Calculating real capacitance for the PTN4050

[McDude]

I've been reading the datasheet for UC2577TDKTTT-ADJ chip (TI) and have a question about calculating "real" or "needed" capacitance based on application, in relation to the PTN04050C boost.

In reading the UC2577 datasheet, there is a section on actually calculating the capacitance and inductance (right word?) needed based on the expected minimum input voltage and desired output voltage (factoring the "duty" of the chip).

The question I have, being an electronics noob is, why is there no such information on the 4050? Is it because the inductor is already figured in? If the chip has a wide input and output range, then I would assume that the external components would be determined based on the application environment, not a pre-canned schmematic that (perhaps?) just assumes a max/high situation? I've been wondering about those 100uF capacitors and what is "really" needed for our application (vaping) vs. a general test circuit (which the datasheets often supply). Oddly, or perhaps appropriately (depending on the answer I get to this question), there is no section in the 4050 datasheet for calculating external component requirements - I found that odd - that 100uF is a "one size fits all" component.

Anyone with more electronics knowledge than I shed some light for me?

 

[Java_Az]

The UC2577-adj is just the bare regulator you have to add quite a few parts to it inductor ,diodes ,resistors ,caps as where the ptn04050c is a module already it has a inductor 2 chips ,caps ,diodes ,resistors. All you have to do for it is slap a cap on the input and output. So with the UC2577 your building the circuit from scratch with the PTN04050c it is pretty much already built. The data sheet on the PTN04050 has a recommended cap list they have one on there thats 180uf .

 

[WillyB]

Well CapeCAD is the only one I know of with SEPIC experience (like the UC2577). He also has the necessary test equipment and knows how to use it. I PM'd him to join in.

there is no section in the 4050 datasheet for calculating external component requirements - I found that odd - that 100uF is a "one size fits all" component.

You make it sound like TI says to just grab any 100uF cap and run with it. Nothing could be further from truth. The exact requirements and various 'terms' are beyond me, but seeing as the Panasonic 180uF cap from the recommended list is readily available and costs only 60¢ makes the choice easy. If I needed a smaller form factor I would defer to Nuck's recommendation, someone who has a lot of experience with the booster in real world use.

Nuck has mentioned that without an input cap the battery takes quite a beating.

Maybe he'll join in.

 

[McDude]

Well, I do think most people tend to just "go with the schematic" sometimes and not read the entire datasheet, so was looking for that more detailed information, ie. the "why 100uF". However, reading the datasheet further, the capacitance table provides the answer. It would appear that 100uF is not required for a<=5v voltage output application. I'm simply the kind of person that wants to know "why" rather than going by a schematic or a recommendation - there's math behind each electrical characteristic of a circuit and the 100uF wasn't making sense.

Here's the tail end bit from the PTN04050's datasheet, it really does seem that 100uF for <=5v is pretty much not needed.

Questioning isn't a bad thing, it makes for better results

 

[McDude]

Of course, not sure I'm reading the table correctly and am completely unsure why different manufacturers of caps are listed in relation to different voltage output configurations for cap selection. There is no "clear statement" in the datasheet about this, but there is information that "alludes" to the 100uF not being a static factor.

 

[McDude]

You make it sound like TI says to just grab any 100uF cap and run with it.

No, not really - I am perhaps saying that TI couldn't possibly design a schematic for a particular application in a general datasheet. I think that's a reasonable assumption, at least an assumption that leads to questioning instead of implementing without understanding (which is my goal, the mods are second to simply learning).

 

[McDude]

Agreed, but given my goal (learning) the UCC2577 actually was of interest vs. using a pre-made solution like the 4050c. The interesting little lesson that I gleaned from examining the UC2577 was "electronic components are integrated into a circuit where the external characteristics are calculated based on application" - this made me consider the 100uF requirement for the 4050c as being questionable or at least deserving questioning.

And yeah, the datasheet has a 180uF which would indicate that capacitance selection is based on application to some degree. So given that we have a particular application, it should be possible to determine exact capacitance. I just have a hard time believing our particular need matches exactly the *example* schematic where every other application would need the designer to reference the capacitance table - those nice coincidences are very questionable to me

 

[McDude]

Although, I do want to stress that I'm not trying to be a "jerk" LOL, I just want to see if the 100uF are required and why - what is the math behind it or characteristic of the component that requires it (such as the on-board inductor).

