C9000 - how is discharge voltage measured?

This discussion began in another thread and is somewhat off topic to that thread, so I figured it is time to start a new thread. This is the thread where this discussion started. Continuing from the last post there...

kxenl said:

My C9000 is from 2010 and it "measures" the voltage during discharge pulse. I can easily get readings from load and no load voltage with my old integrating DMM. Difference is often about 30 mV which would give good idea of IR of the cell if one would know the current.

Click to expand...

Assuming that the discharge current of the C9000 is 1000mA (pulsed)


0.030V / 1A = 30mΩ

kxenl said:

But since my C9000 readings are not very accurate in terms of voltage ( about 80 to 100mV too low on discharge) or even time.. who knows what the actual current is..harder to measure than voltage for sure.

Click to expand...

I was taking measurement last night and made some interesting observations. When measuring the battery voltage of an Eneloop AA cell while discharging with the C9000 I also found that the voltage read low on the C9000 when compared to my multimeter. I calculated about 100mV lower from the C9000 than my multimeter (averaging a half dozen readings.) I was surprise that the C9000 voltage was so much different, so I decided to try something else. I allowed an Eneloop AA to charge until the C9000 indicated "DONE." Then, observing the voltage reading from the C9000 and my multimeter and found that they were very close.


Today, I measured current from the C9000 while charging and discharging. In order to make sure that the power supply to the C9000 was clean DC power I used an Astron RS12-A linear regulated power supply.



This is the output of the linear power supply used to power the MH-C9000


Here are measurements made while charging at different rates:



While taking this measurement the C9000 was displaying ~214mA.



C9000 displaying ~1072mA



C9000 displaying ~2142mA


Using my multimeter I wanted to get an idea of how much noise was on the charge current.

A fair amount of ripple. Not important for charging the cell but important when trying to measure the voltage accurately.


The discharge current was much cleaner.

So, maybe there isn't so much noise to influence the C9000 as it measures voltage while discharging. I still makes sense for it to measure voltage during the pause between pulses so that any resistance at the connections don't influence the measurement.


From what I have read in older posts here the C9000 measures voltage during the pause between discharge or charge pulses. Since then there have been changes to the C9000 such as the fact that when discharging it terminates at 0.9V now, but the first chargers terminated at 1.0V. I am still wondering if one of the changes is when the charger measures voltage when discharging. From what I can see with my multimeter it makes sense to measure voltage during the pause in the charge cycle. You can see the ripple in the charge current probably caused by switch regulated power at 2A. But the ripple doesn't appear to be present in the 1A discharge current. (At least to the extent that my multimeter can measure.) Why would it be advantageous to change discharge voltage measurement from the off time between the 1A discharge pulses to during the discharge pulses?

My question about how the MH-C9000 measures voltage during discharge was answered by a fellow CPF member who pointed out this post:

Maha MH-C9000 SUPPORT / FAQ - continuation

So, only some of the first MH-C9000 chargers measured voltage during the pause or off time when discharging, and the modern chargers measure voltage under discharge load while discharging.


While awaiting CPF to appear back online I acquired a Rigol DS1052E digital storage oscilloscope. My excuse: it's a very early birthday present to myself. While I'm no electronics engineer, I have always wanted such a toy but put it out of my mind due to the high cost. Recently I discovered they can be had for around $400.00 US. Anyway, here is some waveform captures that you may find interesting.


So, seeing that the title of this thread is about the C9000 discharging, here is the discharge waveform for the maximum discharge rate of 1000mA.

You can see that there is a pause about every two seconds. Cursor B, the yellow line toward the bottom shows the load voltage while the charger is discharging. This voltage does coincide with the voltage displayed on the C9000 while discharging. The pause or off time during discharge at 1000mA is 174ms as you can see in this next capture:



The waveform while charging at 2000mA (the maximum charge rate for the C9000) is similar in that there is a pause every two seconds between charge pulses.

The off time during the 2000mA charge cycle is 174ms.



So, it appears that the charger uses a two second charge cycle, with a 174ms pause when charging or discharging at the maximum rate. Remember that the charge current is always 2000mA pulsed on for a shorter time to reduce the overall charger rate. Similarly, discharging with the C9000 does the same thing pulsing the 1000mA discharge current for shorter amounts of time to reduce the overall discharge current for each of the different discharge rates.


So, here is the 2000mA pulse for an average 1000mA charge. In other words, when you set the C9000 to 1000mA it shortens the 2000mA charge pulse to provide an average charge rate of 1000mA.



Here is the pulse with the C9000 set to 200mA charge rate:



When using the break in mode, you set the capacity and the charger charges the cell at C/10 for 16 hours. If I set my MH-C9000 charger to 1000mAh for a break in charge rate of 100mA you can see the 2000mA charge pulse become shorter still. (I haven't measured it, but this is the charge rate I would expect during the two hour top off charge after the C9000 indicates that the charge is done.)



Here again charging in break in mode with the cell capacity set to 500mAh for an average charge rate of about 50mA



The lowest charge rate that I was able to measure was the maintenance charge. When the charge completes, after the two hour top off charge, the C9000 charges at 10mA. Although, it seems to me from the waveform that it is closer to 5mA. (Note: the vertical scale is set differently in this capture that is why the pulses appear taller than the other captures.)



Here is a better look at the 10mA charge rate pulse.



