There are a lot of  questions around equalising Ni-Mh cells.

Following are results from tests using equalising and not using equalising.  

Equipment used:   Pro-Trak BMS,  Pro-Trak Intensive Care System and 6 new second generation GP3300 cells picked at random and made into a pack.

The intensive care system takes the cells down to 0 volts without cell reversal.

Pro-Trak was altered to enable it to switch on the ICS during the delay between cycles.

We alternated the use of the Intensive care system to see what difference equalising made to the pack.

Charge current 6.00 amps
Charge delta 0.05
Discharge 20 amps
Time between cycles 2 hours

Charge data sent every 16 seconds
Discharge data sent every second
Rest data (between cycles) sent every 120 seconds

 

Below are 2 cycles taken from the complete test to show cell data while using the equaliser. 
(Cell 5 is a poor cell but it was poor from the start and is not a consequence of the tests)
At the end of discharge the cells recover to about 1.25 volts but the equaliser slowly deep discharges them to 0 volts. From the graph it can be see this takes about 70 minutes. 
 
 
Below are 2 cycles taken from the complete tests to show cell data when equalisation is not used.
 
 
 

Following is pack information over 28 cycles showing the difference in performance when equalisation is used and not used.
It can be seen that the discharge time (and consequently the discharge capacity) is better when the pack is equalised, but no significant improvement can be see on the pack average voltage or the pack voltage at 5 minutes.
 

 

 
 
 
The fallback of the pack is recorded at 1 minute into charge, the graph below shows that the pack had a fallback of between 0.04 and 0.06 volts during the first minute of charge after being equalised. There was a fallback of only 0.01 during the 1 minute of charge when the pack was not equalised.
 
The graph below shows the pack will false peak when it has been equalised.

 
 
 
The light blue in the following graph indicates that equalising was stopped when the pack reached 0 vols.  This allowed the pack to recover to around 6 volts before the cycle was started again. The discharge time would appear to be the same as when equalising was continuous during the time between cycles.
These tests were with only 2 hours between cycles.  Maybe there would a difference if the cycle time was longer.  We will repeat the tests but with 24 hours between cycles

 
Below are 2 cycles taken from the  tests to show cell data when the pack was equalised to 0 volts, then equalising stopped. You can see from the graph that at the end of discharge the pack is equalised, when the pack reaches 0 volts the equalising is stopped which allows the cells to recover to about 1 volt per cell.

 

We know the above tests are not ideal but some information on equalising is better than none!