Hi Yes! I can't help myself I Keep buying test gear on eBay This one popped up and it is repair so it was sold as not working or not powering on what is it? it's a Stanford Research Sr 650 uh dual Channel elliptical filter designed for uh, audio low frequency. It's basically uh DC to 100 khz. So basically that uh, audio range plus uh, low frequency vibration and shock analysis and stuff like that. Really handy ones.

I've never had a programmable filter in my lab and I've always wanted one, but they've always been pretty expensive. There's a couple of companies that make them uh Stanford Research is one. uh, Rockland is another and there's a few others on the market as well. Um, so you know one of your more obscure bits of kit, But if you need one, you need one one of those things.

So anyway, I'm hoping this will break the Eev blog curse where the faulty products are either just so incredibly simple like a blow input fuse or something, or they're just beyond economical repair. Be so. I Hope it's somewhere in between there and we should be able to fix it. Now the good thing about Stanford Research Gear um, you've seen some Stanford stuff on my blog before is that you can get the schematics for them.

The manuals come with the full schematics and everything else. I Don't have the schematic for it I haven't tried to see if I can download it somewhere, but you can at least buy them. Um, which is great. The other thing is is that Stand Research Gear looks like it's been designed in the you know, early 1990s.

It's all dip uh technology or uses almost always offthe shelf uh parts so very repairable. I'd be surprised if this thing is beyond economical repair unless it's uh, you know the power supply has failed. They're usually linear power supplies inside these things unless it's somehow failed and like taken out everything and that might make it Beyond economical repair. But anyway, um yeah, we should be able to have a good shot at it.

So I picked this up for just over hundred bucks on eBay and they still sell this. uh it is still a current model uh but it was designed back in the 90s or something like that and it goes for 3,300 Bucks new. So Beauty this is the 650 which is the Dual Channel version which means you can do uh band pass filtering as well cuz you can select high and low uh frequency and you can do that band pass so let's crack it open see what we got? Hm H sold as just doesn't power on or something like that? Let's have a look. So yeah, these bits of Stanford Research gear um these type of ones what they what they usually are is just like a little uh, 8bit, uh micro and then basically all just uh, passive stuff and switches and Mxes and stuff like that.

That's what'll be inside this filter. so you know, Really? even if you know, power suppli has failed and taking in a lot of stuff, probably still, uh, probably still repairable, even if you had to replace. you know, a whole bunch of opamps or something like that. Um, might still get lucky.

So really, there's not much else inside this thing and this here we go. This company knows how to pack stuff. Look there we go. They've got the, uh, the molded, uh foam.

the uh, insta pack it's called. So this is a proper, uh, professional test equipment company. It's uh, the Outback equipment uh. company in Gilroy in California So excellent.

Well done. So if you're going to buy stuff from them, well, I would, uh, yeah, look at that. I Mean that is just fantastic. That's how you should pack test gear.

That's brilliant. They've just molded these in here. They they've chosen a box and then just molded those in. Specifically, They got the machine, the insta pack machine to manufacture just the right.

uh, but yeah. basically you put these uh, little. What you do is you put the uh plastic in there and you've got a machine which then squirts the stuff in and then it just molds around the shape of the product. So this is really good so it should have come to me in good condition.

I Don't It looks a little bit crusty. There's a bit of uh I didn't get anything else with it. No manual or anything like that. There's a bit of rust on it.

Yeah, it's not a it's not a new unit that's for sure. Um, and yeah, there's a bit of corrosion on the on. the Bnc's there, that's you. Know that's not terrific, but we can clean those up.

But even if something like the Bncs are ruined, you know you could always replace them in gear like this. So uh, to Scope? Yeah, somebody's written on there. Yeah, thanks for that. Terrific.

Um, yeah. lots of rust on the screws. So yeah, there we go. Quite a bit.

Pretty crusty, but but hey, let's crack it open. Well, no. I'll Power It Up First serial number 28078 and they probably would have made that many. So yeah, it's not going to be serial number 78.

It could be 8,078 but I doubt it. Stanford have made a lot of these things made in the United States of America Actually, there doesn't seem to be a voltage select in there like I think I'll just crack the lid off first and see, make sure there's no uh Jumper in there for 240 volt operation. So as I said, it'll be all dip mostly all dip technology. I'm actually a surprise.

Be surprised if it's not 100% dip technology. probably on one or two big boards. it'll be a processor board and there'll be probably a big analog board. they might integrate them into one.

