Hi all,

I have a design where I'm using the 5V supply from a USB PC connection with a maximum of 100 mA. I have an RFID module that's using these 5V as it's main voltage input. I thought that by putting a 100uF cap from Vcc to Gnd at the module input I would make the 5V ripple-free and very stable. Well, it didn't go as planned. The design works fine without the 100uF, but module fails to work when I use the 100uF. Can someone please explain to me why my design doesn't work with the 100uF? I'm using the following cap. Thanks.

Robert

https://www.mouser.com/Search/Produ...tualkey64800000virtualkey81-GRM31CD80J107ME9L

 

Please provide schematic.
Adding a cap with a random value related to the used current is not the correct way.
the relation between cap and current Vrip=( I x discharge time) / Cap approximation 4 x I / Cap
When output impedance pwrsup is to high in relation to your electronics circuit power will drop due to high cap and system could run into a halt.
However; to verify this way of thinking we need more info like schematic.

Picbuster

 

Some stabilizer circuits at some capacitive load cease to function correctly (oscillation-relaxation is observed). In my development (to avoid this) I applied a resistor connected in series with the capacitor. Try to start to put 1 ohm.

 

This interests me so I am posting so I can follow. I am a newbie and it seems axiomatic that you always put a capacitor at voltage input to eliminate ripple. What was not explained was how to size it.

 

There should be no (well, almost no) ripple in the voltage from a USB port. The reason for adding a capacitor is to reduce the voltage drop caused by the inductance of the supply wires when there is a sudden change in the current drawn by the device. If the current is relatively constant, the capacitor is not needed.

Bob

 

The capacitor may be fooling the USB port into thinking it is shorted. It then shuts off its voltage to protect itself.

 

The capacitor may be fooling the USB port into thinking it is shorted. It then shuts off its voltage to protect itself.

That was my first thought.

PIC-User, does the 5 volts drop out when you insert the cap, or does the circuit simply not work with it installed?

If you have a powered port on that PC try in in there, or some other 5 volt supply such as a portable charger.

 

There should be no (well, almost no) ripple in the voltage from a USB port. The reason for adding a capacitor is to reduce the voltage drop caused by the inductance of the supply wires when there is a sudden change in the current drawn by the device. If the current is relatively constant, the capacitor is not needed.

Bob

Hi Bob, no the current is not constant. When the module is reading the current peaks to 55mA. I'm getting good results with a 10uF cap instead of the 100 uF one.

 

That was my first thought.

PIC-User, does the 5 volts drop out when you insert the cap, or does the circuit simply not work with it installed?

If you have a powered port on that PC try in in there, or some other 5 volt supply such as a portable charger.

Hi, no I have tested that and is not the USB port shutting down. For some reason the module stops working and I suspect it's the ripples in the Vin.

 

Did you actually measure this "alleged" ripple, or did you just assume it was there?

 

Some stabilizer circuits at some capacitive load cease to function correctly (oscillation-relaxation is observed). In my development (to avoid this) I applied a resistor connected in series with the capacitor. Try to start to put 1 ohm.

Did you actually measure this "alleged" ripple, or did you just assume it was there?

@PIC-User I agree with quoted posts cuz SMPSUs can be unpredictable with large reactive loads! So my suggestion is 2 just look across module pwr input with TDO cuz knowing is totally better troubleshooting strategy than speculation

 

Did you actually measure this "alleged" ripple, or did you just assume it was there?

Hmm, no I didn't measure it. I'm just assuming thar thus might be the problem. I tried to measure it with an oscilloscope, but I didn't see a big change.

 

If the current is relatively constant, the capacitor is not needed.

Disagree. The capacitor lowers the perceived output impedance of the power source. This is important for all circuits, and critical for anything involving high frequencies. While there probably is some power decoupling on the module, I'd be surprised if the data sheet didn't require, or at least suggest some external capacitance at the power entry pins.

ak

 

Bob, yes you are right. I see a difference when I put a 10uF cap. My module is way more stable. What u don't understand is why the 100uF cap doesn't work as well. I would think that the circuit would be even more stable, but is not. Ooh well, maybe it requires a large current to get charged and that creates issues.

 

I tried to measure it with an oscilloscope, but I didn't see a big change.

Pic-user To make seeing noise easier plz make sure _scope's_ vertical input channel is set for AC coupling and has BW ≥ 100Mhz. Hope this helps!

 

I would think that the circuit would be even more stable, but is not. Ooh well, maybe it requires a large current to get charged and that creates issues.

Pic-user that's always possible but based on my experiance with SMPSUs I'm betting on Bordodynov's explaination (in post 3)

 

Another thought. Could the 100uF cap have a higher ESR than the 10uF cap? For instance, a 100uF aluminum electrolytic versus a 10uF tantalum.

 

Pic-user To make seeing noise easier plz make sure _scope's_ vertical input channel is set for AC coupling and has BW ≥ 100Mhz. Hope this helps!

Thanks. That helped. Yes, after using ac coupling and a horizontal scale of 10nS there was a ripple that was visible. The peak to peak voltage was about 0.3V.

 

Pic-user that's always possible but based on my experiance with SMPSUs I'm betting on Bordodynov's explaination (in post 3)

I used a 10 Ohms resistor in series with the 100uF cap and the ripple was still there. I'm not sure if this ripple is causing my module to malfunction.

 

Can you take a stab at estimating the frequency of the ripple? To me ripple is an artifact of converting AC to DC and should show up as a periodic waveform at twice the power line frequency. Ripple is what you see from an unregulated DC power supply. Once you run the DC with ripple through a voltage regulator, you don't see ripple any more; what you see are other artifacts and their frequency can be a clue to the source and the solution. A scope picture would help immensely.