PDA

View Full Version : 60 UK 5Volt power supply



mudlark
2007-02-08, 07:52
I've just built one. It sounds OK, I think.

I'll post up some testing results if anyone is interested.

The supply uses a module supply in an ally box with IEC socket. The module is easily put in the box as it has it's own chassis. The output is 5.02 volts rather than 5.2 of the wall wart.

The unit is a Calex 32005A psu from Farnell and the aluminium box is from Maplins. I have had positive support from the manufacturer. It looks better in the pictures than in real life, but it is, I believe, safe. The psu chassis is bolted to the aluminium box. The mains lead is earthed to the transformer core tag. The outlet is a standard two core feed from another power supply. If you are happy to use the old cable from the wall wart this comes free!

The stunning results can also be seen at http://flycatcher.mysite.orange.co.uk/

Mike.

mudlark
2007-02-14, 10:24
Bump?

Is no-one interested in the details?

Jitterbug
2007-02-14, 10:46
In fairness, you didn't sound totally convinced about the sound quality, safety, or even its looks ;->

Skunk
2007-02-14, 10:47
I'll post up some testing results if anyone is interested.


I'd be very interested in some testing results. This looks like an interesting test device http://tangentsoft.net/elec/lnmp/ I've thought about getting, along with a cheap scope to experiment with.

FWIW using good wire with remote sense to the plug, a good plug, and a short DC cable, give further improvement compared to the stock connector I had originally hooked to a similar open frame supply.

mick_w
2007-02-14, 11:46
I'm interested to see if you hear any improvement, I put together a Power One 5v-3A Linear PSU last year and I couldn't hear any obvious improvement in my system.

It looks like I might be local to you, so if your interested and fancy conducting a blind test, I've got a couple of SB3's and a spare linear PSU that would make switching between the various set-ups very easy.

If your interested, PM me (I'm in the south lakes)...

Mick

mudlark
2007-02-14, 13:21
I didn't mean to put people off, I was trying to be realistic in that the cosmetics of this unit are nowhere near "commercial" standards due to a cheap box. I am sure the thing works well, but I am not a expert in determining whether the sound is better or not. I was hoping for comments or to inspire others to make similar supplies and for us to share notes. There are lots of people around with the correct test gear. I am not so lucky as I have no scope or suitable software for my PC.

The unit I made is easy for those like me who are a bit nervous of mains. The module has been TUV approved so is safe of itself. The only dangers I can see are from bad connections to the tranny from the mains cables. If then mains cables come loose then they could make the case live. The design I used has the module on the base so that the screws holding the module chassis are on the base of the box. The box I used requires the mains IEC socket to be mounted on the top section requiring flexible wire between the socket and tranny and good quality soldering, insulation and spacing to ensure safety.

I am not a qualified electricial engineer or technician. I have done some messing with Quad 405 and 303 amplfiers curing hums and fitting upgrades from Net audio without too many problems.

I have a LesW 260a power amp which is beautifully built compared to my efforts and that of Quad. (In my opinion).

mudlark
2007-02-14, 13:32
I'd be very interested in some testing results. This looks like an interesting test device http://tangentsoft.net/elec/lnmp/ I've thought about getting, along with a cheap scope to experiment with.

FWIW using good wire with remote sense to the plug, a good plug, and a short DC cable, give further improvement compared to the stock connector I had originally hooked to a similar open frame supply.

I haven't much in the way of test gear only ears and a digital meter. Sorry if my post wasn't clear. Can you explain what "good wire with remote sense to the plug" means? The module has a sensing system, but haven't fitted one because the output is connected to an old DC cable and I am not familiar with how to connect the sensor to the DC cable withoutb having a junction box adjacent to the SB. This would be a further complication that I didn,t think was necessary. My SB is run wired and I use a digital output so the psu is well overrated as the psu doesn't run the wireless transmitter nor the d/a convertor.

Why does a short DC cable help?

Thanks for your response.

Skunk
2007-02-14, 14:05
I haven't much in the way of test gear only ears and a digital meter. Sorry if my post wasn't clear. Can you explain what "good wire with remote sense to the plug" means?

That's ok, I was just being snarky anyway. Most people *don't* do any real testing of the power supply upgrade, but I would be genuinely interested had you. Subjective opinions are always welcome here as well!

