Idea for portable power supply for TS100 soldering iron

[ealex]

Hello guys,

After trying to figure out how can an electric pushbike charger work with just an LM358 in the secondary and finding these app. notes:
https://www.fairchildsemi.com/application-notes/AN/AN-4138.pdf page 11 and 12, https://www.onsemi.com/pub/Collateral/NCS1002-D.PDF and https://media.monolithicpower.com/document/MP26085_r1.02.pdf, I thought I can use the same thing to charge a 4-5S Li pack, power the TS100 and have a sort of UPS that will maintain the power output when there is no mains ( assuming the battery has enough charge ).
Warning - this is not my domain so i'll use the wrong terms


I've modeled the schematic from AN-4138 in ltspice (draft4.asc, attached):


A short walktrough, from left to right:
- V1 is a 2.5 reference (TL431 that's already in a SMPS)
- V2 is a power supply for the op-amps, aux winding or main winding,TBD
- B1, R11 and D3 are used to simulate main winding: feedback gain, internal resistance and D3 ensures that the main winding can only source current
- B1's variation rule ensures that it will decrease output voltage with increasing input voltage and it won't try going negative, just like a TL431 feedback would work
- R1 is the current sense resistor
- R2 and R3 form the feedback voltage divider
- U1 and it's associated components form the voltage feedback loop, it will try to satisfy V(REF) = V(MAIN)*R3/(R2+R3)
- U2 is the current limit loop: V(ISNS) is negative for the op-amp, so U2 will output a voltage as long as the voltage at it's negative input tries to go bellow 0 - this does not look right, i'll have to build it and see
- D1 and D2 are OR'in the outputs => whoever has the higher output voltage takes over the control loop
- C2 is there to simulate a battery, there is an initial condition that it starts from 12V on simulation start-up
- M1, V3 and R5 simulate the load, that needs to be attached before battery charger current sense, so it's kept out of the current regulation loop

How I've imagined it should work:
1. when there is no load attached: U1 and U2 will act as a CC/CV charger until the battery voltage reaches the set point. I need to check if the current will go to zero, so the battery is not trickle charged. I think the set point should be under 4.2V per cell.
2. when the load is attached: V(MAIN) drops because of the extra load, U1 compensates for that by lowering it's output voltage and the battery will provide some extra current via R1 until output voltage settles.
3. when the primary is not powered, the battery will supply the load via R1, while U1, U2, the voltage reference and the optocoupler will draw some parasitic power if i won't use an extra winding as shown in AN-4138.

Do you thing this will work ?
I have an old / simple SMPS ripped out of a scanner that can be used for testing: tl431 secondary feedback, no extra overvoltage protection, etc.