Help me build a quality, datalogging, charging/discharge bank @ home, DIY.
- Thread starter DogDipstick
- Start date
flippy
1 MW
- Joined
- Aug 12, 2015
- Messages
- 2,351
personally i dont like a single setup with a huge range. accuracy and just general clumsyness of using a huge box to discharge a single 18650 cell is dumb.
i have thankfully been able to spend company money on test gear and have learned the "right tool for the job" after much fails.
semi-pro setup:
single cell test box: EBC X0510, 8 channel test device. not cheap, but very accurate and fully programmable, ideal for lifecycle testing. not recommended for the average DIY guy here due to cost.
single battery up to 150V 400W: kunkin kp184, i use 4 of them in parralel mode to get up to 1200W at 150V. these things are not expensive and have PC control and desinged to be teamed up in groups.
right now i am also using a more janky setup with a jerrycan with a 3 4kW 100V heating elements in the bottom to use as a energy dump to discharge large batteries as it takes many kWh's to heat up all that water, a switch bank controls how the elements work as i can switch them in series or parralel on the switch box for up to 12kW of dissipation at 100V.
i am looking at a less janky setup then the jerrycan but have not found the time for it as i want to maximize compactness. so watercooled resistors, electric eheim pump and a few pc radiatators as they are less delicate then IGBT's and don really have a voltage limit.
i have thankfully been able to spend company money on test gear and have learned the "right tool for the job" after much fails.
semi-pro setup:
single cell test box: EBC X0510, 8 channel test device. not cheap, but very accurate and fully programmable, ideal for lifecycle testing. not recommended for the average DIY guy here due to cost.
single battery up to 150V 400W: kunkin kp184, i use 4 of them in parralel mode to get up to 1200W at 150V. these things are not expensive and have PC control and desinged to be teamed up in groups.
right now i am also using a more janky setup with a jerrycan with a 3 4kW 100V heating elements in the bottom to use as a energy dump to discharge large batteries as it takes many kWh's to heat up all that water, a switch bank controls how the elements work as i can switch them in series or parralel on the switch box for up to 12kW of dissipation at 100V.
i am looking at a less janky setup then the jerrycan but have not found the time for it as i want to maximize compactness. so watercooled resistors, electric eheim pump and a few pc radiatators as they are less delicate then IGBT's and don really have a voltage limit.
DogDipstick
100 kW
Punx0r said:
Did I say I was doing if " properly"...? I did not mean to, if I did.
I thought I said:
I was doing it cheap, diy, and quality, @ home. I can do all those things that the load bank can do, just dont know how yet. Exactly.
I mean, why spend more when you can spend less? I need not meet any standards or certification. I am doing it for me and my DIY customers. Who want to know the specs and tenacity of a cell. Spare time messing around. I cannot mindfully sell someone an assembled pack without a test first. For both record and security.
Yes, the range of those ( professional built load banks, that I didnt even know exsisted, designed by some busines that has to pay taxes and whatnot, increasing cost of product) is a good template. Range from 24v/100A up to 700v/450A So I am right in the middle with 240v and 20Kw.
I mean, I have my own (machine design) expert and I have the same ability to blow air over a calculated Ohm load as the next guy. I dont think I want to waste money o paying for their labor when mine is free.
So I got: The nichrome idea, the strip heaters idea, and the water through a stainless tube idea... All I need is the circuitry.
Most proficient for me I think are the heaters in an enclosure with forced air. ( sorta just like the "proper " installation you just told me about... ).... ( mebbe I'll buy one just to copy it and send it back lol... Nah that would be dirty.. Using someone elses product for copy... I am not buying someone else design, or copying, I am making my own, here with our lil ol' engineering co... )
I mean What the f-u-ck, you guys. Do you want to see me do this or not? Yes Grin makes one. Yes, you can buy a 5,000$ heaterfan in Pottstown.
Yes I have more than a "single setup" with huge "range accuracy and clumsyness" to discharge single 18650. 400w is wimp. 4 in parallel for 1200w is wimp. A SINGLE resister I have is rated at 1250w. I have 14 of them.
Yes My power supply cost 5000$ from AAB. You can buy one new today. On eBay they sit for 2500$. I paid 40$ for two of them, and sold one for 400$. Dont tell AAB. They might get mad. I am budget oriented and have no restrictions requiring me to buy new equipment.
