Matlab decimals causing error 1047

[DuncanFleming]

Hi there

I'm using a Labview with some matlab code, and every time I add decimal digits to certain values I get an error 1047. Here is the code:

format long

c1_0 = 0.00025 % [m] Starting distance of capacitor 1 to ball surface
c2_0 = 0.00025 % [m] Starting distance of capacitor 2 to ball surface
c3_0 = 0.00025 % [m] Starting distance of capacitor 3 to ball surface
T = 20 % [deg C] curent temperature

% Calibration variables - [V] measured at 0 and 0.3 mm
V1_min = 3.7
V1_max = 8.8
V2_min = 3.7
V2_max = 8.8
V3_min = 3.7
V3_max = 8.8
Meas_mm = 0.3 % Measuring range of the sensor in [mm]

% Sorting out Temperature effects - assuming linear expansion due to T
T0 = 20 % [deg C] reference temperature
RdT = (T - T0)*(1.6250e-6/5)  % expansion of radius
CdT = (T - T0)*(1.6411e-6/5) % displacement of sensor tip
R = 0.025 + RdT; % [m] Mean Radius, converted from [mm] to [m]

c1 = (-(Meas_mm*V1_min)/(V1_max - V1_min) + (Meas_mm)/(V1_max - V1_min)*V_c1)/1000 % [m] Measured capacitive displacement of sensor 1
c2 = (-(Meas_mm*V2_min)/(V2_max - V2_min) + (Meas_mm)/(V2_max - V2_min)*V_c2)/1000 % [m] Measured capacitive displacement of sensor 2
c3 = (-(Meas_mm*V3_min)/(V3_max - V3_min) + (Meas_mm)/(V3_max - V3_min)*V_c3)/1000 % [m] Measured capacitive displacement of sensor 3

x1 = R + c1_0 + CdT % [m] x1 position
y2 = R + c2_0 + CdT % [m] y2 position
z3 = R + c3_0 + CdT % [m] z3 position

% Solve for the x, y and z coordinates of the centre of the ball
syms x y z
F1 = (x1 - x)^2 + y^2 + z^2 - (R + c1)^2
F2 = x^2 + (y2 - y)^2 + z^2 - (R + c2)^2
F3 = x^2 + y^2 + (z3 - z)^2 - (R + c3)^2

S = solve(F1, F2, F3, x, y, z)

% These results are the change in the centre of the ball relative to its
% mathematical centre resting place in X, Y and Z coordinates.
x = double(S.x(2))
y = double(S.y(2))
z = double(S.z(2))

 If I change the values of, say, V1_min to 3.175 or V1_max to 8.88 (adding more decimal places) I get a 1047 error. The code as it currently stands works perfectly - no errors. Running the code with any number of decimal places in matlab itself returns no errors - Labview seems to create the issue.

Any help please?

[Craig_]

Unfortunately since this script is executed with an ActiveX Server outside of LabVIEW I don't have the software to reproduce your error.

You're going to have to try several different configurations to isolate what could be causing this. Here's some guiding questions to isolate this:

• Do any of the numeric outputs show correct values, or are they all zero when you see the error?
• If only some of them are zero then does the script still throw an error if you remove the ones that aren't outputting correctly?
• Does a simple script with decimals work properly?
• If you remove the latter half of the script do the intermediary results output correctly?
• If you split the code into two Script Nodes and connect the intermediary data between them does it work properly?

Once we can isolate this further it's usually pretty easy to identify a workaround to avoid the error!

Here's some related documentation:

http://digital.ni.com/public.nsf/allkb/8BEBC0C86541224286257AF300561B5E

http://digital.ni.com/public.nsf/allkb/2B3FF46C8512C4F786256CF30071BE53

[DuncanFleming]

Hi Craig

I've experimented with various things, and the most prefereable setup I've come across is to have the script split up across three matlab script nodes - this seems to work.

When using one node, there are anomalies such as if I end a line of code with a ";" or not (suppressing matlab output) then I get an error. I find this strange. getting rid of the "format long" line also cleared it up a bit.

Thanks for the help. I still dont know why the error happens, but it seems to work for now.

[Craig_]

I'm glad to hear you got it working!