Hyperion EOS 1420i NET3 14s Balance Charger...

[wesnewell]

Nope. If you want to balance charge a 14s pack, you have to have 2 7s balance plugs, or charge the odd one separately. My 14s pack is 5s-4s-5s in series. I made an adapter cable that takes the 5s and first 2 cells of the 4s into a 7s plug, and then the last 2 cells of the 4s and the second 5s into another 7s plug. ha I known I needed to do this first, I'd have configured my 14s pack with 4 bricks instead of 3. But it doesn't matter now, I can just add more of the same since I've already built the adapter cable. If you've got 2 6s and a 2s, it needs to be 6-2-6 in series, then rewire the balance leads to 7s. Or charge the 2 6s together and then charge the 2s separately.

 

[wesnewell]

I did it this way so i could just add more of the same bricks later without having to modify anything again.

 

[deVries]

Hyperion Firmware-Software Issues in Storage Mode -or- Manually Charge for 55% Storage (50-60%) & Does Not Work Accurately ??? (I updated this post with new information I learned. See lower part & look for "Update".) Also, TCS Capacity set at 40% will get you about 3.87v for "storage".

I own two Hyperion Chargers & use the software too. I've used both firmwares 5.0 & 5.1 & any software updates, but I'm getting the same results & issues before & after any updates. (My chargers are the older 720i Net 3, but these also use the exact same firmware & software update files. Firmware just programs & updates for 7s instead of 14s.)

I came to this thread to see if others are experiencing similar lower charge setting issues manually using TCS settings -or- using the automatic Storage Mode, and some "here" seem to be experiencing similar or related problems. I sent an email to support at Hyperion, and I thought I would share that with everyone in case someone can explain what is happening...

Problem is Software is Not Stopping Chargers at Accurate Cut-Off
Points or Charge Related Settings Not Functioning using Storage Modes
or Manually Charging "up" to a stopping point using Memory Editor for
TCS Capacity set at 50-55-60%, for example, or by these other methods:

The Charge Level Percent indicator in Log Viewer or Control Function
Windows Never Updates "itself" often enough to be accurate in Real
Time while charging or discharging or at the end of loading pulse
phases to accurately gauge the "float" offload readings too. This
inaccuracy between load readings & float become apparent after
pressing Stop. This misreading of any float-charge not updating
accurately seems to cause the software to not cut-off the charge
correctly or Storage auto-charge or discharge (up/down) accurately by
always overshooting the mark up (charging well beyond 50-60% capacity)
or down (discharging well below 50% capacity). Only after pressing the
Stop button does the "true" Charge Level Percentage update to the
accurate float level it has charged to, and the charge is always out
of range for what it should be. I use a very conservative 2c
charge/discharge rate for 30c (or higher rated "nano") LiPo batteries.

The automatic Storage method, unfortunately, does not allow the
software to cut the charger on/off & function as needed to get
accurate results in any of these Storage up/down modes or by using my
manual TCS capacity charge-to-storage modes, because when the Load
charge/discharge or Storage charge/discharge "updates" after pressing
Stop "the float" charge reading "jumps" the charge out of its "target
range" with voltage & capacity too low when discharging or too high
when charging for Storage purposes.

For Example: Running Storage Mode from a higher starting voltage of
greater than 3.9v+ going to lower state of charge for Storage in
discharge it always runs the LiPo down to about 3.83v. After pressing
the Stop button the previous inaccurate higher percentage charge
reading "adjusts" downward by jumping to the assumed accurate but too
low discharge "state of charge" of about 43%. This 43% (discharge)
reading, of course, is too low out of range for what should be a
higher voltage for a capacity of 50-60% "state of charge" to get the
correct Storage Charge. No matter how I change settings it always
arrives at this same result when discharging down in Storage Mode to
about 3.83v. Going in reverse using the automatic Storage Mode by
charging UP from a lower voltage, example starting low at 3.6v, the
Storage mode always overshoots the voltage charging way too high above
the correct State of Charge, so it stops well over 60% at about
3.975v. This does not matter whether I ALSO change the State of Charge
using TCS (instead of Storage Mode) to be set at 50%, 55%, or 60%
using Charge Mode instead of Storage Mode. TCS still overshoots this
50-60% range way too high too.

Can you explain what I need to do to get this software to function
correctly for storage settings using Storage mode or a manual TCS
setting?

Also, the TVC adjust voltage does not seem to work when charging TCS
to 50-60% capacity ranges either, when I try to set TVC to -20mV
lower.

