Advantest R6871E-DC 6.5digit DMM Overview/Teardown

[ve7xen]

I've had this for a while and have wanted to do a teardown for some time since there's little information out there on this model. It's a 6.5 digit meter as far as its accuracy and stability is concerned, but can display 7.5 digits. This particular model (-DC) only does VDC and 2W/4W R, the non -DC version also does VAC and IDC/IAC. From the board markings, the design is from the early 1990s, and almost everything in the meter is made in Japan.

First post will be basic information, second post will have the teardown.

If you like videos, I did a overview / basic operation video:



The operation manual (if anyone has a service manual I'd appreciate it, I couldn't find one, this was hard enough to find) is here.

This is a pretty good meter. The specs are fairly comparable with a modern 6.5 digit, it's got a nice display and a fair amount of functionality for an older device, with math, storage, GPIB and so on. After using it for a few months and keeping it on a reference it seems very stable as well. They occasionally pop up on eBay for very enticing prices, which is why I bought mine despite not being able to find much information on it. Hopefully this post will encourage some others to get in on the action .

[ve7xen]

I spent a lot of time on the teardown video going through most of the meter in a bit of detail, so I won't waste a ton of space on words here, if you're interested in my thoughts, watch the video:

Full gallery of photos is here, all pictures below link to full (ie. huge) images.

Any thoughts on why there's a backup battery in here? There are at least two separate EEPROMs in the meter visible in this shot.


High impedance madness! Is that the reference in the machined can? If you look closely at the full-res shot, you can even see a resistor crammed in directly between the pin and the board. What are those black bricks (there are 3 that look similar around the meter)?


Hang on a second... what on Earth is going on with those ceramic things? The curlies had me confused for a while, but they must be for strain / vibration tolerance - not sure what these things are doing though, they don't seem to have any components on them.


Some interesting ceramic transistor carriers.


Haven't seen this sort of isolation before, but they use a copy of it on each board (digital and analog). I suppose they're isolating the same lines on both boards - either that or there are actually 6 lines, with 3 isolated on each board. Also wondering what the paper-like flat part is, marked M003 in this shot, there are a few similar devices scattered in this section.


Can anyone tell me what's up with the capacitor top right (potted looking black thing beside the big hybrid)? Never seen a capacitor potted like that before, what's the dielectric? It's 33nF, IIRC 50V.


These 5x7 dot matrix LED arrays used for units display are fantastic looking, and I didn't know about them before. Very tiny LED dies too. Interesting.


And we'll end with an action shot.


A neat meter and fun to look at inside.

[alm]

Any thoughts on why there's a backup battery in here? There are at least two separate EEPROMs in the meter visible in this shot.

To store calibration data most likely.

Hang on a second... what on Earth is going on with those ceramic things?

You mean the ceramic PCB inserts? For isolation, FR4, especially if dirty, is kind of low impedance compared to the > 10 GOhm input impedance it probably has.

Also wondering what the paper-like flat part is, marked M003 in this shot, there are a few similar devices scattered in this section.

Resistor networks.

Can anyone tell me what's up with the capacitor top right (potted looking black thing beside the big hybrid)? Never seen a capacitor potted like that before, what's the dielectric? It's 33nF, IIRC 50V.

Looks like a film cap to me, especially given the 50 V rating.

[ve7xen]

To store calibration data most likely.

Perhaps my question could be better phrased as "why have battery-backed SRAM and EEPROM on the same memory bus?". Speed, maybe?

You mean the ceramic PCB inserts? For isolation, FR4, especially if dirty, is kind of low impedance compared to the > 10 GOhm input impedance it probably has.

Yeah, I'm thinking of the 3-terminal looking ceramic substrates. I do assume they have to do with the ultra high impedance input, but I don't know what their purpose is. It's hard to tell for sure but they don't seem to have any components on them, and just connect the air wires to the teflon solder pot things.

Resistor networks.

Thanks, that's what I was thinking but the construction seemed odd.

