Thread Tools
Jun 02, 2019, 09:20 PM
ancora imparo
jj604's Avatar
Joe, it's also worth reminding people that the definition "best" is not a constant either.

My thoughts on this are in post #507. They haven't changed in the 6 years since then.

https://www.rcgroups.com/forums/show...&postcount=507

Although that post was really about using IR as a quick and simple empirically derived substitute for proper load testing (and that substitution has proved to be qualified since the advent of the Graphene packs) the initial discussion of the inadequacies of a simple "best" number is still valid.

If people want a Consumer Reports or Choice "Best Buy" they must recognise that there are a myriad of assumptions hidden in the ratings, most particularly the weighting given to particular variables. Choice in this country are in fact a classic example. I assume others publications elsewhere are similar.

Let's say 40% weighting is given to Energy Efficiency in a Dishwasher in assessing the total score. Why? That is a purely arbitrary opinion by Choice staff based on their values. For some people energy efficiency in their dishwasher is really important, but for others quietness is the crucial factor and they literally couldn't give a rats's about how much electricity it uses.

In the case of LiPos there might be several Best Buys based on different criteria for a particular size pack.

Most constant current delivery over the whole discharge
Lowest voltage drop
Lowest temperature rise
Lowest weight/mAh
etc

Then there is the cost/longevity variable.

Is a pack that lasts a single season but delivers outstanding performance better than one that is only average but lasts for several years?
Depends on whether you value performance over all else.
Is a pack that lasts a single season but costs half the price of premium brands better than a premium one that lasts for several years?
Depends on whether you value performance over cost.

It starts to get complicated - but that's the point.
Individuals have to make a judgment based on the data available and their own values. In my opinion, no single number can encompass both since Joe doesn't know what your values are.

It might be possible to have a number of different "value" metrics. This would certainly sort the Thunderpowers and Max Amps out pretty quickly, but beyond that it gets harder. It's also a lot of work!

I also can't help but quote H L Mencken:

"For every complex problem there is an answer that is clear, simple, and wrong."

John



John
Quote:
Originally Posted by MCSGUY
I want to answer in more detail when I have more time. Definitely a value number is a great idea, I have some thoughts. But we also have to figure out a way to deal with the fact even the worst mfgrs put out decent product occasionally .

Joe
Sign up now
to remove ads between posts
Jun 02, 2019, 10:06 PM
HAL... Open the damn doors!
jfetter's Avatar
With any summary you have to factor some assumptions are made, just the sheer fact you are reducing 10 pages of data to a pie chart confirms it. No matter the product, the same condition exists, summarizing is the generalization or data for the ease of consumption, not limited to or made any worse by the fact we're talking LiPo data. Choose your poison, 10 pages or a generalized summary that may not detail specific exceptions...

Jack
Jun 02, 2019, 10:32 PM
aka JetMan Joe
MCSGUY's Avatar
Thread OP
Quote:
Originally Posted by jfetter
Joe,

I guess it's true... if you wait long enough, everything comes around again... including this exact discussion... by me as well?! Chalk it up to waiting a long time and forgetting or getting old and forgetting but this entire topic was raised (by me) and commented on (by you) back in November of 2015. Here is the culminating post IMO that shows the indexed data (weighted by the more important performance metrics) and your own rankings by scoring each data point without any additional weighting... we can talk about it again in a few years <Doh!>...

Jack
It's hard for me to be critical when one time I was running around the house trying to get out the door and asking my wife where's my glasses in a panic because I thought I'd lost them. You probably know how this story ends, I was wearing them

Joe
Jun 03, 2019, 03:43 AM
aka JetMan Joe
MCSGUY's Avatar
Thread OP
I test today because it connects me with the RCG community in a positive way. I also enjoy making it harder for Lipo suppliers to get away with selling bad product. I started testing originally because trusting a "respected" company cost me a bundle and made me realize just how ignorant I was about lipos. I reeeaaallyy do not like feeling ignorant. I like having my trust abused even less.

Posting the data graphs flame proofed me when I started criticizing lipo sacred cows. I study them because I want my opinions on the threads topic to be informed by facts. However Jacks and others inferences the summaries and maybe a rating system are better suited is supported by the fact the graphs garner few views. So In the future Ill only post them to answer questions or to support an opinion.

I agree with John that "Best" is a moving target. In the beginning the best pack was one I could count on not to die and destroy a $5000 jet. It was a revelation that for years apparently nobody else discovered the premier 65C pack of the day failed (each and every time) it was operated at just half its supposed continuous rating. Soon Best became the pack able to maintain the highest voltage (speed) at half its C rating. Today Im interested in identifying the lightest weight pack that will accomplish that.

I made a serious and comprehensive effort to correlate IR measurements with the performance I could expect in the air. The Lipo Tool is wonderful at determining an optimum C rating for a pack at rest. Together with dynamic high discharge tracking its possible for us to also have an idea how a packs chemistry will operate at running temperature. Ive found no other method of predicting how a pack will respond when pushed to its limits and beyond.

