1
\$\begingroup\$

I have a simple (yet probably stupid question, since my knowledge of analog circtuits is very limited) about voltage dividers.

The final goal is to measure AC current for a mains load; to that end, a Hall 5A sensor (ACS724) followed by a RMS-to-DC (LTC1966) conversion circuit is tested.

It is not working, so I'm breaking down the problems I see.

Here's the first one.

Since I'm measuring AC, the Hall sensor's output should be a voltage proportional to the AC current, offset by Vcc/2, right ? Ideally a sine wave at 50Hz; so I'm trying to feed its output along with Vcc/2 to the next component that has a differential input (the LTC1966.)

On the breadboard, I use a 5V power supply (from a USB charger,) followed by a 3.3V LDO regulator (an LD1117V33) and a 1000uF capacitor.

schematic

simulate this circuit – Schematic created using CircuitLab

The voltage divider uses two 10k resistors. On the multimeter, the values look ok (3.3V and 1.65V,) but on a DSO (a Nano DSO,) I get this for the Vcc and Vcc/2 :

Vcc Vcc/2
Vcc supply Vcc/2

It seems very noisy, and I can't see why; everything else is disconnected. I don't expect the value to be mathematically Vcc/2 due to tolerance, and probably the DSO itself is a bit crappy, but:

  1. Why is this happening?
  2. What can I do about it?
  3. Could this have a significant impact on the comparators, or I'm worrying for nothing?
\$\endgroup\$
6
  • \$\begingroup\$ I think you might have to share the board layout. If Vcc is clean then I think chances are high the noise is coming from whatever is connected to Vcc/2 rather than from the LDO. \$\endgroup\$ – Lundin Feb 15 at 7:31
  • \$\begingroup\$ Although a few lower value caps is pretty much standard after any voltage regulator. I don't know this one, but commonly you have some 100nF to 1uF on its output. The huge bulk cap won't help you with higher frequent noise like this. \$\endgroup\$ – Lundin Feb 15 at 7:33
  • 3
    \$\begingroup\$ Try adding a capacitor in parallel with R2 - 100nF would be a good starting point. \$\endgroup\$ – Frog Feb 15 at 8:05
  • \$\begingroup\$ It is possible that the noise isn't actually there and it's a probe or a scope artifact… adding a capacitor on R2 as already suggested is a good thing, you can also try to lower the value of the resistors. As for the comparators it depends on what you need that threshold to. Be sure to add some hysteresis if possible. \$\endgroup\$ – Lorenzo Marcantonio Feb 15 at 8:38
  • \$\begingroup\$ Thank you all. Will try those ASAP. @LorenzoMarcantonio : what puzzled me is that probing the Vcc has very little noise, and I was expecting this to be reduced by the divider \$\endgroup\$ – stefanu Feb 15 at 9:08
0
\$\begingroup\$

Following all the suggestions from Lundin, Frog and Lorenzo Marcantonio (in no particular order, thank you all for helping out), I took the following steps :

  1. replaced the 1000uF capacitor with a 10uF one and added a 100nF one on the 5V side, but got interesting results :
Vcc Vcc/2
enter image description here enter image description here
A much noisier Vcc A far more noisier Vcc/2
  1. added a 100nF capacitor on the R2 which smoothed things out :

enter image description here

  1. tried a different value for R1 and R2, buth lower and higher values, but there were no detectable changes, at least not with my DSO, which for some unknown reason refuses to chart a signal with lower tnan 0.2V/Div scale, so I cannot get a better 'image' of the signal.

My conclusion is that I need to keep the filtering capacitor on R2 and that most likely the DSO and wiring are responsible for some of the noise I see. I hope I can compensate for this later on.

Thank you all for helping out. Time to get to the next question for my project : Hall current sensor noise

\$\endgroup\$

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.