What's inside an EC Apparatus brand constant voltage constant current power supply used for Electrophoresis - the separation of particles in a fluid by using an electric field.
You can get these supplies quite cheaply on ebay if you look around.
A quick reverse engineering of the circuit provides some insight into what's happening.
http://en.wikipedia.org/wiki/Electrophoresis
http://en.wikipedia.org/wiki/IGBT
http://en.wikipedia.org/wiki/Snubber
Datasheets:
http://www.irf.com/product-info/datasheets/data/irg4ph50k.pdf
Forum: http://www.eevblog.com/forum/blog/eevblog-530-electrophoresis-power-supply-teardown/'>http://www.eevblog.com/forum/blog/eevblog-530-electrophoresis-power-supply-teardown/
EEVblog Main Web Site:
http://www.eevblog.com
EEVblog Amazon Store:
http://astore.amazon.com/eevblogstore-20
Donations:
http://www.eevblog.com/donations/
Projects:
http://www.eevblog.com/projects/
Electronics Info Wiki:
http://www.eevblog.com/wiki/
You can get these supplies quite cheaply on ebay if you look around.
A quick reverse engineering of the circuit provides some insight into what's happening.
http://en.wikipedia.org/wiki/Electrophoresis
http://en.wikipedia.org/wiki/IGBT
http://en.wikipedia.org/wiki/Snubber
Datasheets:
http://www.irf.com/product-info/datasheets/data/irg4ph50k.pdf
Forum: http://www.eevblog.com/forum/blog/eevblog-530-electrophoresis-power-supply-teardown/'>http://www.eevblog.com/forum/blog/eevblog-530-electrophoresis-power-supply-teardown/
EEVblog Main Web Site:
http://www.eevblog.com
EEVblog Amazon Store:
http://astore.amazon.com/eevblogstore-20
Donations:
http://www.eevblog.com/donations/
Projects:
http://www.eevblog.com/projects/
Electronics Info Wiki:
http://www.eevblog.com/wiki/
Hi, welcome to Tar down Tuesday And today we have a piece of scientific apparatus you may not have heard of and uh, from a company you will most certainly haven't heard of as well who I believe are now out of Uh business. Anyway, they seem to have dominated this field and a whole bunch of these power supplies are available on eBay and it's from a company called EC Apparatus Corporation. This is the EC 250-90 and what it is is a power supply for a scientific technique called Electr Furus and it's something I've never heard of before and I will link in the Wikipedia article down below. Check it out.
But basically what this Uh technique allows you to do is uh, separate charge particles in a fluid or a gel or something like that at Uh High voltages they're attracted to one end and not the other. allows you to separate stuff out for you know, DNA analysis and all sorts of other analysis and and things like that. And this technique, although very simple, requires a constant Uh voltage and or constant current uh, power supply and the Uh. The value of the electric field you set in there allows you to separate out separate particles and all that sort of stuff.
Fantastic. Anyway, we won't go into the details of that we care about that you can score one of these power supplies on eBay and uh uh. this company Uh, EC Apparatus Corporation uh manufactured a whole range of these ones. This is the EC 250-90 What that is is Uh 250, Vols Uh DC maximum.
Uh, this one goes up to 500 milliamps maximum Set Current Oh I Believe the 90 is the 90 Watts Um so you know 90 WS uh maximum output power. but you can get models of this that go up to 6,000 volts I believe And really Precision currents down to microamp levels. All sorts of stuff. So if you search for the term uh for the on eBay for the search term electr Furis, you'll no doubt find one of these power supplies.
or search for EC apparatus Corporation I Think they dominated the market in these things anyway. Uh, this is a constant voltage, constant current power supply. uh, voltage and current. You can sit with a nice little knob here, turn off and on.
It's got a timer that allows you to uh it either times how long you've been applying the power to, uh, the experiment or um, it can. or you can set the timer so that automatic switches off after a set time period and you can set and display the actual voltage, constant voltage, and or constant current there. So really handy little power supply. usually for high voltage stuff.
