And another bites the dust.
A 34970A switch multiplexer with built in DMM ( 34401 style )
This one had all kinds of problems. The display was all garbled like segments lighting up together , the rotary encoder did not work quite right either and it kept coming up with errors at power up.
The front panel board
![](http://www.siliconvalleygarage.com/eevblog/34970ateardown/34970a-rot.jpg)
The bourns rotary encoder was at the end of its life i gueass. replaced it with a new one. Works fine now. These are mechanical encoders with a cheap plastic axle. i've seen many of these go bad..
The display was another thing... the darn driver logic sits under the glass... luckily it's a single row affair.
![](http://www.siliconvalleygarage.com/eevblog/34970ateardown/34970a-chips.jpg)
I simply bent the display upward. one of the drivers had shorted outputs. these are high voltage 32 bit shift registers made by TI that discontinued them a long time ago . They are now available from a company called SuperTex. Luckily the local 'surplus parts' dealer had a bin of these so i grabbed a few for a dollar each. PLCC44. since there's only two of these i swapped both. After bending the display back it works perfectly fine.
Powerup was intermittent. sometimes it would beep and no display , sometimes it would start fine.
![](http://www.siliconvalleygarage.com/eevblog/34970ateardown/34970a-main.jpg)
THe problem was the semi-rigid flatcable between mainboard and display module. ( left cable above with the blue stripe) these are sincle strand ribbon cables that use a press-fit connector. they are horrible... contacts oxidize over time. so i unplugged it from its connector , pulled the top connector plastic off , put cable back in and soldered each strand to the connector pin. problem solved. works like a champ.
The brains of this thing is a 80196 with the same ASIC found in the 34401 multimeter. Actually this machine is nothing but a 34401 with a scanner front end. The analog section is housed in a plug in module.
![](http://www.siliconvalleygarage.com/eevblog/34970ateardown/34970dmm.jpg)
Essentially the same guts as a 34401 mulitmeter. LM199 refernce diode and the same multislope asics. the problem with the module was a burned inductor ( left of the relays ) little TDK choke at the current input. i guess someone must have overloaded it. The module has no fuses in it and relies on the fuses installed in the multiplexer boards... someone had replaced one with a stub of wire... and fried the meters input.
![](http://www.siliconvalleygarage.com/eevblog/34970ateardown/34970plugin.jpg)
i got 3 multiplexer boards. 20 voltage inputs , 2 current inputs and 2 4 wire ohms inputs. each
the relay counters on some of em read extreme high numbers , like millions of actuations. i ordered these relays from digikey ( they're like 1.25$ each ) i need to swap about 10 in total. the other contact counters are all in the thousands range. these modules have their own CPU and EEPROM that holds the actuation count.
Machine is healthy again and works fine. Another addition to the 'pile'