A quick followup to Bank Note Validator Teardown, to clarify the LEDs used.
http://www.youtube.com/watch?v=qOzP5Vl-CIk
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/
http://www.youtube.com/watch?v=qOzP5Vl-CIk
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. Just a quick follow up on this note validator that I did in the last video. If you haven't seen it. This won't make any sense whatsoever.
and the link it will be linked right here. So just click here and watch it. and uh, a couple of people uh, pointed out that um, the LEDs used in here which I um, sort of assumed were uh, maybe infrared uh leads are actually uh, just a B. It looks like a by color lead and sure enough, I've powered the thing up here and of course I should have checked the pins on the back and sure enough, they do have three pins.
They are a jewel die uh lead so they're at least red. One of the colors is red here. so I've powered it up. So there you go.
even though the Um sensor the photo diode in here is capable of going into the infrared range as I said from 1,400 400 nanometers up to 1100 nanom. so it certainly covers that infrared. uh, Spectrum But clearly they're just using red leads here for the three of them. and I've powered the thing up.
and sure enough, I haven't seen any other color yet. So I can turn the power off there and we can watch it start up. And of course, it's probably not going to pass its power on self test because these leads aren't connected. Uh, over to here so there's no you know feedback from these LEDs to the uh photo diodes in here and I've put down some white paper just to maybe try and get some feedback cuz that op that's an optical path there with the Uh light guide in there to feedback that signal.
But anyway, we can. uh Power that up and uh, there we go. No, doesn't. Oh, there we go I did get them to turn off, so there you go.
but I haven't been able to get it to actually detect a note or anything like that, so clearly it hasn't Um, you know, it probably hasn't pass those S power on tests and things like that. but um, yeah, I might sort of whack that board back onto there so the sensors line up and uh, see if we can still see the colors in there. And also, as many people, uh, pointed out the $5 note I had I didn't see it down there. there it is series 2006 so apparently they did drastically, uh, redesign the note with uh, some people have mentioned better security features in them.
So it's probably no surprise that a 2002 vintage firmware in this note acceptor I've got here won't accept these $5 bills. Now just as a first guess, you might think it's a red green by color lead and uh, well, is it I've hooked up an external uh coin cell battery via resistor here just to give it some current limiting. There's our red and let's have a look. We get nothing out of the other one at all, even though there's 8 milliamps flowing through that other one.
I don't see anything. What? soever and if it was infrared I Thought it should show up on the camera so maybe turn the lights down a bit. In fact, if we turn the lights down a lot. Bingo There you go.
I Can turn that off and on. We can see that that is obviously on the uh, low wavelength side because the camera sensor um, or video camera sensors can easily pick up uh, infrared or not easily. They're not very efficient, but they certainly can. and uh, we can see that that is infrared. but I can't see that at all. Of course, because my eye isn't uh uh, tuned to infrared, but no, it looks like it is. so it's combined red and infrared lead neat. and I'm assuming that the other one on the other board that matches that sensor is going to be uh, identical.
but uh, we do also want to check these ones out here. Well, these are the photo diodes of course, but the leads are on the other board so we'll uh, check those out, power them up and see if they're the same. They are also a bolor lead. They do have uh, three pin on them and they look like an identical ledge.
You can see the Uh three pins down inside there and if you're really got this under a microscope, you'd see a bond wire going out there to each leg. and definitely Jewel Diey. But is it the same wave length? Well, let's find out. or at least close to I mean these could be very specialized.
Uh leads of course, are specifically uh ordered from the manufacturer for a specific wavelength we just don't know and that looks like just the same red to me. Not a problem at all. and I expect Yeah, it is an identical infrared, but as I said, could potentially be slightly different uh, specific wavelengths ordered from the manufacturer. In terms of uh, the infrared, the red looks the same.
but yep, there we go. So they're almost certainly identical leads. So um, basically four identical dual color red and infrared leads at four specific points on the note. And now these two outer leads.
uh with the purple color, which I sort of assumed were um, ultraviolet UV leads. They're also showing up on my camera. no problems at all. Look at that.
