Q: Active full-wave rectifiers

[Paul Andrews]

I want to create a 5V/1A regulated power supply from a 6.3V RMS input. I can get a regular diode bridge in a 4 pin DIP package, but it is marginal and my simulations and actual tests show that the forward voltage drop is too much under even a light load. My simulations show that a Schottky diode bridge works OK, and I can get an SMD 4 pin package with a total forward voltage drop of around 0.8V. However, I have seen designs for active MOSFET full-wave rectifiers, so I figured I might as well try to go that route.

I came across two potential ICs to help. One is the LT4320 and the other is the FDMQ8205. The LT4320 uses external FETs, but the killer seems to be that the minimum output voltage is 9V. The FDMQ8205 has internal FETs, but I can't figure out how I would use it in a regular full-wave bridge rectifier. It has AC inputs and gate drive inputs and I'm not sure what the gate drive inputs should be. In note 4 in the datasheet they show a wiring diagram that connects the AC input to the gate inputs, but I can't tell if that is a topology that would provide full wave rectification, or if it is just something used for testing.

Does anyone have any experience with active bridges at these voltages, or with these ICs? Is there an alternative I should be looking at? Should I just stick with the Schottky diode version?

[jf...@my-deja.com]

If I had to use the 6V transformer, I would make a voltage doubler. 

Then I would use the switching mode regulator in a USB car charger adapter to produce the regulated 5V.   My Dollar Tree carries them.  Sometimes, Fry's has these for under US$1.  Or you can probably find one at your favorite truck stop for US$9.95.  I would physically hack iawy the case to get at the electronic innards.

[chuckrr]

A typical 7805 regulator drops out at around 7.5 volts at the input.
You'll need a little more overhead than 6.3 volts.
That, in my humble observation, is trying to cut it too close.

--
You received this message because you are subscribed to the Google Groups "neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email to neonixie-l+...@googlegroups.com.
To post to this group, send email to neoni...@googlegroups.com.
To view this discussion on the web, visit https://groups.google.com/d/msgid/neonixie-l/09f8e9f7-3f34-4f89-811a-e12901b4d631%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Dekatron42]

In the circuit with the FDMQ8205 that you refer to they are connected in a POE circuit (Power Over Ethernet) where they make use of DC-voltages for the power and AC-voltages via the transformers for the data-transmissions - so the FDMQ8205 actually only work as polarity protection in that circuit and not as full-wave rectifiers.

There is an old Elektor circuit here: http://projectcircuit4u.blogspot.com/2010/04/power-mosfet-active-bridge-rectifier.html that you might be able to use in some combination with the FDMQ8205.

/Martin

[David Forbes]

I'm surprised that you can't just use the Schottky diode bridge, a rather large filter capacitor and an LDO regulator chip. What components are you using, and what's the lowest voltage you see at the input to the regulator?

--

You received this message because you are subscribed to the Google Groups "neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email to neonixie-l+...@googlegroups.com.
To post to this group, send email to neoni...@googlegroups.com.

To view this discussion on the web, visit https://groups.google.com/d/msgid/neonixie-l/1d480f1e-ef4a-4343-8bcd-af999c8d5236%40googlegroups.com.

[Paul Andrews]

@Chuck: It is 6.3VRMS, so more like 8V after rectification, but yes I think in reality it is a bit fine with load.

@nixiebunny: I haven't tried a Schottky bridge, just simulated one.

@jfeng: I hadn't considered a voltage doubler. I tried this one and with a 6800uF cap on the output it produces a steady output voltage.

This leads me to my next observation. I was feeding the output of the 7805 into ESP32 pico, wired up with an SD card reader and MAX98357A class D amplifier driving a small speaker. Even with the voltage doubler solution (and with a nice steady output voltage from that of around 13V), the pico pulled the 5V line down to around 4.7V whenever it tried to read the SD card (or play sound?), which seemed to be enough to cause the PICO to reset. I suspect it is trying to draw more current than the 7805 can supply, but I don't have a way to measure that ATM. A test with a Wemos driving 8 neopixels works fine.

Complicating things slightly is the fact that I have this all wired up on breadboards, and there are definitely losses involved with the wiring there.

The 6.3V is the output of a heater stage of a transformer I am using. I would like to try to use this if possible, but if not then I could use a 5V SMPS module such as this one, but that seems a little defeatist given that I have 6.3V RMS just lying around anyway, and who knows how stable the output of that module is if I get up to around 1A?

[petehand]

I would use four discrete schottky diodes in a bridge, and a low dropout regulator like an LM1117. The 5V version of the LM1117 regulates down to 6.2V input and delivers 800mA.

[johnk]

And I’d ask about mains stability.

Was half-wave ruled out?

 

John K

--

You received this message because you are subscribed to the Google Groups "neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email to neonixie-l+...@googlegroups.com.
To post to this group, send email to neoni...@googlegroups.com.

To view this discussion on the web, visit https://groups.google.com/d/msgid/neonixie-l/1fd82233-c90d-45da-89a0-1d35d17df93c%40googlegroups.com.

[gregebert]

I tend to use higher voltage transformers, in the 8-12VAC range, full-wave rectifiers + filter-cap, then use switching-regulators that are pin-compatible replacements for the popular LDO regulators. They are 90%+ efficient, and therefore generate almost no measurable heat. 

If the current is really low, say 50mA or less, I'll use an LDO.

[jf...@my-deja.com]

If all you want is to get 5Vdc rom the mains, a USB charger seems like a very attractive option.  They are usually less than US$5 and occupy about 1cu in (plus the power plug).  My initial concern would be safety isolation, but some of these will surely be UL listed or CE qualified.

However, if you are doing an engineering laboratory exercise requiring using a 6.3vac source then this is almost interesting.  It should be possible to build a buck-boost regulator that will tolerate a very wide range of unregulated inputs.

[Paul Andrews]

@gregebert: Yes, I ordered a switching regulator. The 6.3V AC just happens to be available on the transformer I am using for other purposes.

@jfeng: I am tending to agree that an isolated 5V power adapter is the way to go, particularly as I will want the logic circuits to stay functioning even if I disconnect the mains transformer. However, as an exercise, this is turning out to be more interesting than I thought, so I will pursue it for a while longer. Not sure if I'm up for building my own buck/boost converter. Yet.

On Sunday, March 24, 2019 at 10:46:22 AM UTC-4, Paul Andrews wrote:

[martin martin]

Sir Nixie Bunny

I have one of your 10 year old Burroughs Nixie clocks.  Is there any chance you would upgrade the FW to include the "slot machine" affect?  I only replaced one tube in about 9 years

I'll pay of course!

martin

[Paul Andrews]

Just to close this up:

The problem with the ESP32 resetting was caused by the quite ridiculous voltage drop across several jumpers.

In the end I am using a small 5V module I can wire into the mains side of the circuit. My design has changed and the 6.3V secondary I was going to tap for the 5V supply will now be biased high.