I have plenty of data logs to show these issues exist if you need some
examples. I am new to battery software management but I am an expert
computer user, so I don't understand at this point if I'm doing
something incorrectly that I'm not aware of. Maybe I need some
guidance & support to get this working correctly.

Thank you for your help!
====================

For those interested in the Hyperion Software features/bugs I found the Software Developer runs a thread about it here:

Hyperion EOS Control Data Suite Software

Click the Link Above :wink:

I think the automatic Storage Mode charges too high & there is no way to adjust Storage Mode. If Storage Mode has to discharge, then I think it stops too low, imo.

Using TCS with STOP "on" (End Action) you can charge "up" using 40% Capacity TCS instead of using Storage Mode & it will stop at 3.875v, which I think is ideal for LiPo storage. (Hey, if anyone reading this has another opinion, then please post your suggestions/ideas too.) In order to do TCS at 40% the battery will need to be at or below 3.810v. Otherwise, the charger won't start the charge & just time-out, because the starting voltage is not low enough.

UPDATE from an RC Forum:
I see that Dave posted this:

NipponDave: The storage termination voltages should NOT be equal on charge vs. discharge. STORE is attempting to achive a resting voltage after the pack sits for a day. Therefore on Charge the voltage must end higher, as it will fall later.; On Discharge, the voltage must end lower, as it will spring back up later.

And either way the exact storage % is irrelevant to the objective, as long as it is somewhere between about 60% and 75%. Like horseshoes, close counts, and we get it real close after a day of rest...

I don't think the Firmware is very sophisticated about Storage Mode, as it drops/discharges to the same voltage 3.83v (which can't be the same as a capacity calculation) every time no matter where the battery was charged to above 3.9-4.2v LiPo. And, if the Storage Mode has to Charge (instead of discharge), then it takes a battery from 3.5v-3.8v up to the same high level charge every time well above 60% Capacity to 3.96v. (After further reading it seems there is no good way to calculate the battery capacity vs voltage & using that capacity estimate for Storage settings because of the mid-range state of charge resting point needed for storage. But, I'm not expert to judge that.)

I do *not* believe these are the two ideal storage set points for maximum battery lifespan, 3.83v & 3.96v, since this is such a big voltage difference high/low of only two Storage set points, but I'd like to see the battery lifespan Storage data that supports this if I'm wrong.

Also, I want to request a detailed explanation about this should be at least in the online instructions about the Charger & software. I think "Dave" said there had been an explanation in the printed information at one time, but it was removed to save space??? Considering the confusion over this I've read from others about this too I think the information needs to be official & provided and not buried in forum posts.

Here is what I learned so far...

everydayflyer:
Storage range is 3.7 to 4.0. I have always stated 3.85 as that is the mid point. Too many get too concerned over an exact voltage or capacity. Capacity can (could) only be determined by a full 100% discharge and a 100% charge both of these at a fixed C rate. Even after doing this as the cells form (when new) or age after use their capacity will change and thus so will the % at any given voltage.

This is not precession Lab quality test equipment. It amazes me that it is as accurate as it is when the cost is considered.

russ skinner:
On a general note I'm finding the storage mode facility of the charger has about .1V tolerance. Think this is a result of the pack capacity estimate. If I run store mode on a depleted pack (resting at just under 3.8V/cell) the cells are charged to approx 3.96V/cell. If I run store mode on a fully charged pack (4.2V/cell) the cells are discharged to approx 3.86V/cell.

Have posted this issue in the main EOS thread - no response so I'd be interested to know other peoples thoughts here if this tolerance is ok?

MrMel:
Storage: Dave has to answer this better (it has been answered before in the EOS thread somewhere, buried deep), yes, it's OK, and within spec, it don't need to be exact.

I think there is probably an ideal voltage that will extend lifespan of the LiPo batteries better than either 3.83v (Storage Discharge) or 3.96v (Storage Charge), but I don't have any data to support this assertion. But! Considering we charge/balance with such high accuracy I'm guessing there is an ideal setting, especially, for the higher quality LiPo with higher C ratings.

 

[GGoodrum]

What exactly is it that you are trying to do? If you are talking about long-term storage for these, the info on RCGroups is correct. The exact value is not all that critical. The cells just need to be somewhere in the 60-80% range.

-- Gary

 

[deVries]

What exactly is it that you are trying to do? If you are talking about long-term storage for these, the info on RCGroups is correct. The exact value is not all that critical. The cells just need to be somewhere in the 60-80% range.

