We have added an excel sheet to the github which allows users to create their own battery models from the battery datasheets.

Regarding the USB error, we have seen this vary from PC to PC. The error is shown when the device receives un-expected data on the USB bus. (For e.g. we check for all zeros, all ones, all constant data, sync pattern error etc ) We will check if we can improve the error correction by skipping the samples in error and just letting the users know this information by appending the log file. That way the GUI does not stop working, but continues to collect the data. If the user looks at the log file, they would know that there was an error at a certain time of the data. If possible, we would show this on the chart as well to indicate that the measurement during a certain time period may have erroneous data.

Regarding the USB error, does it occur during the data collection or when it is idling and just analyzing the data? We have seen the error mostly in slow laptops, USB slot with loose contact, front panel USB connectors on a desktop, a bad USB cable, or very noisy environment. Can you rule out any of these? We will look into the error in the Viewer and get back in couple of days. It should be an easy error to fix. We are able to get the viewer to work without the hardware. So there should be some settings needed. Regarding the addition of a new model, please refer to the user guide. It details the method to model any battery. You only need 4 key parameters for the battery which can be easily found from the datasheet. If you have a particular model in mind, please share it here. we can extract the model and post is online so that others may also benefit. In the coming week, we would share an excel sheet which will make it easy to model any battery. That should work fine for most users. Note that we have moved to Github for all new releases and bug reports. That way, the data is preserved till eternity and it is far more easier to keep track of versions. You may report issues also on github. Please find the link below

@anglercircuits Can you update the latest GUi from github. It has some fixes for this problem. The issue is mostly from a loose USB connector on the laptop or desktop. It could also be due to signal integrity issues on the port. To view the files, there is a IOT power viewer software included in the package. Please run this to view the files without the hardware.

@anglercircuits Can you update the latest GUi from github. It has some fixes for this problem. The issue is mostly from a loose USB connector on the laptop or desktop. It could also be due to signal integrity issues on the port. To view the files, there is a IOT power viewer software included in the package. Please run this to view the files without the hardware.

Well, this was indeed a bug in the front panel design. The objective was to keep the banana posts at exactly 3/4" apart. The holes to mount the banana jack female connectors on the panel was not exactly a full circle. It had a small cutout on one edge. With that cut in place, I made an error in the centre to centre calculation. Error of 1mm. The next builds will have them fixed.

Well, this was indeed a bug in the front panel design. The objective was to keep the banana posts at exactly 3/4" apart. The holes to mount the banana jack female connectors on the panel was not exactly a full circle. It had a small cutout on one edge. With that cut in place, I made an error in the centre to centre calculation. Error of 1mm. The next builds will have them fixed.

We are pleased to announce the release of Ver 2.5.1 of the GUI. The GUI has been completely re-structured in terms of the user interface now and supports a menu driven operation. It supports long span for viewing the data for 180 seconds. Reduced CPU loading and many more stability updates. https://github.com/zscircuits/zs2102a/tree/master/GUI/bin

@Julian Higginson : Thanks for your support in the crowd funding. Our goal is to make this tool available to all IOT developers around the world and help them optimise their designs. Your contribution will go a long way to realise this. Regarding API, this is in one of our lists. Once the hardware is out in production, we can focus on writing the APIs.

The digital pot used to set this has only 256 taps and having 100 Ohms would increase the resolution by a factor of 10. We think 10 Ohms should be sufficient for most of the batteries. Anyway the hardware would be open source and the users may be able to swap few resistors to get higher resistances, if they need it. We would release the documentation on how to do that. In the meanwhile we would check if we can get some kind of non linear function to have higher steps as the output resistance increases so as to get the max value to 100 Ohms, while keeping the 256 steps.

@Julian Higginson : The VI curve for the output simulated for an output resistance of 0.4 Ohms is attached here. Just takes couple of op-amps and a programmable pot to implement this feature. Will check some stability analysis and regressions before adding in the final schematics. With this added, the users could choose the output resistance and the battery voltage and the tool would mimic the exact voltage drop dynamically. Once again, thank you for your wonderful suggestion to make the ZS1100A, a valuable product for the customers.

@Julian Higginson Thanks for your great suggestion. I was able to add in a programmable output resistance to the unit. Works well in simulation. There would be an option to set the output resistance from 0 to 10 Ohms in steps of 40mOhm. This should take care of the battery emulation with depleted batteries. Do you think the range of resistance is good enough?

The current design allows to only drop the voltage in a slow manner based on the total energy drawn from it. It does not allow the simulation of an output resistance. However with slight modification in the current limiting circuit, this could be done. We will give it a shot in the final rev which will go to production.

The feature you have discussed is already on our list of To-Do. Once a battery is chosen from the list, the output voltage and the current limit would be set automatically. A discharge depth can be set from the GUI to lower the output voltage accordingly. Then based on the total energy consumed from the battery, the output can be dropped depending on the battery chemistry. In a way this is like battery emulation and it will be supported in the GUI. We do have a working prototype which meets all of the specs. Few minor changed coming up in the final Rev followed by some intense test and qualification, followed by EMI tests. The plan going forward is to have it sold through distributors to have a wider reach. We would also make the hardware open source so that many manufacturers can make it and sell it. The exact model will be clear in the coming months. It is currently on CrowdSupply for initial funding needs, at a heavily discounted price. Any support is appreciated. https://www.crowdsupply.com/zscircuits/zs1100a-power-meter

No SDK is available for the tool as of now. There is a plan to release APIs within couple of months.

LuisF, Regarding the auto scaling of the Y-axis, did you try out the plot in the Log Scale. This disables the auto scaling automatically.

Hi Luisf. We have added plotting of the top chart with a span of 120 seconds now. Will release that one after some more testing. But just to let you know that plotting of 2 minutes data is time consuming. We would most likely add as an option which the user has to enable to get to the long plotting mode. The default mode would plot only 20 seconds for a snappy response time. Please find the plot below. Now the middle chart also plots the voltage in addition to the current waveform.

luisf We would consider your feedback regarding the long display and to disable the auto-scale feature of the chart. It does not look difficult to implement with the peak detection. For e.g. if you display the data for 2 minutes and if there is some peak in the data, only the peaks would show up from the base floor. That way you could see any activity without panning. Disabling auto scale is a easy implementation. We are glad to hear that you are happy with the purchase and that the tool is being valuable to your R&D efforts. We appreciate the feedback and value your patronage.

In the profiler tab, you could use the coarse slider bar to view the data across time. But at a given time, the data displayed is limited to just 20 seconds as it has huge amount of data. (1 sec = 1Million samples) and if you plot for 1 minute, that would be 60M samples. Displaying this data on a screen with 1080p resolution would mean decimation by a factor of 30K. You will not be able to see any detail in that time frame. The option is to use the coarse slider bar to move across the time and then zoom into any area of interest. There is no time limit to the coarse slider bar and it can span several hours of data. Please let us know if this fixes the problem.

Gregory We have added the voltage plot on the middle chart to show the dip in voltage. Will soon be added in the next release. Stay tuned...