Serial to usb on Kunkin KP184 electronic load

[interflexo]

Yes. The KP184 Modbus Software is compatible with Windows 11.

Please check the drivers of your USB to Serial Adapter for any issues regarding Windows 11.

[rallyhard]

Thank you to the forum admins for maintaining this excellent resource.

I bought Interflexo’s software and find it useful, but have a bit of an issue that I was wondering if anyone else has. I tried asking Interflexo/Manuel, but haven’t heard back yet.

We use the KP184 along with the PC software to perform discharge tests on deep cycle lead-acid batteries. It works well except that during the test, the voltage will occasionally dip severely for a very brief moment and then continue normally. The test bench is not being disturbed when this happens. I’ve tried wiggling the wires between the unit and battery during the test to see if there is a bad connection, but the voltage doesn’t change at all when I do that. It happens randomly during tests whether using remote or local sense. The dip in voltage will eventually cause the test to end early as it dips below the test end voltage, so I have to keep restarting it.


Thanks for any input. 

EDIT: Here are the test reports showing the voltage dips:

https://i.postimg.cc/JtC8kjR3/kp184-Report1.png

https://i.postimg.cc/6qvwqMx5/kp184-Report5.png

[interflexo]

Hi,
No other customer complains about this. It must be hardware specific.
Does this happen with only one specific lead-acid battery or with other ones as well?
Does this happen with a Li-ion battery / LFP battery  / DC Power supply discharge tests as well? Please do some similar basic tests with your system to rule out the device being tested.
What USB to Serial adapter are you using?

Sometimes lead-acid battery internal serial busbar connections get corroded, break and show a similar behaviour.

On one of your test reports the reason for the end test was Kunkin KP184 shut the load OFF (not the software).
On these long tests the Windows system power settings should be configured to not interfere with the on going test.
Can you use an oscilloscope and do a trigger test with a low voltage (Ex: 11V) to make sure the quick voltage dips are not real?

[korjaa]

Hello,


I recently bought one of these KP184s and I've been using it to test Ryobi drill batteries. I have also noticed that occasionally I get spikes, usually downwards, but there have been few upwards also.

Example of downward spike (0A...10A toggling, 15 period):


Example of upward spike (0A...15A toggling, 15s period:


0A...5A toggling causes periodic spikes?



The message checksum is checked, so the data should be OK. I know the checksum is calculated as I spent a while debugging it. The source is available in https://github.com/korjaa/kunkin-kp184, in case anyone is interested. I have not inspected the waveform with a scope to verify if the spike is real.

Does this happen with a Li-ion battery / LFP battery  / DC Power supply discharge tests as well?

Mine are Li-ion batteries, 5s2p

Does this happen with only one specific lead-acid battery or with other ones as well?

Multiple batteries for me.

Update:
Tried scoping the waveform, wasn't able to catch the spike during test. I did however noticed that if the load is small 100mA/10mA/0A, there is actual ~2V..3V dip in the voltage occasionally, maybe every 30s. The spike period matches with my earlier 0A...5A toggling test.


It does not happen if I disconnect the battery from KP184. The downward slope of the spike is related to the current.

Below are some spike lengths measured with different load values:
Output OFF -- 200ms (happens even with the output off)
8mA -- 1ms...6ms
30mA -- 200us...300us
100mA -- 60us...80us
500mA -- 40us...60us
700mA -- 40us
1A -- couldn't get it to trigger in few minutes.

[interflexo]

From the information above the problem does not seem to be related to the data acquisition software as it happens with different software developments around the same Modbus protocol.

The Kunkin serial interface even at 115200 is slow you won't get more than 5 complete data sets every second.
Even the 250us voltage dips should not make much difference on a 200ms data acquisition period.

I think I will add to the KP184 Modbus Software a user configurable minimum amount of time for the acquired voltage to be below the battery test end voltage, trigger the test end condition and call it a day.

I also think this must be somewhat hardware related (KP184 Hw version) because I have hundreds of very long battery discharge data acquisition tests and never seen this happen even once.

Please pay attention to the KP184 4 mm binding posts odd internal diameter that can cause sketchy electric connections. I always use fork crimped cable connections screwed on the back of the binding posts to avoid nuisances.

