Voltage doubler circuit question.

[Ed/KB1HYS]

I have found various references about building Voltage doubler circuits. One shows the center tap & between diodes grounded, others show it not (No Connection).

Which is right?  Safest?  or best general practice?

[kb4qaa]

I can't find the exact circuit quickly, but from comparing them to Orr's west coast handbook I think the first circuit is a full wave rectifier and will give you lower ripple because the output frequency is twice the line voltage.  The second will have greater ripple since the it is a half wave and the output frequency is the same as the input.

At least that is my interpretation.

bill

[wb4bfs]

these are unusual multiplier topologies .... the first one is actually a bipolar supply, once you reference the outputs back to ground .... it would be easy to make a mistake with it  ...the second might (?) work or might not ....both use extra capacitors over the recommended half and full wave doubler standard circuits .... like Bill sez, lookat the handbooks

[Ed/KB1HYS]

They are a bit bizarre. One web page claimed that you would get more than 2x out of the first example.  They didn't seem to follow any normal multiplier circuit so I was wondering...

[The Slab Bacon]

I kinda thought the same thing, those circuits just look wrong. Typically the way a doubler works is to put the rectified outpoot in series with  charged capacitor.
and some pretty serious ripple.

that is why they need lots o capacitance, and are usually somewhat hard on the caps. Doubler supplies also have a big tendancy to sag under heavy loads with long duty cycles. (like AM carrier)

[KM1H]

No matter how I redraw they come up as NFW

[Rob K2CU]

neither circuit is a DC voltage doubler. The First circuit will produce two AC waveforms of opposite polarity and each with a DC offset on the load. The comment that it produces a split supply as in  positive and negative outputs is closest to being right. It is just not filtered DC. You need a second diode in each side between the diode and the filter cap/load.

[Opcom]

The 4x8 power supply - It's not my idea, I learned about it here. It makes sense and diodes are cheap. The secondary must be designed to operate with a bridge. (the bridge from which it evolved is grayed out)

[W2PFY]

The 4x8 power supply

OK, I'M STUPID, WHY IS IT CALLED A 4X8?

[KB2WIG]

" That, Miss Lincoln, is simply my cat. "

Why can't I forget this stuff???


klc

[Tom WA3KLR]

The "4X8" voltage doubler circuit appearing 3 posts above has merit.  I've never seen this configuration before, but it does keep the peak voltage on the secondary end (with respect to ground) down to the full secondary peak voltage.  

The conventional full-wave voltage doubler circuit (see schematic pdf below) has one end of the secondary winding hitting twice the peak voltage of the secondary with respect to ground.

For low voltage supplies, the 4X8 circuit is not necessary.   But if you have a high-voltage secondary around 750 volts RMS and higher slated for voltage doubler service and you don't know the insulation rating or it is old, this slightly more complex circuit looks like a great way to go to prevent secondary insulation breakdown failure.

[Tom WA3KLR]

Ed,

Referring back to your first post, the "voltage doubler" circuits as drawn are not correct. They are missing 2 diodes.  

As drawn, the circuits function as diode clamps (guys, remember the dc restorer?) of the ac signal coming off of the ends of the transformer windings.  So the signal is not really rectified, not filtered, and always has one end of the voltage waveform clamped to ground.  This gives an unfiltered dc with super high ripple.

With the addition of the 2 missing diodes (as shown in the schematic pdf below), the voltage doublers perform as intended.  They are each essentially an independent half-wave doubler with both in series.  They operate on the same half-wave of the ac cycle, so the overall operation of the transformer is half-wave, not a good thing, unless the transformer core has enough flux capacity for the flux-biased situation.

(Later - Rob, sorry, after reviewing the posts I see that I did not absorb your similar posting.)

[Opcom]

As to why it's called that I don't remember. I found the old web page where I explain how it works after studying it, but I didn't come up with it or the name, so sorry I can't answer definitively. I think it came from here but a search didn't find it. Could it be from a book or magazine? A web page that is now defunct?

I think it is called that because if one wished to have a FW bridge and a FW doubler as well so as to have B+ and B++, it required 8 diodes, 4 for the bridge and 4 for the doubler. (If one wants only the doubler, then only 6 are required). The circuit is explained well on the web page:

http://rawfire.torche.com/~opcom/psu/4x8_power_supply.html
http://208.190.133.201/lit/4x8/index.html