3 phase from single phase AC

[rave154]

Hi all, ( me again asking lol )

lets say i have a single AC signal source from my Function generator, i need to be able to produce 3 AC signals that are 120 degrees out of phase ...

im stuck, scanned the web as normal ( which only added to my confusion )

does anybody know how this is done? ( via the function generator if possible so that i can control the frequency easily )

Thanks,

David. D

[CosmicFarmer]

Build or buy your own 3 phase dynamo and use the function generator to spin the rotor's variable speed motor.

@ nvisser - you can try flickr - upload pics to there and then reference them from this website. You will have to use the medium sized image, because the full sized image won't reference properly.

[rave154]

Thanks,

i was actually referring to 3 AC signals produced 120 degrees out of phase.

Thanks for the circuit, i have posted you my email address.

i actually considered a simple motor with 3 magnets on the rim, 120 degrees apart triggering 3 coils so placed so as to give me the 120 degrees phase difference...and then take those 3 signals etc.

any ideas pass em along

[gyula]

Hi all, ( me again asking lol )

lets say i have a single AC signal source from my Function generator, i need to be able to produce 3 AC signals that are 120 degrees out of phase ...

im stuck, scanned the web as normal ( which only added to my confusion )

does anybody know how this is done? ( via the function generator if possible so that i can control the frequency easily )

Thanks,

David. D

Hi David,

You would like to introduce yourself to 3-stage Johnson counters to get a wide band circuit that can give out 120 degree phased AC signals.
See the Practical Applications in this link here and recall what we agreed on earlier wrt CMOS: do not use higher output voltage from you FG to drive the Clock inputs than the power supply voltage:
Johnson counter

Of course you can use a variable 555 timer instead of the FG to drive the Clock inputs, naturally the supply voltages are equal for the 555 and the CMOS ICs.

Gyula

[rave154]

Thanks gyulasun, will check it out

gyulasun to the rescue ( as normal ! )

[rave154]

Gyulasun,

Thanks a lot, that looks like JUST what i need ( again !)

about half way down the page is whats labelled "3 stage square/sine wave generator )

i understand it thanks to your recomendation of the 40106B earlier....however there are a couple of things ive never seen before.

at the bottom left of the block diagram is a component with labels Qa & Qb - disallowed state, which is connected to the middle module, what is this...and also "where" / "how" exactly is it connected to the middle module?

also at the bottom of the bock diagram, the 3 outputs go to a low pass filtar ( can you recomend one to use as ive never used one before ) and also through a component ive not seen before.....a triangle with a little circle on the tip.....what is this?

sorry if these questions seem simplistic....just seemed simpler to ask rather than spending hours trawling the web for answers.

Thanks greatly for your help

David. D

[gyula]

...
at the bottom left of the block diagram is a component with labels Qa & Qb - disallowed state, which is connected to the middle module, what is this...and also "where" / "how" exactly is it connected to the middle module?

Hi David,

I do not know why the disallowed state occurs, nevertheless Qa and Qb outputs of the first and second D flip-flops shown there are summed into a NAND gate and fed back to the Reset pin of the second D flip-flop. So this circuit you ask needs two CD4013 D flip-flops and a two input NAND gate like CD4011. See these links to see some more on these ICs: 4013 D-type flip-flops and 4011 2-input NAND gates See also this link on Counters, and in the middle of this link study Resetting the Counter: Counting Circuits You can see similar AND or NAND gates that resets the counter at a needed count. Once you identify the pinouts of the corresponding gates, flip-flops, consider their tasks, then you can surely go ahead on your own.

also at the bottom of the bock diagram, the 3 outputs go to a low pass filtar ( can you recomend one to use as ive never used one before ) and also through a component ive not seen before.....a triangle with a little circle on the tip.....what is this?

It is a drawing symbol for a linear amplifier like a wide band op amp or a wide band audio amplifier, depending on how high output frequency from the three phase generator one wishes to get. Normally a 2 or 3 stage bipolar transistor amplifier circuit serves well and there are tons of linear ICs you can choose from.
Re low pass filters, perhaps you may wish to use passive LC filters but these need coils in the some mH ranges at least. Here is a reasonable link on low pass filters, you have to choose input-output impedance, input of the filter 'sees' the CMOS IC output impedance, perhaps a single stage emitter follower transistor as a matching stage is to be used that can have a stable low impedance at its emitter output. The output of the filter 'sees' the linear amplifier's input impedance, you have a certain freedom to choose, though normally passive LC low pass filters with equal input and output impedances give 'simpler' L and C values.

This is a vast area to discuss, you could study audio cross over filters for instance, google is your friend too.

