On Saturday, October 22, 2022 at 8:16:46 AM UTC-7, Don Y wrote:> On 10/22/2022 6:34 AM, Fred Bloggs wrote: > > On Friday, October 21, 2022 at 9:39:16 PM UTC-4, Don Y wrote: > >> On 10/21/2022 5:51 PM, Fred Bloggs wrote: > >>> On Tuesday, October 18, 2022 at 5:10:35 PM UTC-4, Don Y wrote: > >>>> On 10/18/2022 8:43 AM, Bert Hickman wrote: > >>>>> Does anyone have any experience swapping out old lead-acid batteries > >>>>> for Li-ion equivalents in a UPS? > >>>>> > >>>>> I see these batteries advertised as replacements, but I'm concerned > >>>>> about their safety in this application. New Li-ion UPS systems seem > >>>>> to be extremely expensive. > >>>> I wouldn't trust them -- unless the UPS manufacturer had blessed them > >>>> (and dubious, even then). > >>>> > >>>> If you are having problems with battery life in UPS, look to see how > >>>> it is being charged. Many "cook" their batteries (poor circuit > >>>> tolerances). Also, they seem intent on recharging QUICKLY (in case > >>>> another outage??) instead of trying to prolong the batteries' service > >>>> life. > >>>> > >>>> Finally, note that many places will pay you for the lead in your old > >>>> batteries. If you're just using 12V7.2AHr batteries, then there's not > >>>> much money there (~$0.20/pound). OTOH, if you have a bigger UPS, you > >>>> can get a fair bit back (I got over $100 for the dead batteries in my > >>>> biggest UPS) > >>> > >>> Do you really think the modern lithium battery is just a dumb collection > >>> of cells. The products going after the SLA market almost certainly have > >>> built-in BMS ( battery management system) that cuts off the charging > >>> current when fully charged. And that BMS is probably made by Analog > >>> Devices. Last time I looked the charge management cuts off the float > >>> mode charge current when it has fallen to 25% or so of the initial Ipk > >>> delivered to the uncharged battery. And there's a sensitivity of 100:1 > >>> state of charge to the temperature dependent open terminal voltage, so > >>> the process requires great precision. Nonetheless, it's simple to very > >>> reliably put the lithium into the SLA charge circuit- within reason. > >> Do you really think a UPS treats a battery as an ideal voltage source? > >> > >> Do you really think a product designed years ago can just tolerate a > >> replacement battery chemistry? That the charger isn't actively monitoring > >> charge current? And determining state of charge by noticing its > >> relationship to open cell voltage over time? > >> > >> How do you think the UPS decides that the battery needs replacing? > >> > >> "Gee, battery stopped taking charge current. Has it become disconnected? > >> Has it failed?" > >> > >> Why are lithium based solutions MULTIPLES of the price of lead-acid > >> solutions? Surely they could just take an EXISTING lead acid design and > >> drop in a "lead-acid emulator" for the incremental cost of the "emulator > >> battery" over the lead-acid original. > >> > >> Maybe there's more to the technology than you think! > > > > Not really. I don't care what kind of battery charge management the UPS has, > > the replacement battery can be made to look identical to the original. > > Cutting off the charge current in float mode does not mean opening a relay > > switch or something crude like that. The replacement battery can present > > enough of a self-discharge current loading to make itself look real enough. > And, you'll discover on one UPS the UPS won't turn on because it doesn't > see a battery; on another, it won't run it's self checks; etc. You don't > know what "firmware A" expects of it's battery that "firmware B" doesn't. > > All of the battery-backed products I've designed predated lithium. Put > a lithium battery in them and they'll throw a fault -- because the > battery isn't discharging at the expected rate; or is charging at > the wrong rate (dv/dt); or, the monitoring circuitry appears defective; etc. > > "Hey, let's make this lithium battery look EXACTLY like a lead acid -- so > it HAS NO ADVANTAGES over a lead acid battery! But, we can charge more > for those non-differences!" > > MSRP $1,865.88 <https://www.amazon.com/dp/B08QQRFQ4W> > MSRP $2,899.99 <https://www.amazon.com/dp/B09JR6ZCQX>1500VA or 100Wh? See first comment. That's certainly expensive. I can get 100Wh for $28.9999
Li-ion UPS battery substitution in UPS
Started by ●October 18, 2022
Reply by ●October 22, 20222022-10-22
Reply by ●October 22, 20222022-10-22
On Saturday, October 22, 2022 at 11:16:46 AM UTC-4, Don Y wrote:> On 10/22/2022 6:34 AM, Fred Bloggs wrote: > > On Friday, October 21, 2022 at 9:39:16 PM UTC-4, Don Y wrote: > >> On 10/21/2022 5:51 PM, Fred Bloggs wrote: > >>> On Tuesday, October 18, 2022 at 5:10:35 PM UTC-4, Don Y wrote: > >>>> On 10/18/2022 8:43 AM, Bert Hickman wrote: > >>>>> Does anyone have any experience swapping out old lead-acid batteries > >>>>> for Li-ion equivalents in a UPS? > >>>>> > >>>>> I see these batteries advertised as replacements, but I'm concerned > >>>>> about their safety in this application. New Li-ion UPS systems seem > >>>>> to be extremely expensive. > >>>> I wouldn't trust them -- unless the UPS manufacturer had blessed them > >>>> (and dubious, even then). > >>>> > >>>> If you are having problems with battery life in UPS, look to see how > >>>> it is being charged. Many "cook" their batteries (poor circuit > >>>> tolerances). Also, they seem intent on recharging QUICKLY (in case > >>>> another outage??) instead of trying to prolong the batteries' service > >>>> life. > >>>> > >>>> Finally, note that many places will pay you for the lead in your old > >>>> batteries. If you're just using 12V7.2AHr