@@GarageTinkering Just looking at the circuit at 15:32 I have some concerns. The output of the LM2596 needs to go to an inductor and the cathode of a Schottky diode. The other end of the diode goes to ground. It's a switch mode power supply. The FB line goes to the other terminal of the inductor rather than ground. Or so it seems to me.
This a video and subject that needs coverage, but you’ve some pretty big gaps in your understanding of the bits. An LDO turns the extra voltage to heat where as the buck converter does not. Instead, it switches the FETs on and off to maintain a chosen voltage. You need to add all of the components in the data sheet and not recommend your beginner viewers gloss over them.
I am. I've actually just ordered the PCBs that I make in the update, so just want to check that they're 100% fine before putting them out there as a good example of what to do!
@@squarewav6047 yeah. I noted the inductor oversight in the description. I've actually updated my choice of components now so it's moot, but yeah that was a mistake. And you're right about the diode. Mouser did me dirty on that one as the schematic they provide has the wrong symbol. Going to release an updated video next week with the new schematic
I've started watching your videos lately and have assumed you knew what you were talking about but using a buck converter as an LDO regulator has me questioning my decisions 😅. Keep up the good work but spend more time on those boring data sheets. The example schematics are there for a reason and typically implemented in commercial products.
6.3v cap is not great for a 5v power line. If you go to the “boring” datasheet you will find a derating curve which will tell you the 4.7uF will be derated to about 60-70% and by time, temp, age variation it gets worst. That’s where unpredictable operations starts and hard to find the root-cause. The fuse protects the network before itself in case of short behind it. Like in the household. That means it protects the cable against melting not the PCB. Love your ideas and videos, I hope I helped you.
Worse than that, one small spike on the voltage line and the capacitor go bye-bye. Always use capacitor rated at least double the nominal expected voltage. So for 5V, at least 10V cap shall be used.
@ what would cause a spike on a regulated output? Failure mode of dcdc converter is a solid short not a transient and also won’t conduct through due to input filtering. Also depends on your spike definition, but caps are good spike absorbers like an ESD spike, but thats again defined in datasheet how much energy it can aborb.
@@bzolitb Faulty diodes on your alternator. Not common to be fair, but still possible. And your instrument cluster may well get zapped if that happens.
KiCad. Lots of the components are already in there, but otherwise you can download schematic / footprint / 3D models from mouser to use too, which makes things way easier.
Wow, terrific video!
TVS sybol you used looks like an inductor to me, it even has the L designator.
Mouser did me dirty on that one... the component that I downloaded came with the wrong symbol and I overlooked it
Just came over from NewTubers, I might need your help designing a PCB for my next classic car build!
I have a guide on full end to end automotive PCB designing coming in the next couple of weeks that will probably be a lot of use for you!
the LM2596S is not an LDO it is a buck converter, it needs an inductor and feedback connected to work!
@@STRATOS13PAO you're right on the FB connection. That should have gone to the output circuit not the ground. Oversight on my part there. Good catch.
@@GarageTinkering Just looking at the circuit at 15:32 I have some concerns. The output of the LM2596 needs to go to an inductor and the cathode of a Schottky diode. The other end of the diode goes to ground. It's a switch mode power supply. The FB line goes to the other terminal of the inductor rather than ground. Or so it seems to me.
This a video and subject that needs coverage, but you’ve some pretty big gaps in your understanding of the bits.
An LDO turns the extra voltage to heat where as the buck converter does not. Instead, it switches the FETs on and off to maintain a chosen voltage. You need to add all of the components in the data sheet and not recommend your beginner viewers gloss over them.
You know, in hindsight I agree and actually have a more in-depth video coming that covers everything
Great video! Are you still planning to do an update video with the new design made discussed in the comments?
I am. I've actually just ordered the PCBs that I make in the update, so just want to check that they're 100% fine before putting them out there as a good example of what to do!
I'm confused. LM2596 is a switching regulator but you don't have the required inductor. Also SMBJ33A is a protection diode, not an inductor.
@@squarewav6047 yeah. I noted the inductor oversight in the description. I've actually updated my choice of components now so it's moot, but yeah that was a mistake.
And you're right about the diode. Mouser did me dirty on that one as the schematic they provide has the wrong symbol.
Going to release an updated video next week with the new schematic
I've started watching your videos lately and have assumed you knew what you were talking about but using a buck converter as an LDO regulator has me questioning my decisions 😅. Keep up the good work but spend more time on those boring data sheets. The example schematics are there for a reason and typically implemented in commercial products.
I'm currently sat here knee deep in the TPS62933 datasheet... it's too late in the day for this nonsense lol
6.3v cap is not great for a 5v power line. If you go to the “boring” datasheet you will find a derating curve which will tell you the 4.7uF will be derated to about 60-70% and by time, temp, age variation it gets worst. That’s where unpredictable operations starts and hard to find the root-cause.
The fuse protects the network before itself in case of short behind it. Like in the household. That means it protects the cable against melting not the PCB.
Love your ideas and videos, I hope I helped you.
Worse than that, one small spike on the voltage line and the capacitor go bye-bye. Always use capacitor rated at least double the nominal expected voltage. So for 5V, at least 10V cap shall be used.
@ what would cause a spike on a regulated output? Failure mode of dcdc converter is a solid short not a transient and also won’t conduct through due to input filtering. Also depends on your spike definition, but caps are good spike absorbers like an ESD spike, but thats again defined in datasheet how much energy it can aborb.
@@bzolitb Faulty diodes on your alternator. Not common to be fair, but still possible. And your instrument cluster may well get zapped if that happens.
Which design software are you using? I am struggling to find one which has the components already in it like you got there.
KiCad. Lots of the components are already in there, but otherwise you can download schematic / footprint / 3D models from mouser to use too, which makes things way easier.