I driving some LEDs at about 114 Hz with a power supply of 9 V and a max current supply of 150 mA, LEDs are running about 60 mA I have some LEDs in parallel. If I increase to 200 Hz it's the same, How can the power oscillations be fixed, why this happens could be because the power supply doesn't like switching loads at this frequency range. Do I need to keep increasing the switching LED frequency? The LEDs are switched by a transistor and resistors to regulate the current.
-
1\$\begingroup\$ Are they actually oscillations or just variation of voltage with load? Show us a scope trace. \$\endgroup\$– FinbarrCommented Aug 31, 2023 at 8:01
-
\$\begingroup\$ Please show your setup. Do you perhaps need more power supply decoupling close to your load? \$\endgroup\$– winnyCommented Aug 31, 2023 at 8:32
-
\$\begingroup\$ Is this actually causing a problem? \$\endgroup\$– FinbarrCommented Aug 31, 2023 at 8:48
-
\$\begingroup\$ I'm modulating the frequency from 0 to 50, perhaps I will place a decoupling capacitor close to the load starting with 100nf. \$\endgroup\$– CitiCommented Aug 31, 2023 at 8:53
-
\$\begingroup\$ I don't see any oscillations. It looks more like a voltage source load regulation problem. \$\endgroup\$– user319836Commented Aug 31, 2023 at 10:06
1 Answer
This would typically happen with poor grounding or high-resistance connecting the circuits. I'm going to assume this behavior only happens while the LED's are switching. I'm also going to assume the magnitude of this problem is reduced by half when only one series string is in circuit. This is to confirm the LED load is the driving factor.
Measure the voltage drop between power points. To do this, Turn up the scope sensitivity (100mV/div to 200mV/div) and clip you ground lead on the regulator's local ground (its pin) to the transistor's local ground (it's emitter). Do the same for the micro or oscillator's local ground.
In that same vein, you can check the 9V+ rail from the regulator's output to the various sub-circuits (this is assuming the scope is floating with respect to your circuit. If your circuit ground is common to earth ground, you're going to have trouble measuring the V+ rail because most scopes tie earth ground to BNC ground and this will short your circuit).
Keep in mind, a healthy voltage regulator does a very good job at regulating the potential difference between it's output and it's LOCAL (reference) ground; not all grounds are the same. This is why high-current power supplies come with kelvin connections - so it can regulate at a remote ground (where it matters instead of its local terminal ground that most small supplies do).