Full bridge synchonous recified power supply up to 3kW.

[Spoerle]

I need a source for galvanizing
Input voltage 230V 50Hz 16A max
output 2V - 60V , current 1A-80A max 3kW
Mode CC, CV CP.
Ability of pulse work , for example 10ms pulse on 200Hz

After a series of vicissitudes, we decided to try our own design .
My current idea to which I welcome every knowledge and idea.
ZVS LLC full bridge with synchnous recifier

Input stage:

In the first phase, a simple mains voltage rectifier. Output 300-325V DC
Later you may finish the PFC, which will increase the output voltage to 350-425V.

Primary side: Full Bridge LLC resonant DC-DC Converter
Full H bridge with STW70N60DM2 N-channel 600 V, 37 mΩ.
MOSFET is driver 2x UCC21530 4-A, 6-A, 5.7-kVRMS Isolated Dual-Channel Gate Driver.
T1 is current transformet 1:200 for measure primary current
Primary DC voltge measure isolater 12 bit ADC over current divider , secondary ADC input measure temperature of heatsing.
Resonant frequency about 150kHz.

Simplified signals primari side
4x Full isolated PWM channel for driving H bridge
1x Primary current analog 0-3,3V ,2,15V for 10A
1x DC primary voltge 12 bit digital 0-450V
1x Tempertuire primary digital 12bit 125C max


Secondary side: Full Bridge Synchronous Rectifier

full briudge Synchnous Rectifier with 8x STP310N10F7 N-channel 100 V, 2.3 mΩ typ., 180 A
MOSFET is drivered 2xL6491 4 A gate driver
Secondary LC filter L1 about 12uH and 16x 10uF ceramic C
Output voltge measure ADC over resistor divider
Output current measure ADC HAL 0-100 40mV/A ACS758LCB-100U-PFF-T
Tempertuire sec. side ADC

Simplified signals primari side
4x PWM channel for driving Full Bridge Synchronous Rectifier
1x Secondary DC current ADC analog 0-3,3V ,3,2V for 80A
1x Secondary DC voltage ADC analog 0-3,3V ,3,09V for 60V
1x Secondary temperture ADC analog
1x PWM for riving Fan
2x analog SR1 SR2 analog for comparator for Driving Synchronous Rectifier


Digital control
ARM M4 with HRTIM STM32F334 otr STM32G4 wit HRTIM


What do you think?

 

[Easy peasy]

do you know ( or calculated ) the desired range of leakage L of your main transformer ? As some one who designs similar all the time - I can say - respectfully - that it seems most likely it will not work as desired

A true LLC needs to run into some very low ESR / ESL caps to function properly - as drawn the sec side mosfets will experience voltage overshoots.

Splitting the series res L and placing on the mid point of each of the driving side totem poles saves a lot of losses in the main Tx and also gives some reduction in RFI.

What is your design max freq ?

 

[BradtheRad]

Your schematic appears to use the advantage of switching mains AC voltage, while Amperes are less. Less stressful on components.

It's also possible to use an autoformer (or auto-transformer) to step down mains AC, and step up current. While having two windings they have a common join at one end. Less bulky, less cost.

With some design effort, its output voltage doesn't vary too much despite varying load.

 

[cupoftea]

Also, i think your vout range of 2 to 60v, is too much for an LLC converter....you could try the phase shift full bridge instead.

 

[Spoerle]

Thank you all for the "knowledge"
I do not design pulse sources practically at all, I have never made a pulse source above 150W.
My thoughts on this project started with a simple robust construction, halfbridge with MOSFET and DIODE in the case ISOTOP, drived one PWM ,
up to megahert construction with SiC transistors that would fit in a larger matchbox.
I rejected SiC, due to problems with G excitation and the complexity of PCB design.
I liked the presented solution because it has less losses than using large MOSFETs and DIODs in ISOTOP.




Paradoxically, I can calculate what parameters a transformer should have, but I don't know how to make it. I've just never seen a transformer design guide for an LLC resonant tank.
I do not rule out the use of external Lr such as here, GS-EVB-LLC-3KW-GS-Technical-Manual-202105
For basic considerations, I considered Lr about 20uH Lm to about 100uH.
As for the frequencies I was thinking about
Start-up switching frequency 350 kHz
Closed loop switching frequency 120-220 kHz
Resonance frequency somewhere in between 150-170 kHz


I will clarify the real requirements a bit
I only need a full 80A in the range of 20-30V, maybe more realistically (18-28V)
If the upper voltage range was only 50V, it probably wouldn't matter
Lower voltage and current limit, what will be real realized .

In short, advise a robust solution that can be managed regularly digitally, will not require long debugging, even HW and software, and let yourself enjoy a lot of output. Is the full bridge LLC ZVC SRC a suitable topology or is there something more suitable?
If so what?

 

[Easy peasy]

Start with a simple H bridge - hard switched - 37.5kHz, lots of snubbing on the output diodes and pri side mosfets - peak current mode controlled with slope compensation - this will save you a lot of grief ....