Parts availability

jimbarstow

Joined Nov 8, 2011
24
I'm trying to build a water tank depth sensor. I started with the design in the Practical Arduino book but decided to go with a design on the Freescale site.

I started to put together an order for the parts from digikey but discovered that for some parts they had a minimum order quantity of 2000.
I tried other sites (mouser, allelectronics) which seem more geared toward non-industrial supply but they didn't have what I needed. (0.2
mf ceramic capacitor, for example.)

Any suggestions on where to go for parts? (I'm in California.)

praondevou

Joined Jul 9, 2011
2,942
I'm trying to build a water tank depth sensor. I started with the design in the Practical Arduino book but decided to go with a design on the Freescale site.

I started to put together an order for the parts from digikey but discovered that for some parts they had a minimum order quantity of 2000.
I tried other sites (mouser, allelectronics) which seem more geared toward non-industrial supply but they didn't have what I needed. (0.2
mf ceramic capacitor, for example.)

Any suggestions on where to go for parts? (I'm in California.)
please post all the parts that have that minimum order quantity at digikey. preferably the digikey link.

are you sure you mean 0,2 milli Farad? Why does it need to be ceramic?

SgtWookie

Joined Jul 17, 2007
22,220
I think it's more likely to be 0.2uF, or 200nF. You might use two 0.1uF/100nF caps in parallel.

However, you really need to reference exactly WHICH design you went with on the Freescale site:
1) Document number, such as AN1950.
2) Page number where parts list is.
For example:
AN1950
http://cache.freescale.com/files/sensors/doc/app_note/AN1950.pdf
Figure 2, page 4 is where the parts list is located for the circuit in question.

or:
AN1950
http://cache.freescale.com/files/sensors/doc/app_note/AN1516.pdf
Figure 4, page 4 contains parts list and schematic

Last edited:

Joined Dec 26, 2010
2,148
I have seen mF or even MF (LOL) or MFD written for microfarad instead of μF. A long time ago that might even have been normal practice, buy nowadays its better to stick to the international conventions for prefixes.

As to the capacitors themselves, it might be worth checking whether their value is critical. If not, you may find that you can use an alternative value such as 0.22(μ?)F.

jimbarstow

Joined Nov 8, 2011
24
I think it's more likely to be 0.2uF, or 200nF. You might use two 0.1uF/100nF caps in parallel.

However, you really need to reference exactly WHICH design you went with on the Freescale site:
1) Document number, such as AN1950.
2) Page number where parts list is.
For example:
AN1950
http://cache.freescale.com/files/sensors/doc/app_note/AN1950.pdf
Figure 2, page 4 is where the parts list is located for the circuit in question.

or:
AN1950
http://cache.freescale.com/files/sensors/doc/app_note/AN1516.pdf
Figure 4, page 4 contains parts list and schematic
The specific article is:

http://cache.freescale.com/files/sensors/doc/app_note/AN1324.pdf?fsrch=1&sr=1

The documents on the freescale site were actually very useful. There was a discussion about different types of amplifiers to use for pressure sensors and they said there was very little difference. The said the precision of the resistors had a much bigger effect than the particular design.

Should have been clearer about the capacitance and used the right symbol; it is 0.2 μF. I'd figured I could use 2 0.1μF capacitors since they are more available but I didn't want to give up yet.

The other part which I'm having trouble spec'ing is listed as a "trim resistor". The article doesn't specify directly what the value is but another freescale article that talks about the same circuit says it is "nominally" 39.2K. Does this mean it is the maximum resistance so a 50K trim resistor would be sufficient?

Last edited:

SgtWookie

Joined Jul 17, 2007
22,220
What I would do is use these 0.1uF caps for both C1 and C2 (read below):
http://www.mouser.com/ProductDetail...GAEpiMZZMsh%2b1woXyUXjzKM/Br94TWB8oLGKWrIMjg=
0.1uF caps are probably the most common caps on the planet; they are made in such vast quantities that the cost is very low.
I suggest buying 100 of them; that represents a $4.00 investment, so your per-cap cost is$0.04/ea. If you are going to build more circuits using ICs, voltage regulators, etc. you will need one of these PER IC, so you will use a lot of them.

If you REALLY insist on using single 0.2uF caps instead, you can use these:
http://www.mouser.com/ProductDetail...GAEpiMZZMsh%2b1woXyUXjwarXwENlZE5C2MmnSjM5Hk=
However, the lead spacing is 10mm instead of 5mm, and they are $1.42/ea. So, that's$2.84 for two caps, or $4 for 100 caps. You choose. As an alternative, you could use this 0.22uF poly cap: http://www.mouser.com/ProductDetail...iMZZMsh%2b1woXyUXjyIv%2bSUPGiajNojKQY3%2bckE= They are$0.16/ea.

HOWEVER!
C1 and C2 are input and output caps for a 78L08 8v 100mA regulator.
While no external components are necessarily required for proper operation, unless you have at least 10nF/0.01uF capacitance on the output, it may become unstable (consider that to be very likely) and oscillate at high frequency. So, I suggest that you use 0.1uF/100nF from OUT to ground, and ~0.33uF from IN to ground; the caps placed as close to the regulator as possible. This is consistent with most manufacturer's recommendations for I/O caps for the 78xx series of regulators.

SgtWookie

Joined Jul 17, 2007
22,220
The other part which I'm having trouble spec'ing is listed as a "trim resistor". The article doesn't specify directly what the value is but another freescale article that talks about the same circuit says it is "nominally" 39.2K. Does this mean it is the maximum resistance so a 50K trim resistor would be sufficient?
From page 4:

FURTHER SIMPLIFICATION

In non–demanding applications the 7 resistor topology that
is shown in Figure 2 can be reduced to 5, by eliminating R6 and
R7. Without R7 the zero pressure offset is untrimmed.
However, in microprocessor based systems it is relatively
easy to read the zero pressure offset voltage, store it, and
calibrate in software. This can be done automatically when the
unit powers up, or as a calibration procedure.
R6 can be
eliminated (reduced to zero ohms) by directly connecting the
R3, R5 divider to pin 2. The output impedance of this divider
then needs to be choosen such that its ratio with R4 = R1/R2,
in other words [R3•R5/(R3+R5)]/R4 = R1/R2. Given the
values in Figure 2, this would mean R3 = 200 k, R5 = 13.3 k,
R6 = 0, and R7 is open. In an untrimmed system, there is no
real disadvantage to doing this, provided that the ratios can be
sufficiently matched with standard resistor values.

The other option is to eliminate R6 and trim R3 with R7. This
situation is somewhat different. The trimming operation will
throw the ratio off, and reduce common mode rejection.
Typically several percent of any change in the sensor’s
common mode voltage will show up as an output error when
this configuration is used.
39.2k being the nominal value means that it is usually the appropriate value for the circuit.

If you want a ~±5% range of adjustment for zeroing, you can use a 300k fixed resistor in series with a 1MEG potentiometer that is wired as a rheostat (connect the wiper to one of the ends; thus reducing the pot to only two connections). However, you should use 2.5% or better specification resistors. I suggest that you use metal film resistors, as they are more stable and less noisy than carbon film resistors.

The other quoted option is not good; you don't want to reduce your CMRR if you can help it.