Some tests and maintenance of my Solartron 7081

[Edward3121]

I bought a Solartron 7081 this month, and I did some tests and maintenance. I don't speak English well, so sometimes I will use Chinese to make the expression accurate. I have found some relevant information, documents and files, which has helped me a lot.

The SN of my 7081 is 000534, and the date code on ICs is about 84, so I believe this 7801 was produced in about 1985.

The Ni-Cd battery leaked, and it has been replaced by an ER14500 battery. Soon after the power is on, the reading will drop a lot. and finally solved by replacing IC302. These are done by the seller.

After shipping to me, I try to use RS232 to control it. But the connector is DB-25, so I try use a USB isolator, a USB to RS232 converter and cables and so on to connect to 7081. It seems worked. But soon I found it was hard to complete a successful communication. Because when I issue the correct command, 7081 always returns E1 or other strange characters. At first I thought this was due to a complicated connection method that caused a problem with data transfer. So I decided to make an Isolated USB to RS232 converter myself, and it will be introduced later. So I can only use GPIB to control it.

I use a MeterCare GPIB to USB/LAN/Wi-Fi/TF, and it work will, and the PC software can draw according to readings and real-time temperature. I connected the 7081 to the 7V output of the LTZ1000 reference. After several days measurement, I found this 7081 looks like an thermometer…… . The Zener of this 7081 need a recalibration. And this 7081 also have a “glitch” every 15 min caused by AZ.

7081 use the “well known” ER3400, a kind of EAROM. This early EAROM has a short life and is easily damaged. So I bought three used ER3400 on Taobao(Alibaba) for testing. The battery on PCB4 cannot use a primary lithium battery, so I also bought three Ni-MH battery for replaceing.

My test equipment and environment are limited, with only one LTZ1000 Reference and two HP3457A DMMs. And I will update the test and maintenance progress one after another.

[picburner]

And this 7081 also have a “glitch” every 15 min caused by AZ.

This flaw had discovered (and even solved) by Mickle some time ago.
Congratulations, your instrument is younger than mine!

[Vgkid]

What is the chiron 7120a voltage reference?

[Edward3121]

This flaw had discovered (and even solved) by Mickle some time ago.
Congratulations, your instrument is younger than mine!

Thank you. I've already downloaded the patch on this website: http://www.ko4bb.com/getsimple/index.php?id=manuals

But I think the AZ bug is not the most critical issue. I decided to try it later and solve other problems first.

So I don't think Solartron has produced many 7081s at all. Think about the HP3458A, it is still in production.

[Edward3121]

I measured the floating power supply. The voltages are:
+28V --> +26.7V
+15V --> +14.5V
-15V --> -15.0V
-10V --> -10.0V
-28V --> -27.2V

The noise are: （See the attachments）

I found the negative voltages have some glitch. But I have no idea about this.

[Edward3121]

What is the chiron 7120a voltage reference?

Chiron 71xx is a series of DIY Reference, which have different options. But the main components are purchased by myself. I use LTZ1000CH, and the Ref board is HSH ver6.3, is a stable verision. I didn't trim the temperature coefficient of the ref board, because I have no conditions(equipment, environment, etc.).

So, I just "think" that its 7V output is stable and the temperature coefficient is low enough.

[Edward3121]

I made an Isolated USB to RS232 converter，using CP2102, ADuM5241 and SP3232. ADuM5241 has an internal isolated 50mW power supply suitable for the load of SP3232 at 9600bps. After the PCB board and other components shipped to me, I found I made a mistake.... I didn't take care of the pitch of the DB-25. But it can still be used. I will made a new one with the right pitch.

Now I can control the 7081 via RS232 port. But I found I still hard to complete a successful communication. The PC can receive right data from 7081, and it seems that 7081 didn't receive the right data from PC. I try to measure the TxD signal on PCB3 IC4 Pin8, and decode it. The signal looks very good, and the decode is right, why does 7081 not behave normally? I noticed that there is a <CR> and a <LF> after the instruction. But 7081 only need one <CR> or <LF>, why there are two?

I use a software called "XCOM", it will automatically add a <LF> after <CR>, but without <CR>, there will be no <LF>. I have to use another software. I tried "Treminal v1.9b"(downloaded from ko4bb), it works very well.

