Alex's Linear Power Supply Build Log

Hey all,

Had to put the power supply to the side for a bit while i was away on work and waiting for parts to arrive. I got the boards back from JLC PCB and the quality is pretty good. I don’t have time yet to get the whole board assembled but I thought I would get the new power section up and running so I could do some load tests on the 5 volt rail.

The voltage is pretty solid throughout a wide input range as expected
5.11 volts with just the LED loading down the output at 2 mA or so. This drops to 4.99 Volts at the rated 500 mA and the switcher IC gets a bit on the warm side, but the diode I’m using is not exactly the correct one as per recommended in the data sheet so I think there is some loss there. I am still waiting on the correct diode as well as some of the digital components

1 Like

another note, here is a scope view of the transient turn on response of the output voltage. currently just running from a 12V input voltage. about 1.53 milliseconds rise time when power is applied.

I’m sure it’s cooler than the previous linear regulator, but I wouldn’t expect it to be too warm at 500 ma. I have a switcher bucking 24V down to 3V3 at close to 1 A and it barely gets warmer than ambient.

Care to post the schematic for your new board?

Sure, schematic is here.

Lab Bench Power SupplyV3.0.pdf (234.4 KB)

I see you have an input filter on your DC-DC converter. If that isn’t designed correctly it can do more harm than good. It might be interesting to put a scope on it and see how it is performing. A poorly designed input filter can cause the converter to be less efficient.

Also, I would use the output of the current control section (Q4) to instead drop VSET instead of pulling down the output of the voltage control section. As it is it might not be very stable at the current threshold as the voltage control section would continue to try to control the voltage.

I did some more looking into the board and aside from the budget wire you saw in the photo for the feedback wire, I think I was also using a blown catch diode. Checked it out with a meter and it looks pretty dead both ways. I re did the assembly with a nex converter as it seems I also killed it during the process. It looks pretty good now, almost no heat buildup at the converter and the as expected 2MHz switching frequency at the catch diode. The output voltage seems to collapse though as I go up above 400mA, down to 4.38V from the nominal 5.03V. I’m thinking its partly my layout not having large enough traces for the current path? I tried to use big copper pours for all of the converter layout though.


Uploaded one of the wrong photos.

As I mentioned above, check you input filter.

Looking at the input to the switcher and prior to the filter on the scope, I don’t see any oscillations or noise getting into or through the filter at those loads. I did however lower the inductor value from the 470uH i originally designed to a 100uH. I have populated the rest of the board and have started testing the rest of the digital components as well.

I see now that converter is only rated for 500ma, but still it shouldn’t be dropping so much at 400 ma. In the photo above D3 appears to be installed backwards, could be how it got damaged and why things ran warm. Most of the parts don’t match what is on the schematic. It was recommended you use the TI WebBench for the DC-DC converter but you ended up using the minimal example from the datasheet of an Infineon part.

It would be interesting to hear how you are getting on with this project.

Sorry about not posting, I haven’t even realized its been 2 months since the last post. I have had to put the project on hold for a bit, due to work schedule and the fact that I just started moving. :tired_face:. the case and board are buried in a cardboard box right now.

Trying to do a little work on the firmware from time to time but I haven’t been able to get any assembly completed on the board or hardware. I had to opt out of using the SD card since I hadn’t read the fine print on the DAC (MCP4922) and left the enable pin unconnected so it was either the card slot or the DAC and obviously the DAC wins. I was able to talk to the DAC pretty easily and had I outputting a varied voltage on both channels through the serial port on my computer’s Arduino IDE.

The external ADC (MAX11612) is sitting on the I2C bus and seemed a little more tricky. I believe I have everything in the setup and config bytes set up and sent to the chip and I’m getting the requisite two bytes back but the output on the serial monitor doesn’t seem to match. I just spent some time earlier today digging through the Datasheet a few time again and will get the board out again this weekend and try some different things that I think I missed.

Forgot,
Here are the datasheets for the parts I’m working with if anyone was interested. I’ll post again soon on my progress with them.

MAX11612
MAX11612-MAX11617-78493.pdf (294.5 KB)

MCP4922
MCP4922.pdf (3.3 MB)

Just started using TI Webench, great stuff for learning different power designs!

1 Like

Well, it’s been a long but I’ve finally gotten time to get back to this project. I’ve actually just respun the new version of the board since I spent some time getting most of the different parts working with test code. the ADC got changed from the MAX11612 to an ADS1115, partially because there is already a working library online for it and that just makes things easier to get it all going since I’m still not great with programming. The new board is actually my first time using a 4 layer board. I just sent the gerbers off to the board house so I should have them in about a week once I finally get back home. other than the ADC part change, I also did away with the SD card. It was just too many things on the board and i wanted to simplify the design before I got to a point where I knew I would just never finish it. I changed the 5V power supply IC as well, I didn’t go with anything from the TI Webench, although I’ve used it on other projects. it is a newer chip from Monolithic Power Systems that is still fairly inexpensive but isn’t at the end of its design life like the other chip i was using was. looking forward to getting back home so can dig into this project again it’s been a while since I’ve done any hardware at all so this will be fun!

Here is the new schematic and the 3D model from Kicad for reference.

Bench Power SupplyV4.0.pdf (227.0 KB)

Nice, very well laid out schematic! Love that annotation :slight_smile:

Do you have a calculator you use for the heatsink to determine the total area you need?

Thanks!

I didn’t use a calculator for the heatsink in the picture that was just a stock sized one I made in Fusion and imported. I’ve used some online calculators before though and usually just grabbed one that I had laying around that was similarly sized. This will have 2 fans moving air directly over the heatsink as well.

There is a section in the Current Sink or Swim project that covers heat dissipation calculations and heat sink selection.