Dear All:

I have bought a 10uF/25V Tantalum Capacitor for use it with a Lm1117 LDO Regulator.
This is the Capacitor datasheet: http://www.ryston.cz/pdf5/ca42.pdf

I want to know the Tantalum Capacitor ESR but i can't find that info in the datasheet.

Is there a way to know the tantalum capacitor ESR?

Will the capacitor i bought will work well like an input and output capacitor on my Lm117 regulator?

Does a 10uF/25V Tantalum Capacitor have a ESR lower than 0.5 Ohm?

Regards.
Alfredo Mendiola Loyola

 

What kind of voltage is the regulator handling? If it's from a line transformer with a 100 or 120 Hz ripple, ESR is of no consequence. If you look at the data sheet, you will see that .1 μF caps are also suggested. One parallels the large electrolytic at the input, and the other is a bypass at the output pin.

 

What kind of voltage is the regulator handling? If it's from a line transformer with a 100 or 120 Hz ripple, ESR is of no consequence. If you look at the data sheet, you will see that .1 μF caps are also suggested. One parallels the large electrolytic at the input, and the other is a bypass at the output pin.

I'm using the Lm1117 to convert 5V DC to 3.3V DC.

Regards
Alfredo Mendiola Loyola

 

Dear All:

I have bought a 10uF/25V Tantalum Capacitor for use it with a Lm1117 LDO Regulator.
This is the Capacitor datasheet: http://www.ryston.cz/pdf5/ca42.pdf

I want to know the Tantalum Capacitor ESR but i can't find that info in the datasheet.

Is there a way to know the tantalum capacitor ESR?

Will the capacitor i bought will work well like an input and output capacitor on my Lm117 regulator?

Does a 10uF/25V Tantalum Capacitor have a ESR lower than 0.5 Ohm?

Regards.
Alfredo Mendiola Loyola

The data sheet for the LM1117:

http://www.national.com/ds/LM/LM1117.pdf

on page 9, section 1.3, says that if you use a tantalum output capacitor, it must be a minimum of 10 μF; you might want to use a larger value. Furthermore, the ESR need not be .5Ω The requirement is that .3Ω < ESR < 22Ω; this means that it should not be less than .3Ω.

As far as how to "know" the ESR of a tantalum capacitor, if you only have one such capacitor, perhaps for a hobby project, you could measure it with an LCR meter. If you are designing a circuit which will go into commercial production, you should buy a capacitor for which you have manufacturer's data guaranteeing the minimum and maximum ESR over the life of the capacitor. You may be able to get that information for the capacitor you bought by contacting the manufacturer.

What kind of voltage is the regulator handling? If it's from a line transformer with a 100 or 120 Hz ripple, ESR is of no consequence.

To the contrary, it's common for LDO regulators to require ESR of the output capacitor to have a value within a specified range.

To quote from the National data sheet for this regulator:

"The output capacitor is critical in maintaining regulator stability."

If you look at the data sheet, you will see that .1 μF caps are also suggested. One parallels the large electrolytic at the input, and the other is a bypass at the output pin.

The data sheet I referenced above doesn't show any .1 μF caps in any of the example schematics. Are you referring to a data sheet for the LM1117 regulator different than the one referenced above?

 

If ESR is still important to you after beenthere's comment, you should be able to calculate it from the dissipation factor. By the way, a handy "cheat sheet" for some of this stuff (attached) is from the manual for the nice old General Radio 1650B impedance bridge. You can get it from here, which is a fast mirror of the bama site (bama could really be slow, at least a few years ago).

 

My old National data sheet (after looking at it again) indicates a .1 on the input and 1 μF on the output. I find no mention of ESR.

 

My old National data sheet (after looking at it again) indicates a .1 on the input and 1 μF on the output. I find no mention of ESR.

This is very curious. I wonder why they have (apparently) changed their mind about the importance of ESR.

Is that the data sheet for, specifically, the LM1117? What is the date of the data sheet? Did you look at the data sheet available at the URL I gave?

 

The PDF has a date of May, 2003, but the drawings are typical of the 1970's. I think the additional material was to show different packaging. It is, specifically, for the LM117.

There was an extensive rewrite in 2006 (looked at the link). I tend to hang onto data sheets forever - what's changed about an LM117? My guess is that the ESR rating comes from regulating a voltage with much higher ripple, like from a switcher.

 

The PDF has a date of May, 2003, but the drawings are typical of the 1970's. I think the additional material was to show different packaging. It is, specifically, for the LM117.

There was an extensive rewrite in 2006 (looked at the link). I tend to hang onto data sheets forever - what's changed about an LM117? My guess is that the ESR rating comes from regulating a voltage with much higher ripple, like from a switcher.

"My guess is that the ESR rating comes from regulating a voltage with much higher ripple, like from a switcher."

