Inside the Stanford Research SR430 Multi Channel Scaler Photon Counter
Unboxing video here: http://www.youtube.com/watch?v=bfp0QnQujvs
Datasheets:
http://www.physics.rutgers.edu/ugrad/389/muon/SR430m.pdf
http://www.elektronik.ropla.eu/pdf/stock/hit/hd63484.pdf
http://cds.linear.com/docs/en/datasheet/100737fbs.pdf
http://www.analog.com/static/imported-files/data_sheets/AD96685_96687.pdf
http://www.onsemi.com/pub_link/Collateral/MC10H107-D.PDF
http://www.datasheetcatalog.org/datasheet/SGSThomsonMicroelectronics/mXqzrvy.pdf
http://www.physics.brocku.ca/~edik/gpib/TMS9914.pdf
http://www.freescale.com/files/microcontrollers/doc/data_sheet/MC146818.pdf
http://www.analog.com/static/imported-files/data_sheets/AD586.pdf
http://en.wikipedia.org/wiki/Emitter-coupled_logic
Forum Topic: http://www.eevblog.com/forum/blog/eevblog-452-stanford-research-sr430-teardown'>http://www.eevblog.com/forum/blog/eevblog-452-stanford-research-sr430-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 Teardown! Tuesday Yes, we're straight onto the Stanford research Photon counter. no Ming around. If you haven't seen the previous uh video where I unbox this thing, click here and you'll be able to see that. um where we unboxed it? Um, scored this thing for 50 bucks on eBay took it out of the box, powered it up eh.

had a quick play around with it, but now we're going to have a look inside. Could be interesting or maybe not I expect it'll be all through hole technology as very common with Stamford research. um, stuff of this, uh, vintage. even the new stuff really? um I Think they're a bit in the Dark Ages when it comes to that sort of thing, but uh uh, nobody could find a um schematic for this thing.

or the full uh service manual. Um, the manual is downloadable which has a you know circuit description and uh, well, sort of a high level circuit description kind of thing. but um, unfortunately we couldn't get the schematics. but yes, I expect a lot of um through hole dip uh, processing stuff like that, um counters and uh, maybe some um.

amplifier front end and uh, that's probably about it. but you never know. Always find something interesting to talk about in tear. Downs So let's take a look at the Stanford research Sr 430 Photon counter or multi-channel scaler.

whatever you know what we say here on the EV blog. Don't turn it on, take it apart. and by the way, uh, some people did come up with some uh, good ideas on what I can uh, do to play around with this thing I like the idea of um, looking at the uh decay of the output of a uh lead I think that one could be, uh, quite interesting. So if you got any um, uh, suggestions for some uh, suitable, uh, you know, photo fast photo diodes that might be suitable for something like that, then uh, please leave it in the comments like a particular model number or maybe even some secondhand links to some secondhand stuff on eBay and uh, we got four screws on the side of this thing.

and by the way, um, a few people have asked about the uh audio. um I have switched back to using the internal mic in my uh Canon Hfg10 camera most of the time because I find it's reasonably consistent when I'm standing behind the camera, which is what I'm doing most of the time. um, it, uh, it tends to work quite well, but now when I'm like half a meter away from it, you might be able to hear that my voice is, uh, dropped. Of course, you know it's attenuated a bit and maybe a bit more echo eous on further, but when I'm standing behind a mic like this, it actually works better than the uh lapel mic.

I think because the lapel mics aren't uh, Perfect by the way, they um, if you move your head from uh, side to side, then um, you get, you know, a similar sort of uh dip effect in the audio. Anyway, that's a little aside. Here we go. Um, that yeah is just going to pop off.

it's unpowered Tada There we go. Yeah, through hole. Uh, couple of through hole boards and four. Looks like there's four boards total in this thing.
Yep, all through whole technology as I predicted power supply, some logic, some front end stuff, and some more processing logic. and of course the CRT stuff around here. So let's take a look if we have a look at the CRT there. it's actually, um, looks like it's very close to the board down at the bottom there, which I don't like.

And the interesting thing is, it looks like it's like collapsed. but it's not. It's actually bent back because the front panel on this thing is actually um, sloped. so everything sloped.

The uh CRT is angled downwards towards the board and the floppy drive on the front's angled uh down. so that whole front panel is uh angled. We have ourselves a date here April 1990 so that uh certainly dates this thing to well over 20 years old. Nothing much to talk about here at all.

Just there. There's the Yol There's the EHT There's the driver board H Nothing terribly exciting there at all, folks. It's got the Uh adjustments by the way. We can, actually, um, adjusting the Uh software on this thing.