You know you're not going to get one of these that only goes from 0 to 10 volts or something like that. Usually, uh, quite high up so this is one of the lower end models. But anyway, um I Thought it might be interesting to take a look inside and at least show you that these things are available on eBay They're not hugely, uh, cheap, but for a couple cou hundred bucks or something like that, you might be able to pick up a you know, a really high power, constant voltage, constant current power supply. So they certainly worth looking at. and uh, and if we switch it on here. Uh, this is a 240 volt model and here it is we can uh, display set and we can set the voltage. There it is. we can go anywhere from uh zero.
Oh, there's something wrong with that Noob Look, you turn it too quick and it, uh, it doesn't like that at all. But anyway, this sucker will go up to 250 OD volts and uh, then you can set the mode milliamps from 0 to 500 milliamps constant current and you can set a timer as well if you really want. but you can, uh, just set your voltage there. Let's set it to 100 volts.
There we go and try and switch it on. and uh. Unfortunately, it gives us an error because there's no load on this thing and by the way, you might think this has four outputs. Here, it's not four separate outputs, they're just paralleled all up.
So depending on on the experiment and how you do it and all that sort of jazz, you may have to hook up multiple things. It's just for convenience, but it is only a single channel, so let's hook up a resistor to this thing and uh, see what happens here we go. I've got a uh Power resistor on there. We may have to recycle here.
We go there, we go. We're now there we go. We're now on and it's going up and we will measure our voltage on the sucker. So there it is, we have ourselves a, uh, the actual Uh current 100 Vols seems relatively accurate.
Uh, it jumps around a lot. I should probably get this on the scope and actually, uh, have a look at this sucker and uh, see how much ripples on this thing? That's probably, uh, worthwhile. but um, by the way, you can't CH just change the voltage like that. it doesn't automatically go, which is rather like track in real time, which is strange.
You sort of have to press that again to sort of make it go down in voltage there. but hey, you know this could be a handy power supply. and I've got the output set to 10 Vols And look at that. we've got some.
uh, we got some Ripple on there. but look, there's lower frequency stuff. This thing just drifts. There's lots of I So what? There we go.
It's sort of like drifting all over the place. So this thing is really not that great like really low frequency stuff. Let me see if I can capture that at a, uh, very slow time base. There There we go.
That's uh, 2 seconds per Division and you can see that it's uh, there's a serious drift issue there. I don't know what's going on. So this is not the world's most uh, regulated power supply, that's for sure, but it probably you know, almost certainly doesn't have to be. And the same thing's going on there at 50 Vols output as well.
and that time period in there? we're looking at 10 milliseconds so that's 100 HZ. So clearly, uh, 50 hzz? uh, full wave rectification in this thing. and yeah, that's just coming through on the output there. so not hugely regulated and I just did the power on waveform. there seems to power on quite nicely there. I've got the scope on Roll mode and uh, there you go, but it is, uh, has a bit of an issue. Wonder if uh, that's like a temperature related or something. all that sort of stuff in there? I'm not really sure but uh yeah, it's bit all over the shop and how does it ramp down? Let's have a look there we go.
Boom straight down, nice and clean so nothing of the playing around. Let's pop this sucker open and uh, see what's inside shall we? I Uh, don't expect a huge amount, but uh, likely all sort of. you know through hole uh stuff. I suspect and uh, like some of these are more advanced than uh, others of course.
So um, looks like we're got to take the front panel off here and we'll see pretty hefty beasts. Uh, by the way. So obviously there's a big, uh, big linear Transformer in this sucker. And of course I expect there to be a fairly significant difference between designs of units.
I Mean this one's relatively low voltage as far as these Electr Fusus power supplies go. Um, as I said, some of these models from EC apparatus go up to six kilovolt, so there's a 3 kilovolt one, and um, all sorts of stuff. So oh, here we go. So I expect there to be dramatic differences between.
oh, that's that's neat. look at that just slides off like that and uh, yeah, it looks like it's all on the front there. Jeez, there's not much in it. There's not much in it at all.
Big thumping linear Transformer and the output board. and just a tiny board here. Jeez, nothing much doing at all there. Notice a tiny little heat sink, uh, down in here, like a T To220 size heat sink and they've just put a big lug on the back.
maybe? uh, you know, big hex, uh, spacer on the back? Maybe to just get a little bit more dissipation in there, but this is rather interesting at first glance. and uh, I think it's going to require me to reverse engineer this thing a bit to figure out what's going on anyway. Main's input coming in here from the IC connector on the back that goes through to our main switch here, which actually switches it. but then we've got some rectification hanging directly off that somehow and then there looks like to be an opto coupler down in there.