I don't think this is the case. It's simply a problem of the stability of the LDO regulator. Whether the LDO is getting its input from a 60Hz rectified supply, or a switcher, and regardless of how much ripple you have or its frequency, the LDO will be unstable with the wrong value of output capacitor ESR. Much has been written on this subject, for example:

http://www.national.com/an/AN/AN-1482.pdf

http://www.pdffinder.com/get/ldo-regulator-stability-and-ceramic-capacitors.pdf

http://www.fairchildsemi.com/an/AN/AN-6037.pdf

http://www.pdffinder.com/get/esr-stability-and-the-ldo-regulator.pdf

If the ESR is too high, or too low, the LDO may be unstable. If you have an output capacitor with a suitable ESR, putting a .1 μF cap in parallel could actually lower the total ESR sufficiently to cause the LDO to be oscillate, so one should be careful about paralleling additional .1 μF caps if they have very low ESRs themselves.

I looked through my cap assortment for a 10 μF tantalum and measured the ESR up to 3 MHz. The ESR goes down with frequency, but it remained above .4Ω When I paralleled it with a .1 μF film cap, the ESR went below .3Ω at 1 MHz and by 3 MHz it was only .08Ω; this could lead to an oscillating LDO in the low MHz region.

Maybe the LM1117 was one of the early LDOs, and National wasn't yet hip to the possible problems with output capacitors, so they didn't discuss the issue in the early data sheet. Whatever the reason the early data sheet didn't mention it, they (and everybody else who makes LDOs) are certainly alerting users in current data sheets and app notes.

 

"My guess is that the ESR rating comes from regulating a voltage with much higher ripple, like from a switcher."

I don't think this is the case. It's simply a problem of the stability of the LDO regulator. Whether the LDO is getting its input from a 60Hz rectified supply, or a switcher, and regardless of how much ripple you have or its frequency, the LDO will be unstable with the wrong value of output capacitor ESR. Much has been written on this subject, for example:

http://www.national.com/an/AN/AN-1482.pdf

http://www.pdffinder.com/get/ldo-regulator-stability-and-ceramic-capacitors.pdf

http://www.fairchildsemi.com/an/AN/AN-6037.pdf

http://www.pdffinder.com/get/esr-stability-and-the-ldo-regulator.pdf

If the ESR is too high, or too low, the LDO may be unstable. If you have an output capacitor with a suitable ESR, putting a .1 μF cap in parallel could actually lower the total ESR sufficiently to cause the LDO to be oscillate, so one should be careful about paralleling additional .1 μF caps if they have very low ESRs themselves.

I looked through my cap assortment for a 10 μF tantalum and measured the ESR up to 3 MHz. The ESR goes down with frequency, but it remained above .4Ω When I paralleled it with a .1 μF film cap, the ESR went below .3Ω at 1 MHz and by 3 MHz it was only .08Ω; this could lead to an oscillating LDO in the low MHz region.

Maybe the LM1117 was one of the early LDOs, and National wasn't yet hip to the possible problems with output capacitors, so they didn't discuss the issue in the early data sheet. Whatever the reason the early data sheet didn't mention it, they (and everybody else who makes LDOs) are certainly alerting users in current data sheets and app notes.

Is there other Low dropout regulators to convert 5V to 3.3V that doesn't need low esr capacitors or tantalum capacitors?

I don't have esr measurement instruments ...

I'm need to use a bluetoot module (3.3V) with a atmega8 device at 5V

My 5V power supply is a Lm2595 150khz switching.

Best regards.
Alfredo Mendiola Loyola
Lima, Perú

 

Is there other Low dropout regulators to convert 5V to 3.3V that doesn't need low esr capacitors or tantalum capacitors?

I don't have esr measurement instruments ...

I'm need to use a bluetoot module (3.3V) with a atmega8 device at 5V

My 5V power supply is a Lm2595 150khz switching.

Best regards.
Alfredo Mendiola Loyola
Lima, Perú

What do the specifications for your bluetooth device say is the maximum and minimum supply voltage for proper operation? You may be able to derive the needed voltage by simply putting 2 forward biased silicon diodes in series with the 5 volt existing supply instead of a LDO regulator; the diodes will drop approximately 1.3 volts, reducing the voltage to the bluetooth device to 3.7 volts.

Another option would be to add a 1Ω resistor in series with your tantalum capacitor. Then the ESR would be guaranteed to be larger than .3Ω, and if the tantalum is of good quality, it will be very likely to have an intrinsic ESR of less than 21Ω, and the total effective ESR will therefore be less than 22Ω and greater than .3Ω.

 

Is there other Low dropout regulators to convert 5V to 3.3V that doesn't need low esr capacitors or tantalum capacitors?

I see that TI has a LDO regulator (http://focus.ti.com/lit/ds/symlink/tps73601.pdf) that doesn't need an output cap, but it's only good for 400 mA. If that isn't enough for your application, search their catalog for a higher current part, or parallel a couple of them with small build-out resistors.