Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

Capacitor Selection For A Dummy

Status
Not open for further replies.

Water550

New Member
I'm truly very clueless about electrical circuits and I'm hoping to get some help. I'm told I have some AC ripple in my DC power lines, so here I am looking for some help. Here is a brief summary of what I have.

I'm using the following part to get a 24V DC output.
https://www.digikey.com/product-detail/en/DRP024V060W1AZ/603-1225-ND/2236868

I'm then taking the 24V DC from this and it is supplied to the input of this voltage regulator.
https://www.mouser.com/ProductDetai...oN%2by1lsBNgdTVGAdL6LsoABrU%2btLt11wkG2xs9A==

The voltage regulator needs capacitors on the input and output lines. I currently do not have those in my circuit and would like to add them. I have no knowledge about selecting the appropriate capacitors and didn't know if someone could offer some help.

I'll be powering up to 10 low current draw sensors with the expected current values 3.4mA on average (100mA peak) with the output from the regulator.
 
If you have AC ripple at the 24Vdc output of that power supply, it is because you are overloading it by drawing more than 2.5A!
 
I have looked to the datasheet of the regulator and I have no idea what it means. As far as overloading the power supply, I can be as the 10 sensors at a max would only be drawing 1 amp if they were all firing at the exact same time. Is there anything else you guys can guide me on to get the correct capacitor?
 
Loaded at 1A with 18V across it it's going to get hot. Have you fitted an heatsink?

The datasheet recommends 0.33uF on the input and 22uF on the output. Anything close to these values will be OK but fit as close to the regulator as possible.
 
They want 0.33uF 50V cap on the input to ground. (maybe ceramic) I have 0.f1uF and would use three in parallel.
The output side; 22uF, 10V or higher. aluminum electrolytic. 47uF or 100uF will also work.
upload_2017-6-23_12-27-39.png

24V in and 6Vout. 18 volts across the regulator (maybe 50mA load) so about 0.9 watts of heat. The part will be good and hot but OK.
 
upload_2017-6-23_12-37-25.jpeg

The above cap is LARGER then the bottom cap. They should look some thing like this.
upload_2017-6-23_12-39-16.jpeg

"through hole"
 
Thanks guys, most other sites just yell at me for asking so basic... you have helped a total n00b and I really appreciate it!

One last question, for the input capacitor it would simply connect to the 24V and goes to the input of the voltage regulator (basically just is added to the circuit)? I'm assuming the same is for the capacitor on the output?
 
Last edited:
Capacitors are a complex beastie. Sometimes lo ESR is good sometimes not. It's all application dependent.
For power supply matters LOW ESR and increasing capacitance (1000uF?) reduces output ripple, but causes a high inrush current on power up unless you use a zero crossing phase switch OR more simply an NTC thermistor (sized to match the circuit) inline with the capacitors.

Capacitors can get hot under certain circumstances (high pulse currents, hi frequencies). Capacitor selection comes into play here, lower esr is key. Using capacitors in combination can get you improved performance such as both electrolytics and ceramic on the same power rail for better noise and transients as well as voltage sag handling. One thing to note with ceramic dielectric is it's capacitance is voltage sensitive so always use units rated at least 2x the max operational voltage.

As you're new to the electronics area, understanding capacitor application is important. When to use the different dielectric types such as ceramic (X7R , Y5V, or NPO) ,electrolytic, solid electrolytic, tantalum, mica, silver mica, teflon, PET, PEN, polycarbonate...etc etc.
It makes good rereading for experienced folks as well as improved manufacturing does inprove cap specs and can change application ranges.

Here are a couple useful refs: http://www.penton.com/webcasts/Keysight_060717.pdf
http://blog.octopart.com/archives/2016/03/how-to-select-a-capacitor

Edit: BTW caps make little explosives if you wire the polarized ones backwards, so pay attenton with that.
 