What is confusing to me is reading that a LiPo storage charge should be between 50-60%. Since the charger can't accurately gauge capacity in the mid-range for a Storage charge, I'm guessing it's the voltage that has to be the reliable target. To get between 50-60% I've read a good Storage charge voltage to aim for is about 3.85v to 3.9v. I've picked 3.875v for a Storage voltage which is exactly 50% of the 3.6v to 4.15v range I charge/discharge to.

The Hyperion Storage Charge goes well beyond this lower range of 50-60% when it charges "up" to do the Storage Charge... It drives the voltage up to 3.97v before stopping.

If the range of voltage for discharge to full charge is 3.5v to 4.2v, then we get 0.7v for the range of voltage. The Hyperion uses 3.97v for its Storage charge when going "up", so this is about 67% of that 0.7v range.

There probably is an ideal voltage number to shoot for, but I am confused between this 50-60% LiPo recommendation vs 47-70% that the Hyperion uses for its two target storage voltages of either 3.83v (discharge 47%) & 3.97v (charge 67%) both are for Storage.

Does anyone know how other chargers work for Storage Mode? What voltages are used & can these settings can be changed to a user setting besides the 2 locked defaults used for Hyperion chargers?

Does anyone know of any LiPo storage research data that supports specific numbers beyond this wide range that Hyperion uses of 3.83v or 3.97v?

I bet there are more ideal or specific voltage numbers to use for extending maximum lifespan of LiPo in Storage Mode, but I don't have that information. Has anyone researched this or know much about this? Thanks for any ideas or information you can provide.

 

[rscamp]

What you really want to avoid is near 100% SOC and a really low SOC. The former increases the rate of degradation and the latter risks over-discharge over time. Everything in between is better than these extremes and variations within this range really offer only very slight differences in longevity. I wouldn't over analyze this or lose any sleep over it. Variations in the cells, charge termination voltage, discharge termination voltage, current draw and storing at different temperature all probably make bigger differences in battery life than whether a battery happens to be stored at 40 or 60% SOC...

 

[deVries]

What you really want to avoid is near 100% SOC and a really low SOC. The former increases the rate of degradation and the latter risks over-discharge over time. Everything in between is better than these extremes and variations within this range really offer only very slight differences in longevity. I wouldn't over analyze this or lose any sleep over it. Variations in the cells, charge termination voltage, discharge termination voltage, current draw and storing at different temperature all probably make bigger differences in battery life than whether a battery happens to be stored at 40 or 60% SOC...

I tend to agree with what you're suggesting. It would be interesting to learn some hard test data about this. The little I know & have found is that temperature might be the biggest factor for long term storage as long as SOC is in a good range too of perhaps 50-60%. Putting LiPo in the Fridg or Freezer might be best if not using for months or years. Just don't charge or use the cold batteries until these have warmed-up to reach near room temperatures.

Someone posted this on an RC forum, so this is an example that there are wide ranging suggestions about how to store LiPo & at what voltage to use...

From Charlie of Thunder Power
We recommend 3.8V (50% charge for 2 to 3 years storage)
3.5V allows 9 to 18 month storage (18 months at optimium condition)
3.3V allows 6 to 12 month storage (12 months at optimium condition)
3.0V bad idea

 

[GGoodrum]

The only reason for picking a storage voltage value in the 60-80% range is to give the cells plenty of "room" for self-discharging. There is nothing magic about actual number used, Nothing at all. All cells will self-discharge a bit over time. It matters not whether the cells are at 50% or at 90%, the only difference is that if the cells start out at 90%, they will have that much longer before they die. There's no need to agonize over what actual value a charger sets it to as a target. Enough said?

-- Gary

 

[EBJ]

OK, I'm a little lost here. Perhaps someone can help.

I'd like to make a single-plug Hyperion set-up for my pack.

I plan on having a 36V 20Ah pack made from (x8) Turnigy 5Ah 5S 20C Lipo Packs.
so you guys call that a 10s4p Lipo Pack right?
So I imagine I'll still have to run ALL of the balance wires (assuming there are 5 in each 5s pack?) that': 5 x 8 = 40 balance wires into the Hyperion charger.
I just don't see how the Hyperion can take in 40 balance wires?

I must be missing something...

