Rectifier using a pic and fet

[Peterae]

OK i have moced this to it's own thread

WW
Yes i also forgot the fet has a body diode DOH
Looks like it has already been done but without using the pic to control the on cycles offcourse his suffers the problems associated with not switching the fets by monitoring cap voltage levels and the incoming sine amplitude.

http://www.thetaeng.com/FETBridge.htm

I am still interested in trying this out of curiosity.

Bearing in mind i am using digital fets that switch on using a logic 5V and also bearing in mind i would be monitoring the sine AC using an A/D to work out when to switch each fet pair on to charge the cap.

If i was to run the drive to each fet from the pic as a pulse stream say 4Khz then will the caps give me the isolation required not to blow the pic but still provide the drive to the fets, i can place 4v7 zeners across the pic for protection if needed.


The whole point of this is to improve the efficiency of the rectifier.

Or should i just give up on the idea  

[BEP]

Optocouplers?

I dunno. The thought of having a FWBR with programmable operating curves is very intriguing.

[Peterae]

OK i don't need a fet bridge, 1 fet will see the whole AC amplitude if the pic gnd is connected to one side of the AC input.

Here's my proposed circuit, the pic is powered from the AC as well so it's a self contained unit.

The good thing also is that it can be setup to regulate to a preset voltage.

So the way it works

Sample VAC
Sample Cap Volts

When VAC is greater than Cap Volts switch cap in circuit with AC
If cap volts is greater than a pre defined voltage don't switch on of if VAC is less than cap voltage don't switch on

Go back round and keep doing this

The only problem i can see so far is that when the cap is connected to the sine i need to make sure i don't pull the sine amplitude below the cap voltage

EDIT OK i have realized this wont deal with the negative cycle as i first thought it would, i need to carry on thinking.
OR will it work, i am having trouble getting my head round the AC input, surely it will only see a varying voltage with no reversal in voltage for the negative cycle because of the pic being tied to one side of the AC?

[BEP]

I just realized something about your intelligent FWBR.

You have almost total control over the forward diode voltage drop. This would be a godsend to free energy researchers(the more serious ones) because it would allow them to rectify even the smallest signals without fighting the weaknesses of zero-crossing diodes.

I'm also wondering about any advantages in radio signal detection stages. Radio receiver circuits with fewer noise generating front-end amplifiers and the ability to pull in signals below the noise floor of other receivers?

[Peterae]

yes exactly my whole point of this, although i need a fair AC voltage for a self contained voltage, unless the pic power rail is powered using a Joule thief LOL

[Peterae]

A while ago i spent some time trying to get a magnetic bridge working without luck TESLA had this working , i was having trouble saturating the core in one direction, i have long appreciated the design of a bridge without forward voltage drops  

EDIT offcourse the only down side of using the pic is the slow sample rate for the A/D which would be about 2.2uS each sample.
A much fast version could possibly be made by using comparators and compare the voltage of the cap to the incoming sine voltage at least then this would also react fast to loading volts drops of the incoming sine.

Infact maybe this is the way to go.

[Peterae]

Here's the latest plan
The 12F629 has an internal comparator and so i will use this for proof of concept, so the plan is for the comparator to monitor the AC and compare this to the DC and pulse the fet line when i need to switch the fet on to chop the power into the dc cap, i should be able to achieve almost the full peak-peak voltage of the sine across the output cap hopefully THATS if my circuit works.

[BEP]

A while ago i spent some time trying to get a magnetic bridge working without luck TESLA had this working , i was having trouble saturating the core in one direction

I'm not surprised. The usual cause is due to the needed reluctance switching. In a fast transition you need a reluctance difference of about 10X to make the field switch directions completely. Saturation on one side would not happen easily after the magnetic connection is made through the other side.

I can't imagine how Tesla did it unless he was using powdered iron cores.

[Peterae]

OOPS did i say bridge i meant rectifier
http://www.linux-host.org/energy/arect.html

[Peterae]

OK built the rectifier as in the 12F629 diagram, instead of using a zener to stabilize my power rail to the pic i used a tiny 5 volt reg 78L05

It all works well but because i opted for a capacitor to block my DC to drive the fet, i forgot one major problem my fet drive phase is inverted  

As a result it is working but regulating the output as a negative voltage and this is screwing with the feedback system.

I guess i need to use a 1:1 isolation pulse transformer, or would putting 2 caps in series negate the phase shift by shifting it again?

EDIT just tried it and it does   so no need for a transformer  

[BEP]

I guess i need to use a 1:1 isolation pulse transformer, or would putting 2 caps in series negate the phase shift by shifting it again?

EDIT just tried it and it does  so no need for a transformer 

It worked?

You must have some sizable inductance in that circuit.

[Peterae]

No inductor at all?

When i scope using 2 probes one on one side of the cap and then a probe on the other i get inversion, but with 2 caps in series i don't, i get in phase.

[BEP]

Peterae,

Having 2 capacitors in series only reduces the total capacitance. It wont invert the signal with just capacitance.

[Peterae]

Cheers WW much appreciated for the help so far.


EDIT dam i keep forgetting about the body diode, that explains my problem and need for a half bridge config.

So in that case now my main problem is driving the gates, i must next do some tests to show if the cap couple drive does or doesnt work, if it does i can use 1 pin to drive the nfet and another pin to drive a pfet.