You can adjust the uh position of the Uh window on the Uh screen so you can actually do that in Um software now. I'm not sure if they're actually doing that on the this CRT board as part of the analog um circuitry or whether or not um, they're actually uh, you know, just uh, doing that by um, uh, not not using the full screen real estate? um I'm not sure, but possibly they are doing it. um, on the board because there's a ribbon Cable in there. If you have a have a look down in there, you can see this ribbon cable coming over to that there.

and, uh, that's a fairly multi-way uh ribbon cable. So um, it' be interesting to see what other connections we've got going to that uh board. That that could be it. So maybe they could be um, sending some sort of, you know, um, signal over to that board.

Some sort of digital uh signal, you know, and it could be, um, then offsetting, uh, the actual position on the screen, physical position on the screen I Don't know. Um, so let's take these uh Boards out. but that's yeah, there's nothing really, nothing else really to, uh, talk about on the CRT board. It's not terribly exciting and this top board here it tells you it's a logic B I'll zoom in on that in a second.

but I like how they've um, put some thought into this that it's going to be um, serviceable and testable from the bottom of the board. they knew they were going to mount it in upside down. So what they've done is, they've put all the silk screen on the bottom of the board. I Mean, ordinarily, you wouldn't do that if this was mounted on the bottom of the board.

Um, as you'll see, we'll probably take the boards out. Uh, from down in. uh, this bottom board down in here and you'll find the silk screen won't be on the uh back of the board. Of course, because you're never going to see it.
Um, but because you can access and test this thing in situe like this to be able to put all the uh, there's all the component numbers U 1000 and you know, uh, something. by the way. Um, yeah, like there's U13 there. It doesn't mean that they're uh, sequential and they've actually got you know, 1,3 chips in this thing.

Often when you start a new board like this, you'll start from um, you know, like a th000 or something. So the first chip on here will be u1000 for example. So um, although I do see here like uh I don't know u304 u1305. So I don't think there's three.

There's so we go from 1,2 to 1305 there. So I don't think there's a couple hundred uh chips on this board clearly. So um, you know, uh, they don't Oh, there's U 806 over there resistor 1,37 so they you know often those was just number these things in blocks those component designators will will be will usually um either start per board. So this board might have say started from a th000 for example, it hasn't because there's 800 down there.

But if it started from a th000 and there's 20 boards on here, you go from U 1,000 to U 1,20 Or often, as looks like might be the case here, they will, uh, set the component designators um, autonumber them based on functional sections of the circuit. So if you have a look at the schematic, this might be a functional C section over here and that might get its own block of Uh component designator codes. um, not only for the chips, but for the resistors and the Um and all of the components, the all of the passive stuff and everything on there. So that's just a common Uh technique.

and once again, power supply board up the top Here they've done the same. uh, they've done the same thing labeled it all on the all on the bottom of the board I Really like it. Soldering looks uh, all pretty good. Hey I like the look at this.

warning I'm going to take a look at that. Check it out. This is the Power Supply board and look. Warning: Never Service This board When this light is on.

there's an arrow and it's you can't see it I'd have to zoom down but there's a there's an LED under there. so with that LED's on, it tells you that the board is H charged up and um, you know there's power still on the power supply board. so don't go touching the damn thing because well, presumably there's uh I'm not sure if there'd be high voltage on that I have to have a look at the value of the Uh caps down there. so I'd be surprised if it was a high voltage uh board.

but uh yeah, it's just saying don't service it uh when that lights on so that's a nice little touch I mean that's you know, some part of the that's a part of the design aspect which really, you know, no the user is never going to see that it's purely designed for servicing Personnel in mind and that's why I was talking about the silk screen access and stuff on the bottom. the um, people who actually uh Design This and by the way, I will uh in fact I might show you now I might power it up and show you who actually designed this thing cuz I found it in the menu after I uh played around after the previous video. Aha, there you have it folks. Look, the LED has turned on there and I haven't switched the unit on I've just plugged the mains in the back so clearly um, this thing doesn't have a real clunk uh power switch down the bottom.
it's um, got a uh, it's got soft power so that board is uh, powered up when you plug the mains in. But anyway, let's power it on and we can find out who actually developed this thing. And here you go folks. We actually found the developers and it looks like there's only four of them: Hardware Designed by Joe Wang Software by Andrew Melon uh Power by Kurt Leman and Mechanicals by Jeff uh Gamesley Gamle Gami sorry if I'm pronouncing that correctly.

