EG1162 Catium Power high-power synchronous rectifying Buck adjustable power supply reduces the voltage of 8 series of lithium iron phosphate batteries to power the Yaesu FT-891 shortwave radio station, with a shell, small size (can be even smaller), the temperature of the shell is approximately fifty to sixty degrees when outputting ≈350W. The cover photo does not show the switch. The main control is EG1162, a high-power synchronous rectifying buck control chip, with overcurrent protection, short circuit protection, and external power MOS.

Design:
Because it is designed to power the radio station, the output voltage of 13.8V is required (when the radio is transmitting at full power), the output current is 23A, so the entire circuit is designed to a limit of 30A. The input voltage is provided by 8 series of lithium iron phosphate, not exceeding 29.2V, and the power MOSFET is preferably selected with a withstand voltage of 60V or higher, with as low internal resistance and junction capacitance as possible.

The author used the HD50N06D bought, not sure if the internal resistance is a bit high or for some other reason, the main heat generation is in the half-bridge part of the tube, the lower tube doesn’t generate much heat, and neither does the soft switch. I strongly suspect that parallel tubes will increase switch losses! If possible, use tubes with lower internal resistance and better performance directly!!
Speaking of soft switches, the negative terminal of the power supply is connected in series with a group of 2 MOSFETs in parallel as a soft switch, which basically does not generate heat, solving the switch problem.

Schematic EG1162 Catium Power supply Buck adjustable

Ripple:
Because it is designed to power the radio station, the output ripple should be as small as possible, so according to the description in the EG1162 datasheet:

Among them, ⁣ I_ripple and ΔI_L Busy refers to the electrochemical current ripples. I don’t know why the two symbols are used in this manual. C_O is an output capacitor. ESR is an equivalent series resistance of the output filter capacitor (you can see the data manual of the capacitor manufacturer), and F_S is the operating frequency.

According to the above formula, select the inductance of the iron silicon aluminum magnetic ring with L = 47uH, buy in Taobao.

The operating frequency was originally scheduled to be 180kHz. The test found that the fever was severe. It was inferred that the switch loss may be too large, and the reduction frequency was improved after 90kHz.

The author choose the two 470UF solid capacitors and several MLCCs. ESR should be relatively low. Take ESR = 15mΩ, V_IN = 25.6V, V_OUT = 13.8V, FS = 90kHz f_s = 90kHz calculates almost ΔIL = 1.5aΔI_L = 1.5A, and then ΔV_out = 25mV, I don’t know if this calculation is right.

For insurance, a π filter is added after the output. The inductance is 2.2UH, which will not be considered attenuation. It is okay to simulate the effect with Multisim 14. Buy in Taobao

There is no high level equipment such as an oscilloscope, the ripples can’t be tested, and the conditional guys can help me try it.

The output draws a bunch of NC MLCC positions, you can look at the addition of welding.

The circuit and PCB here are slightly different from what I use, mainly because the problems found in the debugging process have been modified.
The supporting shell is used, the shell link in Taobao

Voltage Output:
Can be adjusted by adjusting the feedback resistor, please refer to the EG1162 manual for details.

Testing:
When the author tested it, It used a 13.8V output, and connected two cooling fans 12715 and 12708, boiling water quickly (laughs), there is a video at the back. Also,  have already connected it to the radio station and tried it out, the FT-891 can be used continuously in FM mode (100w transmission), barely generating heat, the power supply even does not heat up the radio station itself.

Manufacturing:
Note that a small piece should be cut off from the board at the position of inductor L2, and plug it into the slot reserved on the mainboard.

The power MOSFET is on the back of the board, if running at high power, proper heat dissipation should be done, I used a copper plate + silicon thermal adhesive attached to the shell, which is made of aluminum. Some parts of the circuit are quite compact, such as between inductor L2 and the shell, output terminal blocks, pay attention to insulation.

After installing it into the shell, apply glue first

remember to pile up solder or embed copper wire (also pile up solder on the small board of L2!)

After installing the outer shell, the groove of the outer shell baffle was opened with an electric drill and a hand file. The hole on the top can be used to install a cowl switch. I didn’t connect it here, so I just let the wire out.

Known issues:
However, it was found that short-circuiting the output inductor directly causes a howling sound, is it a problem with the sampling resistor? This diagram is slightly different from the part about current detection in the official data manual, not sure if that’s the reason. In other words, is the short-circuit protection not working?

BOM EG1162 Catium Power supply Buck adjustable

Download files, links, and notes

• Download PCB in Gerber + BOM + Centroid file
• Mirror
• Buy EG1162
• Datasheet EG1162

Source: https://oshwhub.com/azez/lm2596