Yes I have what is needed to do it, and do it DIY. Thanks all of those that have given me constructive criticism, but....you critique is not very rational in my opinion. If you have actual numbers or ideas other that " buy it" or "
DogDipstick
100 kW
Anybody know if where you can learn more about the "Grinspector" firmware? Is it CA2 ONLY? Little help to me, I bet it wont be updated.
I could flash that and try it out maybe? Seems like it is a discontinued product. Oh well.
This should not be to hard to copy. To bad you can only buy one for the CA2.
I could flash that and try it out maybe? Seems like it is a discontinued product. Oh well.
This should not be to hard to copy. To bad you can only buy one for the CA2.
DogDipstick
100 kW
Can I get a confirmation before I brick my CA3? No? Yes? Anybody?
DogDipstick
100 kW
I wouldnt expect a business to support a non-paying customer, nor would I ask for free advice. So a phone call is out of the question.
I repeat, I am not buying a Grinspector. I am developing ( making) one similar in function, that might be able to utilize the Firmware for the device I have already bought ( but not bought directly from Grin)... I do not want to burden people with silly questions. I figured you guys might know.
It would be wrong if I were to ask ( the makers of) "if a Grinspecter firmware flash would work with a CA3?".., then not buy the CA3 or the Grinspector, but use said ( newly gained) knowledge to my own benefit both in education of the hobbyist and financially for profit.
Yeah Not expecting much there.
Who the fu doesnt pay taxes? What does that have to do with anything, John? I pay loads of taxes. I just state theat I like to mind my budget... and that is just what it is, mine.
That was a direct quote from Mr Justin, who else would have said that. This is the extent that I can find on the forum relating to this firmware.
" should work"....
I repeat, I am not buying a Grinspector. I am developing ( making) one similar in function, that might be able to utilize the Firmware for the device I have already bought ( but not bought directly from Grin)... I do not want to burden people with silly questions. I figured you guys might know.
It would be wrong if I were to ask ( the makers of) "if a Grinspecter firmware flash would work with a CA3?".., then not buy the CA3 or the Grinspector, but use said ( newly gained) knowledge to my own benefit both in education of the hobbyist and financially for profit.
Yeah Not expecting much there.
Who the fu doesnt pay taxes? What does that have to do with anything, John? I pay loads of taxes. I just state theat I like to mind my budget... and that is just what it is, mine.
That was a direct quote from Mr Justin, who else would have said that. This is the extent that I can find on the forum relating to this firmware.
" should work"....
DogDipstick
100 kW
Thats great...: I mean dont get me wrong: If experts would like to input, that is up to them: I would be just as amused if the experts keep quiet until I have done wht I think I can do, the critique me.
Goals are well defined and clearly laid out,,: limits are well rated for and shown on the components: there are many ways to do this and the trade off is what it is:
The Grinspector was cumming out when I ( and my inexperienced self on the forums ) and I have no9t heard alot of feedback. I know Im janky. I know Im cheap. I know I can design a good safe build.
The Grin Trip analyzer even has a setting for battery testing study. This is one of the things that gave me this great fledging idea.
Ideal: 3-390+v. 1s to 96s. Rational? 12-240v. maybe. Maybe I should keep it simple and desigh stronger loads under 100v. Load.. Irrational, Active, 1000A and 700V +... rational... (passive), 240v 3.3Ohms? Maybe less?..50-100A test is still a good test for most battery. .... min based on the hardware array.. , ect.... I am trying to brainstorm and learn everything about the limits and design implementations in use and common for this kind of testing.
"Amp-Hours: With an amp-hour plot, you can see your statistics plotted against the accumulated amp-hours from the battery. This function is intended primarily for those using this software to produce battery discharge plots from CA data of a constant discharge through a load. It is a bit goofy for plotting actual vehicle trips, especially if there is regenerative braking and the amps go negative."
Goals are well defined and clearly laid out,,: limits are well rated for and shown on the components: there are many ways to do this and the trade off is what it is:
The Grinspector was cumming out when I ( and my inexperienced self on the forums ) and I have no9t heard alot of feedback. I know Im janky. I know Im cheap. I know I can design a good safe build.