[rallyhard]

Hi, thank you for the reply.

This happens on every battery we test, which has been about 10 different deep cycle lead-acid batteries. I just tried testing a couple Ryobi 18v Lithium-Ion power tool batteries (as korjaa did after your reply), and also got voltage dips. Interestingly, the dips when testing these batteries (I tried 2) were at regular intervals; about once a minute.

We are using this USB to serial adapter:
https://www.amazon.com/gp/aw/d/B076WNJLLV/ref=ya_aw_od_pi?ie=UTF8&psc=1

Yes, on one of the previous screenshots I shut the test off manually instead of letting it complete.

Unfortunately, I don’t have an oscilloscope to check with.

We are using ring terminals to connect the test leads, not banana plugs.

I’m considering buying a second KP184 to speed up the testing procedure when we have multiple batteries to test at once, and am heartened to hear that this isn’t a common issue with the unit.

It would be great if you could implement an option in the software to have a setting for a minimum amount of time below the test end voltage. That may provide a workaround for our issue and allow us to continue using this unit. It is out of warranty and I don’t know how else to approach the problem.

[rallyhard]

Voltage dips on 18v Li-Ion battery test

https://i.postimg.cc/MHT9N0qv/IMG-9844.jpg

[interflexo]

...
It would be great if you could implement an option in the software to have a setting for a minimum amount of time below the test end voltage. That may provide a workaround for our issue and allow us to continue using this unit. It is out of warranty and I don’t know how else to approach the problem.

I will do it ASAP.
Do you have a suggestion for this default minimum amount of time as per your observations?

From your last plot and counting the voltage dips: Did you make sure you are not requesting the periodic internal resistance test during the battery capacity discharge test? You can use the right mouse button on the Batt Tab mouse icon the access the context menu and check for it (user manual page 28). If you were it would look much like that...
Unrelated: Any particular reason for not be using the latest 1.0.66.0 version?

[rallyhard]

Checking the CSV files of tests for the voltage dips, it appears to only dip for one sample. So, it looks like a one second minimum would suffice in my case, or maybe even a 2-sample minimum.

The internal resistance test is turned off.

There’s no reason I’m on an older version; just haven’t gotten around to updating.

I noticed while checking the CSV files that the Li-Ion battery test dips on exactly 1 minute intervals, to the second. I wonder if this could have to do with the battery’s internal BMS. I followed up with another test of a lead-acid car battery and it didn’t dip until 10.5 minutes.

[interflexo]

New KP184 Modbus Software Version 1.0.67.0:

Setup KP184 Software V1.0.67.0

User_Manual

Changes:

1. New options on battery capacity discharge test tab right mouse button context menu for an optional low voltage cut-off 1,5s detection delay. This avoids battery capacity test premature ending on quick voltage dips.

[interflexo]

New KP184 Modbus Software Version 1.0.68.0:

Setup KP184 Software V1.0.68.0

User_Manual

Changes:

1. Some events handling were blocked during right mouse button context menu display. Corrected.

2. CVL (Constant voltage current limited) test mode manual startup load ON event triggers now the automatic data acquisition.

[interflexo]

New KP184 Modbus Software Version 1.0.69.0:

Setup KP184 Software V1.0.69.0

User_Manual

Changes:

1. New software CWL mode with current limiting capability. Operates Kunkin in CC (constant current) mode and a PID configurable control loop holds the configured CW (constant power) setting by permanently adjusting the CC load.

2. Several Python support modules were updated.

[rallyhard]

The new low-voltage cutoff delay seems to be working to keep the test running through the dips! Thank you so much for adding this functionality.

If I buy a second KP184 and serial-to-USB adapter, will the software recognize the additional unit? Will I need to run a second instance of the program, or how does that work?

[interflexo]

Hi. Thanks for the feedback.
The KP184 has no programmatically accessible internal serial ID.
The software is tied to the PC hardware. When required I supply new reg keys.
You can run unlimited instances of the same software application on the PC. Each software instance selects a different USB to Serial Adapter.
I will have a new version to run multiple paralleled KP184s on the same PC software for high power applications as if you have just one really powerful unit.