If it has remote sense there should be four DC output posts. Two of them should be S+ and S-, which are usually shorted together if remote sense isn't being used. You'd have to consult the datasheet for your supply for specifics.

When using remote sense, four wires are run to the DC plug- one from each post. Reason being, there could be a loss in power due to cable length/size, which the remote sense overcomes by sensing the voltage at the plug, rather than at the post.

HTH

Skunk
2007-02-14, 14:13
Why does a short DC cable help?

Thanks for your response.

Sorry, I missed this one. It helps keep impedance low:

From*
Low impedance- wiring: as regards the Julian Vereker comments above, keep impedance low by using low gauge hookup wire. I use 2.5mm2 stuff from maplins but probably anything 1.5mm2 upwards will be fine. You can even strip apart 3 core mains cable with these gauges from B&Q for example.

Low impedance- connectors: Again regarding the Vereker comments, its worth noting that your DIN sockets have approx. 5mOhms impedance, Bundy sockets about 3mOhms and my preferred Neutriks also about 3mOhm. I would therefore solder the umbilical cord from the psu directly and use a cable gland so eliminating one socket. For comparison, 1m of 2.5mm2 copper wire will have an impedance of around 7mohms and a good reservoir cap around 8mohms ESR.

*http://www.acoustica.org.uk/t/naim/powersupplies.html

mudlark
2007-02-14, 14:44
My hook up wire is best quality crap. The mains is an old piece of two core flex. The output wires are a short piece of lawnmower cable and soldered to an old piece of psu cable from a old backup drive. I'll measure the cable resistance which is the best I can do.

if anyone wants to send me some high quality solid silver teflon dielectric stuff I'll try it. PS I don't mind paying for a small amount, but I ain't shelling out a load for 25 metres.

mudlark
2007-02-14, 15:32
In fairness, you didn't sound totally convinced about the sound quality, safety, or even its looks ;->

I have not done detailed sound tests yet. I would appreciate advice about where improvements are likely to be heard. EG The bass on "The Wombles" will be cleansed of any wooliness.

I am sure it is as safe as i can make it be. The module is tested to TUV and the only problems are due to my bad wiring or non conformity regarding proximity of conductors to the case. I am not aware of issues. The cable is flexible stuff from mains installations.

The looks is fine to me, but the Maplins box is made out of slim aluminium (pun intended). The corners and edges aren't brilliant and the long edges have self tappers rather spoiling things. This box has nice proportions, but compared to a good quality box , well it isn't. It cost a fiver (). It is non magnetic.

I didn't want to oversell the project, but it works for me and I thought folks would be interested as it as it is buildable with minimal electrical design as it is made from off the shelf hardware. The only thing that bothers me is the mains. I would rather have used an AC wall wart to a box containing the low voltage electronics, but I couldn't find a single 5Volt circuit board that fit the requirements. Most items were +/- 5V and designed to put out 0.5 to 1 Amp for supplies to clocks and DACs etc.

i am not a qualified electronics expert so I didn't feel able to say the kit is spot on or what. The manufacturer of the module is happy for it to be used in the kit and it does seem designed for such a purpose.

M.

Jitterbug
2007-02-14, 17:45
Mine was just a flippant observation - no offence intended.