The Grin Trip analyzer even has a setting for battery testing study. This is one of the things that gave me this great fledging idea.
Ideal: 3-390+v. 1s to 96s. Rational? 12-240v. maybe. Maybe I should keep it simple and desigh stronger loads under 100v. Load.. Irrational, Active, 1000A and 700V +... rational... (passive), 240v 3.3Ohms? Maybe less?..50-100A test is still a good test for most battery. .... min based on the hardware array.. , ect.... I am trying to brainstorm and learn everything about the limits and design implementations in use and common for this kind of testing.
"Amp-Hours: With an amp-hour plot, you can see your statistics plotted against the accumulated amp-hours from the battery. This function is intended primarily for those using this software to produce battery discharge plots from CA data of a constant discharge through a load. It is a bit goofy for plotting actual vehicle trips, especially if there is regenerative braking and the amps go negative."
Attachments
999zip999
100 TW
Forget the heaters use it to power an alcohol still for your car. Just rejet for alcohol use.
DogDipstick
100 kW
So I guess I am looking at ( within my capability) :
Sub 1-Ohm load, for low voltage packs ( up to 48v) , Nichrome, 10 (1foot long) 1.66OPF coils around a tile or other ceramic fixture, in parallel for .16Ohms or parallel/series some combination for .32 Ohms ( approx, thats a good number for me in these ranges) that can handle the power. This in a bucket of water for cooling.
My Nichrome is .020, or 24g. Looking at over 100A for these tests. ( 5Kw-10Kw goal to start. )
If this does not work for a low Ohm load, I will look toward water heater design for long term continuous load that sheds waste heat into water. I have yet to figure actual lengths, power, and circuit...
For a "higher" voltage pack, and a higher resistance load, like say 1.5 Ohm -3 Ohm load...
My strip heaters. Turns out I have many more than (14) ( 1250) and (6) ( 1700w). I have about 21 of the 46Ohm strip heaters. Using these, (20, 1 for replacement) I can get a 240v max capable load bank.. around 2.3 Ohms total load. This would be good for anythig over 48v of suitable power without even using the 1700W, 33ohm ones that I have. This is extra should I need it for lower voltage packs, and slightly manipulating the load.
I may go ahead with both ideas.
QUESTION FOR YOU GUYS
( This will make for what is called a "passive, constant impedance load" ? Passive, as in you cannot adjust the current, or impedance. Meaning as you decreases the voltage, the current will decrease, with the same ratio also and then the power, at the square of the voltage.
Is this correct?
In an ideal world, what would I want for a battery capacity datalog? Constant power, current, or impedance? Thankyou...
When you guys say " active" load, vs "passive"load, what are you referring to? Actively or passively cooling, or actively or passively manipulating the power, and current, for a certain load? )
Min of the CA3 is 12v. Max is ? I guess should build it as a HV system to start, the CA3 monitor part of it.
Thanks.
Sub 1-Ohm load, for low voltage packs ( up to 48v) , Nichrome, 10 (1foot long) 1.66OPF coils around a tile or other ceramic fixture, in parallel for .16Ohms or parallel/series some combination for .32 Ohms ( approx, thats a good number for me in these ranges) that can handle the power. This in a bucket of water for cooling.
My Nichrome is .020, or 24g. Looking at over 100A for these tests. ( 5Kw-10Kw goal to start. )
If this does not work for a low Ohm load, I will look toward water heater design for long term continuous load that sheds waste heat into water. I have yet to figure actual lengths, power, and circuit...
For a "higher" voltage pack, and a higher resistance load, like say 1.5 Ohm -3 Ohm load...
My strip heaters. Turns out I have many more than (14) ( 1250) and (6) ( 1700w). I have about 21 of the 46Ohm strip heaters. Using these, (20, 1 for replacement) I can get a 240v max capable load bank.. around 2.3 Ohms total load. This would be good for anythig over 48v of suitable power without even using the 1700W, 33ohm ones that I have. This is extra should I need it for lower voltage packs, and slightly manipulating the load.
I may go ahead with both ideas.
