DPL-8 and Storage

[CharlieK]

After much fooling around - - -
Hitting very close to 3.85v per cell involves some piddling around.
It seems strange to me that some of the less renowned chargers I've owned can do a more accurate discharge to 3.85V per cell than either side of my DPL-8.

Using a 4S4000mah Power Wing Lipo,
It looks like the discharge taper as you get close to the desired end voltage isn't quite right! Yes, things can be improved slightly by using different methods, such as discharging to below the desired storage voltage, then charging at a low current until the correct voltage plus a small amount to allow for setting is reached. Or, reducing the discharge current manually for the last part of the discharge to 3.85v.

Why is this of any concern? Well, for one, if I store at exactly 3.85v to start with, it's a lot easier to spot oddball cells in an otherwise good lipo.

[Ace Dude]

Originally Posted by CharlieK

After much fooling around - - -
Hitting very close to 3.85v per cell involves some piddling around.
It seems strange to me that some of the less renowned chargers I've owned can do a more accurate discharge to 3.85V per cell than either side of my DPL-8.

Using a 4S4000mah Power Wing Lipo,
It looks like the discharge taper as you get close to the desired end voltage isn't quite right! Yes, things can be improved slightly by using different methods, such as discharging to below the desired storage voltage, then charging at a low current until the correct voltage plus a small amount to allow for setting is reached. Or, reducing the discharge current manually for the last part of the discharge to 3.85v.

Why is this of any concern? Well, for one, if I store at exactly 3.85v to start with, it's a lot easier to spot oddball cells in an otherwise good lipo.

Which preset are you using and is it a factory preset or one that you customized?

The PL8 is a very robust and powerful charger with lots configuration options so results can vary depending on the many configuration options available.

[Gregor99]

The default termination for storage charge/discharge is C/5. This based on the discharge current you are setting on the charger. Using a higher value such as 10 amps will raise the termination current. Using a lower value for discharge current will lower the termination current and give a result closer to the set point of 3.83v.

Alternatively you can change the termination setting to C/10, C/20 or set a manual current value. Either of these would give a result closer to what you are looking for. As storage charge is considered a "range" and not a precise number, the default settings tend to be a balance of speed and precision. For those that want more speed, or more precision, the tuning options exist.

Originally Posted by CharlieK

It seems strange to me that some of the less renowned chargers I've owned can do a more accurate discharge to 3.85V per cell than either side of my DPL-8.

There are several reasons this could be the case. Most other chargers will not be able to discharge at the same rate as the DPL. A slower discharge rate means hitting the target requires less time in the CV stage. I could provide more thoughts if more details are provided. Discharge current, termination mode (CC/CV, or CC/CC), termination settings of both chargers as well as resulting pack voltages from each scenario.

[CharlieK]

After watching the discharge function as it apples to discharging to storage, even using factory defaults.
First, the discharge usually stops at the selected voltage or close to it.
Then, as the Lipo rests over few minutes period, the voltage per cell will typically rise. How much depends upon the discharge current settings. Even with balance on discharge, (which helps) I end up having to fool with the end of discharge voltage settings. The difference is about the same value as an offset value set by a default in the charge settings. (15mv as I remember) Currently,I'm comparing results from the storage mode and the accurate charge mode, using two 3s2200 mah brand new 50C Lipos
My trusty old 50W Bantam is a little more accurate in hitting the storage cell voltages.
(But not as accurate in hitting the full charge voltage.) (4.2-4.23, instead of the DX8's 4.204 or 4.205 after a few minutes settling time.

As to why do I care - - -
Storing at 4.85v allows me to see what the "bleed down" while stored is, and if a cell in a particular lipo is typical of the other cells, or may have an pending problem. Several of my older lipos are showing typical signs of age, and this, along with decreased capacity helps the decision to retire process.
The main criteria I initially consider
Does the lipo still have the capability to provide an appropriate voltage under load? For how long?
If the lipo carries a 30C (example) rating, can it still provide at least 15C under load? (Or at least twice the current actually needed for the application)
Do any of the individual cells show behavior noticeably different from the others in the same lipo?

Lipos are not cheap, but cost less than replacing an RC model plane that had an unanticipated power problem at a critical point, making it a basket case in fact. Think vertical climb, power loss, low altitude stall,
no time to recover.

An off the subject comment. when looking at the cell voltage graph in expanded (zoom)
some times the graph does not agree with the non zoomed values, or the charger readings. I was also surprised to find that the CC program was evidently written in Visual Basic.

[Ace Dude]

I suggest you re-read Gregor99's note above in it's entirety. It contains a lot of good information on settings you can use to achieve the results you desire.