There you go. They've put their names on here so it looks like only four people have worked on the design of this thing and uh, well, that's you know. I'm rather, uh, surprised at that. I would have expected.

uh, sort of. you know, more people to have worked on this thing. especially like in terms of uh, you know, look poor Andrew's done all the software by himself. That's a lot of work poor.

Joe's done all the hardware by himself uh Kurt Leman uh got off easy with the uh power supply and uh, well. I don't know who designed the uh CRT and U all that sort of stuff. maybe they farm that out and uh, Jeff um, got off. uh, relatively easy with the Mechanicals cuz there's nothing, uh, fancy going on in here much at all.

but there you go. you hardly ever find that and I think it's brilliant. Good on you guys. and if any of those guys are watching and uh, possibly still working at Stanford research, you know, uh, 20, uh, 23 years later, hey leave a comment.

and when I pulled the plug on the back of this thing that lead uh turned out pretty quick, it only uh, stayed on for a second or two. So obviously they got bleeder resistors across this thing to uh, drain that charge off. And here's the power supply when we lifted out and uh I had to do the four Um screws on the back um Grill cuz that the Uh heat sink actually screwed into the back panel and I really like the design and layout of this power supply how it just flips out like that there's the main Uh Transformer on the back there and I really I Really rather like that, especially how the Um heat sink then folds down and mates up with the back panel and uh screws in. Really, rather neat.

It's uh, also sharing a bit of load on the Um heat sink with the Uh aluminium back panel as well. so using a bit of the shazzy and of course they' got the fan on one end here which sucks the air through and uh comes straight out the through the fins and straight out the back. I Rather like it and really, we're chock a blocker there with devices on both sides of these things I think we got a whole bunch of linear Regulators there. That's kind of disappointing I Love all the big um, old school uh Power Molex connectors on here.
but look at this. They beautiful modular wiring for everything with with connectors and the power switch is hardwired onto the board. What the as if the power switch would be an afterthought I Why? So that's rather deceptive. You can see that big clunking power switch down on the front panel there so you might you know you would naturally assume and the date of this thing just being, you know Stanford Research Kind of old school.

You would think that yeah, they'd just be switching the mains, but they're not. and then that switch is going up to this pokey little soft power switch connector and this is actually a um, a soft power supply. Can't believe it. And this is our main output connector here.

and we've got a whole bunch of Uh rails we got plus minus 22 volt so that's uh, like clearly for some large signal level, uh, analog stuff and uh, what else we got. Minus 2 Vols don't -2 Vols was 4 - 5.2 that's Um indicates when you see something with - 5.2 that indicates um, most likely Um Ecl emid emit coupled logic and that's uh, not surprising. They probably using some very fast Um ecl Logic on the Um main input uh board just after the front end. possibly we got -5 volts, oh -5 oh -5 U and -5 we got 3v line and line.

Not sure what's going on there. We got uh, negative reset, negative reset, reset and ref not sure and over here we've got Uh + 5, + 15, U and + 15 and this one here with the Plus plus 5 plus - uh 12. Pretty basic stuff and the reset line that goes to the main processor board under the Uh CRT down in there and the other supplies. We' got our Transformer input here.

This one goes off to the floppy drive that's a plus 5 Vols and 12 Vols for the 3 and 1/2 in floppy and uh, this one here goes with plus - 12 and ground uh goes off to the Uh CRT driver board and you can see the Uh star ground configuration. Got a huge big uh ground plane around here and then just all the various uh uh points for the various boards just tapping off the same point here because uh, the reason that they're doing that is cuz they're obviously doing some sort of um, you know, uh, power supply detection or something like that. Over with this logic here and I rather like the uh placement of these dodes here in this big Dio Bridge right in front of the fan so it's getting all that air flow directly flowing over that, not only through the heat sink, but across your power, your power semiconductor dodes there. And of course, they've kept All of the main caps as well.

Um, fairly far away from the Uh main heat sink of course. So excellent thermal design. And the main capacitors here Arrow M Um, they're actually I believe a subsidiary of Arrow Vox which uh, make huge industrial uh capacitors and uh, stuff like that. So yeah, I'm first time I've seen those in a bit of gear and there's a blast from the past mallerie and I had to Google that one and mallerie.
it looks like a part of Uh Vishe now and along here. No surprises, you know. 7915 linear, Regulators 337 adjustable uh, negative, adjustable positive, and a couple of power transistors as well just for good measure. Some tip 41s and a tip 31 over there.

and of course, all insulated from there with the Uh N washer under there and the seal pad underneath. And here's the rest of the circuitry and it's starting to make a bit more sense. Look at what we have here. an Ad 586 5V Precision reference.

