Maker Pro
Maker Pro

Successful Stanford Research SR530 lock-in amplifier repair

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
Joined
Jan 21, 2010
Messages
25,510
According to the manual I'm reading, there are at least 14 3-terminal voltage regulators in this unit. And I've only found 7.

The manual has a parts list but no schematic. The parts list is actually reasonably useful, listing actual part numbers for semiconductors, and tolerance/stability for resistors.

Almost all the tantalum caps are 2.2μF 35V, the remainder are 1μF 50V. I could probably replace the lot with 2.2μF 50V low ESR electrolytics.

Although... How much fun will it be to track down the actual shorted capacitor?
 

hevans1944

Hop - AC8NS
Jun 21, 2012
4,875
Joined
Jun 21, 2012
Messages
4,875
If Jimi Hendrix is still playing, maybe I can pull his signal from the noise. :)

Ummm. Celestial music (often buried in noise) requires a different technique to recover than a simple lock-in amplifier. You need a correlating detection system plus an original, low noise, recording of the signal you are trying to recover... preferably a digital rendition... and patience to loop the original while attempting to synchronize to the celestial version, which may exhibit some variation from the original. There is also the perennial problem that celestial music is often not a reliable signal, coming and going, surging and fading... it can take years, or even decades, to recover even trivial information, and that's assuming you have chosen the right correlation technique from among auto-correlation, self-correlation, and evangelical correlation (just to name a few possibilities).

You have a fine instrument, possibly in need of a little TLC. Stream Jimi Hendrix, live or dead, from the Internet.
 

73's de Edd

Aug 21, 2015
3,606
Joined
Aug 21, 2015
Messages
3,606
Could your mystical mystery device be associated with the processing of a dual diversity receivers signals .
Mariner 69's telemetry signal processing encroached and surpassed that degree of refinement / enhancement.

Your adjunct package routing also fully answers the mystical . . . .
" somewhere ? Mid way between the " Peking duck " and the Ronne ice shelf " . . .to wit . . .
Perth Awwwww--stray-L- yuhhhhhh . . . . .mate.
 
Last edited:

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
Joined
Jan 21, 2010
Messages
25,510
Could your mystical mystery device be associated with the processing of a dual diversity receivers signals

Well, minus the diversity receivers, maybe.

A lock in amplifier acts as a very narrowband filter that is tuned by a reference signal.

The "narrowness" of the filter can be enhanced by increasing the integration period, at the expense of being able to detect more rapid changes of the input signal. In theory this could be used to demodulate a carrier. Because this is a dual channel device, slow phase changes won't cause the signal to drop to zero with a +/-90 degree phase change (the second channel operates using a 90 degree phase shifter version of the reference signal)

I've seen this done with an RF lock in amplifier to receive AM radio!
 

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
Joined
Jan 21, 2010
Messages
25,510
About to remove a suspicious tantalum cap.

IMG_20180616_081721.jpg

There's the back-lit PCB.

There are TO-220 size pads for the regulators. And seeing that, I've spotted another 2 regulators!

And that cap was fine... Trying the regulator next.

Nope.

Hang on a sec. Does that blue capacitor look unusual? (And no, I didn't touch it with the soldering iron)

IMG_20180616_090213.jpg
 

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
Joined
Jan 21, 2010
Messages
25,510
That cap has a resistor in series with it which measures 1.70Ω. Even if it's that cap, I'm seeing way less resistance. And, it's not actually tied to the output of the regulator.

I'm still reading a low resistance (under 0.5Ω) across the output of the (now removed) regulator.

The good news is, when I power the device up, the current drawn from the mysterious 56 CT 56 winding is now 370mA.

There's obviously a short to ground somewhere. And I need to find it.

And R913 looks toasted. It reads 4.92Ω whereas R914 reads 10.12Ω. They both supposed to be 10Ω 0.5W 5%. I think I'll replace that one. Weird that it went low though.
 

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
Joined
Jan 21, 2010
Messages
25,510
I've traced the short to near an LM311.

And it wasn't that, but it was a tantalum cap near it.

Ok. Time to replace all the suspect parts.

I think I'll replace the LM340T15 (AKA 7815) and the 10Ω resistor, socket the LM311, and I think the two tantalum caps will get replaced with electrolytic capacitors.
 

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
Joined
Jan 21, 2010
Messages
25,510
Well, since I had to get the voltage regulator, I thought I would get a new comparator too.

IMG_20180616_124725.jpg
The new regulator, resistor, and replaced cap.

