Ctek battery charger MXTS 40

[najrao]

Got this only to see what is inside, and repair if possible as necessary. The elaborate case with buttons and display is taken off for later.

First, I don't see any hv bulk cap after the rectifier. The 'top' picture does show 4 unpopulated circles for caps, but these are decapitated by someone who was in earlier. They are bulk (all in parallel), but on the lv side. I must determine if 1000uF 35V 105C ×4 would refill the spaces correctly.

Even with a small internal fan, 40A to rectify and 660W to handle on the primaries of the two transformers together appears to be a tall order for the heat sinks. Let us see.

I shall revert to this after I get time to do some study. If you have seen this charger before, do let me know what to expect. Thank you.

Edit: there is no fan provision. Mistook the the header socket on PCB, goes to external thermistor temperature sensor.
Edit: I fitted 4×1000uF 50V in the vacant positions. Should probably be 2200uF 35V, per my analysis, but had no stock.
Edit: Fuse was blown and shorted by previous genius. Replaced.
Edit: Invisible track fault on PCB had blown a TL2575HV (intermittance causing loss of connection to ground pin). Replaced by stock 2576 not HV.

If any of you have this charger, request check rating of capacitor to determine if more than 4000uF is needed. This is the main and only energy storage and filter. Thanks.

[najrao]

Not interesting to anybody, apparently. But I shall put down my findings so far for posterity, and perhaps once again seek your indulgence to find answers.

The two phase "PFC" is actually a rather neat isolating flyback, using a UCC38060 for PWM. With two separate 'coupled' inductors, aka transformers, it puts out up to 660W of dc at 52.5V (sic). No synchronous rectification, just STTH's to do the job. Regulation is in the secondary part, using three paralleled TL431's, two of which are identical and are indeed set up for exactly 52.5V. Perhaps some funny redundancy to prevent overvoltage. The third is more complex, but does not influence the setpoint of control. All three feed into just one optocoupler(PC817).

So I changed the dc bus caps to 63V 1800uF×3. I also added a zener follower to limit the voltage to the auxiliary switcher 2576, to 37V.
 
There are no further stepdown converters after the 52.5V bus! There are two more mosfets on the heatsink, yet to understand what they do. No more inductors or rectifier diodes.

The charging control by software seems to work as should. So does the pretty display.

I just can't fathom how to charge a 12 or 24V battery from a stiff 52V bus without further conversion. I tested with a 12V 10Ah battery for a few  minutes, and it works! Can anyone?

Edit: Horrible oversight on my part - failed to notice huge inductor with twisted wire 10AWG winding. Straightforward stepdowner after all.