-
Obscure benchtop capacitance meter Nichiei Denshi Kogyo NS-535 MkIII. Any info??
Posted by barijan on 25 Jul, 2020 22:43 -
Recently I've scored a very nice japanese vintage digital capacitance meter, perfectly functional and accurate.
It's fairly easy to use, but its history background is a total mistery.
On the front panel it says - MODEL NS535 MkIII SER. 22425 CAPACITANCE METER - made by a company called Nichiei Denshi Kogyo K. K.
I've searched all over the internet, both in english and japanese, and found absolutely nothing, just very few clues about the company that seems to be defunct for long.
The little monster is very well built - for the standards of its age - with lots of discrete components and socketed ICs. See attached pictures.
Edit: If you're not able to expand the pictures due to the forum's picture bug, I've republished the pictures, further below on this thread.
Althoug it's working fine, nobody knows if it may need some maintenance sometime.
Also, there're two large connectors on the back that seems to be some sort of interface bus, but doesn't seem to be GPIB (IEEE 488) compatible.
Does anyone have any information about this equipment?
Any information or comments will be very welcome and appreciated.
Best regards!
-
At least the company still exists:
https://www.dnb.com/business-directory/company-profiles.nichiei_denshi_kk.2b1e52508c291f3a70ef6ca868e0ee01.html -
At least the company still exists:
https://www.dnb.com/business-directory/company-profiles.nichiei_denshi_kk.2b1e52508c291f3a70ef6ca868e0ee01.html
Thanks tkamiya.
Maybe the remains of the same company... but probably not.
Correct me if I'm wrong, but I've learned that "Nichiei Denshi" is a very common prefix for many japanese electrical related companies, focused on the american market and means something like "american-japanese electronics".
Also "K. K." is a sufix that probably means CO. (corporation).
So, the full company name must include "Kogyo", that seems to be the actual company name.
I'm not sure but I think that a free translation of "Nichiei Denshi Kogyo K. K." would be something like "Kogyo American-Japanese Electronics Co."
I've found this name on the pictures of some other vintage devices, but no info about them as well, just loose pictures found on expired ads from auction sites, as can be seen below.
Thanks anyway.
-
Even without the "Under license from Marconi Instruments, Ltd" label, one can tell the Marconi heritage from the meters with pivot axis on top.
Edit: once again, the famous picture bug showed up. When I looked earlier at the capacitance meter pictures in the first post, they looked normal, but now, after looking at the VTVM pictures in the last post, the first three capacitance meter pictures now load as VTVM pictures. -
I reported a similar issue on the Simple Machines Bug Reports forum:
https://www.simplemachines.org/community/index.php?topic=574166.0
However they closed the Bug as they could not reproduce.
Reported a new Bug ticket today. Hopefully this time they will further investigate.
https://www.simplemachines.org/community/index.php?topic=574216.0
I have also sent a PM to Dave Jones about this, so that he is informed and can follow-up about this.Even without the "Under license from Marconi Instruments, Ltd" label, one can tell the Marconi heritage from the meters with pivot axis on top.
Edit: once again, the famous picture bug showed up. When I looked earlier at the capacitance meter pictures in the first post, they looked normal, but now, after looking at the VTVM pictures in the last post, the first three capacitance meter pictures now load as VTVM pictures. -
I'd have to see the Japanese version of the name to be sure of the exact company, as names and alternate characters are often used in company names, which translate to the same word in English.
Nichiei could be a family name, a made-up word from mashing two English words together creatively or simply a straight Japanese word maybe 'spelt' creatively even, but without seeing the actual Japanese, it's impossible to tell.
The rest of the name is as blueskull says, 'Denshi Kyogo KK' is 'Electronics Industry Corporation' (as in publicly traded company).
This is the closest I've found so far, but they seem to be a parts supplier, not a test equipment manufacturer.
ニチエイ電子株式会社
http://www.nichiei-ele.com/company/index.php -
When I saw this thread, I had to take a double take. The company name is just one letter off from the company I worked back in Japan. It was specializing GPIB.
The abbreviation K.K. indicates it is a "stock funded company". It's more of a legal thing than anything else. I thought it may have been a descendant of the original company with different focus. But who knows.... it's been 40 years.
Japanese company has a reputation for not releasing documents so it's unlikely anything will be found on internet. Booooo! -
Japanese company has a reputation for not releasing documents so it's unlikely anything will be found on internet. Booooo!
Looking squarely at you, Advantest......
Although on the other hand, Kikusui is really good about releasing info, so not all bad.
-
...once again, the famous picture bug showed up. When I looked earlier at the capacitance meter pictures in the first post, they looked normal, but now, after looking at the VTVM pictures in the last post, the first three capacitance meter pictures now load as VTVM pictures.
It seems that the picture bug is somehow intermittent, maybe browser dependant or session id dependant. I've opened the thread on different browsers and the bug showed up on Fireforx and on my mobile, working correctly on Chrome (for now).
Anyway, I'm uploading the capacitance meter pictures again here.
-
Nice score! I am also learning a bit of Japanese on the side.
