Hi I've got a real interesting bit of kit for you today that you don't normally see. It's a bit of special purpose test equipment from Keithley We love Keithley gear here on the EEV blog and I Scored this puppy and to check it out. it's a fairly recent calibration and it was sold as it just powers up. So I Have yet to test the thing, but it's the Keithley 2302 battery simulator.

and what this thing does is that it, basically, as the name implies, simulates a battery any type of battery that you want for use in product design because what it does is actually simulates the ESR the equivalent series resistance, the ionic resistance. I've done many videos on this. I'll linking all my battery video playlists which is quite extensive down below and at the end of the video in the outro as well. and it can simulate that equivalent series resistance output of the battery.

So this is really important in product design, especially for products that require high pulse currents and things like that. So things like you know, mobile phones that will draw big gulps of current when they're transmitting and stuff like that the ESR of the better it makes a huge difference and the chemistry of the battery. There are some products that simply will not work if you power them from a low ESR Essentially a zero output resistance power supply because it's not simulating the battery, the energy can't get dumped back into the battery and stuff like that. This is designed to simulate this and it's a real expensive special-purpose bit of kit it.

I Think it retails for about five thousand dollars or something like that. US dollars It's still a current model. There are different models in the series that have this is only the single channel one. I Think there's like dual channel and other particular models available, but not only can it simulate that ESR it can measure it as well.

so it can actually measure the power consumption of your product under test the current and the voltage. and it can also measure our pulse current as well. So all the stuff people try to cobble together with say a micro current in an oscilloscope and then you integrate it on the oscilloscope and do stuff like that. This can do it or or built-in So really specialized.

Been a kid and it really is quite a big beast like this, but it doesn't weigh much at all. is hardly anything in it, so expect it like there's no big iron core transformer. So I expect like a switch mode supply or something like that. If you have a look on the back, we've got you standard our Phoenix Contact connectors here for your output, two pins for the positive, two pins for the negative, so that simulates the battery and then of course your sense line.

You've got to hook up so it compensates for all of your cable loss because the last thing you want to do is simulate the series resistance of the battery and you introduce all of these series resistance into the cable. That's just ridiculous. And it's got a separate DVM ability and I think it's like I think it's probably just like a fixed 30 volt DC range which is more than good enough. You know it might be like a five and a half digit meter or six knife digit or something like that.

DVM built-in It's got some relay output so this is designed for a you know, production test automation stuff like that. Yeah, obligatory I Triple E for WH GPIB and a remote display thing so that you can stick this in a rack and have the display somewhere which is more out user operator accessible. You where say you were, say say you know we say you're on the eevblog don't turn it on, take it apart. Yeah, all right.

so let's slide this puppy open I Love this and you might be thinking well, why don't you just use a battery in your product design What do you have to buy? Pay five thousand dollars to get one of these to simulate a battery? Well, you can. But yeah, it's just a lot of around to use a real battery and you can't just you know, have like an almost dead battery. You can't just dial it up and say look at, show me what happens to my product when the battery's almost dead and things like that. This allows you to control it.

Get a plot, you know, get all sorts of characteristic performance curves and stuff like that as your battery dies, the performance of the thing. So yeah, and essential being a kit. Really, if you're serious about your product design and we're in like Flynn I like that? Oh geez, look at our a huge board on the top. that's all your power MOSFETs under.

And they've got their little condoms on for our protection so they look like just heat-shrink You might think, why would you put heat-shrink in there when you should be puttin' thermal paste. These would be thermally conductive little tube things. so they just whack them in there. and they're not quite as good as a thermal grease, but they're pretty close.

