Current just keeps on going up...

Thread Starter

jwilk13

Joined Jun 15, 2011
228
Hello again,

I have a circuit that I'm working on, and I'm getting some weird results when I test it. I've attached the schematic for everyone's reference. When testing, I connect a 12V regulated DC power supply to pins 1 and 2 on header J2. When I power on the power supply, the current it supplies is approximately 10 mA. After about 10 seconds, this number starts to increase, slowly going up to 50 mA, then very rapidly increasing upwards of 750 mA until I turn it off. This doesn't seem right to me and I'm wondering if anyone has encountered a similar issue here.

I've ruled out the microcontroller: This still happens when in debug mode in MPLAB without stepping through any code.

I've ruled out the regulators: They still provide constant output voltage up until the current limit of my power supply kicks in.

I've ruled out the BSP762T's: I have them completely disconnected from power, problem still occurs.

I'm going to re-re-re-check everything to make sure it's wired correctly and try to figure out which portion of the circuit is drawing the excess current.

Any thoughts? Pay particular attention to the optocoupler. I've never used them before and I'm not confident I connected it properly. The one I'm using is a FET, not like the schematic shows. The datasheet for mine has both terminals of the FET labeled as "Output". Thanks everyone :)

 

Thread Starter

jwilk13

Joined Jun 15, 2011
228
Update: I removed the opto from the circuit and tested it with pin RC3 of the uC grounded and the problem still occurs.
 

Adjuster

Joined Dec 26, 2010
2,148
You may have an electrolytic capacitor connected the wrong way round. Check this before trying again, because such components can suddenly burst and eject hot (and possibly corrosive) contents under such conditions.

Your opto-coupler output transistor looks wrong (an NPN device is shown, with its collector to GND and emitter to a pull-up resistor to 3.3V) That said, unless "RC3" is capable of sourcing serious current I don't think this is where the amps are going.
 

Thread Starter

jwilk13

Joined Jun 15, 2011
228
Those symbols that look like electrolytic capacitors are actually tantalum, I just used them to specify polarity. I'll double check to make sure they're connected properly.

About the opto, it's a FET design. You can check out the data sheet here if you would like. I was curious about that too, but even if I connect it with collector to a pull up resistor and the emitter to ground, I still get the same output.

Thanks for the input :)
 

Thread Starter

jwilk13

Joined Jun 15, 2011
228
Just checked the tantalums, and both are connected properly. Both of my voltage regulators are working fine and the current being drawn from each is staying the same despite the rise in current being delivered by the power supply. Without 12V being hooked up to the BSP762T's, I'm not sure where else current could be going.
 

Thread Starter

jwilk13

Joined Jun 15, 2011
228
I bought all of the parts to make two of these contraptions, so I'm going to start working on a new one. I'm still open to suggestions as to what may be causing this oddity though.
 

steveb

Joined Jul 3, 2008
2,436
You checked the polarity on the capacitors C2 and C3, but did you check the rated working voltage? Ideally, this should be 25 VDC for reliability, but if they are 12 VDC or less, then this could be the issue. If the rating is correct, you could still have a bad capacitor. Try taking C2 and C3 out and wiring in electrolytic capacitors of the same capacitance and > 16 VDC rating.

Also, do you have sufficient heat sinking on the regulators?
 

Adjuster

Joined Dec 26, 2010
2,148
Faulty tantalum capacitors are notorious for drawing lots of current, not necessarily going to a dead short, so that they may get very hot.

Of course, whatever is drawing this much current must be dissipating a lot of power, so heat may give you a clue.

You must not of course touch anything with the power on. You must also be extremely careful not to touch anything so hot that you burn yourself.

Equally, don't leave excessive current running for long, as more damage may be done, and things may get very hot or even catch alight!
 

Thread Starter

jwilk13

Joined Jun 15, 2011
228
Thanks steveb, that was it!

The tantalums I have are rated 25V, so they should work fine. What you were right about was that one of them was bad, more specifically, C1. This was the tantalum connected from the 5V regulator output to ground. I replaced the tantalums one at a time with 60V electrolytics.

Quick question: Would the capacitance of these capacitors have any affect on what I was experiencing? The electrolytics are 68 uF and the tantalums are 47 uF. The data sheets for those regulators say anything above 33 uF (3.3V) and 2.2 uF (5V) will work.

Thanks again steve.
 

Adjuster

Joined Dec 26, 2010
2,148
Some regulators specify a maximum load capacitance, and some, particularly "low dropout" types, have strict rules about the effective series resistance (esr) of any capacitors used.

Getting this wrong can lead to instability (i.e. oscillations). It may be worth checking the fine print of the datasheets.
 

Thread Starter

jwilk13

Joined Jun 15, 2011
228
Good call.

The only parameter I'm not sure is met is the resonant frequency of the cap. The regulator says it should be above 500 kHz, but there's nothing in the cap's datasheet about it. There's a plot of frequency vs. capacitance, and the capacitance drops off around 100 kHz. It still maintains a high enough capacitance to satisfy the 2.2 uF of the regulator, so I'm not sure this is what's causing the problem.

I'll plug another tantalum in to see if it was just a faulty capacitor.
 

Thread Starter

jwilk13

Joined Jun 15, 2011
228
It seems that the problem was simply a faulty capacitor. I've tested the 5V regulator with the same model tantalum as before with different resistive loads and the problem is not occurring.

Maybe someone can answer this though: What affect does not having a sufficient capacitor resonant frequency have on the circuit?
 
Top