Linear tech/ADI parts have a premium price so one has to evaluate if they are necessary. Also using second source parts have advantages of availability.
I routinely use ST LM217T-DG for the extra pennies they cost, it has slightly better specs than the regular lm317
I routinely use ST LM217T-DG for the extra pennies they cost, it has slightly better specs than the regular lm317
Not to threadjack, BUT.....the LT1963 can be tricked into higher input voltage, much the same way you protect the LM317 in a Maida regulator, AND you can set up "remote sense" -- I believe it's shown in the datasheet.
The fact that TI (and others) keep making them should tell you something about the market demand for these ICs. The LM317 first saw the light of day in 1976 if I recall correctly. I'm sure it's been ported to newer semiconductor processes at least a half-dozen times by now.
Tom
The LM317LITP/NOPB has to-92, soic and dsbga packaging options, talk about flexibility. Note dsbga is light sensitive.
That's cute. 🙂
I built an LM317/337 regulator with ceramic output caps at one point too. The LM317 was stable (or at least I didn't observe any instability) but the LM337 definitely did not like a ceramic output cap. I've used plain electrolytic caps ever since then.
Electrolytic capacitors also have the advantage that they provide the capacitance that's marked on the can even with significant DC bias. Sadly the same cannot be said for X5R/X7R.
Tom
For small currents, a RC output filter is a good way to keep the regulator from going unstable.
Ed
Ed
Here's my first revision of the BBB (Best Basic for a Beginner) circuit. It's handle a range of VA's and voltages. BOM and PCB to follow. C1 is Tantalum.
Regards,
Dan
I'd up the 470 uF to 1000 uF for three reasons: 1000 uF tends to have better availability, it is often cheaper, and it'll result in lower ripple voltage. Better performance, higher availability, lower cost. That's a rare find. Take it! 🙂
Tantalum is polarized, so you'll want the schematic to reflect that. Personally I'd use an 1 uF, X5R or X7R ceramic, but that's me.
You could make it even more useful for audio by making a split/bipolar supply out of this. One could also consider board options to allow the transformer primary to be configured for 230 V for more international appeal. Just suggestions...
Tom
Tantalum is polarized, so you'll want the schematic to reflect that. Personally I'd use an 1 uF, X5R or X7R ceramic, but that's me.
You could make it even more useful for audio by making a split/bipolar supply out of this. One could also consider board options to allow the transformer primary to be configured for 230 V for more international appeal. Just suggestions...
Tom
Tantalums give me alarms ever since I read the following a few years ago. I'd make sure C1's rated voltage is at least twice the maximum voltage that will ever appear on it.
https://web.archive.org/web/2015101...Derating Guidelings for Tantalum 2011 (3).pdf
https://web.archive.org/web/2015101...Derating Guidelings for Tantalum 2011 (3).pdf
As time permits I'm going to implement all of your suggestions. The first two (1000uF and XR7) were easy and are done already. I'm going to build and test the first board and then move on.
Regards,
Dan
That's an interesting document. Thanks for sharing.
Tantalum caps used to be everywhere. Now they seem to have been pushed out by ceramic caps or by small electrolytic cans.
Tom
Or maybe you didn't run them near 100% of their rated voltage. As noted in the KEMET document linked to above, running them below 80% of their rated voltage (or 90% for ≤10 V) makes them last much longer.
With the high-k dielectrics in, say, X5R, X7R, etc. ceramic capacitors you also have to derate the voltage rating. Not for reliability, but rather because of the rather horrid voltage coefficient of those dielectrics. Commonly the 'good' (Class 2) high-k dielectric capacitors are down to ~20% of their marked capacitance at the full rated voltage. Physics is annoying sometimes.
Tom
With the high-k dielectrics in, say, X5R, X7R, etc. ceramic capacitors you also have to derate the voltage rating. Not for reliability, but rather because of the rather horrid voltage coefficient of those dielectrics. Commonly the 'good' (Class 2) high-k dielectric capacitors are down to ~20% of their marked capacitance at the full rated voltage. Physics is annoying sometimes.
Tom