Voltage doubler no center tap

[Phil_S]

I want to do this. Actually, now I need to do it. The chassis is drilled for a particular PT and there is no really opportunity to revise the build. It is a scrap pile build.

Am I right in thinking I should simply reverse the diodes and the caps to get B+ where it says -12V? This PT is about 109VAC end to end on the secondary. I'm hoping for around 300VDC.

Are a pair of 22uf caps adequate?

Thanks.

[matt h]

(deleted)

[sluckey]

You don't need to change any components around. Just swap output labels.

[xtian]

I just built a similar circuit to power 12v LED string with the 5v tap on my PT, using the Greinacher circuit seen here.

[martin manning]

I want to do this. Actually, now I need to do it. The chassis is drilled for a particular PT and there is no really opportunity to revise the build. It is a scrap pile build.

Am I right in thinking I should simply reverse the diodes and the caps to get B+ where it says -12V? This PT is about 109VAC end to end on the secondary. I'm hoping for around 300VDC.

Are a pair of 22uf caps adequate?

Thanks.

Your circuit is a half-wave doubler. See the first circuit shown here for a FW doubler (less ripple and a bit more DC voltage for the same capacitance and load): http://www.electronics-tutorials.ws/blo ... rcuit.html With a 109 VAC input you would get at best 109*1.414*2=308VDC with no load. I think larger caps will be needed, at least 2x larger. You will have only 11uF with two 22uF in series.

[Phil_S]

The exact voltage isn't critical. I figure it will sink below 308 under load and that's fine. Anything north of about 250 will work. I am a bit bashful about revealing the project at this stage. It's nothing special, but inspired by a recent thread. Eventually I'll post a thread whether it succeeds or not, just for the value it may have for others.

Lot's of good input here, an much appreciated. I had some trouble wrapping my mind around how this works and am seeing it more clearly now, so thank you to the several who have pointed me in the right direction.

So, to be clear, in use, the negative side of the output gets grounded, same as the negative side of the first filter cap in an ordinary power supply?

Martin, thank you for pointing out the difference between half wave and full wave. You are right in thinking I want full wave and more capacitance.

What is the difference between the Grienacher and the FW doubler? I can see the difference in the drawing, the Grienacher splits up the caps and the FW has them in series -- that's not the question. From a functional perspective, does one have an advantage over the other or what?

Thanks!

Edit: One more image to add, this one shows a different Greinacher doubler, where the junction of the 2 caps is grounded. This is essentially the same as the one Martin pointed to. The question is where to place the ground and do i just cap the negative output and let it float? I'm feeling out of my depth here on proper usage.

[martin manning]

The first one posted above is just the FW doubler with the ground reference moved to make a bipolar supply. You want the ground at the bottom of the lower cap, like in the third one.

[Phil_S]

Martin: Thanks for confirming this. You know the old saying about the question that isn't asked ;-}

[Jack Hester]

You can look at the 60's vintage Guild Thunderstar Bass or the Thunderbass schematics for a good example of what you want.

Jack

[JMFahey]

I confirm you need the last one, with the correction that they call the output "Vac" which is wrong ... since it's DC

You also need to ground the terminal labelled "- "

On the one labelled "+ " you'll get 2 * 1.41 * Vac (RMS) so for Vac 109V RMS you'll get raw: 2 * 1.41 * 109Vac= 307Vdc

It will end up like this:

EDIT: you'll have uglier and harder to filter ripple so at least double capacitance you expected to use.

To boot, both caps are in series, which halves the effective value ... lose-lose situation (there is not such thing as a free lunch) .

2 x 47uF at least and 2 x 100uF not overkill at all .

Anyway they are smaller and cheaper than you think, since each will have to stand only half voltage.

200/250V insulation is fine.

Also remember that the top capacitor "can" is not ground but sits at over 150Vdc from ground, so mount it accordingly.

If you use a metal clamp, for safety wrap 3 or 4 layers of insulating tape around the cap body, to prevent clamp burrs or sharp edges from punching original very thin plastic sleeve and shorting the PT secondary.

[martin manning]

I confirm you need the last one, with the correction that they call the output "Vac" which is wrong ... since it's DC

...unless the 2V is the expected ripple voltage, which is AC ;^) I see what I believe is your point, though- "VAC" often used to indicate Vrms for sinusoidal voltages. The ripple in the output of such a circuit is the typical 120Hz sawtooth waveform, so I don't think it will be any harder to deal with than usual.

[Phil_S]

Funny how an unclear drawing can be interpreted. I think we all know 2VAC isn't the output. I simply took that to mean it was "doubled" which we all know is sqrt(2) * 2. Ripple won't be a problem here. This project will be getting an involuntary rest while I order caps. I don't have anything larger than 47u on hand and those are earmarked for something else.

Jack: Thanks for pointing to the Guild amps. A picture is always worth at least 1000 words. So many of the drawings floating around have no component values on them, so it is a relief to see an actual implementation.

Phil

[Phazor]

I have an old Standel with voltage doubler but center tap ground. I increased the caps from 50 uf to 100uf and the voltage went from 365 to 405 volts. I'm sure there is a limit but thats under full load. bigger caps will increase the B+. not sure why.

[Structo]

Probably because a larger load doesn't drag the voltage down when using larger caps.

[JMFahey]

Larger caps will lower ripple, so you have extra useful voltage.

Peak will always be the same, but ripple is substracted from it.

Back to the main point, here you have the ripple calculation formula:

http://www.voltagemultipliers.com/html/multripple.html