If this had been done before SY left his old job, he would have had an electron microscope at his disposal.
Oh well.
se
Actually, I still have access to one at UT, but since I'm not disassembling them, there's really nothing to look at.
You can't blame the guy for discovering that Neokorken had the toxic effect of altering the appearance of the user to something from Dr Pepper land.
(ever seen the Team-Yeah pic ? )
Attachments
Hi SY,
Warnings:
I'm sorry for going into painful detail here SY. I'm well aware you don't need this information, but others reading this may.
I'm pretty sure your 3581A only needs a bit of elbow grease to come back to reliable life. It might be a good idea to check / replace the power supply capacitors in there.
Somewhere around here I started a thread on improving the performance of older test equipment. You may find some hints in there, or possibly pose the question for input. Scott Wurcer may be able to help in that department. I believe that I was talking about my 339A in that thread. Those LME49990 ICs may be just the answer I'm looking for in the oscillator and input amp stages for the 339A. Since the 3581 has a lower bandwidth, the LME49990 may be a good solution there as well. Let me know how everything works out for you please.
-Chris
I was speaking with someone else with similar problems with their just acquired 3581C. It think it turned out to be the socket for the input FET. The other issues were related to the front panel switches. The cleaning does not involve a spray type delivery. A cotton swab with a mechanically reduced cotton end. Place your fav contact cleaner in the cotton portion and gently rub each contact ring. Rotating the switch makes this go much easier, but take care not to bend the fingers! After some time of careful, light rubbing will change the contact surface from dark brown to a brass colour.
Warnings:
- The "stick" portion of the cotton swab must be wood or cardboard.
- Make sure not to drip anything into trimmer capacitors or anything else.
- Take your time and go slow.
- Lubricate all bearings and detent balls.
- Clean up anything that may have dripped. Assume it has.
I'm sorry for going into painful detail here SY. I'm well aware you don't need this information, but others reading this may.
I'm pretty sure your 3581A only needs a bit of elbow grease to come back to reliable life. It might be a good idea to check / replace the power supply capacitors in there.
Most probably don't need that. Watch out though, HP designs tend to perform extremely well, and something that looks simple may be difficult to surpass without a lot of design effort on your part.
Somewhere around here I started a thread on improving the performance of older test equipment. You may find some hints in there, or possibly pose the question for input. Scott Wurcer may be able to help in that department. I believe that I was talking about my 339A in that thread. Those LME49990 ICs may be just the answer I'm looking for in the oscillator and input amp stages for the 339A. Since the 3581 has a lower bandwidth, the LME49990 may be a good solution there as well. Let me know how everything works out for you please.
-Chris
Interesting stuff this nano technology. My company spends lots with IBM, so were priviledged to get a talk from Dr Don Eigler, I thought he actually invented the scanning tunnel microscope. This stuff is a wee way off but essential if we want to keep Moores Law relevant. How else can we build skynet😀
SEM, not STM. 😀
Actually, an old friend of mine built one of the first STMs outside of IBM when he was at Columbia. Turns out that (according to him) it was easier and simpler than an SEM. Most of it was constructed with surplus parts, yet it still allowed him to characterize some nanofabrication he was doing on graphite surfaces.
Chris, appreciate the hints. One more warning for the unwary: spray cleaner in the switches can cause the markings to come off the plastic ring that indicates range! I cleaned the switches and reseated the FET, but that made no real difference. The previous owner (John Curl) did a really nice job of maintaining and storing the unit, so everything was very clean inside.
The 49990 has an en of 0.8nV/rt Hz, and that's something the HP guys just didn't have back in the '70s. The main difficulty is that it's SMT-only, so I'll have to try to find an appropriate proto board or adapter at Fry's (our local electronics supermarket) to mount it.
Actually, an old friend of mine built one of the first STMs outside of IBM when he was at Columbia. Turns out that (according to him) it was easier and simpler than an SEM. Most of it was constructed with surplus parts, yet it still allowed him to characterize some nanofabrication he was doing on graphite surfaces.
Chris, appreciate the hints. One more warning for the unwary: spray cleaner in the switches can cause the markings to come off the plastic ring that indicates range! I cleaned the switches and reseated the FET, but that made no real difference. The previous owner (John Curl) did a really nice job of maintaining and storing the unit, so everything was very clean inside.
The 49990 has an en of 0.8nV/rt Hz, and that's something the HP guys just didn't have back in the '70s. The main difficulty is that it's SMT-only, so I'll have to try to find an appropriate proto board or adapter at Fry's (our local electronics supermarket) to mount it.
SchmartBoard|ez SOIC Prototyping Boards
Typical example of their products. They make a very clever adaptation of a PC card where the solder mask between the copper lands is higher than the copper so the chip falls into place!
A big note of caution, as the lands are recessed you must flux them with a liquid flux before soldering.
Oh yeah I did a review http://www.schmartboard.com/schmartboard_at_040109_01.pdf
Last edited:
I strongly recommend a 15 turn 1/8" air core coil and 47 pf cap to ground in front of the input to National's low noise op amps.
Just to be clear, the coil in series with the input and the cap to ground (i.e., lowpass LC filter)?
So just what were you trying to measure with that fancy voltmeter when it died?
(We got nothing else going right now, so may as well bug you about what kind of tests you were running).
(We got nothing else going right now, so may as well bug you about what kind of tests you were running).
Noise. I put the Bybee in series with a resistor (as Scott suggested) and was comparing the Johnson noise to a 0R025 in series with the same resistor, then checking the temperature dependence of that noise.
- Status
- Not open for further replies.