Explicatii
Explicatii
Explicatii
Try connecting pin 37 to the positive voltage, the display will show 1888 as the IC circuit is working
correctly.
Try short-circuit at the input Number is displayed as 000. If this occurs, indicating that the AUTO zero is not
working properly, could be added to increase the capacitor value.
If everything is correct. We will have a digital meter with an input sensitivity of 200mV to use.
1. Hi,
Build an accurate voltage divider with an output of 199.9mV and connect it to the circuit. Adjust "high
Ref" to get a readout of exactly 199.9 (mV).
Voltage dividers for higher voltages should have 1% tolerance maximum. Better to use 0.1% tolerance.
2.
3.
4.
5.
6.
There are high precision voltage divider resistors on the market to be used for DMMs.
Boncuk
In Post #14 qa9b is refering to this site
From this same site the author indicates these values for R3 (between pin 30 and 31) for a list of range of measurements
0 - 2 V ............ R3 = 0 ohm 1%
0 - 20 V ........... R3 = 1.2 Kohm 1%
0 - 200 V .......... R3 = 12 Kohm 1%
0 - 2000 V ......... R3 = 120 Kohm 1%
Does 0 ohm means shorting pin 30 and 31? this doesn't work in my head, does anyone have a different opinion on this
I had already decided that I needed a small power supply for my breadboard for when I didn’t want to use the much larger
bench supply. But I wanted this power supply to be adjustable, and rather than having to get my multimeter out each time to
check the supply voltage, I thought it would be a good idea to integrate a simple voltmeter. A quick look around led me to the
ICL7107 3.5 digit LED Display, A/D converter. I would need the voltmeter to measure at least to +/-15V and 3.5 digits would
give me up to +/-19.99. The ICL7107 is also designed to drive the common anode LED displays that I had.
A bit of research and some time spent with the ICL7107 datasheet and the Intersil Application Notes for the chip led to this
schematic:
During the process, I changed from the LED displays I had to a couple I found in my local surplus store. It took me a little
while to find the data sheets…the nice bit is that the MAN6830 has a dedicated polarity symbol, and also only two segments
for the leading digit. I think that will look a little better. My design includes a test button to check that all LED segments are
working. It requires a 5V power supply (which will probably be drawn from the breadboard power supply itself) and will
probably gain a probe so that not only can the power supply voltage be checked, but it will also be able to check test points in
the circuit on the breadboard as well.
Once I was pretty happy with my design (which owes a lot to the reference design in the data sheet). I thought I’d look around
the ‘net for designs from other people. I found one that looked pretty similar and thought I’d sit down and see what
decisions/component selections had been made in that design. I wanted to make sure that I fully understood the datasheets
and had made rational decisions in the selection of components in my design.
First, the other chap’s design was for up to 199.9V – OK just a scaling factor but looking at the voltage divider on the inputs
suggests that the other design uses 2V full scale (as opposed to my design using 200mV full scale – the chip can do either).
The voltage divider on the chip’s reference pins would provide a Vref of 1V also suggesting that 2V fullscale was desired. It
makes sense and allows the design (with the 1M/10K voltage divider on the input) to read up to the stated 199.9V. At this
requirement of 199.9V, the voltage divider on the input to the voltmeter provides 1.979V to the input of the chip (which should
result in a reading of 197.9V) – an error of around 2%. Maybe the designer got lucky with the resistor tolerances and actually
chose a resistor that was closer to 10.1K in which case the error would almost be eliminated.
As I went further I found that the integrating capacitor and resistor values were for the 200mV full-scale. Maybe it doesn’t
have much of an effect? It might be an area to experiment with when I prototype my design. Other than that there were no
real surprises in the design…
…until…
Later on the page, the designer says that to restrict the readings to 19.99V all you had to do was reduce the 1M resistor on
the input to 100K…
An interesting choice considering that 1M resistor is there to reduce the loading on the circuit under test…so we don’t really
want to play with that. Also if the 1M is reduced to 100K, the voltage divider will have a large effect on the readings! With a
19.99V input voltage, the voltage divider circuit will provide a maximum 1.817V at the input of the chip (potentially giving a
reading of 18.17V)! That’s an error of 19.99-18.17 = 1.82V or around 9%! He would have been better off suggesting the
voltage divider on the reference pins be changed to provide a 100mV Vref and leave the input resistance alone - at least
then the error would have only been 19.99-19.79 = 0.2V!
I have to wonder if this chap had really understood what he was reading in the data sheets…
Of course the moral of the story is that you can’t believe what you find on the Internet (duh!) – and anyone who uses the
schematic I’ve posted should check it for errors!