All right, let's take a look. One of the first things you'll notice is that up here we've got the different model options. The 23:02 which we've got here and looks like shares the same board with the 2306 model, and clearly we have room for a second channel here. This top board is obviously the measurement channel, and we we have room for the extra Phoenix Contact connector there.

plus basically a duplicate of the measurement circuitry here. It looks like it's a bit less circuitry there, so that maybe got some more stuff tacked on here. But yeah, oh no no, it looks like they have it down here. You can see the traces snaking down like this down to this unpopulated area down here.

so hey, that's a bit. That's a bit of a pain in the butt from that like that's probably the PCB designer. When are you've only given me this room? I Need this amount of layout here for all this and you want me to duplicate the channel? I Don't have enough room? You haven't made the case wide enough. so I'm gonna have to split this down here and have the extra circuitry down here.

Bummer, But that might make a bit of sense given that the extra MOSFETs for the other channel are down here. They're not populated of course. and then this is the circuitry for the mosfet that controls all the ESR Of course, that's how you would get a controllable battery. ESR is by using the on resistance of the mic controlled on resistance of the MOSFETs to simulate the ESR.

But yeah, why you wouldn't? from our first layout: PCB Point of view: like whack this in the middle Now look the heatsink in the middle, have the thing, the power transistors either side and then have all the other circuitry going down like this and then this tacked over here. We've got an unpopulated one over here, but like yeah. I Don't know. like did they suddenly design like like at the last minute? or we need a two channel version or let's just squeeze it in I Don't know.

I As a layout person, I wouldn't have put that there. put it right in the middle. Symmetry: I Like symmetry. So yeah, it looks like we've got some power supply stuff happening hand around here.

we'll have a quick look. Looks like we got some digital stuff happening in here. Maybe we've got a big big ass oscillator? or maybe that's the ADC or something like that. Maybe that's the DVM front and we'll take a closer look.

Obviously there's another unpopulated our power supply one here for the second channel as a display PCB down the front. there nothing really doing this. them. you know processor that controls it all.

There's a separate display PCB in there for the vacuum fluorescent display, but you know, apart from that anyway, all the top stuff is the measurement, acquisition and load simulation aboard. the bottom board will be the power supply. Here's here's our power coming in from the bottom board. and also a bit of digital, so there might be some processing.

Yeah, all the audio processing is going to be happening down on the bottom board because there goes your display down on the bottom board, not the top one. You'll notice this little crimped lug in there that's actually a thermocouple. They've got two wires going off there, - a little thermocouple amplifier over there. They're measuring the temperature of the heatsink.

Very nice. But anyway, we're going to have our four MOSFETs there. so that would be four MOSFETs there. They might be paralleling up the pair's or whatever I Don't know.

We'd have to have a look and we'd have to take it off to have a look at what parts they are. But yeah, just big beefy, you know, low voltage in channel MOSFETs and that board just popped out beautifully. There's basically no active stuff on the bottom. it's all just passive decoupling in another.

you know, some resistors and whatnot. and we're down into the power supply. and this has got a few interesting aspects to it and it confused me at first glance. I Didn't know what was happening.

Now look at a nice protective plastic cover here. so that's just what broke it I broke it. look. Anyway, that just comes off.

We've got our main processor down in there. We can have a look at it, but you know, Whoop-dee-doo We've got our mains coming in here. We've got a top-quality mains input filter here. Spared no expense at all.

Really spectacular. Spared no expense spared. No expense spared. No expense.

Oh, look at the beautiful earth down here with a shake proof. Wash it right down on this chassis and crimped properly. Oh, it's just yeah. crotch moistening stuff.

Speaking of which, look at the rod in there that goes all the way back. you know I'm a big rod fanboy and but this is interesting down here. I Mean, look, this is the digital section right down here. Look at how close that is to the mains input switch here.

it's just like I'm sure they've left their adequate clearance and everything else. It just looks so close it's like mum, what are you doing Now If we follow the money here, obviously. Look, this comes in. go through our main switch, then where does it go? This is the output of our power supply.

Look like, where does it go? Here's our mains supply, obviously. Look, we've got our big mains choked up here, so let's flip it around, shall we and have a look what's going on here. Here's our input mains choke that they've got there and our If I choke. They've got some input protection happening there, so that's all good.

But once again, look at this curiously. like +5 volts D which would be digital plus 5 volts digital. They've got a common mode choke here, so it looks like 6.5 volts in five volts out. They've just got a linear rig happening there, but look how close that is to the other main surge circuitry.

