Program that can log from many multimeters.

[nez]

Is there a special step to get the Owon ODP6033 to work in Test Controller?

I noticed that gby contributed support, so I've tried adding it but it doesn't connect (using v1.07).  On the 'Load Devices' tab, it gets added with Type 'Socket', and the Address field empty and highlighted red.  I am connecting the device to my PC via USB.

[I'll attempt to add the screenshot, but forum attachments seem buggy right now]

I can get my BM869s multimeter to work in Test Controller, and I can get the ODP6033 to work in Owon's ODPEasyControl software (although Owon's software has the awkward step of entering the device's serial number every time when selecting Connect->USB).


Great job on this app by the way HKJ!

[MikeLud]

nez,

The device file for the Owon ODP6033 is for IP communication. The address would be the IP address of the ODP6033.

[nez]

The device file for the Owon ODP6033 is for IP communication. The address would be the IP address of the ODP6033.

Ah, makes sense thanks for the quick reply.  When I have time I may try to read into HKJ's documentation to see about a usb configuration.

[HKJ]

Ah, makes sense thanks for the quick reply.  When I have time I may try to read into HKJ's documentation to see about a usb configuration.

If the Owon uses a virtual serial power for USB communication you only need to add the word "com" after the port number in the device definition file.

[HKJ]

V1.08 is up
It adds more scripting commands and extra export options and improves using math channels for device and variables.
   Added: On/off button to the virtual generators (They have always supported on/off)
   Added: Export got a change sign option for each data field.
   Added: Export got a new data reduction method: change of value
   Added: Math type formulaNU (NU=No Update) for use with local variables or extra device data.
   Added: #SaveLog to save log window
   Added: #SaveChart to save a chart
   Added: #ChartY to define the Y-axis
   Added: #ChartSamples to control the first/last sample boxes
   Added: #ChartCurves to define what curves are on the chart

Charts can now be controlled from scripts, i.e. it is possible to make a automatic test that finish by saving some charts.

The export got a new data reduction method, it looks for changes in a column and output a line each time the change is above the threshold from last output line.

I had a small problem, I wanted some data in a format I could directly compile into a program. I had expected to find a tool for it on the inet, but failed. Now TestController can do it: #SaveTable has some extra format options: int/long/float/double that will generate some java code (can also be used in C/C++ with minor edits). The procedure will usually be to export the wanted columns first, then load the exported file and use the #SaveTable command.

[nez]

If the Owon uses a virtual serial power for USB communication you only need to add the word "com" after the port number in the device definition file.

Looks like the Owon uses USBTMC with the National Instruments VISA driver, rather than virtual COM (*similar situation to gby's Keithley).

Fortunately I don't really need it right now, so waiting for eventual support doesn't bother me!

[kharar]

I have not yet had the chance to test if it works at all, but...
At my work we have a KP284 which more or less is just two separate KP184 in one cabinet, but with just the RS-485 interface. Is there any chance the TestController Java program could be made working for both channels at the same time, like eg. expanding the address range to include 1 & 2 ?
We also have another KP284 at the local hackerspace Labitat and I assume it would be rather beneficial to a fair amount of users.

[HKJ]

I have not yet had the chance to test if it works at all, but...
At my work we have a KP284 which more or less is just two separate KP184 in one cabinet, but with just the RS-485 interface. Is there any chance the TestController Java program could be made working for both channels at the same time, like eg. expanding the address range to include 1 & 2 ?

There are a couple of issues in it:
1) KP184: It is not tested yet that it works
2) RS485: I do not have any handling special handling of that (Is it needed)?
3) Multiple address on the modbus.

All issues can be solved without to much work from my side, but I need to know 1) works and if 2) is needed before I will look at 3)

At its simplest 3) can simply be an extra device called KM284 second supply, where I add a tag in the definition to change the modbus address.

[HKJ]

V1.09 is up
The main improvement is more virtual generators.
   Fixed: Help window continue to show help when entering parameters
   Added: Support for a #helpurl tag in device definitions, this will be used when F1 is pressed on the mode or setup popup.
   Added: Pressing F1 on the virtual generator setup page will open a browser on the help page.
   Added: Help page about the virtual generators.
   Added: Pattern generator that can make a sequence of values with either abrupt or ramp change between them.
   Added: User generator that can convert an expression into a curve that can be used to control a device.
   Added: Log generator that can be used to control a frequency generator
   Added: Battery simulation (Virtual battery generator) that can control a power supply.

The help page about virtual generators: http://lygte-info.dk/project/TestControllerVirtualDevices%20UK.html

The new generators add a lot of possibilities to control external devices with a very simple script (It is listen on the documentation page). The most advanced is the battery simulator:





Reflect the voltage to a power supply from it and reflect the current consumption back to the generator and the power supply will simulate a battery (For slow load changes/pulses).
It supports many different types and sizes of batteries.


