3

I recently made a voltage and current controlled supply from a kit I found. It worked for about 15 minutes and then fuses started popping. Unfortunately, I have no idea why, because power supply worked for some time fine then and there were no changes to (just voltmeter connected all the time to output) it when fuse problems started.

At first, I thought that some part overheated and malfunctioned, but everything felt cool. Next idea was that a piece of debris inside the box might have caused a short circuit somewhere. I carefully checked the box and didn't find anything which could have caused short circuit.

Next thought was that maybe IC died. I already asked about that here, so I decided to separate circuit board and transformer and check transformer to make sure it was working correctly. So on one side I had PCB and on the other 50 VA transformer connected to a Graetz bridge rectifier. Here's the picture of the transformer part of the circuit.

transformer and rectifier

I connected ends of the rectifier to my multimeter and powered on the transformer and got around 25 V, as expected. A minute or two later, fuse died again. To me this looked very strange, because the rectifier wasn't actually connected to anything.

Since I run out of required fuses, I decided to connect my multimeter in place of the fuse and see what's going on with current. At first, I connected it to 20 A range and saw that at the time I press the power switch, I get around 0.02 A. In less then a second, that falls to zero. I switched to milliamp scale and got same 20 mA. After power-cycling transformer again, the 200 mA fast multimeter fuse for milliamp scale died.

By this time I was left only with a 10 A fuse, so I decided to do some testing with it. Since the original recommended fuses were of 315 mA fast type, I decided to be extra careful. I checked insides of the box again, checked both switches, all wires and as far as I can see, everything is working correctly. I turned the transformer on with 10 A fuse and did some voltage measurements. I get around 27 V when the ends are serially connected and around 13.5 V when I'm using only one end of the transformer. When measuring DC voltage at the rectifier, I get around 25 V for both secondary coils and around 13 V for one secondary coil. I also get around 9 V AC at the rectifier for both secondary coils and around 3.5 V AC for single secondary coil. I also noticed that when power switch is in OFF position, I get around 3 V AC at the rectifier input and around 3 V DC at the rectified output. These only disappear when I pull the power plug.

After all this, I concluded that there must be something in the transformer/rectifier part of the circuit that is making fuses blow. Any ideas how to find out exactly what it is?

Also, is the behavior of the rectifier normal and should I change the switch to double pole switch which would control both power lines?

EDIT 1

I connected the PCB to the rectifier and turned it on with 10 A fuse. I'm running a small 9 V DC radio from the power source and it seems to be working correctly. Could it be that the recommended fuse current is wrong? The transformer's maximum output current is 2.1 A, so if I'm calculating correctly, maximum input current should be 230 mA. I used 315 mA fuses, so the should have survived full load on the transformer. IS there something I'm missing here?

EDIT 2

Looks like the fuses could be blowing because of inrush current of the transformer. How would I solve that problem? One of the fuses I used and which blew was a slow acting fuse, so they aren't the solution.

