1
\$\begingroup\$

I was in the process of choosing a 3.3V Linear Regulator for my ESP based Project when I stumbled upon the different types of these voltage regulators.

The 1117 regulator had different prefixes, AMS, LD, ZLDO, NCP. What do these mean and how do I choose the right regulator with the right components to go with it..?

To my knowledge, most of these prefixes are abbreviations of the manufacturer name (AMS - Advanced Monolithic Systems) and some of them describe their feature (LD - Low Dropout).

But, a few sources online said that the AMS1117 needs a different type of capacitor than the NCP1117 which raised the question that if the functionality varies with the manufacturer.

How do I decide which of them is the best for my project..?

\$\endgroup\$
1
  • 1
    \$\begingroup\$ Read the data sheets. Beyond that you're asking for an opinion on a purchase, which is doubly off topic here. Likely for your purposes any will work, and if you place pads for the most demanding assortment of recommended capacitors, you can always try various ones comparatively. \$\endgroup\$ Commented Jan 27, 2019 at 6:30

2 Answers 2

2
\$\begingroup\$

How do you decide the best choice?

Define your acceptance criteria based on Vin-out Static Tolerance and Step load error at a minimum with test results using your worst specs.

Then test in a representative PCB layout and verify and record margins at worst case temp for your environment and worst case step and preloads. 0 preload is worst. Cap ESR is a critical tradeoff between step load error and phase margin.

Static source impedance and step load impedance are different thus static load regulation and step load errors are different. Phase margin depends on source and load caps and step loads.

Careful selection and layout matched to your needs may lead to success or failure depending on your specs.

If you don't recognize the differences below or that some do not specify stability margins, then consult with someone who does or read more about phase margin and ask a better question.

PN       Vout tol. Imax  Typ Load Reg. @0.8A  Zo  Vin-out [email protected] Cout
======= ==== ====  ===    ====               ======  ============  ====
AMS1117 3.3V 1.5%  1A     3 mV     3mV/0.8A= 3.75mΩ  1.1V          22µF solid tant.
NCP1117 3.3V 1%    1.5A   4.3 mV  4.3m/0.8A= 5.38mΩ  1.07V     80mV typ 0.1 to 0.5A step @ 10uF

calc...NCP1117 80mV/500mA = 160 mΩ  80mV= = +/-2% error from 400mA steps with 100mA preload  

NCP Instability points

1A step load C*ESR= 50uF*1mΩ=50ns, 20uF*10mΩ=200ns, 2uF*80mΩ=160ns, 1uF*200mΩ=200ns
Static Load >20mΩ 10uF ceramic input , 10uF Tantalum output

\$\endgroup\$
1
\$\begingroup\$

NCP1117

A 10uF ceramic or tantalum capacitor should be adequate for most applications.

AMS1117

The addition of 22µF solid tantalum on the output will ensure stability for all operating conditions.

The requirements are essentially the same, only Advanced Monolithic Systems are recommending a larger value to ensure stability under all conditions, whereas On Semiconductor provide a formula to calculate the minimum capacitance required.

Unless for some reason you need to use the minimum possible capacitance, go with AMS's recommendation and then it should work with both brands (and probably others as well).

As to which is 'best' for your project, that depends on what your criteria are. Is one cheaper, more readily available, come in the package style you prefer? If you can't decide then either one should be fine.

\$\endgroup\$
5
  • \$\begingroup\$ NOT the same. There is no indication the AMS1117 is stable with a (low ESR) ceramic capacitor! \$\endgroup\$ Commented Jan 27, 2019 at 8:05
  • \$\begingroup\$ @SpehroPefhany And no indication that it isn't. I didn't say it was 'the same' only that both will work with the same tantalum capacitor. In reality the only difference might be that AMS didn't test theirs with a ('Low' ESR, whatever that means) ceramic capacitor. I can't see a date on the AMS datasheet, but the file was created in 2009. Were 22uF 'Low ESR' ceramic capacitors commonly used back then? \$\endgroup\$ Commented Jan 27, 2019 at 8:29
  • 2
    \$\begingroup\$ The LM1117 (original manufacturer) specifies minimum ESR of 300m\$\Omega\$, the NCP1117 specifies minimum ESR of 33m\$\Omega\$. The capacitor forms part of the frequency compensation of the regulator so ignoring the minimum ESR is not a good idea. For example, some fraction of them might oscillate straightaway, or perhaps at temperature extremes. A small resistor in series with the ceramic cap will make sure it's okay. Or use a tantalum. \$\endgroup\$ Commented Jan 27, 2019 at 8:50
  • 1
    \$\begingroup\$ Bruce, @SpehroPefhany is correct , It 's the Lead-lag effect of the cap's RC value that determines stability and the Source must be much lower than the Load Cap hence MLCC in to Solid Tant Out. unless your MLCC cap is at least 10x 200ns or 2us which I doubt it will ring with steps which the ESR*C Value Attenuating feedback (f) is cause fo more ringing \$\endgroup\$
    – D.A.S.
    Commented Jan 27, 2019 at 15:27
  • \$\begingroup\$ "Or use a tantalum" - exactly. Use a capacitor type that is known to work with both variants and you won't have a problem - with the bonus of being able to substitute without fear. The 1117 regulator is a generic part that many manufacturers produce, but I bet they all have slight variations that can get you into trouble if you rely on exact behavior. \$\endgroup\$ Commented Jan 27, 2019 at 22:50

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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