IMHO using self-discharge values to determine when to retire a pack isn't the best method available. Sure self-discharge rates was something we looked at with NiXx batteries to determine their health, but there are better methods with LiPos.

With LiPos we can measure the individual IR values of every cell in a pack. Tracking IR values over time is the best indicator of when a pack/cell may be on its way out and it's time to retire the pack. Periodic capacity checks and measuring voltage under load are helpful as well, but measuring voltage under decent loads requires specialized equipment.

There's nothing wrong with VB. There are thousands of commercial software products that have been written in VB.

[Gregor99]

Originally Posted by CharlieK

As to why do I care - - -
Storing at 4.85v allows me to see what the "bleed down" while stored is, and if a cell in a particular lipo is typical of the other cells, or may have an pending problem.

What you are describing is evidence of an internal short in a cell. I applaud your efforts to track and respond to this. Typically the sag is pretty obvious even if all cells do not start precisely at 3.85v.

Your Bantan is likely discharging at only 2 amps or less (50w on 6s). Its easier to hit the target voltage if the discharge current is lower. Re-read my post above. There are some tuning options you can work with to get the result you are after without lowering the discharge current.

[CharlieK]

"There's nothing wrong with VB. There are thousands of commercial software products that have been written in VB."
Actually, VB had some severe limitations involving GUI support, problems with graphing support, and even limitations on calculation accuracy. If the latest and greatest versions resolved this, then not a few of the older VB based programs need to be updated, if they are to meet today's needs.
I'd freely admit that some of the problems with VB were the result of default options that MS made. Others were involved with incomplete documentation. Since I retired almost a decade ago, I've not kept up with the later changes. On the other hand, many of the VB based programs currently out there were written using older versions of VB.

[CharlieK]

[QUOTE=Gregor99;6516312]What you are describing is evidence of an internal short in a cell. I applaud your efforts to track and respond to this. Typically the sag is pretty obvious even if all cells do not start precisely at 3.85v.

"What you are describing is evidence of an internal short in a cell"
Very True! However, I've run across several cells in various lipos that seem to have such a problem, and is only evident when they are stored for more than a few weeks. A contributing factor seems to be that the Mfrs (Not FMA) are not matching cells as well as they used to.

No charger can be everything to everyone!

[Gregor99]

Originally Posted by CharlieK

No charger can be everything to everyone!

I'm not sure if you tried the settings I suggested. But out it sounds like you may have given up to soon. Perhaps, not. Its not clear. If you haven't tried any of the tuning settings, I encourage you to do so. You might find the Powerlab can be everything you need.

[CharlieK]

Yes, I have "played" with tuning settings.
As to the effects, some are not really clear.

[Ace Dude]

In the CCS, if you mouse over the field label you'll get a pop-up explaining what the parameter does.

[zomaar]

I noticed that discharging to storage on the PL 8 was sometimes erratic and the charger was like 'waiting /balancing state but no current was pushed/pulled into the PB's '.
The cells where not balanced during the discharge and with the cell sag , some bounced more then others (old packs) and it tried to correct it without really balancing ,so it ended with the cells uneven

I enabled 'balancing' during discharge in the storage mode and ever since the discharge is consistent and it doesn't get in this ' I don't know what do do state'
I assume this balancing keeps the cells in sync during full discharge. since all cells reach the same target voltage at the same time due to the balancing the PL can stop the storage charge nice and quick.

[Ace Dude]

Many feel it's irrelevant to balance the cells during a discharge to storage voltage as storage voltage is a range of 3.70-4.00V/cell and not an absolute value. 3.85V/cell is the mid-point between the max and min storage voltage values. Balancing during discharge to storage can also add additional time to the process.

[zomaar]

I know ( any balance will add time ) but its like trying to fix everything at end vs fixing from the beginning.
good balanced on storage gives an ease of mind , but i was just much more annoyed with the inconsistent storage charge timings, sometimes it was fast , others it took 30 mins or even more.

If you discharge at 10 A ( might even be more 20A , like in my case , with old pack) The balance during doesn't really make much difference vs the balance it tries to do at the end ( and still fail)
I can only say what I observed , in my case the discharge with balancing is alot faster

[hemp]

Originally Posted by CharlieK

Actually, VB had some severe limitations

Actually, since VB supported COM it had no practical limitations except those of the developers who didn't know how to leverage COM. Of course COM had its own issues.

Since we're making arbitrary judgments based on language used to write the software ... I figured we should at least be talking about the right language: The current CCS was written with the .NET framework. VB.NET has almost nothing in common with the VB of old aside from language syntax.

While that doesn't in itself guarantee that a program is well written (it does have room for improvement), neither would the use of VB6 have indicated that the program was poorly written. Although I'll agree that many, many VB programs were written very poorly.

