1
\$\begingroup\$

I am using RFID aliexpresslink. RFID is suggesting this schematic: enter image description here

The 7809 has this schematic in the data sheet: enter image description here

  1. Why is RFID power design suggesting big capacitors with low resistors? Does it mean that if big capacitors are connected parallel, they don't need input capacitor?
  2. Why do they not add an input capacitor?
  3. Will this design work without problems? (1000uF capacitor design.) I didn't see anything like this schematic on the internet.
  4. If I use only input and output capacitor, will it affect the reading distance? I mean, i I use only 0.33 uF and 0.1 uF, will my read distance be lower?
  5. I read in this forum that the 7809's output capacitor is 0.1 uF because in 1970, capacitors were really expensive. What is the best output and input values in these days? (I searched but only I found this schematic 0.33uF and 0.1uF.)

Current schematic of mine. enter image description here

For freewheeling diode:

enter image description here

\$\endgroup\$
11
  • \$\begingroup\$ About question 1: it seems they want a really smooth DC. Those three RC are 3 strong lowpass filters. \$\endgroup\$ Dec 22, 2021 at 8:57
  • 4
    \$\begingroup\$ I wonder about the competence of an engineer in the 2020s that recommends a 78xx series linear regulator for RF cleanliness purposes. And anyone that calls something like this "as pure as battery"... Not saying the module is badly designed, just that the person responsible for the aliexpress offer was just going around copying decades obsolete schematics from the internet without really understanding them. I have serious doubts this works well – a 1000µF capacitor very likely has a non-negligible internal series resistance, which means that while RF from the power supply might be \$\endgroup\$ Dec 22, 2021 at 12:30
  • 3
    \$\begingroup\$ well-suppressed, now the voltage regulation becomes bad, which might even be worse. \$\endgroup\$ Dec 22, 2021 at 12:35
  • 2
    \$\begingroup\$ Say, your 1000µF has a dissipation factor of 0.16 (pretty typical), so at 134 kHz (which will definitely appear as modulation in the current drawn by the amplifier) this capacitor exhibits 21 Ω resistance; so that the HF has to be fed from the 3Ω between the 7809 and VCC_RF (capacitors mostly useless) . This basically negates any voltage regulation. In other words: Either the board itself has a linear regulator that would take care of all relevant frequencies by itself, so the external filtering is useless, or it doesn't, in which case this thing becomes a nonlinearly oscillating RF nightmare. \$\endgroup\$ Dec 22, 2021 at 12:42
  • 1
    \$\begingroup\$ yeah, go to a proper electronics distributor (Farnell, digikey, mouser, …), and look for things that don't start with LM, have the right in- and output voltages, and then check the noise rejection properties at frequencies that are relevant. You need to first figure out what these would be, I'm not even convinced there's an actual need for a linear regulator with a modern switch-mode supply. As said, in all likelihood that board has its own regulator with its own noise rejection and the people putting together the aliexpress website were just stealing random schematics from forums. \$\endgroup\$ Dec 22, 2021 at 13:09

3 Answers 3

6
\$\begingroup\$

Most regulators will get unstable if the impedance of the input power supply is too inductive, due to long wires or traces. So an input capacitor is added.

You can use whatever value of cap you already have that's equal or greater to what the manufacturer says in the datasheet. If the input supply bulk caps are close, then it's probably not necessary, but a 1µF SMD caps only costs a few cents...

Why is RFID power design suggesting big capacitors with low resistors?

Probably because their board has low power supply rejection and needs clean power. To be honest, it looks a bit overkill.

Personally I'd use a snubber on the transformer to remove rectifier noise at the source, and a large 1000µF output cap because RFID draws pulsed current.

\$\endgroup\$
18
  • \$\begingroup\$ Thank you. If you use 1000 µF output capacitor, will you use 0.33 uF input capacitor. I read Justme's explanation but I just wanted the ask again. \$\endgroup\$ Dec 22, 2021 at 8:20
  • \$\begingroup\$ You need an input cap (0.33µF or whatever value, use what you got) if there is not already an input cap, for example after the rectifier if you use a transformer+rectifier+caps supply. \$\endgroup\$
    – bobflux
    Dec 22, 2021 at 8:21
  • \$\begingroup\$ @Tryingtogetsome 100 nF or 330 nF are advisable because they work better on high frequencies, where large (electrolytic) don't. It is always better, to put a small cap near the regulator. \$\endgroup\$ Dec 22, 2021 at 8:54
  • \$\begingroup\$ Okay, thank you again. I will use 330 nF then. I will keep in mind "It is always better, to put a small cap near the regulator." \$\endgroup\$ Dec 22, 2021 at 9:34
  • 1
    \$\begingroup\$ Thank you so much, have a nice day. \$\endgroup\$ Dec 22, 2021 at 13:58
2
\$\begingroup\$

In general, the presence of an input bypass capacitor does not depend on how much capacitance there is on the output. It depends on many other factors, such as for example how long wires there are to the input bulk capacitor. If 1 cm, most likely a separate bypass is not needed, if 1 meter, most likely a separate bypass is needed.