This looks like our primary Transformer tap and we've got six wires there. three Brown three black, but they're in three separate Uh pairs. Well, you know there's three separate Uh wires there, basically because there's a brid Bridge across those two, Bridge across there, Bridge across there. and uh, so I'm not actually sure what that's doing there.
and uh, then our the other side of our Transformer the Uh secondary output comes in here and just goes through to the ribbon cable here. So then it's doing stuff on the front panel board and then it's coming back here to the output. You can see it coming back out of the ribbon. So obviously our front panel board is what's doing the uh, you know, the reg, the filtering, and uh, uh, regulation in quote marks, um of our output. and the other interesting thing to note is that we've got this sort of card Edge connector over here, which is rather unusual and if you check out the traces here, you can see them going off and then going down into the breakout tabs. Now this is a, uh, common practice. There we go going down in the breakoff tabs. very common practice to when you panelize a board and when you want to test the boards after they've been a assembled but still assembled in the panel.
then you can, actually, uh, get all of the signals out from each particular panel. So what they've got here is, look. there's an extra wire coming out from this tab here, which had come from another board in the panel or some sort of test circuitry or something like that. depends on how they want to do it.
but those two signals go over to the main test connector here. So you have one main big test connector for everything. And of course, this test connector has some of this output, but it's also going through to whatever was on the secondary panel there and it doesn't look like it's an identical board on the secondary panel cuz I don't see any other matching breakouts. um, on the other side for example.
So um, there's some other ones coming out from the back there down to there. So that's uh, interesting in that they've uh, put a bit of thought there into testing these things at the panel. uh, assembly, uh, process and you'll notice they got the same uh, production panel testing stuff on the main board as well. And there's our main board.
Check it out. Bodgy Brothers time. Look at that Axal electrolytic uh cap going over there. couple of series uh 1 N41 diodes under there resist.
They've you know, done it? Okay, They've put you know heat shrink tub in over the uh legs of the those things. but yeah, really? uh, bung that on there. Lm339 Qu comparator, couple of 10 term trim pots, only 324 Op amps, you know, really nothing fancy at all. and uh, this package down here.
they've actually, uh, try and zoom in on that. They've uh, huge big legs on that, uh, transistor presumably down in there and they've just cut it off and then they've wrapped wire over the individual pins and then solder that down in there. Obviously we got a bridge rectifier down in here. Here's our main output: uh Fielder Cap of course, doing the business.
3. Ohm uh, big Power resistor down in there once again with some bodess going on. it looks like is that a Dio down in there with a couple of resistors all bodged on and yeah, um, not that great. Optical uh encoder.
really nice one though you can hear that. Really good quality that's a Burns one you know. spared no expense I'm not sure what these things go for. um, you know, some of the uh, like brand new. Some of the secondhand dealers sell this model for like you know, $1,000 Plus or something like that um processor in here. you know it'll probably be I don't know, a pick or some other lowly uh, 8bit processor I'll take the sticker off in a second, but uh, nothing much else. doing. Date Code of the 49th week 99 there.
So this puts this in a it's a year 2000 model. There you go and that looks like a Motorola MC 68 HC 705 microcontroller. Nothing fancy. and of course there's a E Prom down there as well to Uh store to uh hold the Uh values when you power it off.
All right. I've done a very quick little reverse engineer here. may not be 100% accurate. Uh, so quote me on this, but it should be good enough for the purposes of today's experiment.
What we've got here: 240 volt Mains in our power switch here uh, accidentally Drew those lines on the wrong side. they should be on the switched side of the mains input got our big thump and Mains Transformer with a 12volt p uh tap to Um Power a 5V regulator to uh just power the digital logic circuitry and then we've got a 1.05 ratio tap on here and I measured uh well. my mains here in the lab is actually 249 volts and I was getting out 260. Uh, something out of there I worked out 1.05 so nominal 240 to 251.
there's bumping that that up a bit full wave. Rectifying that here with that big ass dire Bridge with the heat sink on it you saw there and uh, they've got an Opdo coupler here. just uh uh. Reading the Main's Um cycle on the UT and feeding that through to the Uh control circuitries box here is just generic control circuitry which I haven't uh, reverse engineered yet I Don't think I'll bother.
but uh, basically. yeah. DC out. So we're getting our high voltage DC out of there rectified and then, if this doesn't look quite familiar to you this symbol.
it's an Igbt, an in insulated gate bipolar transistor. It's an International Rectifier type IR G4 Ph50 K and uh, that is the main uh controlling element. here. it looks like they've got a 1 ohm uh resistor that looks like it's being used for the current sense.
so there's something being tapped back off there. There's our main big output filter cap here there. that huge axial uh, high voltage one you saw before. they've got a snubber on there.
