A typical buck converter with current mode control has a PWM controller referenced to output ground, a current sense transformer to bring down the current sense signal and a pulse transformer to level shift the gate signal.
Sometimes I have seen the approach to use current sensing in the return path as shown here:
formatting link
The resistor RSENSE seperates the input ground and the output ground, which means that the circuitry connected to the output will bounce up and down with the switching frequency I assume. Ofcourse the output circuit could be reference to earth/chassis and there will be no coupling there, and then the only ground bounce is on the entire switch mode converter input side.
Do any of you have experiences with this? What are the pitfalls?
Generally the current-sense resistor sits where you drew it, in series with the catch diode (or mosfet) but its sense point, node 2, isn't grounded. Yes the node 2 voltage pulses at the switching frequency, but the return for output capacitor C7 is tied to node 0, and does not. There's no "ground bounce" on either side (in theory, at least, ha ha, since typical smps circuits put out plenty of switching noise).
But your circuit has bigger troubles. For a buck converter, the freewheeling or catch diode's sense voltage will be negative, as the current flows from ground through D10 to the inductor. You can't fix the problem by swapping CS+ and CS- pins on the uc1843, because despite your spice model, it doesn't have such pins. You'll have to add an inverting opamp stage, etc.
Hmm, you might check your value for R29.
Also, in real life, a part like IR's IR2110, etc., is much better than their IR2101, whose miniDIP pinout has inadequate creepage.
The IR2113-1, a favorite of mine in the 2110 series, is rated to 600V, has 2A rather than 0.2A gate drive, and it allows the COM mosfet-source pin to have serious ground bounce with respect to the input logic Vss, see IR's warnings in app note DT-97-3 about what can and will happen in the real world (as opposed to the spice world).
I'd throw out that old uc1843 controller, don't you have the feeling with it that you're trying to fit a square peg into a round hole?
Ironically, IR used to make another part that they also called an IR2101, which was a complete high-voltage buck converter with high-voltage power mosfet, just add an inductor and catch diode. But somehow that part disappeared; they forgot about it and reused the part number for a simple HV driver IC, no power mosfet, no buck-controller circuit!
Thanks for the comments and the very good tip of the IR2110
Actually the drawing was wrong, sorry, its like this:
formatting link
As far as the components values, they are no valid. Its just for displaying the general principle and the use of the 1843 is just because I had a schematic symbol for it in PSpice. (should have mentioned that in the original post)
Well, I don't want a specific part as such, allthough something like the UC2845 is available from both TI, ST and OnSemi (third source part).
As I wrote, I am primarily interested in if there is any disadvantages in the use of the current sense in the return path as I have not seen it in any text books, only in a couple of schematics.
One disadvantage could be that not only the current during the On- cycle runs in the sense resistor, both also the off-cycle. The other is that the current sense voltage ripple is present as a common mode voltage between ground 0 and ground 2.
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.