Similarly, you can use active filters instead of passives, though the logic level of 4-5V or higher that are directly available from the CMOS outputs (if you use at least 5V supply of course) may make active filters overdriven unless they are designed to work with that high inputs too.
(Nevertheless here is a link to an on-line passive LC filter designer, of course you can use any Spice based circuit analyser or other free dedicated LC filter designer, use google: http://www.wa4dsy.net/filter/filterdesign.html and some other free filter calculators: Filter - Software: Filter Design )

Notice: if I were to build a signal generator with three phase outputs, I would choose this one: http://www.rficdesign.com/img/04446.png from this link Johnson counter and outputs (Q0=G0 Q1=G1 Q2=G2) could go to the input of the low pass filters and then to linear amplifier stages: it needs only one CD4017 while the one you referred to with the questions needs 3 CMOS ICs.

Gyula

[tjnlsn255]

I am not sure but this sounds like a car alternator output before it is rectified by the diode pack.... just a thought.....

Hopes and Dreams....

Tj

[rave154]

Thanks Gyulasun,

you surely are a fountain of knowledge

making more & more sense to me

[Jbignes5]

I would not trust those sites.

Build Your Own Phase Converter Plans - U.S. Phase Converter Standards Organization

Read that and see why it is not a good idea...

[rave154]

hi,

thanks for all the help, much appreciated.

on a different note and i thought i would post it here rather than start a new thread...

looking at the attatchment, would this give me what i want, two pulsed DC signals of different frequencies down the same wire? or would it just do nothing...or blow the mosfets lol

thanks

David. D

[gyula]

Hi,

Yes it would. Basically you create a mixer and in the coil wire both the 100 and 200 Hz currents will be flowing and their combinations too, including the input harmonics too. This is how the mixer in a heterodyne receiver works: it receives the incoming antenna signal and also the local oscillator signal and normally their difference is the IF Intermediate Frequency, for which a selective amplifier is built and this is what gives sensitivity and selectivity for the heterodyne receiver.

In your schematics you show a coil as the load: it is a relatively wideband load for the mixed frequencies, if you want to enhance any one mixed frequency you may wish to use a capacitor in parallel with the coil to form a parallel tank circuit for that particular output frequency, if you need two mixed frequencies instead of one, you can insert another parallel resonant tank in series with the first tank. Of course it is the application which dictates.

see this link also: Heterodyne - Wikipedia, the free encyclopedia

Note: if a FET blows in your schematics, it would surely blow without the other FET... so no problem with connecting them like that.

Gyula

[rave154]

Thanks gyulasun

would i be able to tag a 3rd mosfet on as well..with its drain connected to the oterh two drains...to be able to send 3 pulsed DC signals at the same time?

[rave154]

hi,

i wired up a circuit using two mosfets as shown in the last attatchment,

ran two different frequencies, one twice as fast as the other....using the PIC32 to generate the pulses & timings..

but on the scope....it looked just like a train of pulses froma single frequency

[gyula]

Hi,

First give one frequency only and no second input to the second FET, note the scope waveform. Then give the second frequency input to the second FET, no first frequency input at the first FET, note the scope wave form again. The two scope shots are different in frequency as the inputs are different, right? In amplitudes, you may see also differences because in case the frequencies are far enough, the inductive reactance of the output coil in the common drains will constitute different currents hence voltage amplitudes in and across the coil.

Now give both input frequencies to the FETs and note the scope waveform. But it may now be difficult to see differences because the "distortion" in the waveforms are often hard to see in the time domain the scope works, a spectrum analyser ought to be used that measures in the frequency domain, its horizontal axis is calibrated in frequency, not in time. New frequencies are generated due to the mixing process; another measuring possibility could be a selective voltmeter, a rare and expensive instrument nowadays. A simple calibrated and tunable tank circuit with a rectifier and voltmeter indication could be also used, but in case of the lower audio band tunability is difficult to achieve, maybe active band pass filters are a possibility.
Look at this link to get some further help:
Mixer waveforms - Microwave Encyclopedia - MicroWaves101.com

and this pdf file if you are not so familiar with spectrum analysis:
http://cp.literature.agilent.com/lit.../5952-0292.pdf

If your FETs were working linearly and not in a switching on/off mode from the input pulses, then by manipulating the input amplitudes you could better see the changes of waveform at the drain. With this I mean this switching mode also makes it difficult to see the resultant waveform on a scope.

A possible idea occurs to me: connect the two FETs in series and then the mixing effect should increase because now two nonlinear switches do the chopping up of the same current via the coil. (I mean one of the FETs is a "block" to the other.) This is how a dual gate FET mixer works. In your application it may be difficult to drive the 'upper' FET because its source electrode is on top of the lower FET's drain electrode, solution is the so called high side driver IC or a wideband transformer like an audio one, though this may limit your freedom in frequencies. A high side MOSFET driver is the best solution in this case.
On the series connected FETs I mean the lower FET is like in your schematic, cut the wire at the drain, the second FET source goes to this drain and the coil is connected to the second FET drain. And the first gate is driven as usual so far, the second gate is driven by a high side driver IC.

Re on your previous third FET with a third frequency input: I can only say a yes.

Now you have another beehive...

Gyula