----- Original Message -----





From: mudlark <mudlark.2m0nsb1171492502 (AT) no-mx (DOT) forums.slimdevices.com>



Date: Wednesday, February 14, 2007 3:32 pm



Subject: Re: �60 UK 5Volt power supply







>
> Jitterbug;180553 Wrote:
> > In fairness, you didn't sound totally convinced about the sound
> quality,> safety, or even its looks ;->
>
> I have not done detailed sound tests yet. I would appreciate advice
> about where improvements are likely to be heard. EG The bass on "The
> Wombles" will be cleansed of any wooliness.
>
> I am sure it is as safe as i can make it be. The module is tested to
> TUV and the only problems are due to my bad wiring or non conformity
> regarding proximity of conductors to the case. I am not aware of
> issues. The cable is flexible stuff from mains installations.
>
> The looks is fine to me, but the Maplins box is made out of slim
> aluminium (pun intended). The corners and edges aren't brilliant and
> the long edges have self tappers rather spoiling things. This box has
> nice proportions, but compared to a good quality box , well it isn't.
> It cost a fiver (�). It is non magnetic.
>
> I didn't want to oversell the project, but it works for me and I
> thought folks would be interested as it as it is buildable with
> minimalelectrical design as it is made from off the shelf
> hardware. The only
> thing that bothers me is the mains. I would rather have used an AC
> wallwart to a box containing the low voltage electronics, but I
> couldn'tfind a single 5Volt circuit board that fit the
> requirements. Most items
> were +/- 5V and designed to put out 0.5 to 1 Amp for supplies to
> clocksand DACs etc.
>
> i am not a qualified electronics expert so I didn't feel able to say
> the kit is spot on or what. The manufacturer of the module is
> happy for
> it to be used in the kit and uit does seem designed for such a
> purpose.
> M.
>
>
> --
> mudlark
> -------------------------------------------------------------------
> -----
> mudlark's Profile:
> http://forums.slimdevices.com/member.php?userid=7151View this
> thread: http://forums.slimdevices.com/showthread.php?t=32565
>
>

Skunk
2007-02-14, 20:55
I have not done detailed sound tests yet. I would appreciate advice about where improvements are likely to be heard. EG The bass on "The Wombles" will be cleansed of any wooliness.
The audiophile forum is full of that kind of advice, if you do a search or two.



I am sure it is as safe as i can make it be. The module is tested to TUV and the only problems are due to my bad wiring or non conformity regarding proximity of conductors to the case. I am not aware of issues. The cable is flexible stuff from mains installations.
I'm not an engineer or tech, but I'll mention some tips I've read. It looks like there is a fuse on the PEM, so that's good. The ground wire should be longer than mains wire, in case it's pulled on (so the ground will be the last to short). Your DC lead appears to have connectors on the case, so that might not apply. If you're considering good cable, those connectors shouldn't be there anyway, according to the second quote in my last reply. Not sure if you should be grounded to the t-former rather than enclosure, but I'd double check. The alumin(i)um of the PS frame should be tightly coupled to the case to disperse heat, and you might consider holes as well. It won't regulate so well if it gets too hot. There's prob other stuff but that's all I can think of, and I'm not sure it's all accurate. Just food for thought.


The only thing that bothers me is the mains. I would rather have used an AC wall wart to a box containing the low voltage electronics, but I couldn't find a single 5Volt circuit board that fit the requirements. Most items were +/- 5V and designed to put out 0.5 to 1 Amp for supplies to clocks and DACs etc.
You could try getting an ALW super reg +5V circuit board or assembled regulator from Patrick Dixon (member here based in UK). The circuit has dropout though so your current supply would be useless. My guess would be a decent regulated wall wart and one of those super regs very close to the SB would be better than any of the off the shelf linear supplies.

mudlark
2007-02-15, 10:37
Thanks skunk for your interest.

I've a very thick skin so any other comments (you know who you are) of a negative nature are no problem.

I've spent more time reading all the stuff about psus. Very interesting. In the past I have bought some power supply cables, interconnects etc that have made a definite difference. I'm doing this PSU for interest, nothing else so I can cope with no improvement. I have bought expensive PSUs on the back of tests in the papers. (Cyrus PSX-R) This PSU is no problem.

My power supply has a slow blow 0.25 A fuse in the IEC socket. This was chosen because the module manufacturer suggested that one should be there, so guess what, I put one in. Putting it inside meant more connections to a fuse holder or similar with exposed 240volts bits plus chance of poor connections etc.

I don't understand why the mains earth lead should be longer than the conductors. How on earth can the earth lead ever be pulled out? It is wired directly to the tranny which has an earth tag screwed to the frame.

On the DC side there is no earth..... There are no connectors on the DC side as the lead is connected directly to the psu output leads. I can't measure the off line resistance of the lead as my meter don't go down low enough.

I can't bolt the psu frame flat to the aluminium case because the tranny and the main transistor on the psu sticks out. The thing is designed to be cooled by convection. The frame is bolted to the aluminium by four bolts to move some heat away from the psu frame. there are holes under the psu frame in such a place that mice cannot get to the 240 V bits. i cannpt account for the health and safety of any spiders that may climb into the unit. If any passing microbes get in then the F+++++ will die and I don't care.

