reiterate, it’s why it might have failed is my only question.
thanks.
 

Alec_t

Joined Sep 17, 2013
14,280
I agree with Tony. If your specified voltages and component values are correct, logic dictates that you must have a fault upstream of the bridge, such that the bridge diodes are seeing more than their rated 50V; or else there is an excessively low impedance load in parallel with the LED and its 16k resistor, drawing more than the rated 140mA.
 

Thread Starter

Notebook64

Joined Jul 13, 2015
33
I agree with Tony. If your specified voltages and component values are correct, logic dictates that you must have a fault upstream of the bridge, such that the bridge diodes are seeing more than their rated 50V; or else there is an excessively low impedance load in parallel with the LED and its 16k resistor, drawing more than the rated 140mA.
Thanks
They have worked for a year or two and there are about 100 say. Failing at about 15%
ok
The supply voltage and also pwm frequency suspect?
Did electronics many many years ago.
As mentioned, the one more distant from the ecu appears to fail more.
thanks
 
Top