I've got 8 milliamps flowing through that sucker and uh, we can certainly see it. So I wasn't aware that uh, video camera video camera sensors could go into the Uh UV range so that's either uh, infrared or this camera can actually see some UV there. Nice. Sure enough, I did a quick check and yes, video cameras can see into the ultraviolet as well.
So uh, yeah, Clearly, of course you know based on the uh purple color of that lead, you know, really, you know, high-end uh, purple. But essentially going into the UV considering that I know for a fact that they use UV detection in these note validators, then it's obviously a UV lead. So we have ourselves four dual color uh, infrared and red leads and two ultraviolet ones on the outside edge of the note. When it comes down to it, there's not a huge amount of security that they're actually doing on this thing.
I Think they? they're probably just doing it more to just validate that you know, just to check that the note is the proper currency. So they're probably, you know, in the scheme of things, not hugely hard to, uh, fool. That's why they have the limit, uh, switches on the sides, so they limit. You know what currency you can, uh, actually detect up to? you know, um, it's up to the vendor. This one can only do up to $20 I Mean, you know, why would you even bother checking the security on a $1 bill for example? It's just, you know, it's just not worth it really in the scheme of things. So I think it's uh, you know this is a basic an old, really, you know, probably a lowc cost model. Not entirely sure. I'm just, you know it's probably not even Advanced for its day.
So really, it's you know, probably one of those ones where yeah, ins say a vending machine. You just, you know no one's going to bother putting a you know, a counterfeit bill in to get you know, a couple of packets of chips or some chocolate or something like that, or an icy cold can of Coke So what I might do is try and actually get myself a modern one, especially like an Australian one. For example, that uh, can detect our polymer currencies and and uh, why I think we'll see you know a dramatic uh difference between the very quite simple uh, you know, sort of rudimentary level technology used in this one and uh, and more advanced uh, modern ones. and also some people wanted to like you know, get me to get the ROM dump and stuff like that and see if we can see the note images and you know the well, the data that uh, you know, knows at what points, but you probably have to do some serious disassembly like like that.
it's not going to have the you know, it's just not going to be obvious when you dump the data. So I don't think it's worth the effort. And it's interesting that although these four leads here are the same, essentially the same dual color Ir and red, the photo diodes are different. These two inner Tracks Of Course angled 45 much larger uh sensor photo D much larger die inside there to sense it whereas these other one these two outer ones are a smaller photo diode and you know, presumably a different type even though essentially operating at the same uh, wavelength.
So obviously these two center lines are doing something subtly different to to these two outer ones here. and some people have asked if I could maybe you know, probe the uh, sensor signals, photo diodes, and stuff like that as the notes passing through. And yeah, you could do that and possibly correlate. Uh, you know you'd have to know the note, the well, the position of the note.
It goes through pretty quick. You'd have to know the position of the note and correlate that timing with the timing on the sensor. but unfortunately, once it's all assembled in place like this, you can't access those sensor boards. All you've got is, um, the outputs of the boards.
uh, presumably from, you know, the Op an B which is probably good enough. uh, stuff like that. but unfortunately I put the thing back together and um, what, um, something? Something is, uh, wrong with this thing. it's just not detecting this at all. Anyway, I don't think you'd learn a huge amount by doing that. These things are pretty darn basic. So yeah, maybe I could have a probe around if I can get it going again, but it's not going to happen at the moment. Damn, don't know why.
the camera sensor can detect infrared light, which humans cannot see, yet the blinking IR LED can be seen as visible light on the camera's LCD screen. why? how does an invisible wavelength suddenly become visible when viewed through a digital camera?