I have rewritten the test controller into page: https://lygte-info.dk/project/TestControllerIntro%20UK.html

[Messtechniker]

Since I now own a Siglent SDG 2042X, I am now upgrading may home grown measurement amplifiers and precision rectifiers to 1 MHz and producing a reasonably fast F/V converter. My VC940 is pretty slow in this area and the Keysight 34465A is only good up to around 400 kHz. The Siglent can, while sweeping, not be queried as to the momentary frequency.

While doing so, I was wondering on a different track why the sound card output can not be used with "Test Controller". Has nothing to do with my 1 MHz drive, but having 16+ bit AD/DA capability to measure voltages and frequencies is in my opinion an opportunity which should not be lost, if not too difficult to implement.

Any ideas?

[HKJ]

Since I now own a Siglent SDG 2042X, I am now upgrading may home grown measurement amplifiers and precision rectifiers to 1 MHz and producing a reasonably fast F/V converter. My VC940 is pretty slow in this area and the Keysight 34465A is only good up to around 400 kHz. The Siglent can, while sweeping, not be queried as to the momentary frequency.

Why not control the Siglent directly from TestController? It already contains script for doing a logarithmic frequency sweep. As for measuring voltage, it is possible to use the voltage measurement from a oscilloscope, TestController has a few (The dynamic range may be too low without changing input range on the scope).

While doing so, I was wondering on a different track why the sound card output can not be used with "Test Controller". Has nothing to do with my 1 MHz drive, but having 16+ bit AD/DA capability to measure voltages and frequencies is in my opinion an opportunity which should not be lost, if not too difficult to implement.

Any ideas?

I do not have any experience in sound analysis, not that it would be that hard to search a recorded curve for zero crossings and calculate frequency and rms values. This will, of course, only work for a fairly clean tone. The level will, of course, be relative, as far as I remember sound cards to not have any well defined calibrations.

[Messtechniker]

Why not control the Siglent directly from TestController? It already contains script for doing a logarithmic frequency sweep. As for measuring voltage, it is possible to use the voltage measurement from a oscilloscope, TestController has a few (The dynamic range may be too low without changing input range on the scope).

I do not have any experience in sound analysis, not that it would be that hard to search a recorded curve for zero crossings and calculate frequency and rms values. This will, of course, only work for a fairly clean tone. The level will, of course, be relative, as far as I remember sound cards to not have any well defined calibrations.

Will look into that script thing. My abilities are certainly not the best in this area . 

Tried doing the voltage measurements using my R&S HMO scope. Works mostly. The problem are measurement dropouts by the scope indicating a "?" every now and then.

Of course sound card inputs and outputs are not calibrated. But "Test Controller" has enough math functions to correct this.

A few hours later. Got rid of that "?" problem simply by shutting down channel 2 and switching off all screen "number crunching". Getting nice and smooth data upon querying the HMO scope by Test controller".
 

[HKJ]

Will look into that script thing. My abilities are certainly not the best in this area . 

You do not need any abilities to use the scripting, the scripts are ready for use:
Connect to the SDG, right click in the log window and select "Waveform generator, Logarithmic frequency sweep"
You might want to adjust some parameters at the top, but it is not required
Press run.

You also have to load the meter you want reading from, but it is not related to controlling the SDG.

[Messtechniker]

Thanks for the hint.  Will try as soon as I get up in the morning. 7 hours or so from now.

[HKJ]

Thanks for the hint.  Will try as soon as I get up in the morning. 7 hours or so from now.

TextController has two scripts for controlling a AWG, if you load the "Virtual Log Generator", there is one where you use the VLG to control it.

[Messtechniker]

OK. The log generator script worked to an extent. Script controlled generator sweeping worked. 
But experienced bad freezes before the graph completion.  Had to restart windows and the Siglent in each case. 

Since Test Controller never crashed before, I think I'll go the F/V hardware route so that Test Controller will have to handle two DC voltages from two meters without using additional scripts. Something it has doing reliably before.

[HKJ]

OK. The log generator script worked to an extent. Script controlled generator sweeping worked. 
But experienced bad freezes before the graph completion.  Had to restart windows and the Siglent in each case. 

Will you try running TestController in debug mode, I am very curious if there are some Java error messages and what they say.

[Messtechniker]

Will look in to this next week. 
Just wont to get my 1 MHz precision rectifier finished first.

[Messtechniker]

What's also been bugging me is that upon starting to log and when changing to Chart,
the X-axis is always set to time and I will then have to change the X-axis type each time. I.e. the X-axis setting is not retained. By the time I have done that, a sweep, for example, has already finished.

