-
Categories
-
Platforms
-
Content
Successful Stanford Research SR530 lock-in amplifier repair
- Thread starter (*steve*)
- Start date
hevans1944
Hop - AC8NS
Steve's new toy does a fine job of recovering a very tiny AC signal buried below the noise floor, provided that signal is coherent with a reference signal. A Jimi Hendrix guitar riff is normally considerably larger than the noise floor, so a lock-in amplifier is not necessary.Just to show my lack of knowledge - will it do a nice rendition of Jimi Hendrix's Star spangled banner?
Umm. The label on the back says: WARNING: No user-serviceable parts inside.
Clearly they didn't have the Stevester in mind when they slapped that tag on it.
I am so jealous! Where did you get that puppy, and how much did it cost?
This little puppy cost me around $300, which is waaaaay cheap if it works. The seller has a second even cheaper unit that's on its way. If either one work is still a great price. If both work then I can probably sell one and cover the cost of both.
If Jimi Hendrix is still playing, maybe I can pull his signal from the noise.
Anyway, this is the inside:
It's not easily visible, but the main board is bowed downward. It looks like the vertical board's edge connectors are not fully plugged into their sockets.
I'm a little disappointed that the main board isn't supported under the edge connectors...
It's possible this happened in transit, bit it seems unlikely.
Another interesting thing is the contrast between the signal input under the long shield on the left, and the long unshielded wires leading to the various outputs on the back panel.
If Jimi Hendrix is still playing, maybe I can pull his signal from the noise.
Anyway, this is the inside:
It's not easily visible, but the main board is bowed downward. It looks like the vertical board's edge connectors are not fully plugged into their sockets.
I'm a little disappointed that the main board isn't supported under the edge connectors...
It's possible this happened in transit, bit it seems unlikely.
Another interesting thing is the contrast between the signal input under the long shield on the left, and the long unshielded wires leading to the various outputs on the back panel.
Manuel would be happy to have one
I thought his timing was interesting. But this only goes to 100kHz, not his 100MHz :-/
I think the board was being bent out of the factory.
I couldn't push the connectors in any further, and the edge of the card touches the top of the case. Is this poor design? I guess a slightly bent board isn't a huge issue, but still, this wasn't a cheap instrument in its day!
Two interesting things here. First the battery measures 3.2V, secondly look at the bodge! The rest of the board is bodge free from what I can tell.
Tests on the main filter caps show they have an ESR of around 0.5Ω, which seems pretty good.
I couldn't push the connectors in any further, and the edge of the card touches the top of the case. Is this poor design? I guess a slightly bent board isn't a huge issue, but still, this wasn't a cheap instrument in its day!
Two interesting things here. First the battery measures 3.2V, secondly look at the bodge! The rest of the board is bodge free from what I can tell.
Tests on the main filter caps show they have an ESR of around 0.5Ω, which seems pretty good.
Ok. Happy to turn it on...
Looks fine and the buttons respond in a way that seems to be correctly.
I think I'll leave further stuff until I've read the manual
Hmmmm.... There's a bit of a funny smell. It's the "transformer" smell. The 12VAC marking on the board (direct from the transformer) reads 15VAC.
I've definitely set the input voltage to 240, and whilst it doesn't say 50/60Hz I doubt that's the problem. My mains currently reads 244V, so it's not that.
Maybe I should pull out the autotransformer?
Looks fine and the buttons respond in a way that seems to be correctly.
I think I'll leave further stuff until I've read the manual
Hmmmm.... There's a bit of a funny smell. It's the "transformer" smell. The 12VAC marking on the board (direct from the transformer) reads 15VAC.
I've definitely set the input voltage to 240, and whilst it doesn't say 50/60Hz I doubt that's the problem. My mains currently reads 244V, so it's not that.
Maybe I should pull out the autotransformer?
Interesting. With a 200VAC input voltage and the 12VAC reading 11.7VAC, there seems to be insufficient voltage to some other rail to allow the front panel LEDs to illuminate.
At this voltage the unit draws approximately 250mA, which is also the recommended (slow blow) fuse rating (and about 300mA at 230VAC).
The 12V winding is supplying 450mA
One 5.6V winding is supplying 1.53A
The other 5.6V winding is supplying 1.55A
The back of the envelope power supplied by the transformer is 23W.
The power supplied to the transformer is approx 69W.
Something is going on.
Setting the input voltage to 120VAC again, and seeing the autotransformer to 100VAC has the device drawing 600mA, and 750mA at 120VAC, which is consistent, if not worse.
Will I have to replace the transformer?
At this voltage the unit draws approximately 250mA, which is also the recommended (slow blow) fuse rating (and about 300mA at 230VAC).
The 12V winding is supplying 450mA
One 5.6V winding is supplying 1.53A
The other 5.6V winding is supplying 1.55A
The back of the envelope power supplied by the transformer is 23W.
The power supplied to the transformer is approx 69W.
Something is going on.
Setting the input voltage to 120VAC again, and seeing the autotransformer to 100VAC has the device drawing 600mA, and 750mA at 120VAC, which is consistent, if not worse.
Will I have to replace the transformer?
Stanford Research SR530
Oh no! But, but, but... I'm going to service it. Does that mean I can't use it afterwards?
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!
Is that 5.6 or 56? But it certainly isn't either (if that means volts)
If you look at the earlier images, you'll see a large heatsink (it's about 80mm tall). This has 3 TO-3 packages, an LM317K, an LM337K and an LM340KC-5.0. the are also 4 other TO-220 3 terminal regulators mounted flat on the board without additional heatsinking. These generate a pair of +/-15V rails.
I wonder if I have another shorted power supply? (But nothing seems to be getting hot other than the transformer).
Is that 5.6 or 56? But it certainly isn't either (if that means volts)
If you look at the earlier images, you'll see a large heatsink (it's about 80mm tall). This has 3 TO-3 packages, an LM317K, an LM337K and an LM340KC-5.0. the are also 4 other TO-220 3 terminal regulators mounted flat on the board without additional heatsinking. These generate a pair of +/-15V rails.
I wonder if I have another shorted power supply? (But nothing seems to be getting hot other than the transformer).
What's the chance? One of the +15V regulators measures 0.35Ω between output and ground. All the others are several kΩ.
And there's some of those evil tantalum capacitors! Is this going to be a case of deja vu?
U904 (the one on the right) measures 0.35Ω from ground to output. Is one of those tantrums grinning evilly to itself?
I think this has become a repair (Yay!)
Edit: I just looked at that part of the board. Count the tantalums!
And there's some of those evil tantalum capacitors! Is this going to be a case of deja vu?
U904 (the one on the right) measures 0.35Ω from ground to output. Is one of those tantrums grinning evilly to itself?
I think this has become a repair (Yay!)
Edit: I just looked at that part of the board. Count the tantalums!
Last edited:
Or did you mean the actual puppy?
That's an approximately 8 year old papion, an ex-grand champion show dog (and not the sort of show dog that rounds up sheep)
She is the dog that my wife just paid $4000 for a knee operation on.
That makes my hobby look cheap!