Community
  • 1
AndrejaKo
  • 23,431
  • 27
  • 112
  • 189
  • The only thing I can think of is that your diodes may be faulty or misorientated. For example, if you connected the top left diode backwards you would have a straight path to the transformer (-) output. One way this could happen would be human error, so double check that; another way could be through choosing diodes with a rating too low, or if the diodes are damaged; they could conduct when they are not supposed to. – Thomas O Dec 21 '10 at 02:06
  • If the fuses are blowing due to inrush current, shouldn't they blow when you turn it on? Remember fuses have an I²T rating. In general, if you double the current you need to drop the time by 4. So even if you aren't pulling 315mA through the fuse, it may blow, it just may take longer. – Thomas O Dec 21 '10 at 02:09
  • @Thomas O I'm using KBL10 rectifier which is a single component with 4 connectors (2* AC, DC+, DC-), so I don't have any discrete diodes. It's rated at 1 kV @ 4 A. I'm 100% sure I connected it correctly, simply because it seems to be working. It could be internally damaged. As for fuses, the behavior is very strange. Most of them die within first 2 seconds of power-on, but couple of them died later. Can you provide me some sources on fuse calculations? I don't really know anything about that field. – AndrejaKo Dec 21 '10 at 02:28
  • It's a bit of a dark art. Your fuse manufacturer will often give a table. But if you don't know the manufacturer, go for a similar type. Here is one datasheet: http://uk.farnell.com/cooper-bussmann/s500-315-r/fuse-quick-blow-315ma/dp/1123085 On the first page we can see 1.5x rated current will blow the fuse in 60 minutes, given that you're less than 1x rated current it's unlikely to be the cause as you are reporting fuses blowing in 2 minutes. The second page gives exact time-current curves. – Thomas O Dec 21 '10 at 02:29
  • As your rectifier is 1kV and 4A rated, I'd doubt it's the cause. Still, you can test it with a multimeter, on the diode test range. If your multimeter doesn't have a diode range, you can do it with the ohms range: expect no reading backwards, and a relatively low reading forwards. – Thomas O Dec 21 '10 at 02:32
  • If inrush current is a problem, then you can also get use an NTC device, which will limit the startup current. There are NTCs designed exclusively for this task. – Thomas O Dec 21 '10 at 02:33
  • 2
    Are you sure your multimeter was on the AC current range? (not DC range.) Also, you can save fuses by wiring an incandescent light bulb in series with the fuse. Now overcurrent will make the lamp turn on... make sure the bulb doesn't draw more current than the fuse rating of course. – markrages Dec 21 '10 at 02:48
  • @Thomas O Kortuk♦ I just checked the rectifier with diode check. It turns out that when connected backward, I get 1.9 V of forward voltage. Usually, I get out of range on diodes when I do something like that. @markrages I'm sure that I set range correctly. – AndrejaKo Dec 21 '10 at 02:57
  • @Thomas O Could you give me a starting point for NTCs? They seem interesting. – AndrejaKo Dec 21 '10 at 03:01
  • 1
    @AndrejaKo See here: http://uk.farnell.com/jsp/search/browse.jsp;jsessionid=ONXBMLNNR24R0CQLCIPZLTQ?N=0&Ntk=gensearch_001&Ntt=NTC+inrush+limiter&Ntx=mode+matchallpartial&suggestions=false&ref=globalsearch&_requestid=49016 These are inserted in-line with the mains and reduce in-rush current. – Thomas O Dec 21 '10 at 11:24
  • 1
    "I also noticed that when power switch is in OFF position, I get around 3 V AC at the rectifier output and around 3 V DC at the rectified output. These only disappear when I pull the power plug." - this needs checking out!? – Linker3000 Dec 21 '10 at 23:54
  • @Linker3000 I think so too! After looking at some devices I have at home, I decided to place switch only on one wire. Should I replace it with double pole switch which would cut off both AC wires? Also, that was supposed to be 3 V AC at rectifier input. – AndrejaKo Dec 22 '10 at 01:05
  • 1
    @AndrejaKo. Do an experiment. Take your multimeter and plug the prongs into the mains on volts AC. You will measure 115 or 230V (you are in Europe, so probably 230V.) Now turn the mains off with the switch. Here, I measure 6V with the mains off, and I think I know why. It boils down to the multimeter and the switch resistance. The multimeter has approximately 10 Mohm resistance. To get 6V out from 230V in you need to divide by 38, so the switch has a resistance of ~380 Mohm, which sounds about right. You will probably find if you connect a 47k 5W resistor the voltage will drop to almost zero. – Thomas O Dec 22 '10 at 01:58
  • Well, I just reassembled the switch (I actually have two, one at the cable itself and one at the transformer), and the problem disappeared. It's not clear to me what I did wrong when assembling it first time. This entire project is getting very very weird. Also, how would I have measurable resistance on a mechanical switch is there are couple of millimeters between contacts? By the way, would 100 ohm, 3200 beta be a good value for NTC thermistor or should I take one form the <20 ohm range? Oh and when I do the experiment, I get 225 V and 0 V now. – AndrejaKo Dec 22 '10 at 02:21
  • @AndrejaKo, to your first question, most switches have insulation resistances in 100's of millions of ohms (simply due to air resistance and resistance of the plastic casing/switches/etc), which is more than enough to be safe. I doubt you could kill a bacterium with 100Mohm impedence. Multimeters are designed to have high impedence, so they don't load down the circuit under test. When you combine the two you get a voltage divider which gives surprising results. – Thomas O Dec 22 '10 at 02:47

