Fixing a laboratory DSC

[vrasp]

Hi guys,

The lab I work in has an equipment called a Differential Scannning Calorimetry. It essentially measures the enthalpy change as a function of temperature.

Anyways, the equipment is made up of several devices and PCBs that have different cuntions. The main fuse connected to the transformer blew and I m trying to figure out why. These fuses go for $50 dollars each and we bought a few. I m going to unplug everything and connect each device until I find what device causes the fuse to blow and go from there.

Here is what I don't understand. The main transformer has an inputs 120v and outputs 230v. It is rated at 2.5 amps and 600VA. Here is a picture of the label.

Two devices (and some small PCBs) are connected to the transformer. One is labelled "input: 230v 6 amps", the other "input: 230v 2 amps". Thats a total of 1840W just for these 2 devices. How could the transformer supply that if it's rated at 600VA?

Thank you!

[petehall347]

Re: Fixing a laboratory DSC

unplug everything .it might be simply a power supply fault .

[eccerr0r]

Re: Fixing a laboratory DSC

Transformers(themselves) won't draw that much power continuously unless its output is drawing that power (or if the transformer is misdesigned for the frequency of operation). So depending on what's drawing power on it will determine how much power is needed for that apparent first stage step up transformer, probably as a cheese around to allow this device to be used in 120V-land like North America.

You say it's a calorimeter of sorts, so I suspect the main use for it is to detect temperature increases and decreases over time to estimate thermal energy change in a chemical reaction. Whether additional heating or cooling is necessary I don't know - perhaps to bring the reaction chamber ambient temperature to a fixed value - would consume the most power. This would be the most stress on that initial step up transformer.

[vrasp]

Re: Fixing a laboratory DSC

The DSC is connected to a heating element and a cooling one. The device that draws the 6 amps at 230v is connected to these elements. Its outputs are 10 amps at 24v for the heating and 6.2 amps at 17.8v for the cooling.

I m not sure I understood. I always thought appliances power should not exceed that of the transformer to avoid overload. How can a 600w transformer not be overloaded with over 1800w of devices?

Thank you!

[eccerr0r]

Re: Fixing a laboratory DSC

Well without a picture it wasn't clear whether that 6A @230V is that device's transformer limit or is it actually drawing that much power? The implication from your post seems to imply that it was another "6A 230V" transformer of sorts and without more clarification or a photo of that input I could only assume what was in the machine.

If it really is a 1400W load, that in itself will overload the step up transformer, hence there's a contradiction here... unless that load is just an underused transformer. If it truly was an undersized step up transformer, then quite possibly that's your problem right there though it should have failed a lot earlier... and you need a new transformer.

BTW if that 6A 230V unit was only to power the 10A@24V (240W) heating and 6.2A@17.8V (111W) cooling, and both of these can't happen at the same time, then unless that 6A@230V unit was really inefficient, it's not going to draw 1400W continuously and the first step up transformer should be just fine. In this case only inrush needs to be considered (which may just be 6A@230V for a short while) and the 600VA step up should be able to handle things after that brief pulse.

Also, since you seem to be blowing fuses, is it instant blowing or blowing after a while? Might reveal some evidence on where to look next.

[vrasp]

Re: Fixing a laboratory DSC

On the photo below is the back label of the device. I assume that's how much current is drawn.

Yes, like you said, it either heats or cools but not at the same time. I thought that regardless of the the ouput of that device (whether it s heating or cooling), it would always draw 6 amps. But if that's not the case it all makes sense.

I don't know about the fuse blowing. I found the blown fuse and haven't been able to test anything as I m waiting for new fuses to be shipped. As soon as I get them I can get on it and try to gather some info.

Thank you for your help!

[eccerr0r]

Re: Fixing a laboratory DSC

That's probably a switching power supply in there but again I'm just guessing based on the large disparity of input and output power - and lack of heat sinking. And the reason why the 600VA step up transformer works fine.

As most BCN people know, switching PSUs have a lot of ways to short out the input causing fuse blows. And the fact that MIDEC appears to have been fly by night, it's highly suspicious of being at fault here...

[vrasp]

Re: Fixing a laboratory DSC

Interesting. I ll take a closer look at that Midec. It does have a fan on the side which I suppose is next to a heatsink inside.

By the way what's BCN?