Just a simple series RC snubber and that's that 3 ohm mod resistor you saw. um I think normally this board was designed just to have the diode output uh, directly from that uh uh, D that filtered DC directly to the output. They've added that 3 ohms and curiously, they've actually uh, tapped that back off with about 1K back to the positive supply of that Uh regulator. So anyway, um, this is essentially how it works.
They using the Igbt to uh, regulate the output voltage and the output current. Now what we're going to do is we're going to probe a couple of things and have a look. And in this case, uh, right now I'm going to probe the uh, gate, uh voltage and I'm going to probe the DC the um the filtered DC output there because that is important I think we'll see something interesting there and uh I don't need to probe the DC output cuz that's just going to be. we've looked at that before. it is just a oh sorry the output here which will be the same here, it's just going to be a fixed uh DC that's going to be the programmed DC output voltage. So basically with that opto coupler feeding back to the microcontroller, uh in here they basically got everything the control of the Igbt synchronized to the AC Mains input here so some probes hanging off there, it's a bit uh, it's a bit hard cuz I've got to connect the membrane uh, wire, membrane, ribbon back to the main uh board back in there so it's all got to sort of sit like that. so I got to probe and then sort of put it back up. but we should be able to still operate it with that and let's have a look at what we've got here.
now. what we've got is uh, uh, Channel One is our gate. uh, voltage. Uh, both of these are 50 volts per division here.
So uh, the gate voltage is currently, um, basically down at zero. We can turn that, turn that up a bit and you can see it's basically zero there. and uh, because I haven't switched the output on yet? Okay, so this is just with nothing with the Igbt Switched Off obviously with the gate voltage just sitting down there and this is our Uh DC voltage here: 50 Vols per division 100, 200, 3 300 So we're getting 350 360 Odd, you know. 365 Vols uh DC or something about that out of our fullwave Bridge rectifier there little bit of Ripple on there.
No big deal, but let's look what happens if I switch my output voltage on. Oh, by the way, I've got a load. I've just got a small resistive load on the output just to, uh, make sure that the thing uh does power on because otherwise it doesn't seem to gives you an error as we saw earlier. Now here we go: I'll switch it on and Wham look at that.
Look what's happened to the output of our Bridge rectifier? Let me trigger on that. Now let's have a look at actually what's going on here. you can see our yellow waveform. The gate voltage there is now, uh, jumped up 50 volts I've by the way, I've set the um, the programmable output on this thing to 50 volts.
That's why we're getting uh, that 50 volt gate voltage plus a little bit more to switch it on. But look, what's happened to the DC input? As soon as the here we go, let me turn. Let me turn it up. Here we go.
As soon as the um, the Igbt turns on like that our the output of our Dio bridge is being clamped right back down to that that programmed or average 50 volt output level there. So if we set them to the same scale, there we go. you can see that they're exactly the same and if we boost that up up like that, so it's so it's slamming that doode bridge voltage down to that nominal voltage. And when uh, the micro switches that thing off, then after a little bit, it starts to ramp back up to allows the doode bridge and um, to charge the output to charge back up to that nominal 360 odd volt output Before bang it does it again. So you can see when the micro switches on the transistor here, it's basically this, um, it's going. It's set for a nominal 50 Vols output here. so it does. The micro Uh controls the gate to deliver a 50 volt average, uh, filtered rectified output voltage there.
But of course because the transistor is slammed hard on, it's dragging the output of this Dio Bridge down to that 50 volt level. So that's why we've got a huge amount of switching. and that's why they they've decided to put a little snubber in there just to take the edge off that. oops, sorry, made a small mistake there.
