6.3vdc form 6.3v ac PSU

[Mick Russell]

Dear all, looking for a bit of helps here.
i have tried 2 different versions of a heater supply from the 'valve amplifiers' book but cannot get the output to go above 5.4V.
The supply uses a single phase rectifier rated at 4A 50v, a 10,000uf cap into a 1085 low drop regultater through a variable resitor (4.7k) and a 270ohm resistor adjust to output then thorugh another 10,000 uf cap.
this only supplies two preamp valves which together only draw 0.6 amps, the transformer i am using has a 6.3v ac tap rated at 3 amps.
ive built two versions now, both in the book, withoiut the valves attached i can adjust the output from about 7 .5 vots down to 6.3v but when i have connected the valves i cant get it above 5.4 v. what's going wrong - i cant have built two duff PSU's or could I ?

 

[DF96]

With a load attached you will get extra voltage drop in the transformer secondary and rectifiers. As the current here will be charging pulses the voltage drop will be much greater than you may expect - a factor of 5 would not be surprising. I would ditch the regulator as you don't have enough headroom for that. Just use an 'adjust on test' power resistor between the two caps to drop the voltage to the required level on load. Alternatively, unless it is an MC input, just use correctly-wired AC heaters.

 

[flatheadmurre]

To get enough voltage you have to use Shottky diodes with low forwardvoltage drop.

There´s also 1085s with maximum output voltage limitation if you got one of those it will never regulate much over 5v with load.

I salvaged some 1084´s from a computermainboard and it took a few rewirings untill i found out about the regs 😱

 

[jackinnj]

You have to get more out of the first stage to clear the over head of the regulator -- consider using a voltage doubler circuit with about 22,000uF you should get average DC of about 7.8V which you can regulate down to 6.3V

 

[flatheadmurre]

Shottky bridge and 1084 works fine.

6.3x1.44-1.2=7.8

 

[Elvee]

It will be difficult, even with schottky's. A synchronous rectifier perhaps?
http://www.diyaudio.com/forums/power-supplies/177458-6-3-filament-5vac-winding-2.html#post2435043

 

[flatheadmurre]

Taking in account that he wants 0.6A out of a 3A heater winding things should work.

There´s some good schottkys not to pricey available.

@Elvee thats a nice solution have to look more at it.

If possible to get away with a 1085 partscount is a bit lower

Just read first post again and you don´t want a big 10000uF cap after the reg a 100Uf is fine.

 

[Elvee]

Taking in account that he wants 0.6A out of a 3A heater winding things should work.

Problem with single phase rectification, it puts a huge stress on the transformer.

 

[flatheadmurre]

Problem with single phase rectification, it puts a huge stress on the transformer.

Sorry but now i´m totally confused instead of singlephase what are you suggesting ?

 

[AndrewT]

bridge rectifier and accept the double diode Vdrop.

 

[flatheadmurre]

😱 !!!!!!

Single Phase is something different from half/full wave i thought.

I work with 3phase/1phase daily so didn´t click.

Even considering halfwave 😱

Of course bridge is the way to go.

 

[Elvee]

In this case, a diode bridge may not be an option, depending on the winding configuration/utilization.

Here is the half wave version of the synchronous rectifier: (C2 should be 12n for 60Hz)

 

[flatheadmurre]

First choice fullwave/bridge if not possible Elvee´s solution.

@Elvee have you tried the circuit IRL, looks interesting.

 

[Elvee]

First choice fullwave/bridge if not possible Elvee´s solution.

@Elvee have you tried the circuit IRL, looks interesting.

Not this exact version, but I'll give it a try tomorrow, it's something you can breadboard in 5 minutes....

 

[paulster]

I do it with Schottky rectifiers, lots of smoothing, and a very low dropout regulator. I've been able to keep in regulation with the mains 10% below expected as well.

You just need to check the specs of the diodes and the regulator carefully and find the lowest possible Vf and dropout voltage for the load current you are expecting.

 

[flatheadmurre]

Workbench is filled right now with an OTL missing its main toroid, but preamp build coming upp then it´s DC heater time 🙂

 

[Elvee]

Not this exact version, but I'll give it a try tomorrow, it's something you can breadboard in 5 minutes....

Tested!
Works exactly as expected, which comes as no surprise given the reliability of LTspice.

For practical use, I would recommend some standard protection measures: a B-E diode for the transistor, a G-S zener for the MOS, plus a gate stopper if there is a substantial length of track.

 

[flatheadmurre]

Good to hear !

Someone with more skills in routing perhaps can make a board ?

This i imagine is something many can have use for.

I´ll try the first design when i build the preamp.

Perhaps with the addition of a 8-10mh choke ?

Unfortunately i´m totally useless when it comes to routing 😱

 

[Magz]

I might find a use for that circuit in + and - versions if there were a circuit board available. Currently using half-wave rectification from a wall wart to get +/-18V in a LF equalizer unit and I can't seem to get that last little bit of ripple out, even with a CRC stage followed by a Belleson super regulator. Not a whole lot of room in the box.

 

[Elvee]

This circuit does nothing by itself to reduce the ripple, it only eliminates the diode drop.

If the ripple problem is caused by an insufficient headroom for the regulator, it could be helpful, it will improve matters by ~250mV compared to a schottky.