I Don't know. this is an eight pole elliptical. Oh yeah, mixture of screws. Somebody's had a go at this man.

We had Flathead on the other side and then we we got a Phillips and then we got a tiny little hex. Unbelievable! So I'm a little bit concerned about the amount of rust on this. You got to wonder what sort of life it's had. and hey, Tada look at that.

that actually looks fairly clean I Don't mind the look of that at all. There we go. we got two big analog boards cuz this is a dual Channel unit. There is a lesser Uh model in the 600 series that only has a single one, so you can combine the Uh two channels and uh, so there we go.

I was right. There's an entirely separate digital board up here and yes, everything is through hole and then a separate analog board for each. Channel Look at that. I mean that is hugely repairable.

We've got the all of the designators all there nicely spaced out, all dip we can just pull that bought out and uh, desolder things if we have to not socketed. none of the well, the memory and stuff over here is socketed. uh, and the ROM but apart from that, yeah, um, that's pretty darn good. I'm happy with that.

and of course, this is exactly what you see inside a filter. We got a whole bunch of just passive uh components and then just um uh, mxes and uh switching stuff like that. So that's pretty much all there is to it because it's just an elliptical filter. It's just using opamps.

So here's all your input opamps and stuff over here and then the rest would be and then you've got all the different range resistors and stuff like that. So yeah, I mean so it shouldn't be hard to get this thing uh, up and running. but uh, well, let's Power It Up Here's what I was uh, afraid of though. look at all the wiring going into that fuse holder over here coming from the uh Transformer Taps primary side here and I love the uh, the big clunk and mechanical switch coming from the front panel.

of course. none of this, uh, soft power rubbish. I like it. So yeah.

I don't see any initial labeling on there. so there could be a 240 volt configuration of this wiring. So before I power it up I probably should consult the manual. Oh, actually, I can just do it the easy way.

I Just got out my 110v. uh Transformer so using one of those weird ass Yankee plugs. So let's uh, let's power this thing on. So it came from the US so presumably um, it's still set for 110 if it's set for 240.

Uh, well, it's going to be under voltage, so not a problem I Found that some Uh dealers will actually know if they're shipping it to Australia We'll actually set it to. 240 volts for you before they, uh, ship it. But anyway, let's go. Nope, he's right.

Nothing dead. What? Well, first of all, I'm uh, before before I measure the power supply. I'm just going to do a quick check of that backup battery in there. There we go.

2.92 Vols That's okay. Um, so that is still Hunky Dory hasn't leaked at all. Although in this, uh, sort of vintage gear, you'd probably replace that. What date code we looking at? 9624? Yeah, 96 seems to be uh, seems to be the latest that I can see.

So yeah, that mid to late 90s vintage. It's got that mid to late 90s smell too. All right. Before I start poking around the power supply.

Let's uh, make sure that the primary of our Transformer is uh, is there. so my power switch is on and we'll just measure that. Well, there's your problem. Look at that.

No, uh, can't measure the primary side of the Transformer So that fuse was good. I had had a look at that before I Um, powered it on and yep, it looks intact and measures intact. So nothing wrong with the fuse. So the primary of the Transformer I'm wondering if maybe there was a uh, sometimes these Transformers will have a thermal, uh, cutout in them so maybe that could be uh, that could be an issue perhaps.

But anyway. um, if we can solve the primary side of the Transformer maybe this thing will come to life and maybe there's nothing wrong with it. I Mean that's ideally what you want in these sort of repair scenarios. If there is act electrically something wrong with it, then usually you sort of hoping for a power supply fault cuz they're the easiest things to fix.

You know if it's something like uh, like the Eom has died in due to age or something like that, or there something else. uh, you know, some other fold or some other failed chip or something like that. Um, but usually it's the power supplies that are going to fail first. So heck, we're not even getting into the main.

the primary side of the Transformer goodness. Now here's something interesting. check this out. Here is the Uh Main's input filter here.

So they've got nice little uh filter module in there with a common mode choke and I'm measuring the output of that and there we go. We got our 35 Ohms. So hello, that has to be the primary of the Transformer. Let's switch it off.

Yep, the input pins on the IC are supposed to just go straight through the filter and out to these two pins. So we measure our primary of our Transformer here. but we don't measure it on these pins. That is strange.