And last, but not least. You may have noticed the noise on the charge pulse, especially looking at the shortest charge pulse of 5.16ms. Well here is a better look at the ripple on the 2000mA charge current.

Oh, fun. I too have had a Rigol in my "gotta get one sometime" list, but since the Agilent 2000 X series came out I'm suddenly thinking the Rigol is the poor relation and maybe I should go up market.

Anyway, with your detailed graphs and timing you could check what the actual time averaged current is when it is set at 1000 mA. The received wisdom is that it will only be about 900 mA.

Thanks for detail photos
It were done several times before for charging
http://forum.ixbt.com/topic.cgi?id=20:27398:3188#3188
http://forum.ixbt.com/topic.cgi?id=20:27398:3193#3193
and 5mA in trickle http://forum.ixbt.com/topic.cgi?id=20:27398:3202#3202

Finally you received correct duty cycle for measuring (91.3%). Thanks again!

Mr Happy said:

[...]Anyway, with your detailed graphs and timing you could check what the actual time averaged current is when it is set at 1000 mA. The received wisdom is that it will only be about 900 mA.

Click to expand...

The measured +duty cycle is 46% when the C9000 is set to the 1000mA charge rate. The 2000mA charge current should be on for 50% of a 2 second time period to produce a average charge rate of 1000mA. 46% of 2000mA is 920mA, pretty close to your prediction Mr Happy.



A method I believe has more accuracy. Measure the charge pulse time, which is 912ms, or 45.6% of the 2s time period for an average of 912mA. (A 2000mA charge pulse for 912ms every two seconds.)



So, that means that the maintenance charge (after the two hour top off charge is complete) being 5.16ms works out to be 0.258% +duty cycle of the 2s time period or 5.16mA average.

Mr Happy said:

Oh, fun. I too have had a Rigol in my "gotta get one sometime" list, but since the Agilent 2000 X series came out I'm suddenly thinking the Rigol is the poor relation and maybe I should go up market.[...]

Click to expand...

Have you been watching the EEVblog?

Those Agilent scopes sure are nice.

So what is the best way to calculate IR reading using the C9000 discharge @ 1000mA?

(Open Voltage - Voltage during Discharge ) / 1A

or

(Open Voltage - Voltage during Discharge ) / .9A


Do you still need to add 50mV for contact resistance since the discharge waveform Curve B voltage agree with the C9000 displayed voltage during discharge? Maha-C9000-readouts-can-we-trust-them

And is this a better way of measuring IR than reading the C9000 charging ICV?

Measure the voltage drop against the contact point of the battery and charger. You may be suprised !

BoarHunter said:

Measure the voltage drop against the contact point of the battery and charger. You may be suprised !

Click to expand...

I measured the voltage drop between the negative end of an Eneloop cell and the C9000 terminal during a 500mA C9000 discharge. Voltage drop oscillated between 0.98mV and 1.71mV on my multimeter. (The 300mV range on my meter is rated at 0.04% + 2) I assume that the 1.71mV drop is during the 1A discharge pulse. That calculates out to 1.7mΩ. Now I know that the measurement can't be counted as super accurate, but it doesn't seem surprising.

ecrbattery said:

So what is the best way to calculate IR reading using the C9000 discharge @ 1000mA?[...]

Click to expand...

That is a good question. I don't think that you can get a real accurate IR using the C9000, but I figured let me see what I can figure out. First, in order to get the IR of one of my new (fully charged) Eneloops, as accurately as possible, I used a CBA II to discharge it at 1000mA and my multimeter to measure the voltage of the cell during the discharge and after it stopped. I checked the current during the 1A discharge with the CBA II and was quite surprised at it's accuracy. It takes a few seconds for the current to ramp up, but when it does, it fluctuates very slightly and equally above and below 1000mA.

The results I got discharging with a CBAII and measuring voltage with a multimeter:
1.283V during 1000mA discharge.
1.323V after the discharge stopped.
That works out to an IR of 40mΩ.

Just for fun I used my oscilloscope to check the voltage drop during a 500mA discharge with the C9000. To do this I carefully probed the positive and negative terminals without placing anything between the battery and charger terminals. I used the 500mA setting because that is a 1000mA discharge with about a 50% duty cycle.

According to this capture the IR works out to be about 33mΩ. Low, but not drastically.


After all that I began to wonder what would be the best method using only the C9000 and a multimeter. I prefer to measure the unloaded voltage after loading the cell, it seems to provide a much more consistent reading. So, I put the Eneloop in the C9000 and set it to discharge. After observing the voltage reading on the C9000 during discharge I quickly removed the cell and checked the voltage with my multimeter.

The results:
1.25V indicated by the C9000 while discharging.
1.337V measured with a multimeter right after removing from the C9000.
Assuming 1000mA discharge current, that works out to 87mΩ. Somewhat high for an Eneloop cell. This method could be useful if the results were recorded and used for comparison later.

Just for fun, let's add 50mV (theoretical contact resistance) to the voltage from the C9000:
Under load: 1.25V + 0.05V = 1.30V
Unloaded: _______________ 1.337V
That works out to a IR of 37mΩ...interesting.

Russel said:

Just for fun, let's add 50mV (theoretical contact resistance) to the voltage from the C9000

Click to expand...

Agree - it's simplest way

Just some clarification about those 50 mV
It's due to Maha schematic ONLY and mistake? in FW
During discharge V on U2 will be negative(-1А х 0.05 Om = -0.05 V)