Hi all,

This is my first post, so please bear with me if I've done anything wrong. I've tried search the forum and have had no luck.

I'm using a Traco-Power TMLM04 switch mode power supply to power a very basic circuit with a PIC16F722. I've been having real problems with the PIC resetting it self due to brown-outs. I scoped the 3.3V DC power line (the output from the TMLM04 module) and it was very very noisy.

The data sheet for the TMLM04 module states that with a constant load, the maximum output ripple should be 250mV. To check that the error wasn't in the circuit I had made, I connected the module to a very basic circuit with just a 15Ohm load resistor. This gives a constant load of ~220mA (max output is 1200mA), and I scoped the output again. The noise from the power supply module is just as bad as with the original circuit and I'm seeing ripple of up to 2.5V at the worst points.

I have about 50 of the Traco Power modules, of which I've tested about 10 in total which span 3 different batches and they're all the same. So I'm sure it must be something I'm doing wrong. I've tried adding a 2200uF capacitor on the 3v3 line, which helps a little, but not much.

Here is a screen shot from the 3v3 line of the basic (15 ohm resistor only) circuit.

The centre line is 0v, the volts per division is 1V and the time per division is 5mS.

There is a copy of the data sheet and extra information available from here:
http://www.tracopower.com/products/ac-dc-power-supplies/encapsulated-modules/

Any help with how to reduce the noise to make it suitable for use with a uController would be much appreciated.

 

Have you looked for any product reviews of Traco power modules?

Have you contacted Traco and asked them what might be causing this?

Those might yield something useful.

 

Thanks for the speedy reply!

I have looked for product reviews, but wasn't able to find any. I have contacted Traco-Power but, as yet, they haven't got back to me.

Does anyone know why a 2200uF capacitor wouldn't smooth out at least some of the noise? I'm wondering if its related to the source being a switch mode, not a linear supply, but I would of expected it to remove a large amount of the noise, but it makes almost no difference

 

What is the input voltage source?

These supplies seem to be operating abnormally. Don't see any obvious reason for that.

 

The input voltage source is UK mains via a MOV to make sure the supply is clean. The input specs to the switch mode module are very relaxed (90-264V and 47-440Hz) and I've tested other mains connected supplies with the same set up and they haven't had any problems.

I'm not sure what else it could be though, if the input is correct and the output load is correct, then that just leaves the module itself?

 

What does the output look like with a smaller (or no) load?

 

I've tested with no load, 220mA, 440mA and 880mA. No load gave by far the best results, although it was still very noisy. And the noise gets gradually worse the higher the load.

 

Looks to me like there is a problem in the Traco supply's input filter section; most likely the filter capacitor(s) are far below the manufacturer's specifications. The supply output should hold up under full load for 15mS after power is removed; yours are not holding up with 1/5 of the rated load for even a millisecond after the input voltage zero crossing.

Hopefully, you purchased them from an authorized distributor; if so you should have no problem in returning/exchanging them.

If they had been sitting dormant for a long time, the electrolytic caps on the input filter could have lost their dielectric properties - or the supplies might be a victim of the Taiwanese "capacitor plague":
http://en.wikipedia.org/wiki/Capacitor_plague
Check the manufacture date of the supplies; there usually should be a MFG date code on the case somewhere.

 

I think you're right and they're going to have to be returned. Out of interest, I was surprised to see such straight lines when the unit would have a decent amount of capacitance to keep the output smooth.

So I put a 2200uF capacitor in parallel as follows, but it made hardly any difference.

I guess that must be because during the time the unit is in error it is providing a lower resistance pathway to ground, which is discharging the capacitor faster. Does that sound about right, or have I missed something?

 

Well, I'm not going to try to reverse-engineer the supply, but if the switch kept closing even when the input was lower than the output, then yes it would drain the output cap.

 

The noise seems to be exactly 10mS period, so is coming at 100Hz which is the rectifier freq from your 50Hz mains.

Is this a fully isolated SMPS module? Maybe the noise is some type of ground loop issue or scope isolation issue.

For a 3.3v DC SMPS to have >1v noise ripple on the output means something is VERY wrong. And if they ALL do it I'm thinking it's something to do with your wiring or measurement procedure.

Is that a PC based 'scope?