I did think about all the matters that you mentioned before making the device....

Cheers,
Mike.

Skunk
2007-02-15, 12:02
I don't understand why the mains earth lead should be longer than the conductors. How on earth can the earth lead ever be pulled out? It is wired directly to the tranny which has an earth tag screwed to the frame.

On the DC side there is no earth..... There are no connectors on the DC side as the lead is connected directly to the psu output leads. I can't measure the off line resistance of the lead as my meter don't go down low enough.
Yeah I reread all that after posting and realized I'm an idiot. Thanks for pointing it out.

The ground wire only applies to captive cords, and my comments on DC wires were just meant to point out the unnecessary connectors.

I won't comment on the safety ground issue because I'm still not sure.



I can't bolt the psu frame flat to the aluminium case because the tranny and the main transistor on the psu sticks out. The thing is designed to be cooled by convection. The frame is bolted to the aluminium by four bolts to move some heat away from the psu frame. there are holes under the psu frame in such a place that mice cannot get to the 240 V bits. i cannpt account for the health and safety of any spiders that may climb into the unit. If any passing microbes get in then the F+++++ will die and I don't care.
Convection works best with active flow, which you don't have the benefit of. There has to be a way to cool the aluminum heat sink (frame). Coupling the frame to an external heat sink (your enclosure), might give convection a fighting chance. I've not seen standoffs used as heat sinks, but whatever works I suppose. The safest bet is a thermometer and data sheet.



I did think about all the matters that you mentioned before making the device....

Cheers,
Mike.
I forgot to mention the gold plated fuse :-)

mudlark
2007-02-15, 12:46
Hi Skunk,
cheers for staying with it.

I also forgot to mention that I am testing some platinum coated terminals on the fuse. (Only joking!)

The screws holding the three feet (to prevent micro rocking) are stainless steel to prevent magnetic effects on the ground plane. (courtesy of Russ Andrews and Les Wolstenholme)(Snigger)
The screws are imperial unit BAs 'cos I'm worried about the ill effects of any foreign influences. ( No offence - I actually am a francophile except the damn Parisiens : on advice from a French person who lived in Britany.) I digress.

EEEEKKK I'm beginning to be convinced by my own bullsh@@.

PS the spell checker on this here program (Firefox 2) can't spell aluminium. It thinks the word is aluminum.

M.

Skunk
2007-02-15, 15:35
EEEEKKK I'm beginning to be convinced by my own bullsh@@.

PS the spell checker on this here program (Firefox 2) can't spell aluminium. It thinks the word is aluminum.

M.

I'm listening to the wall wart right now. A hot soldering iron on the DC post of my linear caused a very faint sizzle noise, and looking in the crevice I could see the trace below bubble slightly when the iron touched the post. I couldn't get the board off the frame (due to the transistor you mentioned) to check that trace on the other side, so I'm scared to hook it back up. The fact that it's dated 1982, and was salvaged, isn't a good sign either.

Anyway, the supplied wall wart is sounding great through the digital out. The combination of the wall wart and my new pseudo-Anticables (magnet wire), has dramatically increased the size of the sound stage. On DSOTM, Run Like Hell- it's now quite obvious that the guy runs across front of the stage, but zags away on the trip back, while before i thought he was just running back and forth. Big difference IMHO. Don't bother trying it over there, as I'm sure the U.S. wall wart is immensely better.

Yeah, that's kind of ethnocentric of Mozilla. I prefer the stuff spelled 'ium', for high speed digital heat sink applications anyway.

mudlark
2007-02-15, 17:02
aloominum! surely.

i was just about to go off on one when i realised that it was nearly tomorrow.

Kan I just say that mr bush really pisses me of with the use of the word newquelar when it should be nuclear.

M.

crooner
2007-02-15, 18:46
Nice supply!
I'm working on a improved Power One supply for the Squeezebox as I type this. Modded with high performance caps and an extra bank of Elna Cerafine smoothing filter caps. Stay tuned!

Skunk
2007-02-15, 21:32
Nice supply!
I'm working on a improved Power One supply for the Squeezebox as I type this.