That's a pretty darn Schmick uh reference. So um, really, that's um, that's obviously aha. why yeah, there's a ref output there. it is down there so that's headed off to the uh other board.

So which is a bit surprising why you'd have it on your power supply board. but still, these LF 35s here 353 here. it's not like we're using Precision resistors in there I mean you know, pretty ordinary uh, carbon film stuff. The odd one there is, um, like you know, a low Tempco metal uh film but and that Lm393 comparator there that's a zeroc cross uh line detector for the AC input Now I Actually read the Uh manual on this thing.

um, the basic, very basic circuit description it's got and it actually tells you that this 5V reference This pretty scho 5vol reference is used to get accurate output voltages from our regulators and uh, these Um Op Amps. When the soft power switch is turned off, they actually disable the regulator. So these are used to actually set a lot of the output rails precisely complete. Overkill Crazy.

So that's what these are used for. Um, the Uh LF 353 circuits here. um, control the pass transistors in here and they can actually uh, switch those rails off. Um, when you turn off the soft power button and it's all and it accurately set by that 5 Vols reference there.

Incredible. Don't get it. Anyway, here's a whole bunch of regulators on the other side as well. They got, um, some slowo blow fuses mounted on the board for each rail, which is quite nice.

Um LM 741 Classic down here. Thoughtfully, they have made these wires for this power switch long enough so that you can put the board outside and poke around in there. Nice. If we check our battery down in here, we're getting just over 2 volts on that 3vt Lithium battery It doesn't look like it's been changed I Think it's probably the original, so that one is a bit low.

So it probably enough though to still retain the uh, the contents of memory perhaps. But uh, you would certainly, uh, change this if you want want, change that If you wanted this to be a serviceable unit and given the age, of course, you would just change it as a matter of course. But um, unlike some multimeters we've looked at recently I don't I suspect that this thing wouldn't uh, lose any calibration values or something like that. Uh, if you uh, took out, just took out the uh battery.
but I stand to be corrected And look what we have here. This is interesting. We have ourselves a thermal cutout for the Transformer there mounted onto the back panel because of of course the Uh Transformers directly mounted on there. so you know it's as good as mounting on the Transformer direct.

really so overhead in protection from the Transformer Very nice. Just uh. switches off the Main's input. Now if we have a browse around our main logic board down in here, you're going to think, take a look at that beasty there that's got to be the main processor there.

it is hooked onto all the memory. there. it is hooked onto all the ROMs And if you go in and you take a look at it, it's a Hatachi HD 63484 And that folks is a CRT display controller, not the main processor. More to the point, it's an advanced CRT controller and uh, it is quite a beast.

Actually, if you have a look at it, it does up to 4,096 by 4,096 pixels. 1 bit per pixel. Mono only, but that's pretty darn impressive it. uh.

Display Control: It can do split screens three displays at once in a window Zoom 1 to 16 time zoom. It can scroll horizontal and vertical in Leed Access mode for flashless display. Super imposition. ah External synchronization very nice.

I Like it. 256 characters uh per line, 32 rasters per line, 496 rasters per screen. Beautiful works at 9.8 Meg off a single 5V Supply and drawing rate is uh 4 8 nond per pixel. oh Color Drawing: It does do color drawing and there's 38 commands.

It does Uh, there's dotline rectangle, polyline rectangle, circular ellipse, paint copy Etc So that's clearly how they're doing all those onscreen menus and putting uh, you know, borders around uh, menu buttons and stuff like that Using this Advanced CRT Controller: It's not being done in the main process. the main process is just sending the commands to um, actually put the text there. But once you've put it there, once you've sent it to the display processor, this display processor just handles it and uh, updates it all. There's the internal block diagram for it for you afficionados of old school.

CR Controller chips of course. um yeah, they don't really have uh, this sort of thing anymore. I mean you know Advanced graphics card? sure, but uh, just you know. dedicated CRT controller chips, timing processor, Mpu interface, Dma control units, drawing processor and then the CR interface which handles all the sinking.

