I want to use the MCP1703 regulator in battery powered projects. The datasheet of this regulator says that the quiescent current is typical 2 uA.

I measured the quiescent current for the MCP1703-5 at 80 uA. For the MCP1703-3.3 I measured 50uA of quiescent current.

For measurement I connected the (6 to 12) volt DC to input and left the output unconnected.

Why is the quiescent current so high ?

enter image description here

  • \$\begingroup\$ In addition to the need for ceramic caps, I am hesitant to trust that type of bread board. It does not necessarily produce reliable low-impedance connections from one part to another. It would be better to solder into perf board, for example, or buy a SOT223 breakout board and put the input and output caps right on the breakout board. Microchip also sells an evaluation board for SOT23 regulators. If you want to try out the SOT23 version of the part. Search for this: "SOT23-3EV-VREG" \$\endgroup\$ – mkeith Apr 15 at 17:20
  • 1
    \$\begingroup\$ According to the datasheet, you measure the quiescent current through the GND pin (center pin and tab on your chip) only. Any other current you measure is not the quiescent current. Your picture doesn't show how your ammeter is connected. \$\endgroup\$ – JRE Apr 15 at 18:27
  • \$\begingroup\$ If the yellow alligator clip is connected to ground, then it looks like your output capacitor is connected between in and out rather than between out and ground. (But it's hard to tell due to the image focus) It also looks like your input capacitor could be reversed, which will also cause excess leakage currents (if it doesn't cause fireworks) \$\endgroup\$ – The Photon Apr 15 at 20:26

It sounds like you didn't connect an output capacitor when measuring the \$I_q\$.

From the first page of the datasheet, this regulator is "Stable with 1.0 µF to 22 µF Ceramic Output Capacitance". This implies it might not be stable without an output capacitor.

If you connect it in an unstable configuration, it could oscillate, leading to much higher \$I_q\$ than expected or specified.

As pointed out in comments, an input capacitor is also required to match the \$I_q\$ test conditions. Whether this actually has an important effect probably depends what kind of power source you use and how long are the wires connecting it to your device under test.

  • \$\begingroup\$ +1 In fact it's specifically stated to be with 1uF X7R ceramic capacitors on both input and output, Tj=25°C, and minimum Vin. The latter variables don't matter much according to Fig 2-1. \$\endgroup\$ – Spehro Pefhany Apr 15 at 15:56
  • \$\begingroup\$ I connected to the input and output 1uf electrolite cap and still the current is high , 50ua for mcp1703-3.3 . \$\endgroup\$ – Ali Apr 15 at 16:06
  • 1
    \$\begingroup\$ @Ali, can you share a photo of your setup? \$\endgroup\$ – The Photon Apr 15 at 16:07
  • \$\begingroup\$ I try to find ceramic capacitor and test again . \$\endgroup\$ – Ali Apr 15 at 16:09
  • \$\begingroup\$ Also, what is the leakage current spec of your capacitor? \$\endgroup\$ – The Photon Apr 15 at 16:11

Typical X7R caps are 1 GΩ-uF or 1 uF = 1 GΩ , 10 uF = 0.1 GΩ or 100 GΩ max

The MCP1703 specs are given in the table for specific conditions and measured by the ground current.

With 12V in and 5V out your load leakage = 80uA

Your DMM sees a 150 kOhm Req load when the spec is 12V/2uA= 6MΩ

Next apply a resistive load of 50mA on output to see if there is any spurious issue with no load. e.g. 5V out/50mA = 100 Ohms Expect Regulator ground current to be 30uA. If your current drops from 80uA to 30uA, then you may have a spurious oscillation in your layout of linear regulator.

So the IC does not appear to meet spec.


simulate this circuit – Schematic created using CircuitLab

Make sure your assumptions the test set does not induce this leakage with jumper wires from In to out in place of IC.

If that proves valid, then.... Consider (1) ESD damage and discount if careful, then...

Write to Distributor about defective parts and get an RMA. Return Material Authorization or choose better parts.

  • \$\begingroup\$ image of my test imgur.com/wmHTpDw \$\endgroup\$ – Ali Apr 15 at 16:49
  • \$\begingroup\$ Now see my root cause (2.) you are using a leaky e-cap. Measure it alone without regulator. \$\endgroup\$ – Sunnyskyguy EE75 Apr 15 at 17:04
  • \$\begingroup\$ I have several of this regulator and each has the same quiescent current. Also I measured the leakage current of capacitors in less than 1 ua . \$\endgroup\$ – Ali Apr 15 at 17:10
  • \$\begingroup\$ Then consider if this batch of parts do not meet spec, contact distributor or email OEM, or measure noise current on a series R of 10k to a scope. or AC measurement using DMM to measure AC and DC voltage drop with short wire jumpers in current loop. to see if it oscillates as The Photon suggested \$\endgroup\$ – Sunnyskyguy EE75 Apr 15 at 17:11
  • 1
    \$\begingroup\$ I connected a 100 ohm resistor load and measured the regulator ground current , still it is 50 ua when the load current is 32 ma . It seems the regulator has problem and I should contact with seller . Thanks a lot for help \$\endgroup\$ – Ali Apr 16 at 21:12

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.