sudomod

Posted: **Tue Sep 05, 2017 8:52 am**

I just spent all of Sunday soldering magnet wire. A microscope would have helped.

: VMU soldering; Screen Shot 2017-09-05 at 10.50.46 AM.png (597.2 KiB) Viewed 15674 times

Posted: **Tue Sep 05, 2017 2:30 pm**

obsidianspider wrote: ↑
Tue Sep 05, 2017 8:52 am
I just spent all of Sunday soldering magnet wire. A microscope would have helped.

Beautiful job!

Posted: **Sat Sep 09, 2017 11:37 am**

satri360 wrote: ↑
Sun May 28, 2017 4:55 pm
TDK VLC5045 is installed on the original PowerBoost and the height is tall as 4.5mm. I'd recommend to replace it with a smaller 6.8uH 2020 (5050 in metric) power inductor so that we have more space in the tiny VMU. I used H=2.2mm 6.8uH Murata LQH5BPN6R8NT0 https://www.digikey.com/product-detail/ ... ND/5797750. WURTH 744042006 could also fit but I haven't tested. Performance results will be posted later.

Power inductors
I searched shorter power inductors in DigiKey library and picked 2 candidates - One is 2.0mm Murata LQH5BPN and another 2.0mm WURTH.
Here is the quick table of critical parameters for them.

: vmu_power_inductor_table1.png (65.34 KiB) Viewed 15615 times

Both Murata and WURTH has lower rated current probably due to the size restrictions. I calculated the peak current using the formula described in the TPS6109x data sheet under the similar design requirement but limited the minimum input voltage from 1.8V to 2.5V. Adafruit LiPo will be dead at 3.0V and I added 500mV margin for safety.
I_L = I_Out x V_Out / (V_Bat x 0.8) = 0.5*5.15/(2.5*0.8) ~= 1.3 [A] where I_Out = 500mA and V_Out = 5.15V.

Rated current of Murata LQH5BPN is higher than this peak current with more than 20% margin. However WURTH 744042006 will saturate at 1.2A and may not function as expected in the extreme conditions when the input voltage is very low as 2.5V. This won't be an issue if we ensure the minimum input voltage at 3.0V though.

Setup
Connected the regulated power supply to the VBAT input of the PowerBoost and wired the USB Fan as an alternative load since I could not find the Electrical load around. This USB Fan has a neat knob to control the fan speed. I have measured the I_Out = 148mA when fan speed is the lowest and I_Out = 345mA when fastest.

: vmu_power_test_setup.png (21.13 KiB) Viewed 15615 times

Results
Output ripple of the Murata inductor is 20% larger than the original TDK VLC5045 particularly at higher input current. This is probably due to the lower saturation/rated current of the Murata inductor and it loses the inductance quickly at higher current.
I also observed original DC-DC was in the discontinuous conduction mode (switch is off) in low input current while Murata is always in the continuous mode. I am not quite sure if this is caused by the higher DCR but this is a good example of the tradeoff of size and the performance. Output voltage is still stable at 5.1V and I conclude that Murata inductor works well based on the measurements.

: vmu_power_inductor_results.png (1.29 MiB) Viewed 15615 times

Posted: **Sat Sep 09, 2017 3:46 pm**

Wow, that's some in depth testing! You could switch the power boost for a generic charge/protect board. They tend to be smaller, the pi is happy with an input as low as 3v, and you gain a little run time too. The pi only cuts the 5v down for the CPU anyway. Apparently the 5v is only really needed for proper USB standard compliance

Posted: **Mon Nov 13, 2017 8:10 am**

Hey how is this project going? This was another project that inspired me to do my own

let me know what you think viewtopic.php?f=13&t=4613

Posted: **Sun Nov 19, 2017 9:16 pm**

kite wrote: ↑
Mon Nov 13, 2017 8:10 am
Hey how is this project going? This was another project that inspired me to do my own let me know what you think viewtopic.php?f=13&t=4613

Kudos to you!

Posted: **Tue Nov 21, 2017 12:48 pm**

satri360 wrote: ↑
Sun May 28, 2017 4:55 pm
Rework
I've realized that ground path is too poor to give stable 13.5V supply for the OLED driver. Add short wire from pin#2 of U3 to the closest ground such as GND pad of R7.
vmu_oled_pcb_rework.png

Added rework instruction for extra ground connections for the DCDC.