Beautiful. And there's a system implementation diagram and aha, there's that dot shif I was uh, talking about at the start of this thing that looks like it's a um separate thing CU This is the main chip uh down in here and it of course we're going to have our main CPU we haven't found yet our main system memory direct memory Access Controller and our up to 2 Meg uh frame buffer just for the CRT uh controller. So um yeah, that dot shifter I guess is um, dedicated external uh circuitry. We might have to find that one too.
Looks like we've got ourselves some socketed mask ROMs up here. 27c 512, 512 uh K bit so that would be 64 kilobytes um each, so 256 uh K total ROM in this thing got ourselves a National Semiconductor DP 8473 going. that's just a floppy disc controller going off uh via the ribbon cable to the floppy drive down next to our battery. there.

no surprises. What's next to a battery in these things? It's a roller. Um MC 14688 uh real time clock ITC chip old school and I found some chips as old as 93. Well, the 24th week 93.

So this dates this thing at uh, just under uh, 20 years old and it's a bit hard to see all that uh logic there under the main CRT uh controller. but it's all 74 HC uh logic stuff. So that's obviously the Uh dot shifter and uh, stuff like that. they're just implementing that using the discrete Uh logic.

So um, you know, maybe they just um to shift physically shift the Uh pixels on the screen, the whole screen image from side to side. Maybe they just, uh, actually delay it by a couple of pixels. um, or started a couple of pixels earlier or uh, something like that. And that's how they shifted across the CRT And there's our do shifting register 74 HCT 299.

And there's our main processor in in 8C 186. Brilliant. Of course, you know it's probably running uh Doss or something like that. given the 3 and 1/2 in floppy it has in this thing.

Probably tells you that in the manual I haven't read it, so that kind of throws you a bit. At first glance, you think this is the main processor cuz look, it's coupled directly onto the memory and the ROMs there. but the memory and ROMs are actually uh coupled through to the main processor. Of course, the um, uh, sh.

the CRT Uh controller just interfaces with the Uh. The same address and datab bus are shared by the external Uh memory so that would have just been like a board uh layout thing. You know, they probably started. Well, let's lay down the memory like this.

Oh, kind of makes sense to put the Uh CRT controller there cuz it's got a map into the memory as well. and maybe oh the oh, we ran out of room on the side I don't know. Whack the processor over here cuz they've done the traditional uh, two layer uh board design here. all of the traces a two layer digital, um, old school board design, all of the top traces going in the vertical Direction And if you flip the board over just like this top board here, you see all the tracers going in the horizontal Direction and tucked right up under our Transformer in there.

TMS 9914 Gpib uh controller. So that all that logic and stuff is just for all the Uh interfaces on the back panel. So obviously I'm not going to go to the trouble to take that board out. We've sort of, you know, seen everything we want to see I'd have to like take out the CRT and staff don't don't really want to do that.
More interested on the Uh boards on the other side. And here's our two main ribbon cables here. Large ribbon cables going off to our main Uh process main. What? What do they call it logic board on the other side? No, they call it a oh, it's an Ecl board.

There we go. So I was right about the Ecl logic. There you go. So we got our Ecl board and then the processing board up the top so we'll be able to take out this uh, processing board I think they call it uh, what is it? No, it's a logic board.

There you go. So that's the Lo This is the processor board, processor board, down the bottom, logic board and uh e, um, input board. There it is. We' folded this uh board out.

We'll take a look at it but as you can see like there's a lot of wasted room. I mean this is, you know, uh, what? 150 mm high? this case or something. There's a lot of spare room in here. They really could have if they wanted to made the instrument like, uh, you know, half the size, stack the boards and and connected the boards using uh, vert, medical.

um PCB uh you know, standoffs or something like that and well, you know room's taken up with a CRT but H Yeah, they could have made it smaller. but when you're designing a bit of scientific um. instrument like this, you know it. Size really doesn't matter and the bigger the better the more impressive it looks.

You know no one wants to buy. You know no one wants to pay their $8,000 and get some little uh Pokey box n you want a big huge box Big CRT lots of buttons on the front and a big twiddly knob. Done it again with the permanent wire connection. solded These Bnc's here.

There's no uh ground on those at all. they're just relying on the shazzy ground and there's a single wire running off directly soldered onto the PCB Oh what the you know. Apart from that, it's quite modular. it's all connectorized and they just go and do that.

I I Don't get it. Oh, by the way, the Uh 64k of Uh video memory for the CRT controller is actually just under there, which you can't see under the two static RAM chips under there. So all of the ram up here is actually uh, dedicated to the Uh processor. It's got nothing to do with the Uh CRT controller.