I can see IR leds light up with my naked eyes
I think these things work by scanning the reflections off of a very narrow portion/line of/on the bill, and coming up with a stream of 1's and 0's that come from that scan. No telling how many scans it's doing per inch, but it is likely going into some register bank somewhere, and then it's simply then compared to some lookup table of known good scan results. I doubt that it is scanning in too high of a resolution, because such scans could render real money unusable due to normal wear and tear, tiny specks, etc. Though I suspect they could also program a certain amount of fault tolerance if they wanted. If I am correct, then the eeprom would likely contain the table of known good bill results, and their corresponding note values — such that it has the equivalent of multiple acceptable scan results for 20's, 10's, 5's, and 1's. And as new bills come out, they just need to add the new ones to the scan tables on the eeprom. Someone could come to the device, pop out the old eeprom with proper extraction tool, and pop in a newly updated one. Take the old one back to the shop and re-image it with a newer updated image.
There we go, can clean money that way a dollar at a time then open a confection shop.
love the engliash launguge. took water bill feeed one those mahcien didn't except it..damn it..
Obey
Pretty sure everyone in America has seen one of this model before, it's the most widely used bill acceptor I've seen.
hi guys, these may well work on opacity too as a form of validation
Why was it so easy (relatively, at least) for them to make infrared and UV LEDS before they could even make visible blue LEDs?
unit looks older..use an older bill..or try a 1$ bill
I think the weird optical sensor is for a fuse, when the unit is disassembled it's deactivated.
They were meant to inexpensive last 3-5 years and be replaced. They worked in bar top game machines , juke boxes , arcade type games ,vending machines ect. Cost was about $100 new. Never ment to be anything special. There are other models that were much more money , and much more complex. I used to work for a company that operated these type of machines in bars , bowling allies , and rest stops.
Bloody hell Dave you did break it :-0
I've never seen a vending machine accept $20 bills. I'm betting that this is from a change machine or an arcade token machine.
Looks to me as if the 2 sensor boards are identical, BUT MIRRORED,… this is so they will recognise the note either way over…………
Cheers, G4ZWI
those bicolor LEDs. The ones read the other ones got to be UV
if I do believe
I can tell you all with absolute certainty there are no UV leds in that acceptor, only IR and visible red.
I worked in the currency validation business for 15 years. The bill acceptor being reviewed was not made by the company i worked for, it is a competetors product. However they all work pretty much the same. As someone already pointed out it is a bit unfair to review its construction by todays standards because it was designed over 20 years ago. Let me say I have much respect for the man doing the review, however he is way off with his ideas about how it actually functions. While it was not the most secure acceptor on the market it is much more secure than he thinks it is.
Hi Dave, I think you could still use an American 1 dollar bill. It still looks the same as it always has.
Would've been cool if you had taken out the Flash rom, read it, and poked around in it 🙂
Doesn't acceptors have something like a dead-man's switch that trips when the vending machine door is opened?
The dark leds are the IR leds i think, and duo leds are red and uv.
Have lots of dark ir leds, and i never see dark uv leds. UV (that I have seen) are always, white, clear, very light blue, …
All 3 color sensors on cameras are sensitive to IR to different extents. They got filters to keep most IR from reaching them, but the filters always let a bit thru. Different cameras see IR as different colors, 'cause their filters filter different amounts of IR and their sensors might have different sensitivities.
Depending on what sensor and filter the camera got, IR can look pretty much any color. The color sensors in most cameras all are quite sensitive to IR; what color the camera "sees" IR depends on the ratio of their sensitivities.
Usually color cameras got a filter to keep infrared from hitting the sensor with much strength; in a sunny day colors look all screwy if you don't have a filter.
You do have to check security $1 bills. Otherwise, people could pass in just a piece of paper, or a xerox copy of a $1 bill.
Generally if the plastic window is missing then the note is invalidated regardless of how much of the rest of the note you have.
I haven't dug deep into electronics for years after getting into computers, but just watching your videos is perking up my interest again big time. I want that new Rigol 1000Z series scope to physically get into it again but I know the wife would kill me.