QUESTION FOR YOU GUYS
( This will make for what is called a "passive, constant impedance load" ? Passive, as in you cannot adjust the current, or impedance. Meaning as you decreases the voltage, the current will decrease, with the same ratio also and then the power, at the square of the voltage.
Is this correct?
In an ideal world, what would I want for a battery capacity datalog? Constant power, current, or impedance? Thankyou...
When you guys say " active" load, vs "passive"load, what are you referring to? Actively or passively cooling, or actively or passively manipulating the power, and current, for a certain load? )
Min of the CA3 is 12v. Max is ? I guess should build it as a HV system to start, the CA3 monitor part of it.
Thanks.
DogDipstick
100 kW
Time to find a good box that fits conduit and has a big killswitch on the cover that prvents the user from opening the enclosure unless it is in an off/disconnected state. Lol.
Cut it up and mount my Ca3. The server rack is perfect for the strip heaters. 14" holes on center already threaded.
Cut it up and mount my Ca3. The server rack is perfect for the strip heaters. 14" holes on center already threaded.
Yes all these are just the raw load dump, a place to dissipate the energy as heat.DogDipstick said:
You of course need the discharge controller part of electronic dummy loads usually used to control the discharge testing.
> When you guys say " active" load, vs "passive"load, what are you referring to?
My understanding is the active loads are these "electronic dummy load" testing units.
Some let you choose between constant power (watts) and CC, but the latter is all I'd use for testing battery SoH. Constant impedance I can't think of a context where that would be useful.
The data monitoring and logging functionality can be separate, but the purpose designed testers usually bundle all three functions,
unlike the Grinspector and Duo-style hobby chargers that do discharge testing via "regenerative mode" from one channel, into the resistors on the other channel.
e.g. iCharger 4010-Duo can do up to 1000W, so 34A for a 7s battery.
DogDipstick
100 kW
My powerlab6 1000W can datalog a constant power discharge, or a constant current. I wonder how exactly it does that. I also wonder what the "Grinspecter" does, and how through its relays driving contactors given situatuin requiring charge or discharge.
I think I want to settle on one, then figure out how to do it. CC or CP
I think I want to settle on one, then figure out how to do it. CC or CP
IMO CP is useless for batteries.
Amps and Ah are the best units in dealing with battery capacity, energy storage and DC usage over time.
Only bring watts into things as required due to voltage conversion, e.g. solar panels, or a 24V bank feeding a 12V converter
Amps and Ah are the best units in dealing with battery capacity, energy storage and DC usage over time.
Only bring watts into things as required due to voltage conversion, e.g. solar panels, or a 24V bank feeding a 12V converter
DogDipstick
100 kW
john61ct said:
Those things are nice but what about .. the power... lol. A battery hast ot be tested for capacity, but also must be tested for power........ IMo.
flippy
1 MW
- Joined
- Aug 12, 2015
- Messages
- 2,351
the only metrics that matter are Wh and discharge amps and temperature at discharge current.
Ah is dependant on voltage so it is a poor inducation for a customer what the acutal capacity is under a certain desired load.
best way to rate batteries is to pull a certain current or wattage and mark the amount of Wh you can pull from it until it hits 2.5V per cell and repeat that for several power settings the battery will see in daily life so you can get a good indication what the practial capacity is.
Ah is dependant on voltage so it is a poor inducation for a customer what the acutal capacity is under a certain desired load.
best way to rate batteries is to pull a certain current or wattage and mark the amount of Wh you can pull from it until it hits 2.5V per cell and repeat that for several power settings the battery will see in daily life so you can get a good indication what the practial capacity is.
The Wh lets people compare the overall capacity of packs at different voltages, yes that is useful.
But when discussing a given pack type, stick to Amps and Volts, much more directly applicable to the critical performance factors, how we actually use batteries, motors and other loads.
The problem with using Wh directly for capacity testing is that the Peukert effect makes results **much** more variable depending on the C-rate used.
Ah more directly relate to the actual chemical reactions inside the battery, Wh are affected much more by SoC and C-rate at that instant.
For example a LI battery might have an efficiency well over 99% using Ah but only 80% in one set of conditions, then 70% in others.
So, best to use CC load test, get your Ah capacity (if secondhand % of original also gives you SoH)
and then multiply that Ah number by nominal volts in order to publish a Wh figure.