So this is our TTL logic board here. and uh, what they've got is that they've got uh, two inputs coming from the Ecl um input circuitry logic board. There's a data input uh here, this, uh, smaller ribbon cable which I've disconnected and this larger uh control cable which uh would go out and uh back and forth and uh, control the Um Ecl counter board from here. so all of our data being clocked from the inputs cuz essentially all this thing is doing is it's this whole box is it's just counter.
It's a big counter, sorts um, and and time, big counter and timer, and sorts things into various bins based on when they arrived, when the pulses arrived, and stuff like that, so you know it's you. It's not terribly, uh, complicated in its operation at all. really. So now, no surprises that we've got some extra memory on here.

So it's clearly, uh, buffering. These are the buffer memories. uh for the input and it's you know, sorting? um, you know the data is coming in and it's whacking those in memory and then it's extracting that uh, later and uh, processing it or something like that. I don't think it's doing bin on here.

There's nothing sort of. You know, there's no um, Fpga or pod, uh, logic or other. um, you know, smart intelligence circuitry on here to do any binning or anything like that. So I think we've just got raw data coming here.

It's being latched and uh, clocked into some fast memory. So on our data input here, we got some 45 nond srams from Fairchild by the looks of it, 81c, uh, 78a And then we've got some uh Sony uh, runof-the-mill Sony uh srams. We're only talking 100 NS there -10 No, that' be 100 nond, not 10 nond on that. So this is supposed to be have like a, um, a a maximum clock rate I think of 100 mahz.

So um, you know I'm not sure how they're uh, sampling this sort of data with 45 uh nond um srams Here, you know we've got our 74f uh series logic. Here's our data you know we got some data latches here, got a 157 we got and then just some slower HC stuff around the bottom. but we've got some highspeed 74f series logic in here. So our data input here, we got some highspeed stuff.

It continues to go along here, just to there. and then we get into slow 74 HC series stuff. On this side, And in terms of clocks, the only thing I see here is a 16 MHz oscillator module up here. so not, uh, not terribly, uh, quick at all.

And of course all the data outputs on these two main connectors at the top and you saw that before that goes off to the main uh processor board. VI via some 74 HC 374 Aha I think I know what's going on here. Um, basically the uh Ecl board up here is the one that's going to have all the fast, um, serial, um input, uh, counting. It's going to have the fast clocks and we'll take a look at that in a minute, have all the fast clocks, it'll have all the serial data input and it'll be uh, doing some serial to parallel uh, conversion in there.

and then we've just got to the slower parallel data coming out here and then being and that being sampled into our Um SRAM down here. So in that case, 45 nond SRAM is going to do the business here and you'll note that there's actually a little Loop of uh, coax in there. it's quite long I Don't know what? Is it? a meter? You know, if you unravel it, maybe a meter long or something like that, go between one of the Bnc's on the uh front panel. In this case, it's the Uh disc out connector there and uh, that is just forming a little uh delay.
So that's just a delay. A coax delay line. Um, going to the board from the front panel connector to compensate for the delay in the input counter circuit. Aha, the manual.

actually. uh, shed some light on this. Pays to read it first, folks. Um, yes, these are our 45 nond uh fast buffer.

Uh Rams here. Here's our buffer. uh uh. address counters here.

the 74, um F1 91s and uh, then we've got our accumulator SRAM over here. And then we've got a separate copy buffer memory. So yeah, our data is serialized on our input Ecl board comes into this fast buffer here which is stored. and then we've got some accumulation uh, memory and an output copy buffer.

Ah, who knew. And here's the magic on our input folks. Uh, we have an ultra fast Ecl comparator. No surprises, uh, whatsoever.

It's a single comparator Analog Devices 9685 with a little heat sink must get a little bit hot and and the rest of the input circuitry uh, just contains um, overload, uh, protection and stuff like that. This uh instrument has a rated input of plusus 300 molts and the inputs are clamped and stuff like that. We got some LT uh, 17, um ultra low noise? um highspeed Op ants down in there LF 412. but yeah, um, basically this thing is just a um, it's just a comparator input because as I said, all it's doing is basically taking an input Um signal.

You know, a low Uh signal level from some sort of Uh detector, you know, Photon uh detector photo diod, something like that and it's uh, just counting those inputs. It's got adjustable uh thresholds of course for the Uh comparator input, but apart from that doesn't do much. And of course we've got our Um Mechel logic here. Motorola Mc10 H Uh 107.

That's just A. If you know your Um Ecl logic, then you'll know that's a triple input exclusive. or Norgate So our two comparators on the input there. they're just our Um input signal discriminators and they, uh, work with uh, hysteresis.

