This repair/restoration is complete. I am posting it in this forum in case it helps someone, in the future.
The meter was purchased at a low price for parts, with a bad display and unable to measure current. Quite a few people on the forums cite first generation 34401A display problems as unrepairable. The display driver is long since NLA.
The original display had leaking segments which made the instrument a little difficult to read but mostly, the extra segments spoiled the luxurious feeling of using a device of such ridiculous overkill for the task. The display was swapped for a Chinese copy, sourced from eBay, and that completely repaired the display. (you can see the original display on the left side of the 3478A in the pic)
The current measurement issue was solved by replacing the current source OpAmp U201. I had watched a YouTube repair video on this issue and it seems to be a common problem (occasionally, a couple of other parts also go).
At this point, the meter was in good shape but the least significant digit was unstable on DC voltages and impedance. Replacing the electrolytics (4) and tantalums (12) made it wonderfully stable.
The 34401A is far more stable at 6.5 digits than the 3478A is at 5.5 digits, so I am currently waiting on a box of fresh capacitors for the 3478A, as well.
A couple of the tantalums are extremely difficult to access with the electrolytics in place, so I pulled the three blocking caps before replacing the local tantalums and then put new electrolytics in place. Fortunately, the main board comes out easily making the job real easy. It took about an hour to tear it down, make a list, and order from Mouser. On arrival, the caps were installed and the unit re-assembled in about an hour.
This is an easy job but I took my time because it is such a nice instrument. I used two irons to remove the tantalums, because there are a lot of devices on this board and I worry that hot air would have caused collateral damage.
Some quick testing with a voltage/current/resistance standard showed the device is extremely close to spot on. If left on for an hour, it gets within 20 micro volts of the reference but but the reference is two years old and I prefer to not further expose this device to my hackery.
The original electrolytic caps were KME, 105C rated caps.
I replaced the electrolytics will mention that nearly all of my capacitor inventory is Nichicon. I mostly work on Japanese audio equipment so this is what I have on hand.
The tantalums are KEMET and it's important to get the correct package size so they fit on the pads properly.
At this point, I have not replaced the voltage regulators, nor do I intend to, as they are close enough (but not right on).
Overall, this wasn't a lot of work and now I have a beautiful meter. If your meter has not been re-capped and the least significant digit or two jumps randomly, you may wish to re-cap to make the meter wonderfully stable again.
The meter was purchased at a low price for parts, with a bad display and unable to measure current. Quite a few people on the forums cite first generation 34401A display problems as unrepairable. The display driver is long since NLA.
The original display had leaking segments which made the instrument a little difficult to read but mostly, the extra segments spoiled the luxurious feeling of using a device of such ridiculous overkill for the task. The display was swapped for a Chinese copy, sourced from eBay, and that completely repaired the display. (you can see the original display on the left side of the 3478A in the pic)
The current measurement issue was solved by replacing the current source OpAmp U201. I had watched a YouTube repair video on this issue and it seems to be a common problem (occasionally, a couple of other parts also go).
At this point, the meter was in good shape but the least significant digit was unstable on DC voltages and impedance. Replacing the electrolytics (4) and tantalums (12) made it wonderfully stable.
The 34401A is far more stable at 6.5 digits than the 3478A is at 5.5 digits, so I am currently waiting on a box of fresh capacitors for the 3478A, as well.
A couple of the tantalums are extremely difficult to access with the electrolytics in place, so I pulled the three blocking caps before replacing the local tantalums and then put new electrolytics in place. Fortunately, the main board comes out easily making the job real easy. It took about an hour to tear it down, make a list, and order from Mouser. On arrival, the caps were installed and the unit re-assembled in about an hour.
This is an easy job but I took my time because it is such a nice instrument. I used two irons to remove the tantalums, because there are a lot of devices on this board and I worry that hot air would have caused collateral damage.
Some quick testing with a voltage/current/resistance standard showed the device is extremely close to spot on. If left on for an hour, it gets within 20 micro volts of the reference but but the reference is two years old and I prefer to not further expose this device to my hackery.
The original electrolytic caps were KME, 105C rated caps.
I replaced the electrolytics will mention that nearly all of my capacitor inventory is Nichicon. I mostly work on Japanese audio equipment so this is what I have on hand.
The tantalums are KEMET and it's important to get the correct package size so they fit on the pads properly.
At this point, I have not replaced the voltage regulators, nor do I intend to, as they are close enough (but not right on).
Overall, this wasn't a lot of work and now I have a beautiful meter. If your meter has not been re-capped and the least significant digit or two jumps randomly, you may wish to re-cap to make the meter wonderfully stable again.