IMG_20180616_124747.jpg
And the socketed LM311 with the nearby tantalum replaced with an electrolytic.

This is the second shorted tantalum I've had in a couple of weeks. I'm starting to get paranoid. I'm considering fuses in the secondary windings of the transformer (because clearly the one in the primary didn't do the trick and I really don't want to reduce it too much.

All three places I checked were at roughly 650mA, but the fault current was in excess of 1.5A. what do you think, a 1A slow blow fuse?

So much for "No user serviceable parts inside". $3.80 in parts (plus what I had on hand) is a bargain.

But are there other faults? I guess I need to do some more testing.

The datecodes on the two replaced semiconductors was 1987, in one case the 52nd week. Looks like it was probably manufactured 1988. Anyone have any idea what these went for? Or even where they might have been advertised?
 
Last edited:

hevans1944

Hop - AC8NS
Jun 21, 2012
4,875
Joined
Jun 21, 2012
Messages
4,875
By the time you get done re-engineering that little beastie with better capacitors it should be worth twice what it sold for when new! Well... it should be, but you know how that goes... I didn't realize those tantalum capacitors were so prone to failure. Mostly avoided them anyway because of cost, but at least with electrolytics you can pretty much expect to replace them every ten years or so... and they usually show some physical signs (bulged cases) before expiring, although an ESR meter will sniff out some bad ones that "appear" normal.

Have you tried calling Custom-Cal here in Florida to see if they can provide you with schematics for the manual? The original manual probably had largeish fold-out schematic drawings that didn't get converted to PDF. And it's also possible that if you contact SRS in California they can provide the schematics. Their website allows a PDF of the manual, without schematics, to be downloaded... but of course you already knew that.
 

73's de Edd

Aug 21, 2015
3,606
Joined
Aug 21, 2015
Messages
3,606
Stir Seeve . . . . .(ooooooops . . . it's my dern lysdexia kickin' in agin' )


By the colors . . . .


Is the left of this RED area, being any more serious than this portion ?

YELLOW . . . . Bet one end of this film resistor is being at nearly 0 ohms to ground..

PINK . . . . .This must be the unit mentioned
now . . .Don't be vague . . . .but aren't the companion ORANGE cased C903 C910 units being Sprague dipped tant units ?
I rarely had problems with those . . .but had extreme failure rates with the particular brand(s) that used progressive depths of dips in paint to mark on their color coding bands.

ORANGE . . . . . I think we found, at least, ONE of your bag of 3 term regulators being used.

BLUE . . . . Yep . . . .that's my collateral damage that you are seeing . . . . . and that I FELT . . .OUUCCCH !

OLD BLUE (C329) now talking . . .I done "toles" him that he shouldst have pulled me FIRST, for him to then have full open access to get to C904.
What with it being installed in double sided lands and its leads being in plated thru holes . . .it would require a broader tip and extra soldering iron heat.

OR, his doing a flipping of the board to be having lead sides on top and then some drops of Rosin flux and a quick addition of new fluxed solder, to amass a large solder drop that fully surrounded both leads , then the part drops down, out and free, with possibly only some tailing solder icicles, then being seen hanging down thru the now opened lead holes.

Thaaaaaaasitt . .

Add on . . . . Just now saw that RED was a clipped lead 7815 . . .and fully cooked.

Don't take my KODACHROME away . . . .

Steves_board.png


73's de Edd
.....
 
Last edited:

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
Joined
Jan 21, 2010
Messages
25,510
And the voltage between CT and 5.6 (the board is marked 5.6 CT 5.6 where the other three wires from the transformer go) measures 17VAC!

The manual states the following:

Power Supplies

The line transformer provides two outputs, 40VAC and 15VAC, both center tapped. The transformer has dual primaries which may be selected by the voltage selector card in the fuse holder. The 15VAC is rectified by diode bridge BR2 and passed to 5V regulator U909. The output of U909 powers the microprocessor and its related circuitry. The 40VAC output is half-wave rectified by BR1 and regulated by U901 and U902 to provide +20V and -20V. These two dc voltages are then regulated again by 15V regulators U903-U908. Each 15V regulator powers a separate section of the lock-in to reduce coherent pick up between sections. U911 and U912 provide plus and minus 7.5V and U910 generates +5V for the analog circuits.

I guess if the transformer does bite the dust one day, this tells me what I'd need.

The manual also says:

OPERATE WITH COVERS IN PLACE

To avoid personal injury, do not remove the product covers or panels. Do not operate the product without all covers and panels in place.

