5V 3A PSU

Ash,
Just a question for Graham, he can say no.

Part completed equipment doesn’t need a Certificate of Conformity. As the assembler I’m the manufacturer in the eyes of the law. On my own head be it.

If Graham’s conscience won’t allow this that’s fine by me. He doesn’t know me from Adam and there are too many deluded amateurs around and people whom don’t take responsibility for their own actions so I would see his point.

As a youth (and even now) I got plenty of jolts simply by not realising that the things on the end of our arms can sometimes have a mind of their own, and even if under control, that in a tight spot (both definitions) manipulating one thing was known to articulate another hand part into touching another.

The introduction of transformer terminal shrouds (where fitted) helped. Then there are other easily forgotten items like capacitor terminals in a switched off thing which are still holding charge... ouch! - not any more.

Obviously, during troubleshooting, you're going to isolate from the mains (unplug it), but that's what you think you did...

And if you think that won't happen, how often do you get into the situation where something simply won't work, and then you realise you haven't switched it on? (the reverse situation)

Well, if I can do it, so can you. When a bit of confusion creeps in, logic creeps out.

And that's just the safety side.

I mean, anybody can build a power supply. I can say the magic word: LM350, and you can cut me out and just build one - honestly, I don't mind.

How will we deal with the heat? If the dropout voltage is 2V5 and the current is 3A the power dissipation is 7W5. The ambient temp is say 20C and we don't want burnt fingers or other things, so a safe case temperature is say 50C (hot to touch). The temp rise is 30, and 30/7.5 is 4 so it only requires a 4 degree per watt heatsink? Except we need to factor in the thermal resistance of the device, that of its thermal insulating pad; and what about the ambient temperatures in summer?

But the transformer also gets hot on-load, and so does the bridge rectifier.

Inside this "slow cooker" we have the electrolytic capacitors which are used to store charge. They can do 105 degrees, which means it can get above boiling point in there (if you want). The capacitors will then last 1000 hours (6 weeks).

The longevity of electrolytic capacitors depends on many factors, heat being a major one, and we really need them to run much cooler so we don't need to unduly worry.

So we need the case itself to give up heat, and things that get hot need to have space around them so they're not heating each other up. Space therefore needs a case large enough.

I am eyeing up an MB6 Takachi aluminium box from RS. It might have the surface area to dissipate sufficiently so I don't need a heatsink. I will have to find that out empirically, and so I will have to build one to find out.

In this post I have not touched on take off points, ground impedance, protection circuits, stability, or a whole host of other considerations.

If it were easy to make a power supply that ticks all boxes, and also allows the equipment to perform and deliver its best possible SQ, I would have done it several months ago.

And no, it can be seen from the above, it won't fit in a PSU1 case.

Should we now cover certification? If I sell it as a kit of parts or a completed unit, as a manufacturer registered as such my responsibility is your safety.

You are unlikely to find those inside hif-fi and please don't short capacitors in the way shown in the video. I am always careful how I discharge the caps when working on my guitar valve amps which don't use the type of high current caps in that video. Most of them run on hundreds of volts DC, which is dangerous and not to be played with. Based on the various PSU's I've built the mains side is the safety issue rather than tens of volts of dc in standard capacitors for a low voltage supply.

Thank you Graham. Your response about your responsibility for our safety is a very good one. Point taken.

Progress, yes. Anything to show, no. Why?

It's called R&D, and if we're going to make things in bigger boxes then it has to be done professionally.

If I put this out to a case manufacturer on the quantities I expect (we're such a well-known brand that you'd expect us to be selling in the quantities of well-known brands, and the next person to say we're such a well-known brand gets it!), then the price to all interested here would be silly - stupid - pathetic - unreal - etc.

So rather than trying to turn a profit (as we have unfortunately to do) on their profit, we're better taking case work in-house.

Which means tooling, and printing (will have to find a painter), and all involved.