I tried that today, until I looked up and noticed half a dozen solder spatters stuck to my LCD display. Luckily it dried instantly enough not to damage it, but took a razor blade to get off!

crooner
2007-02-16, 00:03
LOL!
My "new" power supply is ready. Check it out!

http://forums.slimdevices.com/showthread.php?t=32802

mudlark
2007-02-16, 03:24
Thanks for the interest from across the water.

Crooner can you explain where the extra caps are added?

Cheers.

I've just now read the stuff from Skunk et al from early 2006. Seems a bit like deja vu to you skunk?

The reason I posted is to explain my experience to UK readers as most psu posting seemed to come from the USA.

I will not make any psus for anyone else as I am not a qualified engineer and I don't want to explain this to a coroner. I post to explain my experience and i do not suggest that anyone else should do what I am doing. I do not want to make any money nor do I expect a peerage. (OK Tony)

Also my house has recently been rewired to current (sic) standards and it is fitted with rcds etc.

love to all,
Mike.

crooner
2007-02-17, 01:31
Crooner can you explain where the extra caps are added? Cheers. Mike.

Hi Mike:
The caps replace the large smoothing capacitor on the linear PS. I used three caps in parallel. Two identical Elna Cerafine electrolytics (total combined capacitance: 16,400 uF) and a small 1uF film cap bypass.

mudlark
2007-02-17, 02:13
Cheers Crooner,
The Calex has 2x 4700 anyway so I'll make some efforts regarding listening tests before adding any more capacitance.

I might make up a short output lead and connect up the sensing facility as well.
M.

Skunk
2007-02-21, 01:47
I've just now read the stuff from Skunk et al from early 2006. Seems a bit like deja vu to you skunk?

I will not make any psus for anyone else as I am not a qualified engineer and I don't want to explain this to a coroner. I post to explain my experience and i do not suggest that anyone else should do what I am doing. I do not want to make any money nor do I expect a peerage. (OK Tony)

Also my house has recently been rewired to current (sic) standards and it is fitted with rcds etc.

love to all,
Mike.

Something like deja vu I suppose, but I have always tried to emphasize safety first. The main reason I took mine apart to rewire it was that I had the hot (black in the US) going directly to the transformer, and the white going to the fuse. I'm considering two fuses now, because mis wired receptacles are pretty common.

Ironically, your disclaimer and the need for it is illustrated perfectly by crooner's second link [in his linked thread]- the jaberwockie deal. There was actually a different thread where I suggested he check and re-check safety requirements for this type of device before selling them, liability and such. That thread would be more like deja vu, because of your safety ground issue :).

Sadly, it looks like jaberwockie wired his fuse incorrectly as well (in my untrained, unqualified, and non technical view of how to correctly do this). I've yet to see any on here that look perfectly safe though. My next one will have double shielding (teflon tube) over the mains, two fuses, and any exposed conductors will be well insulated and secured. I might even make the ground wire to the IEC longer, for the hell of it.

crooner
2007-02-21, 02:00
I had mine wired incorrectly as well. I noticed the power transformer was humming more than usual. I inverted the leads and this cured it.

I'm using a 0.5 amp single fuse. I think that's what found in all consumer electronics, including the very expensive high end equipment.

I think I got my PSU safely wired. Runs nice and cool. Reliable little buggers, these Power One units...

I don't plan to sell this stuff either. Strictly for personal use. Jaberwockie did inspire me to use a better enclosure than the cheap box I was using before. I also learned how to use a Dremel tool.

crooner
2007-02-21, 09:30
You know...
All this safety stuff got me thinking...

I'm gonna insulate all the live AC leads with shrink wrap for safety's sake.

mudlark
2007-02-21, 16:25
Shrink wrap as you call it is clearly good practice as is keeping the mains side cables the correct coloured wires.

The mains socket I used has four terminals inside the case two of which correspond to the earth and neutral terminals on the plug side. The other two terminals are one half size which is direct to the positive terminal and the other is full size, but below the other terminals. The smaller terminal is for connecting power on indicators. The large terminal is connected to the load terminal on the transformer. Another way is to take the fuse out and wire the positive conductor to the terminal which is open circuit. Check this with your meter. Check these connections with a multimeter again. If no multi meter available use a battery two bits of wire and a bulb.

mudlark
2007-03-10, 11:24
Hi all if you are there,

I decided to check the temperature of the transformer as I am incapable of leaving things alone for more than ten minutes.

