soalr SMPS part 6: the prototype

 At least we got to this stage. All designing work is done, PCB is designed, parts arrived, time for assembly. As usually it wasn't smooth sailing.

Assembly

When assembling I run out of 22uF 0805 capacitors. Due to global chip (and other electronic parts too) shortage then only models available were 50 cents / pcs. With 12 caps on board that'd be 6 euro just for filtering caps. Not in hackers spirit. 10uF 0805 caps I was able to acquire at 3.5 cents / pcs figuring I can just stack two on top of each other to get better filtering. My chosen MOSFET wasn't available and I had to switch for BUK4D38 which has both higher gate charge and Rdson. Finally silicon wafers' price spiked increasing solar panel prices few times. I hope they'll drop soon but in meantime I want to make it going on panels I already have and what I have is totally incompatible with original project requirements (19V MPPT voltage) so will have to work far away from it's maximum power point.

First power up

I soldered everything, turn the board on and see LED light up. Happy to see it working I measure LED current and I see it's constantly on. How is that possible?! It turns out my synchronous rectifier blew itself up. For some reason both MOSFETs turned on at the same time. Nmos being cheap, low power part gave up first, blowing out short, Pmos permanently shorted itself. Found, Pmos replaced, Nmos removed. For testing non synchronous converter will suffice.

Testing

After this replacement output converter worked. Connecting bench PSU as battery and changing it's voltage I checked that duty cycle changes and output current is kept constant. Shorting jumpers also increased brightness as expected. I couldn't measure efficiency as LED was already in place, I'll have to do soldering next board.

Input section

Next I got to input section. At first I was worried why my LED doesn't turn on when I disconnected solar cell emulator. It turned out I set up input voltage so low that battery voltage coming back to input was high enough to keep input switched on. My obvious mistake, I even pointed out in my notes that MPPT voltage has to be >5V for it reliable switching off charging. After readjusting input trimmer LED started working again, it even kept input voltage at constant level independent of illumination level (tracked MPPT).

Efficiency

Everything seemed find until I tried measuring efficiency and realized that battery current is almost equal to solar cell current. Scoping input MOSFETs switching waveform I found sawtooth waveform with few ms period. Clearly it's not expected hundreds khz switching. Not wanting to debug it on low currents I ramped up solar illumination to full sun and ... it started working!. It turns out that at low input currents input capacitance is too big and loop has such fast response time it over/underflows duty cycle counter making converter behave like crazy. I need to pinpoint the problem and fix it in next firmware revision.

Flashing

I assembled working version and set it up next to front door to make sure there're no other bugs. And sure enough I found one. When LEDs turns off due to too low battery voltage current consumption drops, reducing voltage drop on internal resistance. In consequence battery voltage rises and LED turn on again. This flashing cycle can continue for over and hour until battery discharges fully. 

I took it home for a week long holidays and found that it discharged battery to 2.3V, clearly something is drawing too much power In everyday use case it isn't much of an issue but it means lamp can't be left alone but like most commercial counterparts requires battery removal for winter.

Final thoughts

Will there ever be next revision? Hard to say. I'm not very happy with current performance but it kind of works as a gift it was originally meant to be. Severe lack of time and problem with access to semiconductors is also demotivating. Currently I'm not planing on working any further on this project but maybe one day I'll get back to it.

Comments

Post a Comment

Popular posts from this blog

Active Eurorack Busboard Part 3

Image
THR power supply board   Time travel Let's jump into a timeline a little bit. Before I finish writing about all of my adventures while designing this system I'd like to write about its final form, the one, that is hitting the market right now. Quick start guide Many people probably landed here just to use their board or to judge how hard it is to use them, So here is a simple a guide on how to set up your new THR: Mount the board either on a test bench or in a case. If the mounting is to be permanent consider the thermal requirements first. Connect your power supply to the two-pin connector. Don't search for polarity markings, there aren't any, but the board is reverse polarity protected so if it doesn't work just try to reverse the  connection (just in case - is closer to the edge, + is closer to the capacitor). Make sure that all 3 LEDs are lit up. If neither is lit probably there's a problem with the input power. If some are working and some are not or only b
Read more

External Cavity Diode Laser - part 1: optical path and mechanical design

Image
 Introduction External cavity diode laser (ECDL) combines benefits of laser diodes and DPSS lasers. The semiconductor chip provides easy to pump, efficient, high gain medium while the external cavity allows tailoring device performance.  Wavelength selective elements, like diffraction gratings or etalons can be added to the resonator to reduce emission spectrum, saturable absorber can be added for mode-locking, partially reflective mirror in conjunction with RF modulation can be used for frequency comb generation and many more. In this article, we will try to design simplest, ECDL that'll be used in future experiments so state tuned! How it works There are many great internet sources describing how ECLD works, so let's just quickly recap. Out of the box laser diode has relative wide emission bandwidth, usually around 1-2nm. This is because the gain bandwidth of diode is relative wide and the resonator works with relatively weak coupling (output mirror has reflectivity of 32
Read more

PCB holder 2

Image
 As promised this is update to last post about our PCB holders. Technical difficulties It turns out that if I want tapped hole every 10mm that's a lot of tapping. Additionally tapping steel is quite hard, especially with such small tap. Most companies decided that they don't want to do such job, one that agreed is research institution so I guess they're used to nontrivial requirements. After some negotiations we got price of almost 5k euro for 300x300mm sheet. That's a bit too much for this project. Required changes My changing steel to aluminum and moving holes onto 15x15 grid we considerably reduced required engineering time. After initial testing it turned out that 15x15 hole spacing is more than sufficient, but swapping to aluminum means magnetic mounts are no longer an option. I guess it's low price to pay to reduce manufacturing cost to 740 euros. plates when they first arrived in our lab   Designed mounts We designed few mounts for plates to mount things on t
Read more