Grex Radio Conference

View Responses

Grex Radio Item 28: Instruments and Meters
Entered by rcurl on Sun Apr 23 21:14:27 UTC 2000:

For discussion of electrical measuring instruments.

9 responses total.

#1 of 9 by rcurl on Sun Apr 23 21:40:53 2000:

My Micronta (Radio Shack) multimeter (22-185 - like the 22-183) has lost
its 3M ohm range. The ohm-meter range is selected by a switch that powers
a diode matrix, which presents a digital code to the meter's IC lines
IN6,5 and 4. The code goes 300 ohms, 111; 3K, 010; 30K, 011;  300K, 100;
3M, 101; and 30M, 110. (A 1 means that there is a diode connecting to that
line to the IC, except for 010 and 100, for which a diode is omitted.)

When set to the 3M range, it behaves as when set to the 300 ohm range. 
That seems to mean that on the 3M range, a 101 has been changed to a 111. 
All of the other ranges except the 3M range work correctly, so the problem
is not in the IC. The switch is just SP6T with two diodes at the 3M
position to connect as 101. How can this problem occur? A short between
switch points would affect two range positions, unless it is between 300
and 3M, but those positions are far apart.

#2 of 9 by n8nxf on Mon Apr 24 11:46:54 2000:

Check your diodes and the switch.  You will probably find that the digital
voltages at the IC are marginal.  The problem might even go away if you
replace the battery.  You may have a leaky diode or poor switch contact too.
I love these pop quizzes ;-)

#3 of 9 by rcurl on Mon Apr 24 17:36:16 2000:

The diodes all check out OK. A poor switch contact cannot cause a -x-
connection to be made in addition to x-x. The 'low battery' alarm 
is not showing on the screen. A slightly leaky diode is a possibility
but that would have to jump either -xx or xx-, which would cause other
ranges to misbehave. 

This same matrix also chooses the ranges for AC and DC voltage and
current, with a second switch choosing the function. I have not checked
whether the 3M ohm position is also wacky for V and I, although there is
no AC V or I functionality for the 300 ohm position. Well, I can at
least check the digital voltages and see if those tell me anything.

#4 of 9 by n8nxf on Tue Apr 25 10:14:57 2000:

Check the logic levels going to the IC.

#5 of 9 by rcurl on Tue Apr 25 21:20:29 2000:

The "3000 V DC" range selection is also now functioning as a 0.3 V DC
range, entirely equivalently to what happens on ohms, which confirms
it is the same effect and involves the diode matrix switching.

#6 of 9 by n8nxf on Wed Apr 26 11:42:29 2000:

No surprise there.

#7 of 9 by rcurl on Wed Apr 26 16:10:39 2000:

Checking the voltages on the digital lines only confirmed what I knew -
choosing the 3000K range gives digital 111 rather than 101.  However
checking the reverse resistance on each diode in turn found that the
problem was, as Klaus suggested, a leaky diode. I found it interesting,
though, that the reverse resistance increases as one holds the ohm-meter
leads across it - quickly from 2K (probably from 0) and then more slowly
until it exceeds 2 M. Since the control voltage is only 20 mV, the bad
diode is essentially an open circuit.

Now, the POP QUIZ!  Below is a table of how the diode matrix is wired. The
3000K line has only two diodes, but connects as 111. So, assuming that
there is just one shorted diode, and only the 3000K range is affected,
which one is it? 

      Range       IN6     IN5     IN4
      ^^^^^      ^^^^^   ^^^^^   ^^^^^
        30M       D16     D17
      3000K       D15      *      D14
       300K       dir
        30K               D13     D12
         3K               dir
        300       D11     D10     D9
        
where dir = wired directly, and * indicates where the *apparent*
connection exists. So, which of the other diodes is the shorted one? If
there is a mathematical way to calculate which is the shorted diode?

#8 of 9 by n8nxf on Thu Apr 27 11:15:43 2000:

I figured that out a few days ago but tossed the piece of scrap paper.  There
were two different ways to wire the circuit.  One with pull-up resistors and
the other with pull-down resistors.  It's better to have a schematic to
look at than those weird tables.

#9 of 9 by rcurl on Thu Apr 27 18:21:09 2000:

You are permitted to use scrap paper.. 8^}. 

I don't think pull-up/pull-down matters, except in the IC. The range switch
has a single wire to it from the IC, and the switch contacts are connected
to the diode anodes as shown in the table. Therefore, however the control
is set up in the IC, a positive current is presented to it if there is
a diode connected to that control line. 

I replaced the faulty 1N4148 with a 1N914 I had in my stock, and the
3000 range is now back in service. Thanks, Klaus, for your consultation.

Why is a diode matrix used to switch ranges for everything (v/o/m)?  The
main reason I can think of is to be able to display on the LCD the symbols
for the range being shown (e.g., ohms, K ohms, M ohms). The calibration
resistors are all there for all scales for v/o/m, but are mechanically
switched only for the mA scale (the v/o scales are switched internal to
the IC). 

This has been an educational exercise. Although I've had this VOM meter
for years, this is the first time I've had a close look at how it is
designed.

Response not possible - You must register and login before posting.