 

[wesnewell]

You don't run all the balance leads into the charger. On a 10s4p pack, all the balance leads of each side are commoned and only one set would go to the charger. The easiest way to do this imo is to buy a couple of 1->6 balance cables and simply plug 4 packs into each one. Leaving you room for 2 more in each if you want to go to 6p. I'd also go with 6s packs for 12s4p. That will give you more Wh in the pack for more range, and a little higher voltage for more power/speed.
http://epbuddy.com/index.php?main_page=index&cPath=26_14&zenid=me30rihu7ndk47ndc4v8m5ec35

 

[EBJ]

You don't run all the balance leads into the charger. On a 10s4p pack, all the balance leads of each side are commoned and only one set would go to the charger.

Can you please explain what is meant by 'each-side". Thanks.

Also, if the charger only sees "commoned" balance leads, then how could the charger know each individual cell V ? I was under the impression that each Lipo Cell needed to be monitored. Because some will take to charging differently than others.

 

[EBJ]

Silver,

Don't think of it as a 4S4P pack... think of it as a 4S1P 20AH pack - when properly paralleled, each of the balance taps would also be paralleled from the 4P configuration, these are connected to a balancer as if they were a single 4S pack of 20AH and then you balance.

If your going to ask next, but wait couldnt a failed block in the parallel cause failure of the surrounding cells? Yep it could, but monitoring the cells as a pack is good enough providing you don't over discharge or ove charge them as a pack.

Personally I would like balance taps at each P in the pack, atleast for cell level monitoring but really... cell level monitoring in our terms is more of "Parallel Strand monitoring".

Hope it helps!

-Mike

I found my answer in another thread. It seems when you parallel balance-wires you are also paralleling the cells to be ONE cell (but with a higher Ah rating).

 

[EBJ]

another question:

Gary uses an LVC/parallel adapter to parallel his Balance wires. Is this needed? How is this LVC different than a motor-controller lvc?
Perhaps: This "adapter" is acting as a lvc at the cell-state? where as the controller lvc is acting as an lvc at the entire pack state?

Thanks.

 

[EBJ]

One more concern:

What happens if one of your paralleled balance wires were to disconnect unknowingly (maybe due to a crummy connector or something) ?
Then wouldn't that group of cells (however many are still left connected) be charged at a higher Ah rating than all the others???

 

[EBJ]

bump. hopefully I can get a reply to those last two posts. Thanks.

 

[wesnewell]

I don't know what Gary uses, but it sounds like a cell (actually cell group) level LVC. The controller LVC is by battery pack voltage into the controller. If a balance wire came off for some reason you should notice it. That's why you inspect the pack visually before every charge. Still, it wouldn't cause any problem unless the cell was badly out of balance. The cells don't get charged through the balance leads.

 

[GGoodrum]

another question:

Gary uses an LVC/parallel adapter to parallel his Balance wires. Is this needed? How is this LVC different than a motor-controller lvc?
Perhaps: This "adapter" is acting as a lvc at the cell-state? where as the controller lvc is acting as an lvc at the entire pack state?

Thanks.

Yes, this is so you can have LVC protection at the cell level. This protects against the case where you might end up not as many packs in parallel, for instance, due to a faulty/broken connector. In this sort of situation, the cells will run out quicker than you think, and die before you can do anything about it. Pack level LVC functions are mainly designed for lead acid setups, where the voltage is reduced somewhat uniformly as the capacity is used. Lithium cells are different. Their voltage will stay pretty close to the nominal level all the way until the end, and then it drops quickly (within seconds...). If you don't catch it, the cells can go all the way to zero, which is called cell reversal, and they are dead after that. It is like the voltage drops off a "cliff".

If you have a badly balanced pack, some cells could get down to this "cliff diving" point while the overall pack voltage is still above the controller's pack level LVC trip point. For instance, if you have a 48V controller, and the pack level LVC is set to, say, 40V, you could have the case where 11 cell groups, of a 12s-based LiPo pack, are still at 3.7V but one cell group is at 0V, and still the overall pack voltage will be above the 40V controller LVC trip point.