Of course they've got about 20 MTS of hysteresis on the 300 molt total input and then we're going to have a trigger input um, discriminator, uh, comparator as well and uh, that and they will all have like adjustable uh threshold voltages and adjustable polarity. At first glance, you might think, uh, that there's only a 25 MHz clock on this board, but aha, there's not. If you go all the way down here, well, certainly not. That's the only Crystal oscillator on the board.

But if you have a look at this um MC1 H16 here, we've got an LC tank oscillator down there and uh, that one um oscillates at around about 200 mahz. And then we've got a divide by Uh 64 counter and that gives us a 3 .125 MHz signal which goes into a phase comparator which is u304 there and there we go. Classic 4046. So the output of that phase comparator is then filtered and integrated by U Uh 305 there which is the Lf411 and then that goes down that trace on that on the left hand side of the board.
you can see there down into D301 and that's the frequency control input of our LC tank oscillator down there. So what we get is effectively a Uh Crystal oscillator referenced 300 Uh 200 MHz LC tank oscillator. Brilliant! And that's where we get our high frequency clock for the rest of the circuitry and then all the rest of the circuitry is our Uh Bin clock and our pulse sort circuitry and our shifters. and uh, stuff like that to uh uh, capture the serial data input and then um uh basically uh feed that into an 8bit parallel data which then is fed out via the Uh control connector up there into our main Um TTL uh logic board down here which is then stored in our fast frame buffers.

So wo there you go I mean I would love to have the Uh schematic for this thing that's 74 HC 294, U 412 there. Um, that's one of the keys to this thing. It's uh, a programmable uh divider which actually uh, divides the range and can divide anywhere from two right up to 2 to the power of 15. so that forms the heart of our bin counter.

But there's other circuitry around there which then splits that up into four separate phased clocks. So we've got a quadrature uh clock there. so those four clocks would be used to um form the bin counter. different phases of it.

would it be different operations of the Uh Bin Counting uh system and sorting system so that you know one would uh reset. so one phase would reset the counter, the next phase would uh, latch it in, the next one would count it and you know the next one would store it and uh, so on and so forth. So it cill to have like a system block diagram of this thing. I mean how hard is it to put that in the manual? they've put you know fairly detailed um, top level circuit descriptions.

you know, actually? uh, labeling, individual chips and stuff like that. but they don't give you nice system block diagram to understand. Quite disappointing. Yeah, you can sort of reverse engineer it from the description, but H that'll take quite some time.

So that folks is the Stanford research sr430 Scala what is it multi-? Channel Scala Photon counter thingamabob. And yeah, as we, uh, pretty much knew, and you know, and if you read the manual of, uh, basically, it's just, uh, you know, some comparator discriminator uh on the input. um, you know, fairly. fairly good stuff.

fairly, you know, high speeed and accurate, and uh, programmable uh stuff on the input. but it's just counting and timing pulses, putting in into different bins with some high-speed Ecl circuitry around here, and then outputting that as an 8 bit data into this uh, buffer. and um, uh, you know, uh, buffer and storage board, then going over to the main processor and displaying it. and, well, that's you know, that's pretty much all it is.
I Mean the key with this thing is what you what? You're actually measuring your detectors and stuff like that and you know you don't need. um, you know. Essentially, this is not, uh, rocket science here. but the stuff you attach on the front end of this thing is the rocket science part of it.

All the physics involved. This thing just counts it and analyzes the data and just, uh, pre presents it in a, you know, a digestible format. pretty much. Oh, and I forgot to mention the Uh E board up here.

four layer board. Of course. you can tell by the uh, dark patch here. they've obviously got the power and uh, ground planes in there and not over this.

They've just got it over the um, all of the ecl. um, digital part of the stuff right over. It looks and like they haven't split the uh, ground plane. Going across the comparative input, they've just put it right over that.

and of course, much more critical when um, like decoupling? obviously decouple, uh, uh, cap on each right next to each uh chip. Much more critical then just the um, you know, just this, uh frame. You know, just this, uh, buffer board down here. you can get away with a two- layer board there.

but all this, uh, really fast and high power Ecl stuff does, um, take a lot of current when it, uh, switches. So the inductance of the ground planes, uh, the, you know the inductance of the power system. The ground plane is going to essentially work as one distributed capacitor over there. very quite low imp, quite low impedance, quite low inductance, and, uh, absolutely essential.

This probably almost certainly wouldn't work on a two- layer board. just too quick. Edges are too fast, they gulp too much current. Forget it.

So I hope you enjoyed that. It's some basic uh technology, but you don't get to see e, um, stuff, uh, that often these days. and uh, they still sell this thing and I'm sure it's still exactly, uh, the same design. So um, if you want to discuss it, jump on over to the Evev blog forum.

