# Cheap homebrew lab PS; need verification of magnetics for toroid

A sixteen state state machine made with an sram and max7004 cpld is a 200khz osc that drives the FETs for the chopper in four switchers. A low end PIC can modify the contents of the SRAM to change output voltage or foldback limit.

**Need to verify T2. Making this cheap revolves around winding T2.

Using Pri multi wound L=0.8(rN)(rN)/6r+9l+10b Sec single wound L=(rN)(rN)/9r+10l r radius l length b depth Pri 4 layers 800 turns/layer 26 gauge magnet wire Sec 1 layer 16 turns 14 gauge magnet wire Fair Rite 5943000911 toroid form**

Since pic chips and cplds are free at school (within limits), A lab grade bench PS can be had for cheap if I wind a transformer. Plan is to initially make only 16 states in state machine to drive four separate switchers. Once basic PS functional, add extra inputs to state machine to get feedback inputs for adjustable voltage and short circuit folback limit. Since state machine is implemented with SRAM instead of eprom, pic chip can change state machine values to tweak performance. 4Mhz clock from micro divided in cpld to make 200khz. A-D outputs drive the chopper FETs to energize PRI on T2 Ultimate goal is 4 variable 0-30V 0-3 A with adj foldback current limit. A lot of trouble but should be comparable to PS costing hundreds.....

Can someone verify that the description of T2 will be adequate for a 200 khz switcher for outputs of 0-30v at up to 3A? Thanks

• Max7004? CPLD?? – Andy aka Nov 21 '13 at 8:30
• Let's see your flux density calculations for the transformer. – Brian Drummond Nov 21 '13 at 10:23

Your secondary is 14 ga, which has a cross-sectional area of $2\ \mathrm{mm}^2$. You spec 16 turns, for a total secondary area of $32\ \mathrm{mm^2}$. The area in the center of the toroid is $\pi({2.91\over{2}})^2\ \mathrm{mm^2}$, or $6.7\ \mathrm{mm^2}$.