-
Here's some update about this capacitance meter.
Since I have found no information about it yet, I've decided to do some basic reverse engineering, just for fun.
There's no CPU whatsoever. Just ADCs, lots of CMOS logic gates, counters, dividers, encoders, decoders, op-amps, and so on. Discrete electronics on its most.
I have learned that it uses two separate A/D converters, Teledyne TC8750 for Dissipation Factor (D) and Intersil ICL71C03 working together with an ICL8052 for capacitance (C).
Both A/D converters output data in BCD and requires BCD to 7-segment decoders and drivers.
Fortunately the datasheets of those are available on the web.
It's an overkill since they are not using all digits/counts these ADCs are capable of.
Attached is a close picture of the A/D section.
Notice the weird little brown resistors that look like capacitors. Never had seen any like these before.
All ICs are dated as 82, 83 and 84 so it probably has been built in 1984. Most of the logic ICs are from Motorola, made in Korea, but there're also some japanese Hitachi, Toshiba and Oki. The op-amps are from JRC.
Fortunately most of the trimmers are clearly labeled, although I had no need to touch them. This beast showed to be very accurate already.
I still intrigued by the data connectors on the back.
Due to the lots of buffers and optocouplers they seem to be output only. Probably BCD.
I'll post some more details and pictures later on.
-
I noticed some clues that may explain why nothing about this equipment could be found on the web so far. It seems to be a very limited serie.
The main clue was the fact of the serial number to be etched on its front panel, altogether with the model number and equipment description. Yes, every unit is personalized with it's own front panel lettering. Also, the serial number is replicated on every board inside.
I found that this number were exclusive because the guy who sold it to me had another identical unit, but with another number on the front panel and boards.
At first, when I saw "SER. 22425" at the front panel, I wondered -"No way, it cannot be the serial number". But it actually is.
It seems to me that these units were built by demand and, at that time, they could be customized with some options and add-ons, that's why the front panel indications may be different as well, depending on the options installed.
The I/O connections on the back, the triggers and comparators indicates it might be used for some kind of automated component sorting.
Definitivelly, it's not an "one-size-fits-all" consumer product.
I'm wondering about the original price tag of such things. Probably very expensive.
Note its accuracy. The display shows exactly 0.0050nF (5.0pF) when connected to a precision 5pF capacitor. It has a resolution of 0.1pF, obviously focused on small capacitances, since its max range is 1uF (up to 2uF using a feature of the D/A converter to extend range a bit).
Besides it's rarity I still hope to get some more information about this nice piece of the electronics testing history.
-
With the thumbwheel switches on the front panel to set test limits, this would appear to be a custom unit for a capacitor manufacturer to verify production.
-
With the thumbwheel switches on the front panel to set test limits, this would appear to be a custom unit for a capacitor manufacturer to verify production.
That's precisely right.
Other details that corroborate with this observation are the high speed of the measurements (about 300ms or less) and the fact it uses the industry standard parametes (under pure sine wave @ 1kHz x 1V RMS) frequently used for quality control of general purpose capacitors below 1uF.
BTW, I'm very impressed with the precision of this sine wave, less than 0,02% error on both voltage and frequency, even not using a cristal oscillator.
Besides, I noticed that it has both opto-coupled as well open-collector outputs, triggered when measurements are under or over the values set on the thumbwheels.
These can control actuators to sort components to different trays or packaging machines. For instance, high tolerance capacitors to tray one, low tolerance ones to tray two and defective or out-of-specs ones directly to the trash bin - or to the chinese sellers on eBay and Aliexpress.
Parallel BCD data is also output, that may used for printing, counting or statistical purposes.
Such units normaly where installed on testing-sorting machines like these on the pictures below.
-
Given that, I'm surprised that the measurement connections appear to be only two BNCs, instead of the 4-terminal measurement often used for external component fixtures.
-
Given that, I'm surprised that the measurement connections appear to be only two BNCs, instead of the 4-terminal measurement often used for external component fixtures.
Yeah. I've noticed that as well. Probably it's for not so critical tolerances, besides, it's from the early 80's.
Despite that, even with only 2-terminals + ground-guard it still very stable and precise if using quality shielded cables.
Also, there's a vernier potentiometer to compensate eventual stray capacitance over the cables, zeroing it. I noticed that even with a half-meter cable it requires only one quarter of a turn on the vernier to correct it. Note that the vernier does 10 turns.
Here's also a picture showing the measuring a 15pF capacitor, but in this case with no cables, therefore no zero correction necessary here.
On the 100nF range and over, the stray capacitances over the cables are pratically negligible.
-
I have an ancient Wayne-Kerr admittance bridge, with decade ratio transformers, hermetically-sealed standard capacitors, and dual magic-eye tube null indicators that also has two BNC connectors. With the ratio-transformer circuit, it is straightforward to null out the parallel capacitance, and the admittance bridge (by definition) measures the parallel capacitance and parallel admittance (loss) of the DUT. After replacing a dodgy replacement power cord, it still works like a charm.