Just one further idea: Why can't the generator which is to be controlled be selected through the VLG popup?

[HKJ]

What's also been bugging me is that upon starting to log and when changing to Chart,
the X-axis is always set to time and I will then have to change the X-axis type each time. I.e. the X-axis setting is not retained. By the time I have done that, a sweep, for example, has already finished.

Maybe I need to fix something here.

Just one further idea: Why can't the generator which is to be controlled be selected through the VLG popup?

Two reasons:
1) Devices are independent, i.e. they do not know anything about other devices.
2) Flexibility, you can control any parameter on a device, also parameters that are not directly available in TestController (i.e. needs a SCPI command)

[gby]

HKJ,

I just updated to V1.09 and I think I found an obscure debug mode bug.  If I have one device checked in the Load Devices tab I get a happy
   ;; Found Kollmorgen AKD1G Servo on 192.168.1.173
connection when I start TestController.  If I then go to the Load devices tab and click the Reconnect button it happily reconnects.

If I start TestController in debug mode it connects properly when I start the program but when I go to the Load devices tab and click the Reconnect button it does not successfully reconnect.  This problem is repeatable.  Even though TestController says it did not connect, the native device communications software can not connect until I close TestController as if the port was being held by TestController.

Below is the result for the cmd prompt debug window from startup to end of pushing the Reconnect button.

Code: [Select]

Starting
;; Start thread for: 192.168.1.173 - Kollmorgen AKD1G Servo
;; : Tx <txrx? DRV.INFO>
;; 192.168.1.173: Tx: <DRV.INFO..> 44 52 56 2E 49 4E 46 4F 0D 0A
;; 192.168.1.173: Rx: <Drive model           : AKD-P00606-NBEC-0000> 44 72 69 76 65 20 6D 6F 64 65 6C 20 20 20 20 20 20 20 20 20 20 20 3A 20 41 4B 44 2D 50 30 30 36 30 36 2D 4E 42 45 43 2D 30 30 30 30 0D
;; Found Kollmorgen AKD1G Servo on 192.168.1.173
;; Start thread for: 192.168.1.173 - Kollmorgen AKD1G Servo
;; : Tx <txrx? DRV.INFO>
;; 192.168.1.173: Tx: <DRV.INFO..> 44 52 56 2E 49 4E 46 4F 0D 0A
;; 192.168.1.173: Rx timeout 1000ms   In buffer:
;; 192.168.1.173: **Device do not match** <Unknown>
;; : Tx <txrx? DRV.INFO>
;; 192.168.1.173: Tx: <DRV.INFO..> 44 52 56 2E 49 4E 46 4F 0D 0A
;; 192.168.1.173: Rx timeout 1000ms   In buffer:
;; 192.168.1.173: **Device do not match** <Unknown>
;; Stopping thread for: 192.168.1.173 - Kollmorgen AKD1G Servo

Odd that everything seems fine in normal mode and the problem is only in debug mode.

[gby]

HKJ,

I am working to start writing scripts using the #math script statement to record extra data from the device.  Inside the script I need to refer to the logged value.  Before, for values defined in the device file with #value and polled in #askValues, I used <Device name>.<signal name>.  For example, like AKD2G.A1_Current.  When using #math with formulaNU for extra log data how do you refer to the value in a script?

I tried math.<signal name> but got errors.

[HKJ]

I just updated to V1.09 and I think I found an obscure debug mode bug. 

Please check on the dos window if there are some Java error messages.

[HKJ]

I am working to start writing scripts using the #math script statement to record extra data from the device.  Inside the script I need to refer to the logged value.  Before, for values defined in the device file with #value and polled in #askValues, I used <Device name>.<signal name>.  For example, like AKD2G.A1_Current.  When using #math with formulaNU for extra log data how do you refer to the value in a script?

I tried math.<signal name> but got errors.

All regular values are created as <Device name>.<signal name> before starting on the Math, but the result from Math are not create as variables. This also means you do not have access to them in a script or other Math expressions.

[gby]

I just updated to V1.09 and I think I found an obscure debug mode bug. 

Please check on the dos window if there are some Java error messages.

Unfortunately there were no Java error messages.  I started TestController in debug mode, saw proper connection, clicked Reconnect, and saw failed connection.  The dos window copy I gave in the previous post was the entire contents of the debug dos window.

I saw this issue with two different Socket type devices.  I also tried an LXI type device (Siglent SDG2042x) and that worked as it should.

I can say that the two devices that have this Reconnect problem both provide multi-line response to the query used as *IDN? for device identification.  So far TestController receives the first line, seems to ignore the follow on lines, and just works.  Perhaps the extra lines of info affects reconnection in debug mode but not in regular mode where the Reconnect button just works?

Can you think of anything else I could do to help get debug information on this issue?