What the individual cell LVC circuits do is simply monitor the voltage of each cell, or group of paralleled cells, and if the voltage dips too far down (3.0V for LiPo and 2.1V for LiFePO4...), an optocoupled switch is thrown. All of these "switches" are paralleled together and then connected to the throttle signal and ground. The net effect is that if any cell's circuit trips, it will pull down the throttle signal, which cuts the load. This will cause the cell voltage to recover above the cutoff, which resets the circuit and turns off the "switch". If you still are full on the throttle, it will cause the low cell circuit to trip again, which again cuts the throttle. What I've found with a typical 10Ah LiPo pack, and also with 10Ah LiFePO4 setups, is that the first time the LVC trips, it "hits" the throttle, like a big tuna strike. If you back off the throttle about halfway, you can go another couple miles, or so, (depends on your usage, of course...), until giving it any throttle at all will cause this hit and recover cycle, which happens at about a 1-2 Hz rate. I've checked the capacity and usually when I get the first "strike", there's about 10% left in the "tank". If I let it go to where any throttle at causes the stuttering, it will take the full capacity of the pack on a charge.

I found it convenient to also include the parallel adapter right on the same board, which simply saves one extra set of wires. Some people make use of it, some people don't. Also, what I offer is a kit, which requires that you solder the parts to the board (3.5 parts per channel...). You can also buy a preassembled LVC-only board from Geoff, in the UK. You'd need to do the parallel adapter separate.

Another option for cell level LVC protection is to use something like a CellLog unit, which has a programmable LVC function which then turns on a buzzer. As long as you can have the CellLogs close enough to hear the buzzer, while you ride, this will work fine.

-- Gary

 

[EBJ]

Thanks for that info Gary. About how long does it take to solder-up everything in the "kit" ?
Also, can you please post your website again (as for some reason the link in a previous page wasn't working for me) *perhaps put it in your sig?
EDIT: Just noticed it's in your info once I click on your name.
Double EDIT: Website does not seem to be working for me. IE and Chrome can't display page.

Thanks!

 

[GGoodrum]

It takes about an hour, or so, to build up two 6s LVC/parallel adapter boards. Less tiMe, if you've done one or two before.

Yes, my site has been down since yesterday. I'm in the process of moving it to a new web hosting service. The one I've been using for some time has been down a lot for the last six months. Time to move on.

-- Gary

 

[EBJ]

Ok, So I found some 5A rated 25 pin D-Sub connectors from Digi-Key:
http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail&name=225MER-ND

I already have them in-hand. I was looking for some Bulk-Cable and only found 24AWG 25-conductor Bulk Cable from Fry's (for a good price too).
So I have a 10S pack, which means I will be using 12 of the pins for the balance leads. That means I can run 6 pins for the positive charge cable and 6 pins for the negative charge cable. And be left with 1 extra pin. Or I could run 5 pins for the positive and 5 pins for negative and be left with 3 extra pins in case I wanted to add a temp-sensor (currently don't feel like it is very useful)

I don't think 5 or 6 wires of 24AWG will be a problem for the main charge cables,
But I was wondering about the balance leads spanning 5 feet w/ 24AWG wire. (Right out of the Hyperion the stock balance wires are 24AWG)

What do you guys think, will 5 feet of 24AWG cable be a problem for the balance wires?

 

[hillzofvalp]

I doubt it will cause issues.. I was actually wondering this myself after I setup up a 26AWG serial (db25) cable a little less than 5 feet. I tested an individual wire and it got a little bit warm around 4-5A. Ran it for 2 minutes with no problems. I don't see the charger balancing on one cell for that long. Does anyone have any real objections? What is the max balance current on the 1420i?

Also, what do you guys think of the new 1000W hyperion 720 duo3. I was thinking about getting one for faster charging my 14s5p pack but don't like that it can't be linked with another charger for 28s... (not sure if that's really a game changer anyways)...

 

[EBJ]

Thanks. That 24AWG bulk wire is very cheap at Fry's in case anyone else is planning a D-Sub connector.

 

[jonathanm]

Ok, So I found some 5A rated 25 pin D-Sub connectors from Digi-Key:
What do you guys think, will 5 feet of 24AWG cable be a problem for the balance wires?

I'm no expert, but according to what I have seen on the web, the discharge current (while balancing) could range from 0.1A to 1A. your voltage drop over 5 feet of 24AWG cable could therefore range from 0.06v to 0.6v (according to some online calculator I just used)

Assuming the balance current draw ramps down as balance comes close, then it should be fine....but I would test it to make sure.

 

[hillzofvalp]

what is an acceptable voltage drop?

 

[jonathanm]

I don't know...the thing is the hyperion seems to be very sensitive....if the voltage detected via the balance leads is too much off what it detects via the main discharge leads, it might complain about it.

I would just give it a try and see what happens....

if the cells are not too much out of balance the current draw will be small and the voltage drop too, so it might not even be an issue....