(I remember an episode of Frank's Place on TV, where the eponymous hero asks a New Orleans voodoo doctor if her anti-witch powder works, to which she replies "like a charm".) -
I have an ancient Wayne-Kerr admittance bridge, with decade ratio transformers, hermetically-sealed standard capacitors, and dual magic-eye tube null indicators that also has two BNC connectors...
Wow. Post some pics, pls....(I remember an episode of Frank's Place on TV, where the eponymous hero asks a New Orleans voodoo doctor if her anti-witch powder works, to which she replies "like a charm".)
LOL
-
Finaly I've found some more consistent information about the company behind this capacitance meter.
It was a joint venture - created in 1961 and now defunct - between the japanese company Cornes and british Marconi, hence the name Nichiei Denshi Kogyo, that means Japanese-British Electronics Company, also known as AJEC - Anglo-Japanese Electronics Co.
Japanese name: 日英電子工業（株） = Nichiei Denshi Kogyo K. K.
There're other companies with very similar names in Japan but not related at all.
Cornes still active - and big - but no longer manufacturing such devices, therefore no available documentation or support anymore.
The Marconi's influence on this capacitance meter is noticiable on the style of the etched lettering on the front pannel, including serial number, very common on their instruments of that era. On the other hand, the internals are very japanese-like, considering the extensive use of OKI, Hitachi and Toshiba parts.
Here's an excerpt from the "our history" page from Cornes website:Quote1961 - 日英電子工業（株）（AJEC）を英国マルコニー社と設立。
スポット信号発生機（Spot Signal Generator）の販売を行うとともに、測定器4機種を国産化。この頃、日本には同等の製品、技術は存在しなかった。
60年代の日本の造船活況期、コーンズは船舶用無線機やレーダーの輸入販売も開始し、船舶機器の市場をリードした。
1961 - Nichiei Denshi Kogyo K.K. (AJEC) is established with Marconi Corporation in the UK.
In addition to selling the "Spot Signal Generator", domestically produced four other measuring instruments. At this time, there was no equivalent product or technology in Japan.
During Japan's booming shipbuilding industry in the 1960s, Cornes also began importing and selling marine radios and radar, leading the marine equipment market.
The picture below shows another interesting analog LCR bridge/meter made by this company, notice both AJEC and Nichiei Denshi Kogyo logos.
Also, some detailing on Marconi's characteristic etched lettering including serial numbers, making "personalized" front panels for each instrument.
-
Because of this pandemic thing I'm having plenty of time for playing with my collection, so I've decided to spend part of this time doing some reverse engineering in order to figure out the functions of the connectors on the back.
Although I don't have any documentation about this device and its obscure connections, at the end it was a very satisfying task, since this piece of equipment is all glue-logic based, assembled on thru-hole boards. No microprocessors, no ROM or RAM, instead, only a well-designed logic altogether with a great analog section.
All I need was a continuity checker, some IC datasheets and a bit of patience. An oscilloscope and a BCD decoder helped also, speeding up this task, but could go through without them.
Basically, the 50-pin centronics connector (OUT-1) has several buffered 4-bit BCD outputs that mimics the display.
The 24-pin connector (OUT-2) is not GPIB. I wish it was. Actually it's a bunch of opto-coupled control lines activated by the thumbwheel comparators on the front panel. Probably to control the actuators of a testing and sorting machine. Although all the opto chips are there, many of them are inactive, reserved for additional options.
The only input is an secondary external trigger, TTL level. Another external trigger connector is available on the front panel, but differently, this one on the front panel may be triggered by a simple momentarily switch.
Here's the full pinout.
-
During the process I've found that an additional digit for dissipation factor could be easily added.
A disabled LED 7-segment digit was sitting there on the display board, as well the connections. I just had to jump a trace that has been cut for that digit, add one more MC14042 (4-bit latch) to an unpopulated socket and bingo! One more digit of resolution, almost for free.
There’s only one minor drawback. The comparator will consider just the two most significant digits for D.
It's perfectly possible to enable the additional comparator for the third digit and install the respective thumbwheel, since there's unpopulated space on the panel pcb for it. But I think it would be overkill.
My next step is to build and program an interface, probably using an Arduino Mega since it has the sufficient data lines for the more than 30 parallel BCD bits and other flag bits needed. With such interface I will be able to send the data both to a more comprehensive and user friendly alphanumeric LCD display and also directly to a PC via USB/Serial, making data logging possible.
Anyway, keep in mind I'm doing this absolutely for fun and curiosity, besides learning a lot about such old piece of technology.
-
This is so relatively easy to add that it is almost a password crack on modern equipment
-
Nah, adding that thumb wheel ain't overkill, it's part of the fun of the hobby!
-
Nah, adding that thumb wheel ain't overkill, it's part of the fun of the hobby!
Indeed. And I'm very tempted to do it. 
The difficult part would be finding one with exactly same vintage style and dimensions. Specially here in Brazil.
Most of the modern ones are thinner and use push-buttons instead of wheels.
-
Seeing as it's Japanese, I'll keep an eye out when I'm in Akihabara, might get lucky.