Hello all,

For my project I need a 5V - 1.5A fixed voltage regulator. (Input voltage ~12V)

At first I wanted to use LM1086 , but Google told me to better avoid those "older generation" LDO-s. Reason is that back in the day they were designed to work with tantulum caps and also need certain ESR on the output cap, in order to stay stable.

Now we have awesome ceramic caps (tantulum = bad), but they have extremely low ESR and so the story goes..

So, I searched for "newer" LDO-s that should be designed to work with low ESR and found these two: LD29150DT50R and LD1086DT50TR.

Problem is that their datasheets do not say much about capacitor requirements. Only have a picture:

What do you guys think, do I need to use polarized (electrolyte) caps here or are big (>10uF) ceramic caps a safe choice.
Also are here any ESR considerations that I should worry about.

 

Problem is that their datasheets do not say much about capacitor requirements. (...) What do you guys think, do I need to use polarized (electrolyte) caps here or are big (>10uF) ceramic caps a safe choice. Also are here any ESR considerations that I should worry about.

I think the information you're after is given in the charts on the two data sheets: see Figure 18 on page 14 for the LD29150, and Figures 21, 22 and 23 on pages 19 and 20 for the LD1086. These show regions of stability for ESR vs. output capacitance.

 

I use LM1086 without any problems, or any other LDO linear regulator.

The question that need to be asked is are you planning to use the TO-220 package or the SMD package.
In general, SMD breadboards will give you the best performance.

As a general rule, simply follow the filtering recommended by the manufacturer.

The input side should have a reservoir capacitor (100-1000μF aluminium electrolytic) if being supplied by a fullwave rectifier circuit.

For the output side, 0.1μF ceramic and 10μF tantalum electrolytic is commonly used.
The important thing is to put the capacitors as close to the pins as physically possible.

 

I think the information you're after is given in the charts on the two data sheets: see Figure 18 on page 14 for the LD29150, and Figures 21, 22 and 23 on pages 19 and 20 for the LD1086. These show regions of stability for ESR vs. output capacitance.

Thank you !
Yes, they are what I was looking for. All of them seem to show that still at least 1 Ohm ESR is needed.

I looked at a ceramic capacitors ESR graph now (on second page). It shows that ceramic capacitor has lower than 0.01 Ohm ESR at some frequencies.

Not sure if it makes them unusable for LDO requiring atleast 1 Ohm ESR now or are only the lower frequency (<10kHz) ESR values important (where ceramic capacitors ESR is over 1 Ohm).

 

I use LM1086 without any problems, or any other LDO linear regulator.

The question that need to be asked is are you planning to use the TO-220 package or the SMD package.
In general, SMD breadboards will give you the best performance.

As a general rule, simply follow the filtering recommended by the manufacturer.

The input side should have a reservoir capacitor (100-1000μF aluminium electrolytic) if being supplied by a fullwave rectifier circuit.

For the output side, 0.1μF ceramic and 10μF tantalum electrolytic is commonly used.
The important thing is to put the capacitors as close to the pins as physically possible.

Thank you for the answer!

I will be using SMD (TO-263) package, so guess that's good.

Looked that there are even 100μF ceramic capacitors - do you think it would also work as a reservoir capacitor for the input. (Just curious.)

And if we put a ceramic and tantalum in parallel, could it somehow decrease overall ESR value to a point where LDO is not stable anymore (as two resistors in parallel). Not sure if it works that way for ESR tho (again, just curious).

 

I believe you might be overthinking this problem.

For the input capacitor, you need to look at your current load and voltage ripple. For medium applications, say up to 200mA load, 470-1000μF aluminium electrolytic works fine.

On the output pin I use 0.1μF ceramic SMD.

Further along the supply rail at linear ICs, I have 0.1μF ceramic SMD and 4.7μF tantalum SMD in parallel at the chip Vs and GND pins.
For digital ICs, I use just 0.1μF ceramic SMD.

I don't bother with the 10μF as recommended in the datasheet.

 

I believe you might be overthinking this problem.

For the input capacitor, you need to look at your current load and voltage ripple. For medium applications, say up to 200mA load, 470-1000μF aluminium electrolytic works fine.

On the output pin I use 0.1μF ceramic SMD.

Further along the supply rail at linear ICs, I have 0.1μF ceramic SMD and 4.7μF tantalum SMD in parallel at the chip Vs and GND pins.
For digital ICs, I use just 0.1μF ceramic SMD.

I don't bother with the 10μF as recommended in the datasheet.

I believe that you are right and I am overthinking - that is why I turned to people with more experience .

I shall settle down with LM1086 for now. If for some reason it does not work out, I can switch it to LD1086 (they have more or less the same footprint and pinout).

Guess that tantalum is safe enough from overvoltage, if I use 20V tantalum at 5V output.

It would be interesting to read tho, if someone would answer the questions I had previously.

 

Hello all,

For my project I need a 5V - 1.5A fixed voltage regulator. (Input voltage ~12V)

At first I wanted to use LM1086 , but Google told me to better avoid those "older generation" LDO-s. Reason is that back in the day they were designed to work with tantulum caps and also need certain ESR on the output cap, in order to stay stable.

Now we have awesome ceramic caps (tantulum = bad), but they have extremely low ESR and so the story goes..

So, I searched for "newer" LDO-s that should be designed to work with low ESR and found these two: LD29150DT50R and LD1086DT50TR.

Problem is that their datasheets do not say much about capacitor requirements. Only have a picture:

What do you guys think, do I need to use polarized (electrolyte) caps here or are big (>10uF) ceramic caps a safe choice.
Also are here any ESR considerations that I should worry about.