Oh noes!!! I failed to follow this rule too!


Is the left of this RED area, being any more serious than this portion ?

I think that's just a pattern in the thermal compound that was under the 7815.

. . . . Bet one end of this film resistor is being at nearly 0 ohms to ground..

Well, it was once a 10Ω resistor that now measures 4.6Ω

. . . . .This must be the unit mentioned
now . . .Don't be vague . . . .but aren't the companion ORANGE cased C903 C910 units being Sprague dipped tant units ?
I rarely had problems with those . . .but had extreme failure rates with the patricular brand(s) that used progressive depths of dips in paint to mark on their color coding bands.

The only markings they have are 2.2 35, and a black bar with a + at the bottom of it. The markings are upsidedown if you expect the leads to be at the bottom. The parts list just describes them as "Capacitor, Tantalum, 35V, 20%, Rad"

. . . . . I think we found, at least, ONE of your bag of 3 term regulators being used.

Yep, that's a good one :)

. . . . Yep . . . .that's my collateral damage that you are seeing . . . . . and that I FELT . . .OUUCCCH !


YOU
touched it with the soldering iron? (in any case, it tested fine)

OLD BLUE (C329)
now talking . . .I done "toles" him that he shouldst have pulled me FIRST, for him to then have full open access to get to C904.
What with it being installed in double sided lands and its leads being in plated thru holes . . .it would require a broader tip and extra soldering iron heat.

C904 almost fell out. That one was easy :) The regulator was a bitch, the leads were a tight fit in the holes. Really hard to desolder.

his doing a flipping of the board to be having lead sides on top and then some drops of Rosin flux and a quick addition of new fluxed solder, to amass a large solder drop that fully surrounded both leads , then the part drops down, out and free, with possibly only some tailing solder icicles, then being seen hanging down thru the now opened lead holes.

There was quite a bit of flipping back and forth. :)
 

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
Joined
Jan 21, 2010
Messages
25,510
And it looks like I have yet another problem.

The reference input works fine, but the signal inputs don't do anything at all.

If you remember from the very beginning, the input circuitry is under a large shield, so off that comes!

Here is a portion of what is under that shield, near the signal inputs

IMG_20180616_172631.jpg

There's a few interesting things.

The first thing you'll probably note is the missing capacitors. Well they're not missing, they're mounted on the other side of the board. In what looks like "someone has been here", these capacitors are 2200μF 16V capacitors which are much larger than the empty spots indicate.

IMG_20180616_173800.jpg

The original C107, C108, C125, and C126 are specified as wow, 2200μF 16V. I can't imagine they were ever small enough to fit on the board. Maybe they were always meant to be like that, or maybe it was a relatively late change.

Ok, the second thing you'll note are the 4 tantalum caps, and the two TO-92 devices that just happen to be regulators.

It makes sense that these would power the input stage (otherwise why put them here). Is it possible that one of these tantalums is bad too?

Is it too much to ask?

Some of the ic's under the shield look to be difficult to obtain, so I'd rather not have to try (and then have to calibrate stuff).

Wish me luck.
 

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
Joined
Jan 21, 2010
Messages
25,510
The voltage across the various big caps are:
  • C107, 1.5V
  • C108, 2.5V
  • C125, 1.135V
  • C126, 0.87V
Well, that's not going to stress 16V caps.

What about the regulators input and output voltage (btw, they're not hot)

And the voltages around the 78L15 are close to zero (actually -0.6 at the output)

But the voltages around the negative regulator are fine (what is it about these positive 15V regulators today?).

Where is the +18V going?
 

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
Joined
Jan 21, 2010
Messages
25,510
R138 looks suspicious. And doesn't measure 20Ω any more (it appears to be open)

Did one of the tantalum capacitors fail which probably caused almost the full 18V to appear across this (or have about 100mA flow through it).

Let's see... 100mA would be 0.2W (and it wouldn't have fried - is 1/4W), 18V would be almost an amp and that would fry it. And this also explains the short between input and ground on this regulator.

Excellent, this also means the regulator probably survived, although I'm not sure it would have liked it's output pulled negative, and the output tantalum wouldn't like that either.

I'll start with both caps and the resistor.
 

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
Joined
Jan 21, 2010
Messages
25,510
I didn't have a 0.25W 20Ω resistor, so I put two 10Ω 0.125W resistors in series.

This is what it looks like

IMG_20180616_194614.jpg

Now for a bang test :)
 
Top