Voltage drop.

[davelectronic]

The output voltage isn't likely to climb above the 12.60 Volts, the no load voltage. My intended load will function from anywhere between 11.40 Volts and 14.40 Volts. The AC mains input on the primary side will fluctuate I've no doubt about that, also the transformers secondary side AC output will also fluctuate. But the margin between the rectified filtered voltage and the regulators output is reasonable, not perfect, but ok. The load current will be just shy of 8 Amps, and an approximation of a 50% duty cycle is fairly typical with HF linear amplifiers. So it should be ok i think, i have three other projects very similar, but those will run 6 x power transistors. And i would like to try and reach a target voltage of 13.20 Volts, or close to it. But i will probably use darlington transistors, i have used them before with good results. The two transformers in the psu above are right at about the limit of current i would want to draw from them. The 12 Amps might as well be a peak maximum, i'm unlikely to hit that much current with a HF 100 watt linear amplifier with a 4 watt drive into it, there are always losses in these mosfet driven amplifiers. The manufacturers always hype up the specs. In reality i expect it will deliver between 75 to 85 Watts after losses, but that's fine.     

[james_s]

Yes that's what I mean, the output of the *transformer* will fluctuate with the mains voltage and the load, and will only match the rating under the specific conditions the transformer is rated for. The output of the regulator should not sag, that is the whole point of a regulator. The transformer needs to be rated such that the voltage into the regulator stays within an acceptable range throughout the range of input and output conditions it is designed to tolerate. If it dips too low the output of the regulator will sag due to insufficient headroom and if it is too high the regulator will dissipate an excessive amount of heat and may be damaged, but as long as it stays within the safe operating area then everything will be fine.

[davelectronic]

I have tried to keep the input voltage at a reasonable margin, so not to high, but try to avoid voltage drops. Transformers can be so difficult from each other, regardless of what the ratings say on them. Decent transformers are very expensive, and the auction sites specials are a lucky dip if you get a good one. It's a shame MOT's are so wasteful of power, as repurposed and rewound they have a whopping amount of current available. But run singly they are to wasteful and to hot. Before I'm done with linear power supplys, I would like to try an LM723 themed project.

[james_s]

There is always the option of winding your own. I don't know what suppliers are in your part of the world but in the past I have bought cores from Bridgeport Magnetics. I wound a 600VA toroidal power transformer by hand and it was not all that difficult, you can even buy cores that have a primary on them already. For easier winding you can get cut-C cores that don't require passing the wire through the hole in the core with a shuttle. Surplus transformers turn up too, and the toroidal kind are particularly easy to modify to suit your needs.

[davelectronic]

I did look at winding transformers in the past, mostly vids on the tube. I'm sure we have a couple of small businesses that cater for custom transformers, but beyond glancing of the web site, didn't go any further. Sounds tedious winding by hand, I've only ever rewound a couple of MOT secondary windings, and though that was tedious. Look at a really old method of two metal rods, and two cross pieces and windings on each rod. Bit easier as the rods can be spun to get the windings on. But I don't know how efficient that type of open frame transformer would be. I'm close to finishing the PSU project with two rewound MOT transformers, both in series primary and secondary windings. Be interesting to see how the voltage secondary holds up, and with the primary windings in series there is a significant voltage drop under load. But from memory and testing with load there was enough to use a L7812, or LM317 voltage regulator. And the biggest PSU project is nearly hardware complete, that has a pair of 225 VA toroidal transformers in parallel for the primary windings and secondary windings. I will probably stop there using three terminal IC linear regulators. But would like to try the LM723 linear regulator.

[xavier60]

The unloaded output shouldn't be 12.6V. It should be much closer to 12.0V. Has it been checked for oscillations?

[james_s]

Well the advantage of winding a transformer on a cut-C or E-I core is that you can wind the bobbin(s) before assembling the transformer. I made a little jig for winding a bobbin that let me spin it with a cordless drill so it greatly reduced the time. The larger the core, the fewer turns per volt you need so winding a big transformer is a lot easier than a small one. It's really handy being able to make a custom transformer with whatever ratings I need, it's one of those things I always assumed was harder until I did it.

[davelectronic]

Yes I wouldn't mind giving it a go, winding a transformer. As you say you can Taylor the exact specifications you want. I think there are very few supplies that sell a EI or just blank toroidal frame. Making a bobin and winding via a wood frame on the a bobin would be fairly easy. The Voltage is 12.60 Volts DC, as there is a silicon rectifier diode in line cathode anode in line with the negative volts rail. The diode pushes up the voltage by approximately it's voltage drop I believe.