[Vgkid]

The high impedance section is very interesting, can I get some of the component ID's if you have them. I'cant tell if the to-72 packages are fets/bjts.

[SeanB]

Those memories are EPROM, not EEPROM. You need 12V and a programmer to write to them after UV erasure....

Nice displays, I have some just like that on a modem, but those of mine have a few dead pixels from old age.


IMG_1323 by SeanB_ZA, on Flickr

[ve7xen]

Those memories are EPROM, not EEPROM. You need 12V and a programmer to write to them after UV erasure....

There are two windowed EPROMs with the metalized labels, but also two 28C16 parallel EEPROM, one on the same bus as the EPROM and one over on the other side, maybe for the display/button controller.

[ve7xen]

The high impedance section is very interesting, can I get some of the component ID's if you have them. I'cant tell if the to-72 packages are fets/bjts.

Most of the TO-72s in the high impedance section seem to be 2N4117, including the 3 on those riser boards, but the markings are really hard to read since they're printed on the side and the transistors are packed pretty tightly. 2N4393 is also represented in TO-72. There's also a single TO-93 2SC505 in there, you can just see it in the very bottom left of the second photo. Just past the high impedance frontend there are some odd "tall" TO-92 package 2SC1573 parts that are about twice as high as a normal TO-92, these are seen near the middle of the 6th pic.

[digital]

Thank's for the teardown some very interesting circuitry that instrument must have been expensive to manufacture.Thank's for sharing

[HighVoltage]

Thank for the teardown.
But what happened to your hires pictures?
Could you please upload them again?

This link is dead:
https://www.ve7xen.com/gallery/v/advantest-r6871e/

[ve7xen]

Oops. I got rid of PHP on that server and moved to a static blog generator. The gallery got lost in the crossfire, since it ran on PHP.

I've thrown them back up with a basic directory listing: https://www.ve7xen.com/gallery/v/advantest-r6871e/

Unfortunately the thumbnails in the OP are gone and I won't be able to easily recreate them.

[HighVoltage]

There are some very interesting pictures.
Thank you for re-uploading

[kipp]

Your post was many years ago now, you probably don't care anymore or have figured it out, but since it hasn't been answered here, in case someone else is reading it:

The battery is not for calibration constants, as you say the instrument has EEPROMs for that.  The battery keeps enough RAM powered to store the instrument's state across power cycles.  It maybe powers one chip.  E.g., averaging configuration, math functions, whatever, however you had it set up, the battery is for saving that.  You can pull it and replace it without worries (mostly), you'll just have to re-enter your config.

You should be able to live without the battery at all if you don't care that the instrument is reset to default settings every time you power cycle it.

The only hiccough is that when you put in the new battery, or power it without the battery, the RAM will power back up with its contents scrambled (not cleared, random).  There is a good chance this will confuse the instrument (the data will not correspond to a valid configuration) and it will refuse to function.  It'll get stuck on one of the messages it displays while booting.  In that case you need to find the instrument reset key sequence in the documentation and use that to manually wipe it back to default settings.  Advantest has a table of the reset sequences for their instruments somewhere online.

[ve7xen]

Thanks kipp! Very useful and interesting information. I guess they chose to do it this way to save write cycles on the EEPROM. They were thinking of longevity of this instrument ! So far I haven't done anything with the battery and the instrument still seems to be working fine, but now I don't have to worry about losing cal constants if it ever goes out.

[beanflying]

Thanks for the videos and info. I just scored one of evilbay. They are HUGE compared to my 34401a 

Did you ever do any good with the service manual? In particular mine came set for 100V operation while we run 240V locally so an idea of links or jumper settings to make this happen is what I am after specifically.

[DavidKo]

I know it is quite late answer, but I expect that you need only rewire the cables on the white thing in rectangle. You can measure voltages upfront (I use autotransformer for that) to see the voltages between each pairs. I my switching unit are two gray cables with 3 wires connected in parallel (if I remember that correctly it was white and red). I had used the ve7xen picture for example.