Howto properly filter/smooth stepped up voltage

[ranger81]

Hello,
I have a digital radio which is powered by four AA batteries or with an external power supply which provides 6V DC.

I would like to build an adapter to use the radio with a regular USB connector. Since USB only delivers 5V I used a small step-up module (based on MT3608) to step up the voltage to the desired 6V. The radio powers up but it can't detect any radio station. I guess stepping up the voltage induces some high frequency noise which is disturbing the radio somehow.

Checking the output of the module with a scope I found out, that there is a Vpp of approx. 1.8V at 1.1MHz (near the switching frequency of the MT3608). This was during a test setup with only 60mA load attached to the output of the boost converter. Please check the attached scope screenshot.

My question is, how to filter/smooth this properly. I tried different things using obviously a low ESR capacitor on the output, some inductors I had laying around and so on, but nothing really helped. Could you please give me an advice?

PS: Just in case you are wondering if it's possible to power the radio directly from USB 5V: No, the signal strength of any station is very low and therefor the radio performs very bad compared to battery operation.

Thank you very much!

[danadak]

Google "ceramic disk esr curves" and look at the images. Its a challenge
without the right components to get good filtering. Coupled with that is
the minefield that the same technology from one vendor to another can
have dramatically different ESR performance.

First recommendation is grounds have to be solid, very tough to do this
design in a proto board.


http://www.ti.com/lit/an/sbva012/sbva012.pdf

http://www.ti.com/lit/an/sbva012/sbva012.pdf

http://www.analog.com/en/technical-articles/designing-second-stage-output-filters-for-switching-power-supplies.html


Regards, Dana.

[ovnr]

The short version: Filter the output as well as you can, then drop that into a low-noise linear regulator with excellent PSRR at the problematic frequency, then filter it again.

The TPS7A49 series generally has good performance; in this case, it has a PSRR of around 30 dB at 2 MHz.


It may be worth it to find a boost converter with a frequency that's easier to deal with.

[vk6zgo]

Have you tried the radio direct off 5v?

Radios often will work quite we'll on reduced voltages---after all that's what happens as the batteries flatten!

[Siwastaja]

Any 4AA device runs from 5V directly unless completely broken by design - this solves your problem, no boost converter needed.

Well designed devices will run down to, or even below 4V.

With rechargeable AA cells, the nominal voltage is 4.8V.

[ranger81]

@Siwastaja, @vk6zgo: Have you read my entire post? Then you would know that I have already tried that without luck. The radio signal strength is very weak when powered with only 5V.

[vk6zgo]

 Chuck that digital heap of rubbish out & get a real radio! 
Seriously, though, maybe it uses a bit more current than the USB port can supply.
I have heard unconfirmed reports that digital radios are a bit hungry.

[RoGeorge]

You don't need to raise the voltage to 6V. Just plug it to the USB at 5V.

5V is exactly the voltage when the batteries are at about half of their life.
The radio will probably work just fine even with 3-4 Volts.

[Zero999]

As mentioned above, I doubt the radio doesn't work because the USB is only 5V. Did you measure the voltage across the radio, whilst it was being powered from the USB port? The USB port is probably cutting off because it can't, or hasn't been properly configured to provide enough current to power the radio.

[suicidaleggroll]

The short version: Filter the output as well as you can, then drop that into a low-noise linear regulator with excellent PSRR at the problematic frequency, then filter it again.

The TPS7A49 series generally has good performance; in this case, it has a PSRR of around 30 dB at 2 MHz.


It may be worth it to find a boost converter with a frequency that's easier to deal with.

LT3042 has a PSRR of nearly 90 dB at 2 MHz, I would consider it a much better option than the TPS7A49.  That said, 1.8V of ripple on a 6V regulator means something is wrong.  Did you make this step-up module yourself or did you buy it somewhere?  If you must use it, try increasing the output voltage to more like 7V and stick an LT3042 or similar in to drop it back to 6V.

Of course the other posters here are correct too.  4xAA only has a nominal voltage of 4.8V, the radio should run just fine on 5V.  Check to make sure you're actually getting 5V from your supply, current limiting may be dropping it down.

[Siwastaja]

@Siwastaja, @vk6zgo: Have you read my entire post? Then you would know that I have already tried that without luck. The radio signal strength is very weak when powered with only 5V.

AA cells, especially cheap ones, have high internal resistance. If the radio wasn't working well at 5V, that would mean it would only work with completely fresh, expensive AA cells which are kept warm.

As the AA cells are discharged even slightly, they drop well below 5V under load - and the radio most likely takes some peak power which is not insignificant.

"5V not enough" requirement would make the radio totally unusable with any type of AA cells - NiMh or Alkaline or whatever.

While such "completely broken by design" products do exist, I really think the problem is most likely elsewhere. Try with an adjustable lab supply, first set to 6V, then set to 5V, and verify the actual voltage at the terminals.