Once again, they've probably had added adequate clearance according to the you know, this, the creepy, just standard, and everything else. but it's just. it's just amazingly close and it's obviously not coming through here. I Thought at first it might be some, you know, capacitive divider or something powering that.

but no, it's all just part of powering the digital stuff over on this side of the heatsink here. So obviously what they're doing is routing the mains from the power switch probably under the back of the board, the back side of the board and popping up into here because you can see some dark traces on the bottom through the bottom of the board there. But that's it's my layout. I Mean somebody just wasn't thing keen about the layout and just hasn't adequately separated.

I Like from a technical point of view, it's fine I think but it's just rather puzzling why they did it like that. Strange. Anyway, that power supply is no doubt going to be absolutely first-class Oh I might be able to take it out and take the cage off and have a squeeze, but not hugely interested. Um, it looks like it's a fixed plus 12 volt DC output here.

So they're just using the Joule Why is there for extra current handling capacity? So it looks like 12 volts is powering everything. And then of course that goes up to here, which is our main power supply generating the different rails. Obviously they've got a second one here for the two channel model and they're absolutely identical. You can be able to see the differences, but it basically generates minus 15 volts and plus 15 volts.

B This one's labeled A and this one's labeled B. So minus 15 plus 15 plus 25 and minus five. Over here down here, it's just beautiful. Gotta HRC fuse down here.

Sprite Capacitors niche Econ capacitors on the output over here Thing of beauty? No. I Don't know how much that puppy would have cost, but it's worth it. Spared no expense in this thing. It doesn't matter when you're building a $5,000 instrument, you don't worry about cost, it just it essentially doesn't matter The instrument.

you know, the instrument bomb. Cost is what it costs. Oops I forgot. It also generates +5 as well.

So plus -5 plus minus 15 and 25. So why have they actually rolled their own here instead of typically putting that into the mains power supply over here? Well, they'd be doing that for. Probably that they can control the efficiency better in this thing. They can control the pulse current capability and they just want it.

This is the one that you have to engineer very, very nicely because you want you know, an incredibly low output resistance and it needs to be engineer really a lot. Nice right? This is a precision battery simulator so you don't just want it. You know, left up to some third-party supplier is designing your mains power supply over here to do that. Yeah, bugger that.

Now look at this. This seems to be a after fall thought look we've got now. this is interesting. Check out this wiring here.

It's all neat. They've cable-tied it, they've heat shrunk it everything else. But it seems to be an afterthought because look, they've sold it in The wires down to like a surface. Is that a resistor down there? Yeah.

I Want one too? Is it? or is that that one? I Don't know, but yeah, like they've They've punched in that and they're taking that over to the mains power supply. Why? I Have no clue why that is puzzling. Of course they didn't muck around on the heat sinking on these two. MOSFETs Here did they look? It's just go in the full length.

they're nice and of course the thermals are very good on this. As you'd expect, the air is sucked in through the vent holes on the front here. so it comes from like, you know, the clean air in the lab at the you know, because this thing had probably be rack-mounted for example. So it's sucking the air in the front and then also goes over the top of the board like this and then over through the thin fin heatsink which has got the channel out and the fins going in that direction.

Nice. and then it blows it out the back. So yep, thoroughly. it's very nice and the Motorola fanboys go wild.

And the National Instruments ones. - you've gotta have the National Instruments chipset for the GPIB Chien Yuen Nice fake rubbish Oh Dings Ilambda thank you very much. Certainly spared no expense on the third-party provider of the power supply. Look at that.

Oh, it's just beautiful. Okay I love the angled a heatsink on the back. that yeah, they've got the odd cap stuck between the heatsink, but they know what they're doing. This is Lambda for goodness sake.

Anyway, that's beautiful, that's you know. I wonder how much that would have cost them - you know they weren't penny-pinching when they got them to design that mains power supply. Gorgeous and nichy con Main: DC filter cap of course in the output ones Nippon Chemi-con Thank you very much. Best you can get.