Bear in mind that we don't know why the circuit is like that, ask the one who provided the circuit. Internet is full of circuits that look random and made by hobbyists and tinkerers instead of engineers.

  1. The regulator may omit the drawing of the input bypass capacitor, if it is expected to be near a power supply bulk capacitance which is not drawn in the circuit. In general, it is best to have the 330nF input capacitor that to omit it. The fact that there is a lot of output bulk capacitance does not mean you can omit the input bypass cap. Take the circuit as a block diagram type of example circuit, not as exact design.

  2. They may assume or expect that the power supply with its bulk capacitors is within short distance from the regulator input, and it is just an example picture, because it might be different depending on which regulator you use. It may not resemble the exact way how you should do things.

  3. It might work. Or it might have problems. It is impossible to say just by looking this circuit in isolation. If you have a random 12V power supply, there is no way to prove it works with it or it won't. And there must be better ways to do it than with three high value caps. The first problem might be that the inrush current is so large that a power supply might hit a current limit and shut down or blows a fuse on the supply output. So it may require a power supply with large output current, even if the circuit consumption is very small.

  4. How would we know how it affects it. If you remove bulk caps and add input bypass, it might even work better depending on your power supply which is unknown. The point is, if there is no specs what the RF device needs from the power supply and regulator, it is impossible to know what power supply and regulator is good enough for the RF device.

  5. No the values are not 330nF and 100nF because capacitors were expensive. The manufacturer just guarantees that these values of high frequency bypass caps have proper impedance range to keep the regulator stable. The user of the regulator is responsible for using larger bulk capacitors on input and output if necessary depending on the circuit.

\$\endgroup\$
1
  • \$\begingroup\$ Thank you for huge explanation. \$\endgroup\$ Dec 22, 2021 at 8:16
1
\$\begingroup\$

You may want to check that 78xx schematic. To the best of my knowledge, pin 2 is GND and pin 3 is OUT. I stumbled upon this looking up the pinout on google and it might confuse some folks, causing premature release of The Magic Smoke. If you want a relatively stable and quiet power supply with one of these, simply throw lots of capapcitance at it, say a 470uF plus a 100nF in parallel on the input and a 1000uF plus 100nF in parallel on the output. The more the better, depending on the current draw of the application and surge-load capability of the supply. For good measure use ferrite chokes on the in and out, a donut type coil one on input and a rf blocking barrel on the output. Any rf pickup by the circuit or its wires will make it into your device, consider using a clip-on ferrite close to the device on the supply wire. That will kill any remaining junk right at the device side.

\$\endgroup\$
5
  • 1
    \$\begingroup\$ Be careful with inductivities on the input. Depending on the situation they might form an oscillator together with the input capacitance or lead to really high transient voltages when current consumption after the 7809 changes. I've made that error once and killed some electronics before realizing my mistake... A freewheeling diode solved the issue. \$\endgroup\$
    – kruemi
    May 16 at 8:37
  • \$\begingroup\$ Which pin is which depends on how you number them. The datasheet from where this numbering is copied from uses number 3 for the middle pin, which means, middle pin is still ground regardless of the pin numbering scheme. \$\endgroup\$
    – Justme
    May 16 at 8:53
  • 1
    \$\begingroup\$ Kruemi, The typical input rf chokes are only a few turns of thick wire on a donut shaped core, they block rf but are very low inductance. I use whatever I can scavenge from old equipment, any ol' ferrite core will do, snip off the original wire and make a few loops of your input wire through it. Justme: Have a look at any datasheet for the 78xx series regulators, They are always 1-In, 2-Gnd, 3-Out. The datasheet you are referring to is a very odd one, where does it come from? To avoid mishaps there would have to be a drawing that shows the odd numbering clearly. \$\endgroup\$
    – DoctorZ
    May 16 at 9:56
  • \$\begingroup\$ Thank you DoctorZ for your comment. @kruemi I also want to thank you for your comment. I added photo for freewheeling diode on question. Is that what you say ? I can add diode like this. \$\endgroup\$ May 23 at 5:22
  • \$\begingroup\$ @DoctorZ I've meade the mistake of just putting in the same Component i've used elsewhere on the board. Which was from the switching regulators. I've killed two 7805 until I realized my mistake. Sadly, the 7805 failed short which put 24V instead of 5V on the components. Luckily, I had not place the expensive ADC and analog switches. Also I had not placed the opamps. But I killed two Arduino (I did not like +24V on Aref) and one I2C Isolator in the process. That's why I put out this warning. Coils can be mean beasts! \$\endgroup\$
    – kruemi
    May 23 at 7:47

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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