SMPS Buck Regulator Experience?

Status
Not open for further replies.
Would anyone be interested in updates as I build and test smps regulator prototypes?
Mike
Sure! I'd like to know your opinion on that regulator.

I like the MC34063A and I often used it as a buck or a boost regulator. It's available in the PDIP8 package, so it's easy to use it on veroboards for prototyping. Here's the datasheet if you want to take a look.
 
Last edited:

I feel, once Mike has those chips and decided to use them, let us see how to balance part. the ESR caps have ESR printed on them.
Of late I heard somewhere, that in olden days capacitors have not been made with that quality level, and that in modern days the capacitor manufacturers rae well ahead and most capacitors are really low ESR according to that article. So most caps, example Philips make,may perhaps be useful, unless one works around few hundreds of KHz if not 1MHz.
 
Well lets see.... at digikey $1.56 for the chip, $0.40 for the diode, $0.60 for the PFET, $0.86 for the inductor, and $0.60 each for the input and output 10uFs (unfortunately they are making you get 10 at a time )
You're not going to post a pretty list with part numbers and links like I did (grin)?

Any pictures of your working circuits?

Mike
 
That part number seems familiar (used in a couple ICD2 clones maybe?).

How are you using it? As a 5v/500ma regulator?

Mike
 
Last edited:
I used it to step-down 12V to 5V @ 0.4A and I've recently tried to boost 9V to 15V. I like the datasheet and related app notes, they provide design formulae which you can use in an excel sheet for different values of the input voltage, etc. But the higher switching frequency of your regulator looks interesting, I'd like to use smaller inductors.
 
eng1,

Yes, the higher switching frequency and smaller inductor values on the '4575 are what caught my eye originally.

There's also has a 3 amp version, the MIC4576, in the same TO-220 and TO-263 package.

Thanks for the info'.

Mike
 
Last edited:
Mike, I am not expert on driving LEDs, but it apears to me that if you are going to build a signboard, most of the current will go to the LEDs and control portion is going to take comparably negligible current, right?

Why don't you built your buck regulator for LED Forward Voltage + 0.1V (or whatewer spread of your LEDs VFW is) and use just linear LM78L05 for control circuitry? It's just 20-50c extra for linear regulator but your appliance will be supergreen and producing no heat.
 
I don't really understand why the C1 (150F) need to be low ESR. If you procure aluminum electrolytic capacitor for C2, they should serve your needs.
 
You're not going to post a pretty list with part numbers and links like I did (grin)?

Any pictures of your working circuits?

Mike
You want links? Sheesh!

http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail?name=LM3485MMTR-ND
http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail?name=587-1337-1-ND
http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail?name=SS34-E3/57TGICT-ND
http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail?name=513-1458-1-ND
http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail?name=FDC610PZCT-ND

I am expecting the hardware shortly, but the version I am building is actually very low current for a battery operation. That inductor and capacitor with a 0.022 ohm ESR resistor will give you 300KHz 1.5A supply with 0.6Apk-pk ripple current in the inductor.

I do not currently have a requirement for normal output currents, but feature hungry customers will get it to that point in short order.
 
Last edited:
I don't really understand why the C1 (150F) need to be low ESR. If you procure aluminum electrolytic capacitor for C2, they should serve your needs.

The input current turns on off very fast. If you have any length of wires from the power supply then the change in current in C1 will be just about the same as the change in current in the inductor.

C1 needs a low ESR or the change in current will make the voltage oscillate quite a lot.

I would also add a ceramic capacitor or two on the input of the IC.
 
OOPS . I regret my mistake. I meant 159µF. Having said that I realised a problem. Normally when we type a letter the curser goes to next location. but when you select a symbol it is still on the selected symbol with blue background. As usual, if we type the next character, the symbol gets cancelled and new character takes over. The site designer may please see to it.

Additionally, all of us know well that any character after a full-stop and "i" , a single character word are generally in uppercase. but the package used by the site owner sustain it in lower case which appears meaningless. I wonder whether any mod can be offered for these even !!

@diver300, thanks for the explanation.
 
Last edited:
159µF isn't a standard value, use 150µF or 220µF.

Yes I know it's a typo, you meant 150:mu:F, sorry I couldn't resist.
 
Last edited:
159µF isn't a standard value, use 150µF or 220µF.

Yes I know it's a typo, you meant 150µF, sorry I couldn't resist.
Thanks Hero999.
Oh again a printer's devil
May be, I should double check before posting . I shall be more careful from now on.
 
It works!!!

After being side-tracked this last week, I just finished wiring up the prototype MIC4575 switching 5v 1a regulator and hooked it up to an LED Clock project and it works great.

The back-story

I normally use 7805 TO-220 type 5v 1a regulators on my projects which are powered from surplus 12-15vdc 1-3 amp laptop power 'bricks'. This works great on low power projects.

Some recent projects are starting to use a bit more current and the 7805 regulators are getting extremely hot trying to dissapate several hundred milliamps or more at the 9-12v drop from the 12-15v power 'bricks'.

The switching regulator prototype

I put a couple single-in-line machined pin sockets on the prototype which allows me to connect the regulated output to my solderless breadboard using a couple solid conductor wires.

This prototype board will eventually include another MIC4575 regulator for 3.3 volts as well as an ICSP header, DB-9 serial connector, etc., all of which will use the same single-in-line machined pin sockets along one edge of the board for connection to a solderless breadboard using solid conductor wires.

Quick-n'-dirty executive summary

I don't have a way to measure ripple voltage or quantify efficiency and performance. I'm just delighted that the regulator IC seems to be about room temperature (without a heat sink) while using the same power 'brick' and powering the same Clock project which used to practically burn up a 7805 regulator after only about 30 seconds.

I sampled the MIC4575 IC in both 3.3v and 5.0v versions and in both TO220-5 and TO263-5 SMT packages. The other parts required to build a single regulator cost around $3 (USD).

All in all, I'm delighted and already have plans to include a surface mount Maxim current sense IC and (high side) shunt resistor and an 8-pin PIC with 3.5mm serial jack on the next prototype. I will use the serial connection to Hyperterminal or to my serial LCD when I want to display target circuit current usage down into the nano-amp range.

Regards, Mike

<added>

anyone want to see a picture of the prototype?
 
Last edited:
anyone want to see a picture of the prototype?

Hell ya! You know you didn't need to ask.

I collected the parts to build one years ago, but never assembled it. I don't know where the chip has gone now.
 
@ Mike,
Ofcourse any one would love to see the nice mod done.

Having changed over to 4575, you may need to check the noise level at the output. Yes, the mod brings a lot of improvement in efficiency, but the thermostat effect for the crystal inside the clock is lost perhaps.
 
Last edited:
Status
Not open for further replies.
Cookies are required to use this site. You must accept them to continue using the site. Learn more…