All right, let's go and have a look at the more detailed circuitry here. Let's just zoom straight into this section. Over here, this little isolated section with the big crystal oscillator. See what's what? Alright, this looks for all the world.

Keithley Custom moldy slope integrator like you see in the Keithley high-end DBMS Except this one's probably not going to be as good. It just doesn't have the same huge requirement. but you can see the PLCC package up there. It's some custom part number for Keithley so it's probably like a custom embedded PLC the PLD type array which is implementing their moldy slope integrator circuitry and you can see the voltage reference down in the bottom right corner as well.

Multiple packages there, but this one just uses a buried zener reference and they're just the Zener diode reference, but they could have put a more high specter unit in there so though they'll just hedging their bets there. Alright, let's jump straight in. take a look at these power. MOSFETs Here and there we go.

We're in like Flynn Look at this. I RFP 91 forties. a genuine International Rectifier nominee, so these wouldn't be fake rubbish. You can buy an ebay no siree.

Bob These are P Channel MOSFET skits, but the other two are going to be on the other side. They're going to be probably N-channel MOSFETs So these a hundred volt jobbies 23 mm it's point One One Seven Ohms not the lowest RDS on. but they don't have to because this unit is capable of going from basically zero Ohms I es are right up to one Ohm actually in ten milli Ohms steps. I Believe it is based on the specs, so how can you use a Point One One Seven Ohm one even if you parallel these two, which they're probably doing to get that zero on output impedance? Well, it depends on where it is in the feedback loop for that.

just like a regular power supply PSU can have a zero, effectively zero output impedance because it's where it is in the feedback loop. It's It's just compensating for that. So there you know this will be in there. There'll be some sort of step response that allows them to progressively increase the on resistance of this thing, just basically based on the voltage drop across it and of course the other matching.

MOSFETs Yep, they're in Channel Of course these are Vishay ones ir F, P2, 40s and once again, not particularly low on resistance. You know, not the in the industry, but they're nice big beefy power transistors so you can't blow them up. And once again, I Believe they'd be like operating these in parallel to get extra power handling capability. Well, there you have it.

they're not in parallel. If you take a look at the bottom side of the board, you can actually see the arrangement. or at least you can see the bottom side arrangement there. They're clearly in series.

Both the P-channel ones are in series: a source connects to drain or drain connects to source on one of the PD on the P channel pair, and same for the N channel pair as well. Curious this is. and then you can see that they directly connected into those Khattak to-220 package power resistors there. And let's go into near the inputs over here and we'll find that along with the MOSFETs.

Of course we've got to have the current sensing resistors for this thing. Very, very important. So expect to find some pretty schmick ones down here. Let's have a look.

Doesn't get any more than this. Spared no expense. All right. look at this.

It's a Isabel and hood I Can't pronounce that. it's German These are German for terminal resistors like a 10 watt job. So they've got the sense terminals built in and the low ppm. everything else and manufactured by nude virgins right behind that beautiful one.

We've got some CAD X Look at that for Kanak ones. They make good current sense resistors as well. Point 2 ohms each. My guess is that they're going to with that oddball value.

They have two of those in parallel so there are only 1% jobbies. but it doesn't matter about the accuracy. As I explained many times before, it's all about how many ppm they are. and there's one little current sense resistor left.

that's that Brown looking day or one their day. I'll make some great resistors as well, so no expense spared. I'm at this hundred ohm Joby Been a hundred ohms that would be for the five milliamp current range whereas the other ones at Beefy your 5 amp range and we've also got some input protection. You've got to have that.

so reverse protection with the big diode leader axial diode there and also this Adi 5 amp bridge rectifier as well. and we've got a couple of power hidden away there. There are IRF Zed 44 for those playing along at home. a lo a RDS on job.

They're probably switching in their 5 milli ampere range or something like that, but yeah, there's a couple of extra MOSFETs here and there. And of course, the rest of the supporting circuitry is going to have a lot of Schmick Op amps like that Dip package there. It's prominent why they use Dip. Well, it's an Lt1 double to 300 big current feedback amplifier only available in Dip package for the wind.