And if you've got a link to the schematic for this thing, please leave it in there. And if you are one of the original designers who, uh, put their name in there, please let us know. We'd love to hear about the design of this thing. Catch you next time.


Avatar photo

By YTB

20 thoughts on “Eevblog #452 – stanford research sr430 teardown”
  1. Avataaar/Circle Created with python_avatars Brooksie1of1 says:

    Made in my hometown!!!!

  2. Avataaar/Circle Created with python_avatars Trains & Tronics says:

    Nice video. I would like to see a 2nd part where you could make it without blinking.

  3. Avataaar/Circle Created with python_avatars Doug Gale says:

    An LED is a photon counter. Just reverse bias it to charge it up with electrons, turn off the voltage and time how long it takes for it to the voltage to drop back down with a high impedance input. Generally 20000 microseconds in shadow, but much less time in light, and a very long time in the dark. Yellow LEDs are the quickest, blue are the slowest. Photons cause the electrons to be knocked out of the cathode.

  4. Avataaar/Circle Created with python_avatars Jeffery Rowan says:

    A possible solution to the sound problem could be a lavalier on the brim of a cap. It would be like a boom mike attached to your head.

  5. Avataaar/Circle Created with python_avatars Wagner Garcia says:

    HELLO
    I'm with the problem of Overload in Lock-in SR830 !
    You know how to fix this problem ?
    I have reseted the lock-in , but it worked .
    thank you

  6. Avataaar/Circle Created with python_avatars Steven Hardy says:

    Very interesting post….if you get bored with it or don't want it taking up space…I am willing to bet there are some colleges in the Philippines that would love a piece of equipment like that (I know that Australia sends much assistance to the Philippines)

  7. Avataaar/Circle Created with python_avatars Karol Piotrowski says:

    Well, it's piece of gear made for experimental physicists, which means people that love to tinker in things, so it's a very nice design for audience intended:)

  8. Avataaar/Circle Created with python_avatars Nathan Ell says:

    Comments about the P9324 date code, missed the P9248 date code on a chip previously looked at, P9218 date code right above the P9324 😛 Absolutely love the teardowns though!!

  9. Avataaar/Circle Created with python_avatars Josh says:

    but you did turn it on…

  10. Avataaar/Circle Created with python_avatars SID from Australia says:

    Run the receiver of the lapel mic through a Compressor. Altronics sell a good one that suits a single XLR mic, RCA outputs, 19" rack mount and it's cheap.

  11. Avataaar/Circle Created with python_avatars Dan Colmer says:

    I just saw one of these at a particle accelerator lab where I'm doing work experience. I instantly thought of the EEVBlog. 😀

  12. Avataaar/Circle Created with python_avatars Petertronic says:

    The "dot shifting" part may have been to slowly oscillate the display on the CRT to reduce phosphor burn-in.

  13. Avataaar/Circle Created with python_avatars hobbified says:

    Second-sourcing. Intel wasn't always the super-giant that it is today, and people wouldn't put Intel chips in their designs unless they could get equivalent chips from at least one other manufacturer.

  14. Avataaar/Circle Created with python_avatars Yong Meng Sua says:

    Hi there,

    Where do you bought this broken SR430 to teardown?
    I'm interested to get one for myself to have fun.

  15. Avataaar/Circle Created with python_avatars Alberto Vaudagna says:

    Thanky for your presence. If you rember how much you took to design this unit?

  16. Avataaar/Circle Created with python_avatars 3DCoolStuff says:

    Maybe you could upload some videos of your recent works, we could learn something.

  17. Avataaar/Circle Created with python_avatars avidrandomer says:

    AMD had (and technically still has) a license agreement with Intel to the architectures of chips up to and including the Intel 486. So it's an AMD manufactured 186.

  18. Avataaar/Circle Created with python_avatars David Miguel Saude Piçarra Piçarra says:

    that shit got na 186 amd prossesor
    GOD THEM!

  19. Avataaar/Circle Created with python_avatars katta says:

    Fantastic Job if you are the Joe. I think Dave would be happy to get a schematic if it is not proprietary after 20years. Do you have any suggestions on good experiments for Dave. This still cost USD7950.00

  20. Avataaar/Circle Created with python_avatars Bud Ski says:

    Did some research, it appears that Intel and AMD used to have a Technology Patent Agreement, in which they both shared and collaborated with their Architecture designs… Very interesting, as both companies are major Processor Competitors now adays.

Leave a Reply

Your email address will not be published. Required fields are marked *