Originally Posted by CharlieK

Since I retired almost a decade ago, I've not kept up with the later changes.

I'm not even close to retired. But I'm old enough to have done my time with VB ... and COBOL .. and numerous other horrible atrocities which at one time were considered acceptable.

[CharlieK]

Getting back to the original Storage Issue. It looks like one of the contributors is the difference between the set voltages.
Charge X.xxx Discharge X.xx This may partially explain why I can usually get better results by discharging below the desired 3.85v, then charging to reach it with a "normal" charge preset to the 3.85 reduced set voltage.
I also wonder if some of the charge settings also apply to discharge, when there is not a specifically called out comparable setting in the discharge mode.

The majority of "hard core" programming I did was for instrument and equipment control in industrial and military test systems some time ago. Think of an electronic test system/station that had stimulus and measurement capability from DC to tens of Ghz. And with accuracy that was such that full calibration and accuracy required the use of flown in transfer standards that were "one off" the national standards.
The last form of basic I played with (last year) was a non MS one. It's a modernized version of the old Qbasic intended for use on 64 bit windows and other systems. Why? a friend wrote a stock market related program package years ago that originally used an old Lotus version and Qbasic.
The modern replacements for discontinued Lotus did not have the ability to do certain things involved in graph scaling. He was able to use the basic 64 version to convert (with minor problems) and update the older program versions, thus avoiding a major amount of time and effort. The last released lotus version he has does work as well with win 7 & 8.

[Ace Dude]

Originally Posted by CharlieK

Getting back to the original Storage Issue. It looks like one of the contributors is the difference between the set voltages.
Charge X.xxx Discharge X.xx This may partially explain why I can usually get better results by discharging below the desired 3.85v, then charging to reach it with a "normal" charge preset to the 3.85 reduced set voltage.
I also wonder if some of the charge settings also apply to discharge, when there is not a specifically called out comparable setting in the discharge mode.

The majority of "hard core" programming I did was for instrument and equipment control in industrial and military test systems some time ago. Think of an electronic test system/station that had stimulus and measurement capability from DC to tens of Ghz. And with accuracy that was such that full calibration and accuracy required the use of flown in transfer standards that were "one off" the national standards.
The last form of basic I played with (last year) was a non MS one. It's a modernized version of the old Qbasic intended for use on 64 bit windows and other systems. Why? a friend wrote a stock market related program package years ago that originally used an old Lotus version and Qbasic.
The modern replacements for discontinued Lotus did not have the ability to do certain things involved in graph scaling. He was able to use the basic 64 version to convert (with minor problems) and update the older program versions, thus avoiding a major amount of time and effort. The last released lotus version he has does work as well with win 7 & 8.

Keep in mind the default values in the storage preset are optimized for getting a pack to storage voltage as quickly as possible w/o necessarily ensuring every cell is precisely at the same voltage at the completion of the preset program. Gregg has made several suggestions on how to achieve the results you are trying to accomplish.

Certain business requirements will require specialized systems/programming languages to achieve the level of precision and accuracy required. VB was never intended for the scientific and engineering communities. However, it's certainly adequate for the CCS.

[CharlieK]

Actually, some of the most difficult tasks I remember from the "good old days" were finding ways to convert measurement methods from the old "classic" laboratory ones to something comparable that retained accuracy, yet could be automated.and eventually "computerized" This was in the days when the 741 Op Amp was the new kid on the block, and RTL IC logic was going into large scale use. Automated and repeatable accuracy down into the millivolt and milliohm areas was not something easily obtained! Being able to mass produce something like a modern digital lipo charger and sell it for today's prices was unthinkable!
Not to mention the difference in size from a relay rack/cabinet or two to something that you can hold in one hand!

[Ace Dude]

That's certainly before my time, but I do remember having to swap floppies several times to compile my Apple Pascal & FORTRAN programs on Apple IIe and Apple 2+ computers back in the early 1980s. Single sided floppies, 48K memory, and no hard drives. Even upgraded a few of them from 48K to 64K one day. Ah, the good old days....

[CharlieK]

I remember writing a Midi utility for the Apple II+ in those days. It accepted a memory dump from a Yamaha keyboard, placed the raw data in (of all places) Hires video memory, and then wrote to a file.
The file could then be sent back to the keyboard, or used by a midi editor/utility. The utility was written in a combination of assembly and machine code. It had to fit in a very small memory area that was not disturbed by either DOS or Basic. At about the same time, I also ended up writing some printer drivers for CPM running on a Apple II Z80 card.
The more or less demise of the Apple was due to disk drive and interface card problems.