[eccerr0r]

Re: Fixing a laboratory DSC

BadCaps.Net

[vrasp]

Re: Fixing a laboratory DSC

Of course! Haha

How do you know Midec is unreliable? Or are you saying this because of the switching power supply?

[eccerr0r]

Re: Fixing a laboratory DSC

just a hunch because they went out of business it seems (website is gone?!) and switchmode PSUs tend to have their own set of failure mechanisms that aren't shared with linear PSUs.

But need to look at that PSU and confirm, it's only a speculation at this point.

[stj]

Re: Fixing a laboratory DSC

series lamp-bulb time.

btw, can you show us this amazing $50 fuse??

[vrasp]

Re: Fixing a laboratory DSC

series lamp-bulb time.

btw, can you show us this amazing $50 fuse??

The fuse is a KTK-5 Bussmann. 5 amps 600VAC. It s actually $50 canadian.

Can I really use the lamp bulb method in this case? The line is 230v. If I understand correctly, I would need a 230v*5A=1150w bulb? That sounds like a pretty powerful lamp bulb. Maybe as expensive if not more than the fuses. Correct me if I m wrong.

[stj]

Re: Fixing a laboratory DSC

no, your using it to limit the current - not allow it all
a couple of 100w 120v lamps in series will do for now

[vrasp]

Re: Fixing a laboratory DSC

no, your using it to limit the current - not allow it all
a couple of 100w 120v lamps in series will do for now

How do I know how much current I m limiting? Wouldn't I want enough bulbs that the current doesn't go over 5A so I dont damage anything? Also the voltage on the line is 230v.

Thank you!

[eccerr0r]

Re: Fixing a laboratory DSC

F**K those KTK fuses, that's those stupid fast blow midget fuses they use in multimeters and we had that other rant thread somewhere on BCN.

The reason why light bulbs "limit" the current is that it's connected in series with the load. Now unless your device somehow has negative resistance (I'd love for someone to give me a resistor with negative resistance... things that are piecewise negative don't count) the amount of current going through the bulbs can only max out if your device was completely shorted out.

Now you probably don't need a whole bunch of bulbs in parallel - yet. Unless you're sure the short is in the cooling/heating circuitry. However since this is a KTK fast blow fuse, it might be troubling and more pictures of how things are connected and what it's driving would be interesting. Normally you do not use KTK fuses on SMPS's because of inrush!

As a rule of thumb, since 120V at 1A is 120W, a 100W bulb is a bit under 1 amp. You can make a rough guess at how many bulbs you need in parallel.

So where is the fuse actually, is it in the 230V side or 120V side, I thought it was 120V wall plug - switch - fuse - step-up transformer 230V - that MITEC PSU that needs 230V.

[stj]

Re: Fixing a laboratory DSC

the limit is the lamps - so in that instance, 100w
at 240v that would be about 400mA

your just trying to stat the power supply's and not run the whole unit at this time.

[vrasp]

Re: Fixing a laboratory DSC

My mistake. Yes, the fuse is before the transformer so it s at 120v. I have more pictures of the setup but it s quite messy. Lots of cables everywhere. The Mitec is on the left.

So if I have let s say, hypothetically, five 120w bulbs in parallel instead of the fuse, and there is a short to ground somewhere. This is what I understand: The current drawn will try to increase until something breaks and the circuit is open. Let s say each bulb only allow 1 amp. Will the rest be dissipated as heat? Or will the bulb break?

[eccerr0r]

Re: Fixing a laboratory DSC

Think of it this way: one bulb straight connected to 120V... will not burn out
Two bulbs connected parallel straight to 120V... will not burn out, and more the same... but more current is being drawn with each bulb added.

Now connect two bulbs in series. Note that each bulb will be dimmer than before. Measure the current flowing through the two bulbs in series. It's necessarily lower than before! Double up on one of the bulbs. Note that the parallel bulbs are now dimmer than the single bulb...which will now be brighter! Now short out the parallel bulbs... The parallel bulbs will go out but you won't trip a fuse, and the single bulb will be full brightness.

Don't know if that blurb above makes sense but that's how we're taking advantage of series bulbs to limit the amount of current going through a circuit.

---

Hmm... that thing... now you might have to open it and look inside...

[vrasp]

Re: Fixing a laboratory DSC

It s more clear but there is still something I dont get.

Will bulbs in parallel limit more current then in series?

120v with 100w bulbs is about 0.83 amps. Should I put 7 of these in series to test the 6A Midec?