I Don't think I've ever seen one of those filters die before. Unbelievable. And you can see the circuit there. I Mean it's just a standard um.

input filter. They've got the common mode choke there and the suppression caps, input and output going to ground, but that's it. Like it's not like it has any internal fusing, so really for that? Uh, so one side of that common mode choke has to be broken in order for there to be no continuity between the pins, the input pins over here, and the output pins here. So let's measure that.

No, nothing there. We go there, we go. So we got direct continuity on that pin there. Let's try the other pin should be this one.

and of course, no, um often these things will be potted too. So um, I don't know. Well, let's try. this one's not.

though this one's not so we'll might have to. It's probably welded shut, but we might have to rip that thing open and have a look. but I can sort of bypass it cuz I can just take off I can. Oh no they? no they.

Well, they're solded on I could desolder those and feed the mains directly into there to, uh, bypass it. but jeez, I wouldn't have thought. the Main's input filter go figure. I Just noticed something out of the ordinary.

Look how all these chips are solded in. Yet, this one is socketed. Why? It's almost as if that board has been repaired. I Mean that's just a 74.

What is it? 74 Hc4? Yeah. H So that filter just uh popped out of there these. uh Clips were just, uh, holding that in there so we should. Yeah, there it is.

See, it's solded. shut there so we can get the iron on that and uh, whip that sucker open. Maybe there like some internal wiring that's just broken off or something. Maybe there's a board.

Maybe it's a right angle? uh IEC connector or something. and yeah, just the internal wire into that is broken. perhaps? So I'm having some luck on this. It looks like this whole top cover here can slide off.

Hey nice. There we go. We're in and check it out. This isn't exactly professional construction on your filter here.

There's the common mode choke wrapped in electrical tape and there's your uh, there's your Main's rated caps tucked away in there. Three of them. Oh goodness. Okay, well that's embarrassing.

I Missed the Volum selection switch I Knew it must be in there somewhere. With all his wiring going into there, it didn't make sense just to sort of terminate it in there and like an and solder them in the correct position. and yeah, it's tucked. it's slid all the way down in there and people watching this in in uh, probably you know was screaming at me all along.

Oh there's the voltage selection switch Yeah I missed it I'm watching the um screen of the LCD here. so I missed these sorts of things. Anyway, yeah, it was in the 120 volt position, so I can just uh, whack that around and slide it back into the 240 volt position. But that doesn't explain why our fuse is good and we're not getting Uh from the output terminals of the filter through to the I because of course the I the I input will be going through the fuse and then Um through to the Uh filter side of it.

so the fuse has to be in that path there somewhere. I think Okay, now I'm fairly convinced I've figured out what's going on here. Here is one of the I Well, here's the I input connector here and of course it's not connected through to there and it should be, but it is connected through to the fuse there. and of course that goes through the fuse and I can see a trace running onto the other side and then that has to run back to this common mode choke over here, but there is no connection between those at all that fuse and the other common mode choke.

That's why we're get no continuity from the I through to there whereas the other one the I other I input pin is here and we're getting continuity just fine. So I've I've actually resold it I reflowed both of those and added new solder through. but because the connection is on the other side of the board I can't physically get in there. um to check that, but I'm pretty sure that is supposed to connect through to there and it doesn't.

And by the way, in the 240 volt position? uh, we're getting of course double the primary resistance 52 ohms whereas we'll get in uh, 25 ohms in the other position. Actually, I'm beginning to think somebody's had a go at this. There's a nut missing out of here. There's only actually one nut holding the Transformer on and well, I don't know.

It kind of looks new. And check this out. Here's the U secondary and the secondary has this wire. sort of.

You know, hanging out here like this. It's not like the others and it hasn't been cable tied with the others. So I'm not. you know? I'm not entirely convinced that uh, somebody hasn't been fiddling with this thing.

Um, I don't know. I just get the just get the warm fuzzies. That's the case. All right.

So I put a Mod wire in here. but I don't know I'm getting the heie jeebies. There's a reason why that Trace under there and I'm pretty sure that's the trace is broken I tried to, uh, desolder the pins and lever it off, but uh, it was just all too awkward. Couldn't get in there.

uh, it was pain in the ass that wasn't going to work at all. Okay, I've put a half a fuse in there and run it through my Uh three-phase power meter. I've disconnected the Uh secondary. uh from there just for now I Just want to power it up and get the primary.

So let's give it a B and 1 Point 1.5 Watts There we go. That's not too bad at all. Okay, here we go. Moment of Truth We've pluged the secondary back in and uh, let's give it a bell.

see if anything Powers up. Woohoo! Look, it's alive. It's alive. It's doing something.