4 Answers4

5

If you think it is happening due to inrush current, then it could probably be solved by a choke coil. They are there specifically to stop inrush current.

I would check to see if one of the diode's is fried. If you had one fried it will open-circuit for one half of the phase. When it open-circuits the coil will just become a floating load. This will cause a much higher current.

Test it by replacing the diodes. try measuring them to see if one has a higher resistance than the others.

Someone correct me if I am wrong.

Addendum:

more about choke coils.
Here are some:
http://industrial.panasonic.com/www-ctlg/ctlg/qAGL0000_WW.html
I googled choke coil to find the best explanation, I read the first paragraph of this and it sounded like it would be of help, so here you go:
http://www.wisegeek.com/what-is-a-choke-coil.htm

Kortuk
  • 13,412
  • 8
  • 62
  • 85
  • Could you give me a pointer for chokes? Wikipedia doesn't mention inrush currents in choke coil article. – AndrejaKo Dec 21 '10 at 03:02
  • @andrejako, have you checked your diodes? – Kortuk Dec 21 '10 at 04:51
  • A common mode choke is used to stop common mode eneergy from propagating. This will allow differential signals(what you want), but block common mode signals(what you do not want for EMI reasons, not really a problem for you, just extra info). – Kortuk Dec 21 '10 at 05:10
  • @Kortuk♦ It turns out that when connected backward, I get 1.9 V of forward voltage. Usually, I get out of range on diodes when I do something like that, so the result is a bit strange. Should I replace the rectifier? I'll check on common mode chokes. – AndrejaKo Dec 21 '10 at 11:19
  • @andrejako, test the resistance through the diodes in your rectifiers, one my have fried. Try replacing the rectifier, this might make a large difference. – Kortuk Dec 21 '10 at 15:20
  • @Kortuk♦ Just checked diodes and I get around 4 MΩ in the right direction and out of scale in the wrong direction for all diodes. Should I replace the rectifier? I'm still reading on chokes and NTCs. – AndrejaKo Dec 21 '10 at 16:12
  • @andrejako, 4MOhm seems a little high in the forward direction, but I could be wrong. I would try replacing your rectifier and see if your issue is mitigated. – Kortuk Dec 21 '10 at 16:14
  • @Kortuk♦ OK, I'll try doing that. – AndrejaKo Dec 21 '10 at 16:20
3

A 50VA transformer will take about 0.21 Amps when correctly loaded (VA / Input voltage)

So a fuse of about 1.5 x the input is suggested and should be Anti Surge (Usually marked T or TT - T stands for "träge" which is german for Lazy or slow) - so 315mA A/S or 400mA A/S

If your fuses are vaporized and cover insides with remains of the wire - This indicates a major short...

What type of transformer are you using - is it a toroidal - if so have you got a shorted turn (if you mount a toroidal transformer incorrectly you can add an extra winding which is shorted out - this is creates by mounting the transformer with a conductive clamp which is bolted down in the middle - if you are using a toroidal - try removing the clamp...)

It is possible that you have a faulty diode in your bridge - I have seen diodes that measure OK when tested with a meter, but when either loaded, or subjected to a higher voltage, break down and become shorts, or leak - the easiest way to prove is to replace ALL the diodes, as I have found if one is faulty, it usually subjects others in the bridge to stress, which may make them more likely fail, and for the cost of 4 diodes of 1N400X or 1N540X - I usually use 1N4007 or 1N5408...