 

[CapeCAD]

Disclaimer: My experience with the SEPIC is limited to my own design and reading hundreds of references.

With SEPIC or BOOST converters there is an inductor at the input so the input current is continuous and there is little ripple current at the input capacitor. Input capacitor is used to prevent impedance interactions with the input supply (batteries).

SEPIC and BUCK converters have an inductor on the output so ripple current must be taken into consideration, though we are a using a purely resistive load in relatively short bursts, so ripple current should not be as critical as it would be if we were powering something sensitive to the ripple.

With that said, if TI says it is "Required" then they do not have enough capacitance on their board and expect you to provide it externally if you want to meet their published specifications. If you do not need to meet their specifications, then it's up to the end user to determine how much ripple can be permitted on the input and output.

I personally wouldn't go with less than 10uF (preferably ceramic because of the lower ESR) on the input and output of any mod, even if it's not "required".

 

[CapeCAD]

Of course, not sure I'm reading the table correctly and am completely unsure why different manufacturers of caps are listed in relation to different voltage output configurations for cap selection. There is no "clear statement" in the datasheet about this, but there is information that "alludes" to the 100uF not being a static factor.

Each of the capacitors listed has a different ESR, this has to be taken into consideration along with the capacitance. The capacitors you highlighted are ceramic and are listed for lower voltages because ceramic capacitors are not rated for the same voltages as the tantalums, but have a lower ESR so higher capacitances are not needed.

 

[WillyB]

Of course, not sure I'm reading the table correctly and am completely unsure why different manufacturers of caps are listed in relation to different voltage output configurations for cap selection. There is no "clear statement" in the datasheet about this, but there is information that "alludes" to the 100uF not being a static factor.

Well there is always a lot of LOW ESR talk when discussing caps from TI, which basically narrows the choices quite bit, The RS caps are definitely out.

One thing about these forums is after a while you realize the posters worth paying attention to. Many have extensive vaping and electrical expertise, and some also have sophisticated test equipment.

What kind of test equipment do you have?

 

[Java_Az]

Well, I do think most people tend to just "go with the schematic" sometimes and not read the entire datasheet, so was looking for that more detailed information, ie. the "why 100uF". However, reading the datasheet further, the capacitance table provides the answer. It would appear that 100uF is not required for a<=5v voltage output application. I'm simply the kind of person that wants to know "why" rather than going by a schematic or a recommendation - there's math behind each electrical characteristic of a circuit and the 100uF wasn't making sense.

Here's the tail end bit from the PTN04050's datasheet, it really does seem that 100uF for <=5v is pretty much not needed.

Questioning isn't a bad thing, it makes for better results View attachment 26199

View attachment 26199

They also say to use 5 of these caps for the input . 5 on the output with a see note#2

(2)
The maximum voltage rating of the capacitor must be selected for the desired set-point voltage (VO ). To operate at a higher output
voltage, select a capacitor with a higher voltage rating.

In other data sheets i have read the caps are required for performance reasons. mainly they say something about lag( slower start up time) but i would also worry about efficiency without them on the circuit.

 

[McDude]

What kind of test equipment do you have?

I have next to nothing in terms of test equipment. A cheap multimeter, a "lab power" box that I created - that's about it - soldering instead of breadboarding at the moment. So, really, I'm a complete electronics noob - but instead of creating mods based on diagrams and not understanding it, I took it as a "goal" to learn basic electronics about 4 weeks ago. For two weeks I basically studied up on resistance, capacitance, and current behavior as well as running simulations and crating test circuits both in hardware and in simulators. Once I got to the point of being able to design a mod circuit that I both created and understood, I decided to start making the "real things". My background is in computer programming, so "logic" isn't alien to me - that helps - and I have worked with hardware at the software level (programmable controllers), just never worked with the physics of electronics before. I'm finding it absolutely facinating

I was reading the datasheet more and came across the ESR stuff as well, so that's my next point of research. If I'm going to design a boost mod, then I'd like to know *exactly* how the circuit works and what exact part each component plays in the whole (and why) which now will include the ESR factor decoupled from capacitance needed to achieve certain results.

I appreciate the constructive information, very very much (WillyB and Java_Az).