There you go. Nothing smoking so we're at least getting yeah, yeah, the micro and everything else. uh AC DC Coupling overload. Yeah, who? Well, who knows these are fed input uh things? I believe Often the no.

There we go. it's just resetting. There we go. So hey, the micro is working.

Everything's hunky dory so something blew the ass out. Well, or there's an intermittent track in that Mains uh uh. input filter there H Go figure. One thing I'm definitely going to do is replace that crusty, noisy fan with a modern Silent One Definitely do that mod.

Now one of the good things about this amp is that it's got differential inputs too. Here it is. It's uh, you can either select a source A or Source B as single ended, so you can actually use it as a switch to switch between two input signals or a minus B is, of course, uh, the difference. So you get a differential input and of course we got those on both channels.

So two totally independent uh channels. as we saw two totally independent boards on there. Theyve uh, they'd have their own isolated power supplies. Everything so fantastic.

And then you've got your uh input. uh, gain of course for each one. Um, so you can gain it from anywhere from 0 to 60. DB Very nice.

and your output gain as well. So beautiful. All that's left to do is uh, feed a signal in and see if it filters cuz that's all it does. It's a filter.

All right to test this. I've got my DSA my Dynamic signal analyzer. Why? Well, because I can and I Love Dynamic signal analyzers. Um, you don't need a DSA to test something like this.

Of course all you need is a function gen a sign source and your multimeter or your scope and you can sweep it over the frequency range and determine it. But anyway, the DSA will allow us to sweep the frequency up to 100 khz which coincidentally is the uh limit of this thing. So fantastic. Um, what I'm doing is the source output here I'm generating 1V RMS of random uh noise.

So there it is 1V RMS there. and uh, the reason I'm doing random source is cuz this doesn't have a swept sign uh generator. So what we can do is we can Generate random noise. So it generates random noise over the entire spectrum and then over time you can average that out and get a flat Spectrum response.

So let's actually start that: I've turned average in on I've got 100 averages there. What I'll do first is I'll actually show you that noisy input signal now. it's working like a scope now, so there's our input signal there. It's just garbage.

Okay, so we can, uh, turn that Source off of course and bang. It drops away to nothing. So there's our 1vt RMS Random noise and random noise is actually very useful, especially if it's uh, proper random noise. So then you can actually average it out over cu the uh, cuz the power in each frequency bin over time is going to be uh, the same so you can average that out.

So anyway, Let's uh, turn that back to our spectrum and then let's start Here we go and you see how it starts out noisy, but then it eventually gets flat. So I'm just feeding the source directly to the channel one input. Okay, so it's flat over the entire frequency range. Now what we'll do is, we'll uh, put our filter our Stanford research filter in series with that and we'll set it to 50 khz smack in the middle and we should see it go flat and then drop off like a brick.

Almost like a brick wall. cuz this is an eighth order filter. Really high order filter so it should have a really sharp fall off right at 50 khz. pretty much smack in the middle there.

All right, we're in the low pass side of it. The other side is an identical Channel but it's a high pass uh filter. So uh, it's Auto automatically defa to 5 khz at the moment? Well, let's just ramp that up. You notice how it just shows up.

enter like it's it. Don't believe it changes. uh, like dynamically as you go, you got to hit enter before it actually applies it. So that's why it's warning you there with the big enter thing.

So the dynamic speed there isn't bad. There we go. Spot on: 50 khz. we'll enter that in.

so that's our low pass filter. Our input gain: uh zero DB our output gain: zero. So we should be hunky dory. So let's just go back to here and uh, run this again.

and where Curs is at 51.2 khz at the moment. So just before that cursor, we should see that drop off if this sucker works. So let's press start again. Hello.

look at that. There we go. It's not exactly a brick wall. it's uh, it's dropping off.

But there we go. it works. What if we lower the frequency down to say 10 KZ and see what happens? There we go. I Won't bother zooming in on the DSA There we go.

Let's run that again and we should see it drop off about here. Ah, calibration. Come on. give me a break.

Come on, you can do it. Come on here. We go Average in progress. There we go.

it's dropping off. Tada Like the proverbial brick wall. this thing as a first order approximation pun intended um, works. I'm very happy that is a winner.

Awesome. Well, at least the uh, low pass filter stage. Let's try the high pass filter. So what we'll do is, we'll just move this over here.

