0
\$\begingroup\$

I just tried it and my power supply makes this unusual whining sound (not there normally), that doesn't seem right.

I am using an MPS mEZD41503A-A for my power supply, it is stepping up an 18650 3400mAh to 5V @ 3A

I tried to connect 2, 3.3F 3V super caps that were wired in series (6V), to the output of the power supply in parallel.

The reason is to give it more current draw ability.

So now I'm trying to think of other ways to increase current draw ability.

This power supply can take in two voltages and provide two boosted outputs... I think it would not make sense for me to connect the same battery to both voltage inputs on the power supply... and pull even higher current at the expense of the battery's capacity/shorter runtime.

The context is powering a robot with 12 servos. I had enough power before until I upgraded my servos which pull more current. The result is on startup my robot flips out/digital servos going all over the place.

I already tried/setup a high-side MOSFET switch but the microcontroller and the servos share the same power supply so I think the servos reset the microcontroller when they start up causing the erratic behavior. I think an "easy" fix would be to slap on another power supply(cheapo 2A) that'll power the microcontroller on its own, pull from same battery.

Edit: for context the servos are MG90D vs. standard blue 9g servos

update

I ended up adding another boost converter (MT3608) to the same battery and this one specifically powers the microcontroller. Ideally it would power the other low-power electronics but due to the existing wiring/hard to update... this was the simplest fix.

\$\endgroup\$
2
  • \$\begingroup\$ If the supercaps were discharged initially, it's a wonder the supply survived the inrush current of connecting them at all! \$\endgroup\$ Sep 5, 2022 at 19:02
  • \$\begingroup\$ Well that's why it's $15 each vs. $1. But yeah so far it's working fine after I did my blunder. \$\endgroup\$ Sep 5, 2022 at 19:04

3 Answers 3

2
\$\begingroup\$

The mEZD41503A-X is designed to run with 3 × 22 μF as on-board output capacitors. That's these I've marked with red dots: -

enter image description here

That tallies with the data sheet for the MP3429 which is used by the mEZD41503A-X. Is the dark grey/black chip in the middle that I've marked with an orange circle.

And you want to increase the output capacitance by adding 2 series 3.3 farad capacitors. That's an increase of 25,000 times over the standard circuit board value and, you expected it to work.

Because of the added capacitors, the chip will likely go into current limit and the large inductor will squeal at you and you might be fortunate if the device survives. But, you might also kill the supercaps because, charging with such a supply, it might imbalance one supercap capacitor and exceed its voltage rating.

You have to be careful when charging series supercaps to prevent these imbalances.

Footnote: the data sheet for the controller IC does say it can be used in an extended soft-start mode and this might clear one problem (the squealing) but it may not clear the supercap imbalance problem: -

After the IC is enabled, the MP3429 starts up with a soft-start (SS) control. The SS signal is controlled by charging SS from 0V and compared with the internal reference voltage. The lower value is fed to the error amplifier to control the output voltage. After the SS signal rises above the reference voltage, soft start is completed, and the internal reference takes charge of the feedback loop regulation.

I had enough power before until I upgraded my servos which pull more current.

Get a more powerful power supply or add another of the existing power supplies and split/share the load.

\$\endgroup\$
1
  • \$\begingroup\$ "..will squeal at you..." yeah I think that's what's happening. I thought it was like the flash cameras charging the capacitor but the sound was coming from the power supply. Thanks for the annotated diagram helps. \$\endgroup\$ Sep 5, 2022 at 18:10
2
\$\begingroup\$

"So now I'm trying to think of other ways to increase current draw ability"

Caps will not increase the amount of current you can draw from a regulator. They will just smooth out voltage dips from transient current demands (peaky currents).

I tried to connect 2, 3.3F 3V super cap

3.3F is way too much for a typical regulator. Regulators can only tolerate limited amounts of capacitance on their output. It's a stability/control systems thing.

The context is powering a robot with 12 servos. I had enough power before until I upgraded my servos which pull more current. The result is on startup my robot flips out/digital servos going all over the place.

You need to narrow down whether the problem is startup currents, peaky currents during operation, or continuous currents.

This power supply can take in two voltages and provide two boosted outputs... I think it would not make sense for me to connect the same battery to both voltage inputs on the power supply... and pull even higher current at the expense of the battery's capacity/shorter runtime.

So where do you expect the extra energy for your more powerful servos to come from? Thin air?

\$\endgroup\$
2
  • \$\begingroup\$ Thanks for providing context. Yeah I was expecting it to cause a problem somewhere eg. less runtime. But I think I was more concerned if the battery has a max 4A draw, then 6A (attempted) is too much. Interesting about the 3.3F being too much. \$\endgroup\$ Sep 5, 2022 at 18:06
  • \$\begingroup\$ Note that if the actual problem is transient demand and not sustained overcurrent, such that a capacitor might help, then it needs to be connected via a soft-start circuit that allows it to charge slowly, instead of having the regulator see all that capacitance during startup. During startup, that capacitor is making your overcurrent condition much, much worse. In fact the problems you're seeing at startup with the new servos might be related to their capacitance and inrush current. \$\endgroup\$
    – Ben Voigt
    Sep 5, 2022 at 18:06
1
\$\begingroup\$

It sounds like extremely bad idea to put 1.5F worth of capacitance to a DC/DC converter output.

Especially when the datasheet does not specify maximum allowable capacitance, it does not mean you can add any amount.

A wild guess would be that you are exceeding capacitance value by a factor of 1000.

The capacitance is so large that the converter tries to start up but the voltage won't rise quickly enough and the converter might shut down due to overcurrent protection or short circuit protection.

So you can't pass the 3A limit of the converter, even if there are supercaps on output a 3A load still draws 3A from the converter.

\$\endgroup\$
3
  • \$\begingroup\$ Yeah in the past I was using capacitors as batteries, so I thought I could charge them up/use them in parallel with the power supply output. Thanks for the info. \$\endgroup\$ Sep 5, 2022 at 18:08
  • \$\begingroup\$ Well you can't charge empty supercaps by just connecting them directly to a power supply. They'll pull as much as current they can get and bring the supply output down until they are charged to the output voltage. For example it takes 3 seconds at 1A to charge 1F to 3V. \$\endgroup\$
    – Justme
    Sep 5, 2022 at 18:21
  • \$\begingroup\$ I see. I was using them as "batteries" for a solar panel project (which failed) so they would slowly build up over time as the panels dumped power into them. The issue was the microcontrollers did not like the slow build up of voltage up to the desired 3.3V or so... needed more electronics to manage that or use batteries. \$\endgroup\$ Sep 5, 2022 at 19:00

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.