Every once in a while; a magazine article pops up suggesting that LDO regulators can suffer instability if very low ESR caps are used (Like MLCC or maybe even tant).

Its a bridge I haven't had to cross yet, but the advice I've collected so far is that various manufacturers have published supplements to their application note that deal with this problem in some detail.

 

Every once in a while; a magazine article pops up suggesting that LDO regulators can suffer instability if very low ESR caps are used (Like MLCC or maybe even tant).

Its a bridge I haven't had to cross yet, but the advice I've collected so far is that various manufacturers have published supplements to their application note that deal with this problem in some detail.

Thank you for the suggestion. I searched for manufacturer published materials now and came up with this: ESR, Stability, LDO

Theory gets complicated really fast with pole/zero cancellations (I think..) and more fun - more than I have time for at the moment.

But it did show me what to look for and how to test the instability of an LDO.
It is a shame tho, that LM1086 does not have the needed ESR/Output capacitance plots.

 

If the input voltage is 12 Volts, why not use a LM7805 or LM340-5.0 type?

Capacitors required is modest, at the input 220 nF and 100 nF at the output, both of which can be multi-layered ceramic.

Because they have a NPN transistor as series element, they actually often better dynamic properties (As Ripple Rejection and Load Regulation) than an LDO type.

 

If the input voltage is 12 Volts, why not use a LM7805 or LM340-5.0 type?

Capacitors required is modest, at the input 220 nF and 100 nF at the output, both of which can be multi-layered ceramic.

Because they have a NPN transistor as series element, they actually often better dynamic properties (As Ripple Rejection and Load Regulation) than an LDO type.

The TS needs 1.5A, with that in/out difference; a 7805 would be pretty close to thermal shutdown at its alleged 1A.

Most application notes describe how to boost a 78xx regulator with an external bypass transistor - a current sense resistor is inserted in the input feed, the voltage developed across the resistor biases on the B/E junction of a PNP bypass transistor. A few appnotes describe for even higher current with a moderately rated PNP driving a much heavier NPN bypass transistor.

 

If the input voltage is 12 Volts, why not use a LM7805 or LM340-5.0 type?

Capacitors required is modest, at the input 220 nF and 100 nF at the output, both of which can be multi-layered ceramic.

Because they have a NPN transistor as series element, they actually often better dynamic properties (As Ripple Rejection and Load Regulation) than an LDO type.

I have understood that linear regulators act as a dynamic resistor and have to dissipate all of the excess power.
At times my schematic could draw 1A+ so at 12V it means 7W power to dissipate. My plan was to keep things small (SMD) and it looks like LDO satisfies those requirements.

But if you say that it can be achieved with LM7805 or LM340, I will definitely want to use it.

 

I have understood that linear regulators act as a dynamic resistor and have to dissipate all of the excess power.
At times my schematic could draw 1A+ so at 12V it means 7W power to dissipate. My plan was to keep things small (SMD) and it looks like LDO satisfies those requirements.

But if you say that it can be achieved with LM7805 or LM340, I will definitely want to use it.

A 7805 probably won't do the current you stated.

Look at the appnotes for how to add an external bypass transistor.

 

There are many 3-terminal 5V regulators that can handle greater than 1A.

http://www.st.com/content/ccc/resou...df/jcr:content/translations/en.CD00000449.pdf

http://www.ti.com/product/LM340

http://www.linear.com/product/LT1085-Fixed

 

There are many 3-terminal 5V regulators that can handle greater than 1A.

http://www.st.com/content/ccc/resou...df/jcr:content/translations/en.CD00000449.pdf

http://www.ti.com/product/LM340

http://www.linear.com/product/LT1085-Fixed

Thank you for the suggestions. Unfortunately I am afraid that they all suffer with one shortcoming (for my requirements).
In their datasheets they all mention "If adequate heat sinking is provided, they can deliver over 1.5-A output current."

For this project, the extra effort of big heat sink still makes me prefer LDOs.

What Ian Field proposed should work:

But I have decided to stick with the LM1086 for now and play with the caps a little. I will come and rant about how bad it is if I have a bad experience.

For now I thank you all for your input, appreciate it!

 

Which ever way you look at it you will still have to dissipate 7W (1A @ 7V).
How is a LDO regulator going to help you here?

 

Which ever way you look at it you will still have to dissipate 7W (1A @ 7V).
How is a LDO regulator going to help you here?

Yes, you are right.

I had not done my research..

Just realized that LDO is just a linear regulator with a low voltage drop on the regulator. For some reason thought that some switching is involved with it.

Now I do see that 7805 etc are actually a good idea. I will think about it, maybe even go with the external bypass transistor option that Ian Field proposed.

 

Yes, you are right.

I had not done my research..

Just realized that LDO is just a linear regulator with a low voltage drop on the regulator. For some reason thought that some switching is involved with it.

Now I do see that 7805 etc are actually a good idea. I will think about it, maybe even go with the external bypass transistor option that Ian Field proposed.

I wouldn't do that. Your bypass transistor still has to dissipate 7W, or more like 7V @1.5A = 11W

What are you using as your power source and what are you supplying with 5V @1.5A?
If you reduced the input supply to 7.5V and use a LDO regulator, you will be dissipating 1.5A @ 2.5V = 4W.
You will still need a heat sink though a smaller one.

 

See

 

See
View attachment 121966

Thank you Bordodynov, but I am not sure what to think of it. (Have not used LTspice myself.)
Seems like second regulator has extra load in parallel and third has polarized cap, but their purpose and if other caps are tantalum or ceramic (for example) leaves me wondering.