I took the opportunity to measure the on-load and off-load voltage. The device measures the same in either state. (Wired only).

I touched the transformer and it surprised me that it was very hot. The frame almost too hot to touch. I decided to fit a fan. In went a small 50 mm fan from a computer connected to the 5 volt supply. The fan is supposed to need 7volts to start and 10 volts to run according to the spec, but it runs anyway?! No voltage drop detected. I can't think why running the fan at 5 volts would be a problem. If anyone could let me know why I might have a problem running it at this voltage I would be grateful.

The fan sucks air in at the back of the unit and I have drilled a few holes at the front for exhaust. Not perfect design, but who cares? I can feel a draft coming from the outlets. I'll report back after a few hours.

After two hours the thing is cool, but noisy. The fan, although new, is complete rubbish. Who was the idiot who bought that? Oops, that would be me! The case has been opened and the tranny was really OK temp wise. Clearly some active ventilation is a good idea if only for the longevity of the capacitors. It would appear that without cross ventilation stratification is occurring within the case. S'obvious really I should have listened to Skunk one of my American correspondents.

Skunk
2007-03-11, 15:46
It would appear that without cross ventilation stratification is occurring within the case. S'obvious really I should have listened to Skunk one of my American correspondents.

This can be fixed passively IMO. Just move your standoffs so that the largest area of the frame is in tight contact with the external enclosure, which will continue the heatsink effect to the outside of the box.

Consider flipping the frame upside down, which IMO would actually be right side up, so the transistor is on top and can dissapate heat into the air as well as into the aluminum frame. They even make To3 heat sinks, which are designed to combat stratification IIRC.

I forget if you said it was used or new but it could be a bad supply or poor design. The power one seemed to stay really cool, but my enclosure had tons of ventilation holes..

mudlark
2007-03-11, 16:24
Hi Skunk,

I quite like the idea of a fan actually. I'm trying a quiet version. If I turn the box on it's lid the ventilation out of the bottom is sufficient to keep things really cool so I don't think that the psu is badly designed. All it needs is some movement of air.

I'll reappraise the layout and let folks know what my opinion is. I've already put the fan in what might not be a good place.
cheers,
Mike.

And the conclusions.. even a quiet fan is disastrously noisy. That idea has gone on the back burner for now. The latest discovery is that I might have wired the transformer primary incorrectly. You know who you are..... I took the plunge and rewired in accordance with information found on the web and common sense. I am not amused as the transformer is now running quieter and what's more much cooler. Hmmmm. I must contact my legal adviser.

mudlark
2007-06-06, 14:09
At last i made the effort to do something about the temperature of the PSU.
I have tried a quiet fan, but this was a disaster as the enclosure, being rather thin is a fabulous transmitter of noise and vibration. Even quiet fan mounts did little to help with the noise levels. This makes using a 50mm fan impractical. IMHO.

Today I removed the fan and drilled several holes along the top edge of the power supply box side panels. The bottom has lots of holes that were put there to allow air in and out for the cooling fan.

The drilled holes are about 3mm or 1/8 inch. The holes are at the transformer end of the box. The holes do allow access to the mains side of the transformer if you wanted to stick something into the case. I will have to study whether this is an acceptable risk or not. I would welcome comments!

I am amazed that these little holes work really well to keep things cool. I didn't expect ventilation holes would be effective using convection.

Skunk
2007-06-08, 18:21
The holes do allow access to the mains side of the transformer if you wanted to stick something into the case. I will have to study whether this is an acceptable risk or not. I would welcome comments!


Hot glue metal screen on the inside of the vent holes, for a nice look to boot.

Cheers!

mudlark
2007-06-08, 18:58
Thanks skunk, but I wasn't thinking of attacking the thing now I have got this far.

lol,
Mike.

mudlark
2007-07-21, 08:40
I have now finished my power supply.

