1
\$\begingroup\$

A uController circuit is powered with an AC-DC converter. This circuit in turn acts on some relays to turn on and off some loads. We have two bi-stable relays and another simple relay (Omron G5DN-1A) schematic. The bi-stable realays cause no problem, but once the Omron relay is connected, power on the DC side randomly drops every and causes the uController to reset. It is not necessary for the relay to act, doesn't matter whether its open or closed or even if the relay has or has not any loads at its output. Just with the relay being connected on the PCB. This happens with a cheap AC-DC converter we bought on Aliexpress but not with other more expensive converters like VSK-S1 ac-dc converter. We have added some capacitors at the 5V output and and input filter and the voltage drop is reduced but not eliminated.enter image description here. Any ideas? We think maybe the converter used is too high output current for our consumption (700mA vs 50mA). the other we bought and work are 200mA.

\$\endgroup\$
4
  • \$\begingroup\$ Where is your uC in the schematic? What’s your decoupling situation around it? \$\endgroup\$
    – winny
    Apr 4, 2019 at 16:02
  • \$\begingroup\$ These images do not tell the whole problem. How random? What scope V scale, where, when? How much ground shift? What triggers this "random event"? Can you shunt nodes to rule out some sources? \$\endgroup\$ Apr 4, 2019 at 16:24
  • \$\begingroup\$ It is very random. It also happens when connecting the programming cable. The curve is at 5V and drops to 2.8V. It lasts 48ms until it recovers. What do you mean by shunt nodes? We can try anything that can lead us to fixing it \$\endgroup\$
    – solartec
    Apr 4, 2019 at 16:40
  • \$\begingroup\$ uC is decoupled and we have a good ground plane well isolated from the AC side. However we have ruled out noise being induced that can do the reset as we have seen that the actual voltage output of the converter is dropping to 2.8V \$\endgroup\$
    – solartec
    Apr 4, 2019 at 16:41

2 Answers 2

1
\$\begingroup\$

Your last comment is confusing/incorrect. The power adapter current rating is how much current ( amps ) the supply can provide.

This happens with a cheap AC-DC converter we bought on Aliexpress but not with other more expensive converters like VSK-S1

I think you have your answer. Those cheap adapters from China are not always what they claim! Here is an EDN article, showing the shoddy and dangerous things coming from China.

Teardown--12V-AC-adapters---The-Horror

There are several blogs showing cheap-o supplies rated for 1Amp can only provide 0.5Amps!

Also, there is more to power supply hold up than just more caps. The supply has a transient response based upon its control loop. If that is not well designed, then adding bulk caps is just a band-aid. Stick with a quality supply and your woes will be far less, and your product will work better and last longer.

\$\endgroup\$
2
  • \$\begingroup\$ The ac-dc adapter schematic is a copy of the LNK362 recommended circuit. I also suspect it has to do with the control loop and wandering how it can be fixed. The expensive converters we bought are overpriced for this project and it cannot afford it \$\endgroup\$
    – solartec
    Apr 4, 2019 at 16:38
  • \$\begingroup\$ We have tried with another cheap 12V adapter with 300mA output rating and that one has no problem \$\endgroup\$
    – solartec
    Apr 4, 2019 at 16:43
0
\$\begingroup\$

It's not the current sourcing abilities of the AC to DC converter that is causing the problem, it is how fast the converter can source current. Apparently the converters source impedance is high at high frequencies.

In some datasheets this is listed as response time, but output characteristics on many supplies are not listed and this becomes problematic if there is a current source issue.

I've dealt with a problem similar to this with relays and current through a transformer, relays are a large inductive load when switched on because of their inductive nature.

If you are forced into using that supply from a cost or regulatory perspective, the only way to solve the issue would be to add energy storage to cover the gap when the relay is thrown on. I have used capacitors in the past, but at some point it interferes with the supply and the turn on time, most supplies have a capacitance limit.

You may be able to use an inductor (or inductors) to separate the supply from the relay and microprocessor.

Other than that a power supply is with enough sourcing capability is the best option.

\$\endgroup\$
4
  • \$\begingroup\$ ty. We are not forced into using this converter, but yes a simple and cheap option. Trying to understand the problem with this circuit so we can build our own, or have another circuit option with other components. What component of our AC-DC circuit do you think causes the problem? The LNK364 with the opto isolator and feedback circuit? I must add that we have isolated the relay trigger and triggering with another isolated laboratory power supply and the problems persists. So the relay is only connected to our circuit on the AC side. \$\endgroup\$
    – solartec
    Apr 4, 2019 at 17:08
  • \$\begingroup\$ The most likely problem is the supply cant source current fast enough and the voltage drop is causing the microprocessor to reset. If Vcc on the microprocessor dips below 5-10% it can cause metastablity (check the micro datasheet). Another problem might be with the ground layout, you may want to move the ground on the scope around to different points and see if the voltage looks the same or if the dip changes, if it changes that means there is problems with ground currents/common mode noise. \$\endgroup\$
    – Voltage Spike
    Apr 4, 2019 at 17:29
  • \$\begingroup\$ We also think that the supply can't source fast enough. Vcc is dropping almost 50%. Ground plane seems to be ok and not the problem. We are still however not sure what part of the supply is causing the problem. The LNK364 maybe? \$\endgroup\$
    – solartec
    Apr 8, 2019 at 8:18
  • \$\begingroup\$ @solartec Usually its the control loop on a DC to DC converter or AC to DC converter \$\endgroup\$
    – Voltage Spike
    Apr 8, 2019 at 14:55

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.