These Bnc's are a bit crusty and uh oh, overload. That's not good so that doesn't sound good. 50 Nice round number and okay, no, don't like the look of that overload. uh.

okay, so well, let's run it. No, not getting a good vibe there and of course we should see a flat line here and then drop like that. But no, no, we got nothing. this thing.

No. so we could have a fault in unless it's a pebcak and I'm using this wrong. Um, uh, because I haven't used one of these in Donkey's years. Absolutely Donkey years.

and I haven't read the manual of course. So no, those overload indicators aren't good and you can whack the filter in and out too. It can, just, uh, bypass it. That's very handy.

but no, no. I don't like the look of that at all. Input: A No, no, we could have a loser on that channel. Woohoo! Excellent.

More stuff to troubleshoot, but this channel over here, the low pass one looks like it works. A treat. Beauty Now I'm back on the low pass filter again. and curiously, um, I've selected the filter out so I've pressed the filter out button and well, as you can see, the filter is still in.

so I don't know there could be an issue there, but no big drama. And of course, the reason we're seeing a relatively, uh, gentle roll off here is because we're in linear axes mode. So uh, go to your more traditional log axes and you'll see it drop off much faster. There we go.

it's got like 80 DB stop bed attenuation and if we wanted to, we could try and measure all that sort of stuff. but I'm not going to do that today. I'm quite happy that at least this low pass filter section seems to be working. Awesome! There we go.

I Auto scaled that and we can start to see the rolloff performance. Switched it down to 10 khz here. so that's 100 khz. That's 10 khz cuz it's a decade cuz we're a log scale and 10 khz this would be 1 khz so I could drop that down to 1 k HZ just for entertainment and uh, who just overshot there I Got to get it spot on.

Here we go start boom and we'll start to see it drop off at that 1 khz Mark Beautiful. No wait, hang on I Just tried that high pass Channel again I just repowered it and uh, the uh overload leads have gone off and I set it to 10 khz and it's bang on. So I'm not sure what the hell was happening before, but that works. 10 khz high pass? No worries whatsoever.

Wow, this thing's fully working. Okay, uh, well. fully working. Haven't done complete tests on it yet, but hey I'm happy with that and what I've done here is put it in band pass configuration.

so my source is coming into here where set a high pass filter frequency of 10 khz so anything above 10 khz it's going to allow and then the output of that goes over to the input of the low pass filter and then the low pass filter set to 20 khz. so anything below 20 khz is going to be let through. So Bingo we've now got a band pass between 10 khz and 20 khz and the output comes from here. Back over and look at that.

Woohoo! There is our ban Pass filter characteristic: 10 Kilz to 20 khz. Bang on! No! I just switched this to AC coupling mode and just permanently. uh, set it into overload. Not sure how to reset the uh overload on this thing.

might have to read the manual. um, but curiously switching from AC to DC coupling as you'd expect I heard a relay but this channel over here. the low pass filter. there's no Reay I don't hear it and same with filter in and out I Would have expected that possibly to be a Reay bypass, but I can't hear it.

So anyway, there you go. That's a repair. In quote marks of a Stanford research: Sr 650 dual channel uh filter and it's a very nice instrument. If you can pick up one of these like I did for just over a 100 bucks.

uh, and looks like it was just the Main's input. uh, blowing then really, there's not much else that can go wrong with this. Um I believe this uses the fet input just like the uh pre. They Stanford research have a pre amplifier and it's notorious.

uh, the front end on that for blowing the input fits, but you can get replacement uh fits for those and uh, actually repair one of those pretty easy I can't remember the model number of that offand but uh, the SR 650 I believe uses that basically the same Uh front end on that. So could be a similar story with the FED inputs, but it seems to be working. An absolute treat. So anyway, I'm going to, uh, repair the um I can't put the cover back on the Main's filter up there with that bodge wire on it.

Doesn't sort of work. Maybe I'll sort of replace it. replace that noisy, horrible fan in there I got to clean up the Bnc's on these things and um, yeah, just generally do some more testing on it. But hey, that's a winner.

That's band Pass. There we go. That's exactly what I wanted. So now I got myself a really powerful uh, dual Channel programmable filter.

A very handy Edition for the lab and if you need something like this, you need it. They're great to have lying around. you don't use them all the time, but valuable bit of kit for your lab if you can pick one up like I did. Very nice, all right I Was hoping for something a bit more.

You know, an electronic type repair but hey eh, it's a Main's track. eh. whatever. Catch you next time.