I read out the calibarate data:

CALIBRATE,DUMP

OK
Vdc calibration constants
-3.39746475E-06
 163.963581E-03
-1.31130218E-06
 1.63743720E+00
-774.860382E-09
 16.3751919E+00
-1.19209289E-06
 163.774825E+00
-834.465026E-09
 1.63783555E+03
Vac calibration constants
 325.469603E-03
 3.25636314E+00
 32.4725567E+00
 324.885288E+00
 3.24942276E+03
kOhm calibration constants
-3.39746475E-06
 165.954848E-03
-1.31130218E-06
 1.65959544E+00
-774.860382E-09
 16.5967363E+00
-1.31130218E-06
 165.976621E+00
-774.860382E-09
 1.65985627E+03
Top Ohms range calibration constants
-774.860382E-09
 9.90320171E+03
 602.445125E-03
Ratio terminals calibration constant
 16.3757144E+00
Zener current token value
091

The zener current parameter is 091，and I will change this parameter to see what effect it has on 1N829.

I also try to read out the NVM data using the MONITOR command, but after transfer to the floating processor, 7081 doesn't respond normally, I tried to change S1 on PCB5, but there was no result. I will try later.

[Edward3121]

I did some measurements on different locations in 7081 using HP3457A, and unless otherwise specified, the LO terminal of the 3457 is connected to the TP305. I found that +10V Ref or -10V Ref has very low frequency oscillations, but 20V(+/-10V ref) is not, 0V(IC306 Pin2 to TP305) is not either.

HP3457A (Divider Buffer), S7081(7*9 short input) +10V Ref, Powered ON, Temp=1N829, RoomTemp≈19C


HP3457A (Divider Buffer), S7081(7*9 short input) R351 left side, Powered ON, Temp=1N829, RoomTemp≈20C （this pic may have some scale mistake）


HP3457A (Divider Buffer), S7081(7*9 short input) Volt drop of R351, Powered ON, Temp=1N829, RoomTemp≈20C


HP3457A (Divider Buffer), S7081 1N829 Volt, Powered ON, Temp=1N829, RoomTemp≈20C (The time this one took was too short)


HP3457A (Divider Buffer), S7081(7*9 short input) Volt drop of R304, Powered ON, Temp=1N829, RoomTemp≈20C


HP3457A (Divider Buffer), S7081(7*9 7V input) IC351 Pin7, Powered ON, Temp=1N829, RoomTemp≈20C


HP3457A (Divider Buffer), S7081(7*9 7V input) +/-10V Ref, Powered ON, Temp=Not Care, RoomTemp≈19C

[Kleinstein]

In old meters the opto-couplers are a common source of failure. This could effect the reference current setting.

[Edward3121]

In old meters the opto-couplers are a common source of failure. This could effect the reference current setting.

Yes. IC302 has already replaced by a "new" one. "new" means a used one brought on Taobao(Alibaba), not a brand new one.

The current on 1N829 looks very stable now.

[Edward3121]

I measured the battery on PCB3, and it is dead. This battery is an inorganic lithium primary battery, and its nominal voltage is 3.6V, so I believe it is a Li-SOCl2 battery. I happen to have some of these batteries and they are the same size. So I replaced it.









I bought three used ER3400 on Taobao(Alibaba), I'm not sure they are in good condition. After I received them I did some testing.  I removed the original ER3400 on the PCB4 and installed a "new" one. Power on, 7081 displayed "NVM FAIL". (Because the data in NVM is not related to 7081 at this time, the checksum cannot be passed, so NVM FAIL will be displayed.)

bought on Taobao(Alibaba)


Original ER3400


RESET Swith OFF and enter:

MONITOR<CR>

and

I<CR>

Wait for tens of seconds, the 7081 responds as follow, means the ER3400(NVM) has been initialized. (Caution: This will clear all calibration data on the NVM, please be cautious.)



Reset 7081 （RESET Switch ON), it will  display "CAL INCOMPLETE". (Because the initial NVM writes the chequer board pattern and the checksum, CAL INCOMPLETE is displayed instead of NVM FAIL)

But one of the three ER3400s always display NVM FAIL, try to initial it again is the same NVM FAIL. So I believe that the life of this ER3400 is exhausted.



The ER3400 is a very old type of ROM that, unlike modern EEPROMs, consumes life even when reading data.（ER3400是一种非常古老的ROM，与现代EEPROM不同，即便是读取数据也要消耗寿命）In order to protect the original ER3400, I use these 2 good ER3400s for Zener current adjustment, other calibration data is temporarily unimportant.



I failed to read the data in the NVM directly out through the RS232, but read out the calibration data. But there is no instruction to write the calibration data directly back to the NVM, only recalibration, so I am considering whether to make a PCB for ER3400 data copy. I will first try to successfully access the NVM using the MONITOR command.

[guenthert]

What is the chiron 7120a voltage reference?

It's mentioned here: https://www.eevblog.com/forum/metrology/volt-reference-diyer-ltz-10v-pwm/msg1250869/#msg1250869