Products > Test Equipment

Replacement for Fluke 700013 IC (quad SPST analog switch)

<< < (5/16) > >>

Kleinstein:
The DG411 is low resistance, but also has more leakage. The DG211 should be a lower leakage counterpart - AFAIK with the same pin-out  :-*. Leakage can be a problem with the higher resistors.

softfoot:

I am following this project with great interest,  and look forward to a working solution.
Regards,
David.

Noopy:
Hi all!

TiN had sent me a 700013 and I have taken some pictures of the die.  :-/O






Well you know what it does...




The die is 2,6mm x 2,3mm.




A 1982 Fluke design.
QAS-B seems to be the name of the switch. In the servicemanual of the 8842A the chip is named "IC CMOS, QAS, PLASTIC, HIGH, B Grade". Sounds feasible!  :-+




Fluke used eight masks to manufacture the switch.






Some test structures.
On the right side there are three different type of the resistors.
On the left side it seems like there are complementary MOS transistors. You can see the small spot in the middle of the cross where the thin gate oxide is placed.
In the middle perhaps some bipolar transistors?




Let the music play!  ;D




The four switches are easy to spot.
You can see that the left control side is supplied with +5V and 0V while the right switching side is supplied with +7,5V and -5V (of course other voltages are possible).




Some protection at the input pins.




On both sides of the analog switches there are metal lines to isolate the switches. On the left side +5V/0V and on the right side +7,5V/-5V.




The inputs without the latches are fed into a kind of a push-pull-stage before reaching the control circuit.




The control circuit contains the latch, the AND gate and the chip select.
The upper area is connected to +5V, the lower area is connected to 0V.
On the right side of the picture there is the transition from +5V/0V to +7,5V/-5V.




To control the analog switch quite a big differential push-pull-stage was integrated.
It seems like the transistor areas were optimized for ideal switching behaviour.






Now that is interesting: There are two small MOSFETs doing the switching of the analog channel. But there are two additional bigger MOSFETs connected to two unused bondpads. It seems like you can choose which transistor you like: a big one with low resistance but with a high leackage current and parasitiv capacitance or a small one with a higher resistance but with lower leackage and capacitance.  :-/O


https://www.richis-lab.de/aswitch01.htm

 :-/O

JoergR:
Hi all,

This is my first post on the EEVBlog - pls be gentle. ;-)

I believe I too have a Fluke 8840 with a faulty 700013 - but before I venture to desolder everything, I wanted to know:

How did you ascertained that this was indeed the faulty chip?

Background:
All DC readings are garbage, no autoranging. Display works. I removed the AC unit and placed it in another meter where it works just fine.
All self tests except 11 fail. Voltages appear within specs, so the power supply seems ok.
Following the manual, I can get correct voltages displayed when placing a voltage directly between TP103 and common low and I get a zero reading when shorting TP103 to common low. The front panel switches allow me to switch the ranges of the display, i.e. I can display 1.3V as 13V, 130V etc. So the ADC and the Voltage reference appear ok. With a short on the input, all ranges _do not_ display zeros. The manual says:

If zero is not displayed for r3 and r5 with HI and LO shorted, then Z301 or U302-D is probably bad.

Now, I also checked track and hold. The voltages on the chips are ok, but the waveforms are off - They don't extend below 0 far enough. The shape is similar, though.

Hence, my question: Can you suggest another procedure to test whether U302 Chip, the Fluke 700013, is really at fault? I haven't given up the hope that the fault lies somewhere else ;-)
Thanks in advance!
Joerg

Kleinstein:
One should be able to test the input amplifier part up to the S&H stage separately. So set a range and check the result from a signal at the input at TP302 and directly at the S&H input (e.g. pin 17=18 of U301). There can be a little offset, but should not be much.

When seeing which range does not work and how it fails one could than do a more specific test.

Navigation

[0] Message Index

[#] Next page

[*] Previous page

There was an error while thanking
Thanking...
Go to full version
Powered by SMFPacks Advanced Attachments Uploader Mod