Hello everybody!
I'm new to the forum so let me start greeting everybody.
I'm a young electronic engineer and starting to build experience in circuit designing and developemnt on my own, since where I work I'm "the expert"... the only electronic guy...
Ow well, I see along lonly road ahead...
Now the question: voltage regulator stability.
I used quite some times before the classicla LM78XX devices, always with no issue. Now I have to develop a circuit that takes a not-so-stable 12V input (could be as high a 19 in spikes) and bring it down to 5V. I started with the LVT1117-05 and after some time of the circuit being on, I hear acoustic noise...
From there I start to dig.. and the ceramic capacitor can have that issue... ok, got it now... but now the other questions arrived...
A lot of regulator, mostly LDO seems to not be declared stable from the datasheets using them, the ESR is too low. I was planning on using LM340-0.5 but on that datasheet I was not able to find any reference on capacitor caramic or not.
I don't feel safe in using tantalium becaus of the possibility of the high spikes.
Can anybody shed some light on this topic? well, both on the specific device LM340 and if possible in general on the stability/instability issues and design good practice for supply solutions?
Thank you a lot in advance!
Best regards

 

how about a buck down converter lm2596 ready made on the Ebay.

 

Thanks for your quick reply!
That device is a buck converter, but i'm trying to be on the cheap side as much as possible.
Thing that I didn't mentioned before (sorry) is that I'm going to need just a very low currennt output (50mA - 70mA at most) so I see it a bit excessive to use a switching solution for such a low current

 

I've had no trouble with the LM340-05 (7805) in the past. If extreme low cost is an issue, then I'd just put about 100 uF @ 35V aluminum electrolytic at the input, and 10 uF @ 16V ceramic at the output. The 10 uF ceramic is to supply quick transient current. If the load current is steady or only slowly changing, you might get by with even less, like 1 uF, and save a few more pennies. If you are concerned about surge current into the 100 uF capacitor, you could put a resistor in series with the input, or more elegantly, an inductor.

If this is an automotive application, you might consider using the LM317AHV, designed to withstand high transient input voltage. That's a $1 part, though, and if you're really cost sensitive, might not fit in budget.

 

Thanks for your quick reply!
That device is a buck converter, but i'm trying to be on the cheap side as much as possible.
Thing that I didn't mentioned before (sorry) is that I'm going to need just a very low currennt output (50mA - 70mA at most) so I see it a bit excessive to use a switching solution for such a low current

A simple resistor/zener shunt regulator is adequate for that specification.

That the input voltage can dip down as low as the required output voltage is a worry, but a hefty reservoir electrolytic can take care of short dips.

For a linear regulator, a LDO type is the absolute minimum you might just get away with.

If output voltage integrity is very important - you might have to jump past a basic switcher and go for a buck/boost type.

 

so use an Lm317t, or lm7805

 

so use an Lm317t, or lm7805

If the voltage source regulation is as bad as the OP states - even a LDO might not do.

For 50 -70mA, a resistor/zener is plenty, and a decent electrolytic might keep things going if the source dips low.

 

Thanks for all the answers!
I?ll dig more into them.
Still the question about all the issues with stability and ceramic capacitors is there.
I was considering the classic approach as suggested, but I wanted also to give a try to SMD design, that pretty much brought out the topic about the ceramic capacitors issues