Then the impact of Peukert's Law is much less a wild card.
Consider V and A as your "realer" foundation variables to measure, power in Watts is just a calculated value derived after the actual measurements.
But when discussing a given pack type, stick to Amps and Volts, much more directly applicable to the critical performance factors, how we actually use batteries, motors and other loads.
The problem with using Wh directly for capacity testing is that the Peukert effect makes results **much** more variable depending on the C-rate used.
Ah more directly relate to the actual chemical reactions inside the battery, Wh are affected much more by SoC and C-rate at that instant.
For example a LI battery might have an efficiency well over 99% using Ah but only 80% in one set of conditions, then 70% in others.
So, best to use CC load test, get your Ah capacity (if secondhand % of original also gives you SoH)
and then multiply that Ah number by nominal volts in order to publish a Wh figure.
Then the impact of Peukert's Law is much less a wild card.
Consider V and A as your "realer" foundation variables to measure, power in Watts is just a calculated value derived after the actual measurements.
flippy
1 MW
- Joined
- Aug 12, 2015
- Messages
- 2,351
i present battery performance to my clients at either
capacity in Wh = constant x number of Amps
capacity in Wh = constant x number of Watts
some clients have devices that regulate on amps, others on watts and i test my batteries based on that. its useless to test in constant watts when their machines only care about amps.
usually this metric has 4~5 different lines of different power numbers to give a nice chart of how amp or watt load reduces the usable amount of Wh they can pull from the battery.
my batteries usually run into the kWh, so Wh is more usable metric in those cases and the industry standard.
capacity in Wh = constant x number of Amps
capacity in Wh = constant x number of Watts
some clients have devices that regulate on amps, others on watts and i test my batteries based on that. its useless to test in constant watts when their machines only care about amps.
usually this metric has 4~5 different lines of different power numbers to give a nice chart of how amp or watt load reduces the usable amount of Wh they can pull from the battery.
my batteries usually run into the kWh, so Wh is more usable metric in those cases and the industry standard.
DogDipstick
100 kW
Do you guys think I can throttle the load with a Curtis motor controller for a load regulator?
The Curtis 1207B programmable motor speed controller provides efficient, cost-effective, and simple-to-install control for a variety of small electric ve-hicles. Typical applications include walkie fork/pallet trucks, mini personnel carriers, and sweepers. The microprocessor-based logic section combined with a proven MOSFET power section gives the 1207B controller high power and advanced features in a simple, compact package. The optional programmers (Curtis 1311 and 1314) enable the user to set parameters, conduct tests, and obtain diagnostic information quickly and easily."
Maybay? I got these out the yingyang. Can afford to fry a few in testing honestly. Wouldn't this regulate the (input) voltage to a current with ramps and whatnot?
This would give me 10Kw in the 36v range. Lotsa amps. Maybe I can find a higher power/voltage/current controller, but I got these 24 and 36v 250A jobbies like candy.
The Curtis 1207B programmable motor speed controller provides efficient, cost-effective, and simple-to-install control for a variety of small electric ve-hicles. Typical applications include walkie fork/pallet trucks, mini personnel carriers, and sweepers. The microprocessor-based logic section combined with a proven MOSFET power section gives the 1207B controller high power and advanced features in a simple, compact package. The optional programmers (Curtis 1311 and 1314) enable the user to set parameters, conduct tests, and obtain diagnostic information quickly and easily."
Maybay? I got these out the yingyang. Can afford to fry a few in testing honestly. Wouldn't this regulate the (input) voltage to a current with ramps and whatnot?
This would give me 10Kw in the 36v range. Lotsa amps. Maybe I can find a higher power/voltage/current controller, but I got these 24 and 36v 250A jobbies like candy.
DogDipstick
100 kW
flippy
1 MW
- Joined
- Aug 12, 2015
- Messages
- 2,351
DogDipstick said:
it would be very hard to control the power and its extremely likly that the controller will simply error out and shut down if it does not see a turning motor.
PWM controlling a couple phat IGBT's directly to a a couple reisitors would be a better solution. control the whole thing with a shunt and a simple arduino and a couple IGBT driver chips so you dont blow up your arduino.
also: isolate the fudge out of this setup. control side must be competly galvanically isolated from the power side.
