A SOC measured by pure voltage isn't very accurate at all when speaking of Ni-MH cells. Using voltage is the only method to monitor SOC because the pack is in constant use. It a different story with Li-Ion cells with which it's a very good SOC test.

To prevent "voltage depression" or the common used term "memory effect", your pack does a deeper discharge/charge cycle once in a while.

Even if the cells lose capacity with age, you only use a narrow portion of it anyway.

If there isn't any spot welded straps, how are the cells connected in series then? Don't tell me that they use spring clips!!!!

 

Someone can send a PM to Rich Krueger (moderator on this site) for input. He is a journalist who was at the 2008 Detroit International Autoshow and reported Johnson Controls Saft approached Ford Motor Company about a battery swap. If I remember correctly for the reported 30% improvement in (mpg) mileage, the internal cooling and heating items would be removed, higher density plus more battery cells added and a software updated added. Cost? Rich "quoted" around $3,000.00.

Rich might be on his way the 2010 Hybridfest where he competes against other hypermilers.

 

Isn't that a battery change to Li-Ion cells? Obama is in Michigan this week about the opening of the new LG Chemical plant to make Li-Ion cells destine for the next generation of Ford's Hybrids. This is about going to a higher capacity Ni-MH cells.

 

"A SOC measured by pure voltage isn't very accurate at all when speaking of Ni-MH cells. Using voltage is the only method to monitor SOC because the pack is in constant use. It a different story with Li-Ion cells with which it's a very good SOC test."

So how do you measure it? Sample the electrolyte? Count the coulombs in & out of the battery? If you do the latter how do you account for the aging effects that ever so slowly alter the chemistry of the battery?

Roger the deep discharge for rejuvenation in the software but it will never recover all capacity lost due to aging, just most of it... but the aging will always have cumulative effects.

In the end the electro-potential inherent in the chemistry of the battery will be the best, and likely only, indicator of capacity and state of charge.

...And I'm not talking of Li-ion batteries.

 

What I'm saying is that in a Ni-MH cells monitoring SOC without a load will only result in a approximate value but and in FEH case, does it really matter? The "only" true way to test for capacity is to discharge the cells to .9V/cell @ .1C. There are testers around like the ZTS Pulse Load that do a SOC test without discharging a cell but on a pack because cells become unbalanced is difficult. Once unbalanced, only one or two 16 hr charges @ .1C will cure that. Since the FEH only uses a small portion of the pack's capacity it seems that the once in while deeper discharge/charge cycle is all that is needed.

I know that your niot talking about Li-Ion cells!

 

Roger the test discharge issue. I've done my share on really large batteries. But you know, these test discharges monitored individual cell voltages and were terminated when a given cells ICV dropped to its limit.

In the end, when you criticize SOC as measured by voltage or voltages, you demonstrate a lack of understanding of the underlying issue of electro-potential, cell chemistry, and how to measure it.

Using an algorithm to determine SOC from calculated amp-hrs vs voltage(s) is a reasonable methodology given the computing power available today. But in the end you're captive to a real world, non-software based, "analog" battery chemistry issue.

 

Here we go again! If one doesn't agree with Mr. KnowItAll, one must have a lack of understanding of whatever the topic is!

It's no wonder your arrogance has caused menbers to leave this forum for good!

Have it your way, I'm done with this thread.

 

Voltage is ONLY a good indicator of charge at the very high, or very low end.
It is a poor indication when a battery is at ~50% SOC.

The Ford likes to keep the battery at a ~50% SOC all the time.
It discharges only a little ( 5% - 10% ) then charges a little by the same amount.

So during normal day-to-day operations, it DOES use columb counting ( or amp-hour counting ). Whether it is really moving 48% to 53% or actually moving from 40% to 45% is unimportant. All the computer needs to know it is used up 5% and it needs to put 5% back in.

Obviously, this system will drift over time. To account for self-discharge, heat losses, etc. I'm sure it put's 1.x Ah charge back in for every 1 Ah used.
With enough data ( and you can bet Ford's got it ) you can get very good at this.

As a safeguard, voltage is monitored. If the voltage gets lower than "expected" it charges more and vice-versa.

I watch SOC on my SGII. Recently, I saw it jump from 44% to 10% in about 1 second. What I surmise happened was, the voltage hit a low trigger, (safeguard) and the computer said "Oh Oh... the SOC is not where I thought it was... it's not really 44%.... so it jumped to 10% ( based on voltage), and then slowly charged back up to ~53% which is the "nominal" charge.

I saw this happen for the first time at 115,000 miles.
However, I've seen it go from ~50% to 90% and work its way down about six times.

For what it's worth, my Honda Hybrid does this about every 300--500 miles.
( SOC adjust ) Ford does it about every 10,000-20,000 miles.
However, my Honda was built in 1999 (2000MY) so Ford may have learned some things in the 5 years since.

HTH,
-John

 

My issue here is simple: how do you measure SOC in a live, operating battery? ...And in real time. SOC does not jump around, it doesn't somehow go from 44% to 10%. It can't possibly do that inside the battery.

So you're stuck with how does the system measure SOC and then how does the software manage it.

What kind of errors can occur with the measurement system. In several posts I have seen an attitude that seems to indicate there is, somewhere, a little window that tells one a spot on SOC.

My experience with really large batteries (Pb-Acid) tells me that ICV is the only really useful way to tell what's going on. This is an issue of measurement of an analog thing, not how the digital software manages it once measured.

Finally we have to take notice of the effects of aging. The chemistry changes over time, capacity drops because of this, and remember that per cents are a relative, dimensionless number. When you translate 10% early in life it might mean a given number of amp-hours, but late in life it will, probably, mean measurably fewer amp-hours. So don't get fooled by per cents.

In the end I am inferring things here from extensive Pb-Acid battery experience. As more NiMH battery lifetime experience is gained the picture will change and I will learn from it.

What won't change is: How is SOC measured in the real world? I still believe that ICVs are by far the easiest and most accurate. When combined with the computing power available today it gives a powerful ability to maximize battery capability & lifetime.

 

Bill-
That's why I said percent reported on the screen is unimportant, as long as it is quantity in, followed by quantity out.

I think you mis-understood.

I did not say the SOC jumped from 44% to 10% in an instant.
I said the software jumped from 44% to 10% in an instant.
I think you are smart enough to understand the difference.