| Electronics > Projects, Designs, and Technical Stuff |
| Single phase to three phase sinewave converter |
| << < (12/16) > >> |
| Yansi:
So in fact you will not feed any coils, just the measurement circuit, then small current will suffice. I have just wanted to be sure I understand your intention correctly. Do you know how clean (THD, upper harmonics) the signals should be? What is the limit, what is still usable for the test? I think unfortunately, that the digital solution may still be the easiest one. There is not much code that is needed to make a threephase inverter bridge to produce the voltage and to sync it to any other signal. However, I think you could probably still make it happen in good old analog fashion. I have remembered, that one of my friends have built some fully analog controlled VFDs, so I have asked him if he could share the schematic of the 3ph VCO, and sure enough, here it is! There are four OTAs used (2x LM13700, still should be available) and the frequency is voltage controlled. Use the VCO input together with a PLL, for example a 4046 to lock the frequency of this to the mains input. The output amplitude at the RST terminals should be stable, due to the AGC (on the left). The phase stability and accuracy is given by the timing capacitors - all must be the same in capacity and stable in temperature. (Use good quality foil types) Then use three analog amplifiers, something like TDA2030 to drive three small step-up transformers, to up the voltage from those few volts up to the desired size. I think this analog amplifier + step-up transformer solution will have the advantage of producing very clean outputs, and is probably good enough for some tens of watts of output. I can't think of anything simpler than this. It is just a couple of OTAs, the PLL and three power amplifiers, plus the three output transformers. (Just use backwards connected small mains transformers) |
| Yansi:
In principle, the circuit above is an oscillator with a three phasing RC filters, except that the R part is created by the OTA, which serves as voltage controlled resistor. Each RC circuit provides 60 degree of shift, 180 degree the whole loop (oscillates). That's why there is the inverting amplifier to obtain the third phase, which is 60 + 180 degree away, to make the third phase of 240 degree shift. |
| sanwal209:
Indeed its a good solution. I was thinking about another solution Mains AC 220 -> 5V -> Buffer -> 120 Phase sifter -> 120 Phase shiftier -> TDA2030 -> Step Up Transformer | | | |__ TDA2030-> Step Up Transformer |__TDA2030-> Step Up Transformer It will be more simple and we dont need PLL right? |
| Yansi:
I think crating 3x 120 degree shift is not that easy, as 3x 60. You will in the end need probably more active parts in the circuit, than in the diagram above. Also your solution will provide incorrect phase outputs, if the mains frequency gets offset away from the nominal, where the phasing networks have been calculated. I think the solution with 3x 60 degree loop will be more robust and not frequency dependent (i.e. less phase error). :-// If you change your mind you want to test 50 or 60 Hz system, you won't need to change all phasing networks. The PLL will just lock at different frequency and the output phase relations will be still correct. :-+ I am not sure, why are you afraid of a PLL. I think PLL is easier to implement, than to wrestle three 120 degree phasing networks. My solution IC count: 2x LM13700, 1x opamp, 1x 4046 (or other similar device), plus 3x some monolithic power amp. I don't think it is that bad. |
| sanwal209:
I simulated in Tina, Have a look at the attached schematics and Waveforms. |
| Navigation |
| Message Index |
| Next page |
| Previous page |