Power supply help

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
So this is my first "complicated" power supply I'm making and I don't know where to start. Its for a mono version of this amp.

It needs to be able to supply 410V and about 125ma (I'm simply halving what the transformer he used in his schematic, a hammond 372JX, could provide in ma). Its a bit of over kill but I want to use my 4B32's to rectify the supply (half wave each and have about a 12v voltage drop, 16v MAX. Page on them. ). I want to go with a simple LC filter as I have had problems in the past with capacitor input when using gas rectifiers.

The part I really need help with is making it a tube regulated supply. I don't have the foggiest idea of where to begin (though I know I want to somehow fit a 6C33C in the design for the irony). If it makes any difference I plan of having the transformers custom made. This amp is for a sub woofer as well.
 
Sorry for the double post but I saw no option for editing my previous post.
To get the basic layout of the power supply I decided to use a program (PSUD2) instead of doing all of the math out by hand (mainly because I lost the formulas my friend gave me and I didn't feel like deriving my own). I'm not 100% sure I'm using it right (I had never considered the resistance of the mains transformer windings and the filter capacitor). I also couldn't find a model for the 4C32 so I used one for the 3B28 (next best thing as far as I know).
As far as I can tell from the program a 245-0-245 transformer (490VCT) into a hammond choke (10H, 82ohm) with a good valued capacitor will give me ~410v depending on the load (was assuming the primary impedance of the output transformer a good value for Rload? I've never designed a power supply this way. I've always been a "trial-error-eh?-meh...good enough" kind of guy).

Is this a good starting point or did I screw something up in the program?
 
If you really want to experience new things, how about an off-line flyback? It's far more efficient in every respect (stray fields/EMI, cost, size, weight, output regulation, current limiting, losses, life expectancy...) And you only need 50W...
 
If you really want to experience new things, how about an off-line flyback? It's far more efficient in every respect (stray fields/EMI, cost, size, weight, output regulation, current limiting, losses, life expectancy...) And you only need 50W...
Hell no! I've seen fist hand what flyback transformers can do...joe still twitches every once and a while. I stay away from those things...
 
That was not worse than the harm that 490V AC or a 10H choke charged to 100mA can do ;)

I have no experience with tubes, but if this class A choke loaded output stage is properly biased, it should be conducting slightly more than 410V/4800ohm~=90mA when idle (I suppose the pots are intended for bias adjustment). Current consumption may double during negative signal peaks and may drop to 0 during positive peaks, so net current draw remains at 90mA.

BTW: A good 400V flyback SMPS would actually have four 100V outputs in series to allow to use 200V ultrafast diodes. This has nothing to do with the 25KV to 50KV secondaries found on TV flybacks...
 
This is what I've been playing with but I don't know what certain values should be. Any help would be appreciated.
 

Attachments

  • untitled.png
    untitled.png
    80.6 KB · Views: 241
OK, like I said, I don't really do tubes, I did them in school years ago, but it's been silicon ever since...

I had a play with PSUD2 and I looked at your circuit. From first principles the topology you have chosen will produce a peak output of <346V with 245v secondaries (rms volts * sqrt(2)). Therefore the value you should enter in the transformer properties is 245 not 490.

Entering the values for the choke and cap as in your circuit gives me an output voltage of ~200V with a 5k load. 5k is not an unreasonable figure, but you can backfigure the load by dividing your target voltage by the anticipated current draw 410/0.125 =3280ohms. V=IR.

Messing with the capacitor resistance value (ESR) affects mostly the ripple, so this is a parameter with a minor influence on your output voltage. I'm a bit surprised that so much voltage is lost as a result of the tx resistance and choke, but you live and learn.

To get the voltage you want with the 2-tube fullwave rectifier you will require a much higher voltage secondary, about twice what you have now.

Maybe somebody who's a bit more up to speed on this will chip in.

w
 
So 245-0-245 secondaries don't give you 490VCT?
I know I'm using choke input which tends to drop the voltage but the original schematic I'm replacing with my own used a 300-0-300 supply and SS rectification. The filtering was a bit more complex as well. I can't imagine the secondaries having to be more than 300-0-300.
 
These secondaries will give you the voltage you want if you connect them in series and use a full 4-diode bridge. This will give you peak voltages of ~690V.

In the configuration with the centre tap the voltages are halved and only one half of the winding is active at any time.

w
 
Ok so I doubled the secondary voltage like you said (it was the same that I was working with before but know I understand whats going on and such) so the transformer will have 490-0-490 for the secondaries. Now the amp should draw just a little under 125ma max and with the simulation using a 3.28k load it idles at just under 125ma it looks.
So the whole problem was me not knowing the program and not realizing just how much of a voltage drop the choke caused.

So now that I've finally figured that out I'm wondering if it is worth the effort to make it a tube regulated supply or if I should keep it simple like this.
 
lochness said:
So now that I've finally figured that out I'm wondering if it is worth the effort to make it a tube regulated supply or if I should keep it simple like this.

Unhhh...

As a general rule power stages are designed with a good PSRR and run from unregulated supplies.

A regulated supply will require higher voltage still.

The regulator will obviously affect the transfer function of the amplifier, whether the influence will be entirely benign is a moot point without the benefit of experience or a simulation. It might be inaudible.

On balance I'd go for simplicity in this case, I'm not a minimalist.

w
 
This is a rough sketch I drew up. For a few different reasons I'm just going to use the 3B28 as the rectifier. The reason the heaters are each driven by different transformers is that I had 2 center tapped 2.5v@5A transformers (i really don't know where they came from :confused: ).

I did it up like a similar power supply I saw a long time ago but don't know what the value for the 2 capacitors should be.
 

Attachments

  • scanimage001.png
    scanimage001.png
    5.3 KB · Views: 145
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.