None of that surface mount rubbish. And here's something that I didn't expect to see. look at this other Dip packages a micro or now owned by Microchip the Mick 5021 This is a high side current sense MOSFET a driver so specifically designed for current sense applications like this. but it's evolved.

you know I wouldn't have expected it in this type of application is voltage is 12 to 30 volts operation. it just it doesn't seem to fit and look at this physical location. It's a way away from the current, central and current sense resistors and the MOSFETs over there. So I'd be curious to trace that run out and get a schematic for this.

I Can only imagine that they're using that. maybe for a gross current overload. you know, maximum over current protection thing for the MOSFETs or something like that perhaps. And another part I Didn't really expect to see a video up Ampere Analog Devices ad8 1/8 You can see three of those in there with the prominent white silk screen there.

So yeah, maybe it was just the designers. you know, Jelly Bean, Go to Op Amp for some reason And hello old friend Down in the bottom left-hand corner there the device is 8620 instrumentation app I've used that one lot and it's a nice little instrumentation. A bit low cost in quote marks, but yeah, for a precision instrumentation amp, it is I Surprises for finding an amp in here. And of course, as always, in a design like this, you're gonna find a smattering of ultra precision operational amplifiers.

We got the O P 1 7 7 s scattered in various locations across this. So yeah, no surprise is you're going to have a bunch of those and you're going to have a bunch of yeah, there's LM three, nine threes and there's other, you know, jellybean type stuff. So I won't look too much further and I'll just leave you with a yeah choice. Smattering of high-res photos here for the other power supply section.

And as always, the high-res photos are down on Eevblog Comet linked down below if you want to check them out yourself. And because I actually use a macro camera as I'm tearing something down I Use a macro lens camera, take some photos and I always upload them on Eevblog comm. Anyway, fun look at all this. It's wonderful.

You can smell it, smell the quality. Oh yeah, so let's just power this puppy up. It's like a jet engine taking off. Jeez, that fan is not going to need temperature control on it.

The vacuum fluorescent display is a bit low. There you go. This puppy's a little bit old, but you can see that we've got our voltage and our current. They're displayed and then we can choose display type, actual voltage and current.

So there you go, and we've got our DVM input separately. In this, the pulse current I was telling you about. If you want to measure the pulse current of your product and you can do long integration, you can set the integration time over or that it actually calculates the current and stuff like that. So it's very simple.

we can set the GPIB address, view, or change. We can set our current range. so let's go in there and 5 amps auto 5 milliamps so we can really get right down if you do it doing low power products and we'll definitely have to do a video on this showing you you know, like actually measuring the current and maybe compare it to a integrated oscilloscope value perhaps I am. Anyway, so it's only got the two ranges 5 milliamps and 5 amps.

so it would be a real bugger if your product drew 10 milliamps for example pulse current and then you had to go to the 5 amp range. You lose all that resolution. Unfortunately, there's just nothing you can do about it. It's yeah, it's really.

you know you really need a a current meter like this with many many different ranges. so number of power line cycles that it takes into account average readings. Savoring a corset up pair on set up, calibrate fold each protect current limit mode so you can protect your product under test Revision number: VFD Brightness: Oh yeah, thank you very much. Full already on full dammit Anyway, unfortunately we won't see anything if we operate this because it'll just show random voltages because we haven't connected anything.

The sense line isn't hooked up so it's just high impedance, floating, etc etc. But hopefully we can play around then with this isn't a future video, we're gonna have some fun. So anyway, there you go. Hope you found that interesting.

That's a teardown of the Keithley 23:02 battery simulator. Definitely not something that you get to see every day anyway. high rest hair down photos down below on Eevblog comm LinkedIn And also check out somewhere over here at the end of the video will be all my battery playlist videos. I've got a whole ton of them and if you liked the video, please give it a big thumbs up.

Catch you next time you.