The first question is what is always Best for everybody all the time? Feel free to correct me if you think Im wrong but I believe the most important feature is consistency. Otherwise its Are you feeling lucky Punk? Im not inclined to buy lipos from Dirty Harry or his Chinese cousin. Unfortunately its impossible to be sure but subjectively, we do have over six years of accumulated history that can be used to come up with the QC Integrity Rating of a supplier (consider me the odds maker). Lets look at a possible A rating scale:

AAA.......Preferred Supplier, Best QC practices
AA..........Safe Supplier, Decent or improving QC practices
A.............Average Supplier, QC practices inconsistent
F..............Unsafe Supplier, buyer beware

Its a start maybe. At least food for thought and discussion.

Joe
Jun 03, 2019, 06:07 AM
HAL... Open the damn doors!
jfetter's Avatar
Oooh, I like that, definitely a good start! Pair a decent performing LiPo with an "F" and it means a lot more all of a sudden!

BTW, Dirty Harry translates to "Dirty Plunder" in Chinese... an appropriate name for his questionable cousin ;-)

Jack
Jun 03, 2019, 01:49 PM
big ignore list/drama is dumb
brushless55's Avatar
I like this, keep it going!
Jun 03, 2019, 06:40 PM
big ignore list/drama is dumb
brushless55's Avatar
Joe, I got my batteries a little bit ago
thank you!
Jun 03, 2019, 07:06 PM
aka JetMan Joe
MCSGUY's Avatar
Thread OP
Another metric that I've gotten supportive feedback on is the power to weight ratio of a pack (Average Watts/Gram). In fact for some that's a primary factor in their choice, aside from cost. Watts increase with load so in order for the number to have a practical value it should be tied to a standardized "C" level. I suggest 30C because I believe for the majority it represents their maximum requirement. And most non-heavy graphene packs can achieve that number but start to over stress at 35C. IMO anything less then 30C and the pack isn't working hard enough to imply much about it's grunt.

Linking cost with W/G in some way doesn't feel right to me. Typically suppliers try to justify their premium price with longevity statements. Some packs are so cheap it's obvious they're not produced with a eye on cycle life. The majority fall in the middle, generally near each other so really we have 3 price categories: High, Average and Low. I would classify anything at $150 and above as High and $90 and below Low.

If this is a track worth continuing down I should see feedback from more then just two or three guys.

Joe
Jun 03, 2019, 07:07 PM
aka JetMan Joe
MCSGUY's Avatar
Thread OP
Quote:
Originally Posted by brushless55
Joe, I got my batteries a little bit ago
thank you!
Excellent! That didn't take overly long for ground shipping.

Joe
Jun 03, 2019, 08:56 PM
Registered User
atreis's Avatar
Got mine today. Thanks.
Jun 03, 2019, 09:01 PM
Destructive testing expert!
Bill Sanford's Avatar
Thanks for all your great data and work Joe.

I'm a lurker in this thread, but very much appreciate the detailed charts and data. That said, I've worked in the broadcast industry in technology and engineering for 40 years, so understand electronics and the great data you have provided. I wouldn't be happy just seeing a simple rank or number. I love the gory details, so keep it up!

Again, many thanks for all your excellent work. Obviously this is a passion for you!
Jun 03, 2019, 09:33 PM
HAL... Open the damn doors!
jfetter's Avatar
It is that simple, my chart had two different rankings, generated two different ways:

First, take every measurement you have for every pack in the same class and rate them 1 to however many there are (for each data point). Add them up for each pack and divide by the number of data points and now every pack is ranked using all data points from 1 to however many there are. This system is UNWEIGHTED.

Second, choose the data points you feel are important, use a multiplier for any that are "more" important and use that formula to calculate the WEIGHTED score of each pack in the same class. You can add cost, ignore cost, or generate a pure performance number followed by a "bang for the buck" ranking, this is where a little consideration for what's important and generating the formula to plug into Excel come in but in the end, once you have the formula, you apply to all packs in the same class and there you go...

Jack
Jun 03, 2019, 09:53 PM
tic
tic
thunderscreech
tic's Avatar
Outstanding work by Joe and others on this thread. More people should be aware of it
Jun 05, 2019, 01:46 AM
aka JetMan Joe
MCSGUY's Avatar
Thread OP
Spectrum “smart” packs are now available however when I started to place an order for their $160 “100C” 6S/5000 there were two serious red flags. The weight was only 670g and the spec listing indicated 12awg power leads!

I have a call into tech support to see if the power lead size is correct or a typo. If it’s only 12awg I’m going to pass. We know it will self destruct if I apply a serious load. Also unless they have developed some revolutionary chemistry the pack will blow up like a balloon at 35C.

My suspicion is Spectrum’s using their name to sell the a big lie to unsophisticated 20C users who’ll never know they’ve been served the kool aid. I hope they prove me wrong but I don’t see how. 100C , 500Amps, what a joke

Joe
Jun 05, 2019, 01:52 AM
HAL... Open the damn doors!
jfetter's Avatar
Are you sure it wasn't "Awg"? The capital "A" stands for Awesome and is totally different than a lower case "a"

Jack
Last edited by jfetter; Jun 05, 2019 at 06:16 AM. Reason: Typo...


Quick Reply
Message:

Thread Tools