Uh, this resistor doesn't go back to the output of the voltage regulator, it goes back to before the input of the voltage regulator which is the Uh filtered 12volt rail in here which also Powers uh, the LM You know the comparators in the opamps and the other analog circuitry which we saw in there. and now I'm looking at the output voltage which is the Blue Wave form you can see in there, that's the 50 volts output voltage up and you can see the see the UN ation, the real low frequency undulation in there. Still not. uh, sure of the cause of that, but you can see basically.
what I wanted to show is that there doesn't appear to be any missing pulses in there. Oh, and the yellow one is still the gate voltage. By the way, it doesn't appear to be any missing pulses. any skip pulses in there whatsoever.
at least with this particular load. And what we've got here is our gate uh Drive voltage again with our Um output from the Opto coupler, the blue waveform there, so you can see how the timing is totally synchronized to the AC Mains input via that Opto coupler and what we've got here is once again the gate voltage at 10 volts per division there, and the blue waveform this time is the rectified Uh analog positive output voltage where that Opto coupler was tied to. So that's the 12vs Uh nominal output veil. By the way, all this is referenced to the negative output of the Uh power supply on the front panel.
by the way. So as you can see, they're floating that isolated 12volt Uh rail up to the positive Peak there of the Gate Drive voltage. So that blue waveform was that Uh reference point in there which goes off and Powers all of our analog I should draw that actually going into there powering now analog control circuitry and it's no Sur surprise if we probe the uh, the zero vo the zero volt I won't say ground because it's uh, you know this is effectively our ground Point Here, this is the reference. oh sorry, the output. Uh, that I'm using as the reference for the scope here. So if we probe the negative output of this 12vt rail, Surprise surprise. Tada Instead of being up here, it's level with the output voltage. So that regulator there and all of the control circuitry in there is floating uh, well, and some of the control circuitry in there is floating relative to the output voltage.
And if I hook a 10 ohm load on the output of this thing and set the current limit to 250 milliamps with uh, you know, like 10 volts output voltage. Uh, yellow voltage is the gate voltage and the blue voltage at 100 volts per division. yellow voltage is 20 volts per division. By the way, the gate and the blue is 100 Vols per division.
So we're already at 100 200, 300 that 360 odd volts that we got before. let me switch the output on so it'll It doesn't instantly go into current limit. it's got to ramp up the output voltage before it will attempt to and it'll ramp up to that current so you'll see it takes some time. Check this out there.
We go. start it out. You'll notice that when it gets to that voltage, it now then raises entire voltage up to 10000 300, 400. We're getting about 440 odd volts there now on that uh uh output of the bridge rectifier into the Igbt.
We've also got some pretty serious overshoot on that that's 100 volts per division, folks. Ouch. And we'll find of course that it's not actually the output of this uh, 250 volt tap on the Transformer actually going up. There's no mechanism to actually, uh, make that happen.
but it's the uh, effective DC output voltage on this bridge caused by the converter that's going up. So what I've got is I've got my Uh RMS voltmeter here hooked up to that main. That 250 volt tap on the Transformer. As you can see, I'm getting 258 Uh volts here and there's our voltage.
There's our 360 odd volts. I'll switch that current limit back on and we'll see it ramp up and you'll notice this thing won't change at all because it's low impedance and it's fixed. It's coming directly tapped off the mains. so I wasn't expecting that and uh, of course I was using my standard uh scope probes in Times 10 configuration.
These only have a maximum DC voltage rating of 500 volt. So really, we were uh, exceeding the operational range of the probe there. Oops. Should have got my high voltage probe out so that is very, very interesting.
and unfortunately I've got to leave it there I um should do some more work actually reverse engineering. Uh, exactly what's going on here I could have a slight uh thing out of place I need to be a bit more thorough in that. but if anyone has actually got a schematic uh for this sucker, then uh, please leave it in the comments I couldn't uh find one after a uh, half reasonable search? Or it maybe even a um, like a similar model doesn't have to be the 250-90 There may not be huge differences between the uh individual models. Heck, couldn't even find a user manual for this thing. I Mean jeez. So there you have it, that's a tear down of one of these EC apparatus Corporation Electro Fesus power supplies. and well, they're pretty crude uh things ultimately, but you can get yourself well, at least a half decent. um, uh, constant current, constant voltage, programmable high voltage power supply by searching for that term Electro Fesus which I think is my favorite new term I think I've got it down packed.