The aluminium case remains unpolished or painted. I have looked at other pieces of equipment and I have decided that the current arrangement is safe enough. There are two sets of holes on the bottom face. Those arranged in a rough circle were holes designed to let air get to the fan which was mounted horizontally above the holes. This fan has now gone as it was too noisy. The holes towards the front of the bottom surface were put there to allow air out from the tranny and other hot areas. This is now the way in for air that rises up through the case and exits out the new holes in the forward parts of the sides. These holes allow the case to breath sufficiently to keep the temperature down to warm. More than adequate, although I wouldn't have thought so.

Another picture of the final thing can be seen on my pathetic website. http://flycatcher.mysite.orange.co.uk/

More pictures are on the psu page.

I have done it again. I've spent even more money on the power supply. I found a new case courtesy of some kind sole. This case is fantastic and available from a company in Italy to lucky people all over the world. It allows the fitting of the frame psu like a glove. The psu fits without having to drill any holes in the new case. All you have to do is cut holes for the power supply cable and fuse holder and the supply cord. The result is better than the usual commercial quality.

mudlark
2007-08-01, 04:38
If you are now cursing me I apologise for bumping this post, but I have found a much better case. This case is available from Italy to all over the world. Some folks may be interested as this project now is close to commercial quality.

http://www.autocostruire.com/catalog/product_info.php?cPath=43&products_id=336

Mike.

jwcurtin
2007-08-01, 05:50
This case is available from Italy to all over the world.

So are you going to post a URL?

Cheers

John

killie99
2007-08-02, 00:32
Mike

The website says the height dimension of the box is 80mm, is this the internal or external dimension? If it's the external dimension then what is the internal dimension? Looks like a very nice box, much better than any of the standard offerings from RS/Maplin/Farnell.

Thansk, Stuart



If you are now cursing me I apologise for bumping this post, but I have found a much better case. This case is available from Italy to all over the world. Some folks may be interested as this project now is close to commercial quality.

http://www.autocostruire.com/catalog/product_info.php?cPath=43&products_id=336

Mike.

mudlark
2007-08-02, 04:38
Mike

The website says the height dimension of the box is 80mm, is this the internal or external dimension? If it's the external dimension then what is the internal dimension? Looks like a very nice box, much better than any of the standard offerings from RS/Maplin/Farnell.

Thansk, Stuart

Hi Stuart,

The inside dimensions are height 80mm height(tight)x 104 wide x 168 long.

The top and bottom are about 1.5 mm, the sides 8mm heatsink form and the ends flat 3mm aluminium.

The front plate is 85 mm high x 128 mm wide.

the rear plate fixes inside the top and bottom plates and on the front of the side plates and is 80mm x 124 mm.

My project is a frame power supply which fits to the width 100mm and the side plate has holes which allow the frame to be mounted about half way up the case.

The side plates are reversible allowing the mounting of components by screws and nuts at 25 mm and 75 mm or 40 mm.

My frame power supply is mounted at 40mm by screws and nuts. The nuts slide into channels in the side plates so there are no holes on the outside. If you go onto the website and enlarge the small picture of a 187 the enlargement clearly shows the construction method.

I hope that covers everything!

I really am enthusiastic about these cases because all the faces are held in place with screws so all faces can be removed for the fitting or maintenance of components. Even with 16 screws holding it together it still looks good.

Mike.

killie99
2007-08-02, 05:34
Hope you don't mind me asking, what was the total price including shipping?

mudlark
2007-08-02, 05:39
The total bill was 38.80 Eur. including postage. Unfortunately the postage was 14.00 plus vat at 20% which is a bit of a shame. The postage rather saddens the excellent base price.

killie99
2007-08-02, 05:46
Still a very good price considering how well the unit is made. I'm looking at a GX287 but I'm concerned about the internal height not being adequate. The supply I've aquired has a very big transformer on it and there would only be a hairs width to spare, need to have a look at my PSU to see if there is anything I can do.

killie99
2007-08-03, 03:27
Ordered a GX287 case about an hour ago and had an email saying it has shipped, most impressive! Total price was 28.90 which I'm not grumbling about as it's a fine looking case, much better than the 24 aluminium box that I found on Farnell.
Now I just need to look out a shoe horn to get the supply into the case .....

killie99
2007-08-04, 07:10
Just want to check I'm conneting the earth lead correctly to my box. The earth from the mains lead should be connected to the chassis, which in turn is bolted to the heatsink, which in turn is connected to the ground of the transformer by a wire? Correct?

mudlark
2007-08-04, 08:54
Hi,

I am NOT a qualified electronics engineer so I am not prepared to advise you on this. All I can do is tell you what I do. You should get advice from a qualified and experienced electrical engineer.