Wonko The Sane
  • 539
  • 2
  • 2
  • @Wonko The Sane I'm using a toroidal transformer. I don't think I have a short in it. The transformer itself is covered in around 1mm thick layer of transparent plastic. The mounting equipment which shipped with it consists of two rubber disks of same diameter as transformer, a plastic disk and a screw. The screw itself goes through the rubber disk, transformer, second rubber disk and ends inside the plastic disk which has a threaded hole for screw. I'll try today with new bridge and T fuses. – AndrejaKo Dec 23 '10 at 10:01
  • 2
    T400mA fuse solved the problem. – AndrejaKo Dec 23 '10 at 14:27
  • Toroidal transformers have a very high surge / inrush current compared with 'standard' transformers. – Wonko The Sane Dec 24 '10 at 11:32
  • 1
    The problem with mounting toroidal is to make sure you don't create a conductive loop which will act as a extra winding that is shorted out. - to avoid this either use a plastic disk, plastic screws or plastic cable ties... – Wonko The Sane Dec 24 '10 at 11:37
  • 1
    @andrejako, if it used to work, started popping fuses now, and you have replaced the fuse with a larger one, and it stops popping fuses, it means it is pulling larger current. Why it changed is not fixed by replacing the fuse. – Kortuk Dec 25 '10 at 17:53
  • @Kortuk♦ "if it used to work" is itself strange. First fuse of 315 mA used to work for maybe 15 minutes and the rest of them died immediately, so I think that someone miscalculated fuse type, but I don't understand why first fuse worked for around 15 minutes and why it later died. I'm also thinking of putting 5 33µH coils on input and output to lower the inrush current next time I disassemble it. Would that work? – AndrejaKo Dec 25 '10 at 21:54
  • @andrejako, I thought you said it worked for a long time then suddenly started having issues, I would not worry about it then. – Kortuk Dec 25 '10 at 23:51
2

As transformers take a surge when turned on, I would suggest an Anti-Surge Fuse on the input.

Also does the fuse blow when you operate the switch to change ranges - some rotary switches are make before break - this means that when switching the output is connected to both inputs, this will short out one winding on the transformer and this will blow the fuse.

How does the fuse look after it has blown, on clear fuses you can see the wire inside, if the wire has a 'small' gap this means it is a 'minor' overload, if the wire has a large gap, or has vaporised and covers the inside with a dark stain, this indicates a 'major' overload.

A minor overload is either an long overload just above the rating of the fuse, e.g. 2 x fuse rating.

A Major overload is e.g. 10-20 (or more) x rating of fuse.

Wonko The Sane
  • 539
  • 2
  • 2
  • One fuse has cut wire, others have vaporized and cover insides with remains of the wire. As far as I can see, there were no problems with range changing, but only first fuse survived long enough for me to change range. I tried with slow-blow fuses, but they didn't help me (T series, I've seen other slow fuses too, but didn't have any available to try out). – AndrejaKo Dec 21 '10 at 22:24
  • Actually, it turned out that the slow fuse was a 160 mA fuse, so they might work. I'll do some more testing with them. – AndrejaKo Dec 21 '10 at 22:29
0

Use a higher rated fuse. The transformer will require magnetising current to set up its field in addition to the current the load will take. Dont forget that the fuse is there only to protect against a 'catastrophic failure' of the downstream kit, not to give you a precision current limit

BullBoyShoes
  • 928
  • 4
  • 5
  • How big fuse should I use? Rating is my biggest problem because 315 mA fuse I used, as far as I can see, is already too much for the transformer and won't be able to protect it in case something downstream fails. I already have "precise" current control in L200CV, so I'm not using fuse for that. – AndrejaKo Dec 21 '10 at 21:53
  • 1
    @Andr: Fuses are meant to guard against fire, not really anything else. – Nick T Dec 22 '10 at 18:31