Took me a while to get my tongue around that anyway. I'm sure the more scientifically literate out there will correct me if I am saying that long. Say it three times quickly. Electroforesis.
Electroforesis. Electroforesis. Goodness sake. Anyway, if you want to discuss this, um, jump on over to the Eev blog Forum that's the place to do it as always.
I'll link um stuff down below data sheets for various things um, or up there depending on where you're viewing this and I hope you liked it and if you did, give it a big thumbs up. Catch you next time.
These work great for reforming high voltage capacitors
good for runnin neons???
ะะปะฐะณะพะดะฐัั
it's a peace of sheet
the way you pronounced it at the beginning really irritated me, but at the end you started pronouncing it properly. the important thing is the emphasis on the e.
One hell of a shit supply.
Well, after all those years the schematic was not found yet? Anyone please?
Hello EEV BLOG
thanks for the electrophoresis reengineering video
please can you do a reverse engineering on DY-300 (china) electrophoresis power supply video
Thanks
Hello, I have some video with a protection circuit design against overvoltage crowbar type, but it is better not put the positive ande negative terminals in the short to work with currents from 10A to 50A? Tks
The world's worst regulated power supply..
How is the base of the igbt regulated from a digital source?….. I could get my head around a variable resistor. But how the heck does a semiconductor chip produce differing voltages to control the base?.
man i want that transformer lol
wa
Hi Dave, great video–I love these teardowns. ย The issue you see with the knob not working when you spin it quickly is most likely because the micro isn't sampling the encoder fast enough. ย It probably decides that the position hasn't changed when you move it several "clicks" (usually 4) between samples. ย I bet the designers decided it was OK because who would ever spin it that fast anyhow ๐ ย It's a "feature."
the 're' syllable has a "long" e sound, e-lek-tro-fo-ree-sis NOT so very difficult
Please do a video on how to create a variable power supply from a pc power supply
A colleague and I have created a couple of videos from electrophoresis simulations to illustrate the concept a while ago. You can check them out on his channel (ohickey100).
Electroph..ff..f..what?!?? ๐ nice video! Thanx!!
One of the best teardown vids on this channel! ๐ I wish Dave would take the time to explain the circuit in all of his tear-downs like he did in this video. Usually he will point to some component or sub-circuit, and say "nothing interesting there", and then my hart sinks. Love the channel! i would love to see more explanations like this.
Don not turn it on! Take it apart! ๐
Very nice desulfator for battery charger ๐
Well those are all possible problems with any poorly designed feedback loop and not specific to having a capacitor there. A well designed loop can minimise or avoid those problems or a poorly designed loop can have those problems even without that cap. If you can't even work out how to compensate a feedback loop properly, you shouldn't be designing power supplies. The feedback loop in the SMPS must be stable, regardless of what's "prior to its input".
i guess there is going to be an IGBT fundamentals friday soon…
Dear Dave, I'd appreciate that
greetings from Germany,
Mรคander
EEVBLOG is more for EE, not just for hobbyists. Nothing personal. IMHO EEVBLOG is more for the purpose of understand what is happening, why the LED is blinking and not just make the LED blink.
an additional "benefit" would be overshooting and/or very high peaks with spike errors and may some low frequency ringing may occur as well.
Adding more low freq-gain to the error correction is not the right way.
Stabilisation must be prior to the SMPS input.
A 90W 250V flyback converter would be small, simple, cheap and probably work perfectly well for electrophoresis.
If you want good DC regulation in an SMPS, stick a capacitor in the local feedback loop of the error amplifier, giving it essentially open loop DC gain. More gain in the error amp means better regulation at that frequency. Most well designed SMPS have this.
The load on this is a block of polyacrylamide gel. As long as the voltage is vaguely constant the particles will move at a fixed speed through it depending on their charge. Allow them to move part way through the gel and remove the voltage and you can work out their charge from their position. Noise on the voltage doesn't make much difference and the gel isn't exactly going to breakdown from the spikes.
Be very careful with electrophoresis equipment. Many labs still use ethidium bromide to visualise DNA and stuff. EB is very nasty carcinogenic agent.
Great stuff… I love reverse engineering some mystery hunk of junk. Keeps you sharp! and you usually learn something new. Often, now NOT to do something ๐