Your input voltage of 24V is much too high for the 6.3V low dropout regulator. It causes a lot of heating in the regulator that might cause it to briefly shut down on the current peaks.
 
src.
your d/s : PARD (20MHz) < 240mVpp (here it says PARD's spec.-d @/against full rated load ???)

LDO d/s : Ripple Rejection NJM2396F63 VIN=VO+2V ein=0.5Vrms, f=120Hz min. 52 / typ. 60 / max. - dB (see pg.4 for cap. values)

the particular LDO is a collector load one -- those things are prone to oscillating -- the quieting and cancelling of such is somewhat tricky and extremely a particular design specific ... since the **broken link removed** does not specify enough detail ... i'd contact them directly on this subj. OR chose a regulator with more data avail on it

+ some tests
_Draft_CL-LDO_noise.png _Draft_CL-LDO_noise_stmx.png + apx. the OP case _Draft_CL-LDO_noise_simple.png
how to interpret these results ...
  • as it was already denoted above in this thread the +24V might be a bit high starting point
    • usually the twice the output is good for linear regulators
    ---- if the near max. the src. current is consumed
    ---- if not the ceiling might be lower or as low as the 6.3V + dropout voltage
  • the +24V supply looks a quality one !!! if the PARD is for max. rated load
  • if the LT's substitute for NJM2396 is not too idealized and won't perform significantly better the C-load LDO appears to be quite stable
    • about -- the first graph -- ramp up from SC -- at 35ms
  • the graphs look messier than the probable real life situation would be coz they tested at tight timescale , tight evt. schedule and near their capabilities
  • the ultimate filter in between +24V and LDO has 4700µF cap that enables SC current to stay high longer -- it also elongates DC drifts before LDO
  • the tests were carried out mostly to find out the probable instability of the LDO (1-st) and the "general power-setup" (2-nd)
    • it was not targeted for any specific application neither the OP's case
    -- so the inductor and capacitor values must be re-chosen to suit the actual target application !
  • other than that ↑ -- a "proven" concept is to use step down or buck before LDO
    • one of the causes is electrolytic capacitor voltage ratings
    • also having a "noiseless" input to regulate the varying output
    • e.c. . . .
since the more realistic test didn't show up any potential issues -- there might be a different problem https://www.analog.com/media/en/technical-documentation/application-notes/AN-581.pdf
 
Last edited:
I agree that the datasheet of the Japanese low dropout regulator is missing the details about the mandatory output capacitor as shown on datasheets from National Semi who are now part of Texas Instruments. A translation problem?
 
something about -- from past
 

Attachments

  • __pstca.com_spice_ripple_ripple.pdf
    128.7 KB · Views: 353
A 12 V / 2 A power supply will cost less and generate way less heat in the linear regulator. If this is not an industrial application or environment, a 12 V wall wart will work fine.

Depending on how fast the sensors kick up to 100 mA, you might need a larger output capacitor on the regulator to help with the transient current.

Depending on how long the wires are from the power supply to the regulator, you might need more input filter capacitance close to the regulator input pin. That little 0.33 uF cap assumes that the DC source is very close and has a fairly large output filter capacitor.

I recommend 100 uF / 50 V and 0.33 uF / 50 V in parallel at both the input and output of the regulator. 50 V is a minimum voltage rating for long-term reliability in a 24 V circuit; higher is ok. If there are long wire runs to the sensors, then add the same capacitor pair at each sensor.

What are the sensors and why do they need 6.3 Vdc? Link to data?

ak
 
Hi water, one thing not mentioned I'll chip in on, put the capacitors reasonably close to the regulator, esp the input one, within 10mm if you can.
 
Hi water, one thing not mentioned I'll chip in on, put the capacitors reasonably close to the regulator, esp the input one, within 10mm if you can.
Depending on how long the wires are from the power supply to the regulator, you might need more input filter capacitance close to the regulator input pin. That little 0.33 uF cap assumes that the DC source is very close and has a fairly large output filter capacitor.
 
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top