In my case I have wired the transformer earth tag straight to the earth terminal on the IEC socket. I think that insulation breakdown on the transformer is the greatest risk so I make this the "best" earth. With a separate power cable it is very difficult to put stress on the internal cables. I use a socket with integral fuse so no extra wiring is needed. No lights or switches as this is more to go wrong. All joints are supported and insulated with heat shrink. A wire breaking free on the mains input side is probably the next risk if this grounds onto any metal that is connected to the outside. The same is true if earth cables become disconnected and touch the mains conductor! I also keep the mains wiring away from the hot transformer. My mains wiring in the house is brand new with EL circuit breaker protection. There is no total safety.

Clearly if you can either double insulate things or prove connection to earth.

Don't take any risks. If you aren't comfortable with making these connections then don't do it. Get a mate to check the result.

Mike.

killie99
2007-08-06, 04:46
Checked with a spark at work. Connection needs to be:
1. earth from mains lead connected to chassis
2. flying lead from chassis to heatsink (do not rely on bolting the heatsink to the chassis)
3. flying lead from heatsink to ground of the transformer.

mudlark
2007-08-06, 10:32
Checked with a spark at work. Connection needs to be:
1. earth from mains lead connected to chassis
2. flying lead from chassis to heatsink (do not rely on bolting the heatsink to the chassis)
3. flying lead from heatsink to ground of the transformer.

The problem with this approach is that the chassis has six parts, so does that mean you are going to earth every part?

Which part of the chassis is the main part to earth?

My view is that transformer failure is the greatest risk so that's the earthing point.

With these cases the only way to earth the whole chassis is to drill the case, remove the coating and bolt an earth tag to each component..............real nice......

The safest approach is to use a plastic case?..........

killie99
2007-08-07, 01:05
In that case I'll just get one of my children to touch it everytime I want to use it just incase it's gone live .....

I know what you're saying, you can never be too careful.

mudlark
2007-08-07, 05:56
Oh I forgot to mention,

If the transformer overheats, the plastic case could easily set fire and the house will burn down taking your children with it.

Perhaps a thermal sensor inside the plastic case which shuts off the power if the thing gets too hot.

etc etc...?? ...!!!...

Mike.

PS Please sleep easily in your bed ....if your house hasn't burnt down yet.

PPS What I am saying is that it's all a matter of risk assessment.

ermine
2007-08-07, 06:12
I put a 25W toroidal transformer in a plastic case when I was at college, and powered something which presented a little bit too much of a load. The transformer overheated, and the plastic melted in the heat. The sharp part of the core cut through the enamel insulation of the windings, shorting a turn that caused the fuse to blow.

Took some careful prising to remove the melted mess from the wooden desk but no Armageddon occurred. Guess it confirms the value of a 3A plug fuse :) Before any wiseass jumps in, I do know the UK plug fuse is to protect the cable, and the equipment should have its own fuse but I didn't know that then, and heck, students are always short of $

FWIW many small EI core transformers have a non-resettable thermal fuse to cover the eventuality - it is placed just inside the primary windings. Most wall-warts that aren't switchers that I've dismantled have used this arrangement.

mudlark
2007-08-11, 11:24
Ermine,

Thanks for the information.

The power supply I use here has such a thermal cutout device. As the transformer is matched to the regulator and the whole lot is tested as a unit this gives me some confidence.

I use a plug fuse and a slow blow fuse in the IEC socket of the correct rating. If a fault occurs and the tranny overheats it is scrap.

I have measured the resistance between the case and the earth pin on the IEC socket. The side that is screwed to the frame is about 2 ohms as is the aluminium front. The other sides of the case have a large resistance. Does this mean I should fasten earth leads to all case sides? I can't recall any piece of equipment that has earth leads to all case components.

Personally I'll be taking the risk as my kit is wired up.

PS The cost of this power supply with the new case is nearer 75.00