batteries, then there's not > >>>> much money there (~$0.20/pound). OTOH, if you have a bigger UPS, you > >>>> can get a fair bit back (I got over $100 for the dead batteries in my > >>>> biggest UPS) > >>> > >>> Do you really think the modern lithium battery is just a dumb collection > >>> of cells. The products going after the SLA market almost certainly have > >>> built-in BMS ( battery management system) that cuts off the charging > >>> current when fully charged. And that BMS is probably made by Analog > >>> Devices. Last time I looked the charge management cuts off the float > >>> mode charge current when it has fallen to 25% or so of the initial Ipk > >>> delivered to the uncharged battery. And there's a sensitivity of 100:1 > >>> state of charge to the temperature dependent open terminal voltage, so > >>> the process requires great precision. Nonetheless, it's simple to very > >>> reliably put the lithium into the SLA charge circuit- within reason. > >> Do you really think a UPS treats a battery as an ideal voltage source? > >> > >> Do you really think a product designed years ago can just tolerate a > >> replacement battery chemistry? That the charger isn't actively monitoring > >> charge current? And determining state of charge by noticing its > >> relationship to open cell voltage over time? > >> > >> How do you think the UPS decides that the battery needs replacing? > >> > >> "Gee, battery stopped taking charge current. Has it become disconnected? > >> Has it failed?" > >> > >> Why are lithium based solutions MULTIPLES of the price of lead-acid > >> solutions? Surely they could just take an EXISTING lead acid design and > >> drop in a "lead-acid emulator" for the incremental cost of the "emulator > >> battery" over the lead-acid original. > >> > >> Maybe there's more to the technology than you think! > > > > Not really. I don't care what kind of battery charge management the UPS has, > > the replacement battery can be made to look identical to the original. > > Cutting off the charge current in float mode does not mean opening a relay > > switch or something crude like that. The replacement battery can present > > enough of a self-discharge current loading to make itself look real enough. > And, you'll discover on one UPS the UPS won't turn on because it doesn't > see a battery; on another, it won't run it's self checks; etc. You don't > know what "firmware A" expects of it's battery that "firmware B" doesn't. > > All of the battery-backed products I've designed predated lithium. Put > a lithium battery in them and they'll throw a fault -- because the > battery isn't discharging at the expected rate; or is charging at > the wrong rate (dv/dt); or, the monitoring circuitry appears defective; etc. > > "Hey, let's make this lithium battery look EXACTLY like a lead acid -- so > it HAS NO ADVANTAGES over a lead acid battery! But, we can charge more > for those non-differences!" > > MSRP $1,865.88 <https://www.amazon.com/dp/B08QQRFQ4W> > MSRP $2,899.99 <https://www.amazon.com/dp/B09JR6ZCQX> > > $1000 premium for lithium battery pack. Same exact run-times, > same load capacity, same networking capability, same display > (literally the same module). Ah, but the more expensive one has > a shiny front panel (and isn't intended to be floor-standing)! > > [Latter has shallower aspect ratio and weighs less -- though at the cost > of an additional rack unit. *Might* be important in a datacenter; > doubtful a SOHO user would bother with it as he'd likely not have > many such devices, many batteries to replace, etc. Note the number > of lead-acid offerings vs. lithium -- why is that?]You're becoming hysterical. Batteries are not precision components. Any design that's looking at voltage time drift or charge rates to make any decisions is deluding itself. All you end up doing is set up the user for quirky indications they'll just learn to ignore.
Reply by ●October 22, 20222022-10-22
On Saturday, October 22, 2022 at 2:05:28 PM UTC-7, Fred Bloggs wrote:> On Saturday, October 22, 2022 at 11:16:46 AM UTC-4, Don Y wrote: > > On 10/22/2022 6:34 AM, Fred Bloggs wrote: > > > On Friday, October 21, 2022 at 9:39:16 PM UTC-4, Don Y wrote: > > >> On 10/21/2022 5:51 PM, Fred Bloggs wrote: > > >>> On Tuesday, October 18, 2022 at 5:10:35 PM UTC-4, Don Y wrote: > > >>>> On 10/18/2022 8:43 AM, Bert Hickman wrote: > > >>>>> Does anyone have any experience swapping out old lead-acid batteries > > >>>>> for Li-ion equivalents in a UPS? > > >>>>> > > >>>>> I see these batteries advertised as replacements, but I'm concerned > > >>>>> about their safety in this application. New Li-ion UPS systems seem > > >>>>> to be extremely expensive. > > >>>> I wouldn't trust them -- unless the UPS manufacturer had blessed them > > >>>> (and dubious, even then). > > >>>> > > >>>> If you are having problems with battery life in UPS, look to see how > > >>>> it is being charged. Many "cook" their batteries (poor circuit > > >>>> tolerances). Also, they seem intent on recharging QUICKLY (in case > > >>>> another outage??) instead of trying to prolong the batteries' service > > >>>> life. > > >>>> > > >>>> Finally, note that many places will pay you for the lead in your old > > >>>> batteries. If you're just using 12V7.2AHr batteries, then there's not > > >>>> much money there (~$0.20/pound). OTOH, if you have a bigger UPS, you > > >>>> can get a fair bit back (I got over $100 for the dead batteries in my > > >>>> biggest UPS) > > >>> > > >>> Do you really think the modern lithium battery is just a dumb collection > > >>> of cells. The products going after the SLA market almost certainly have > > >>> built-in BMS ( battery management system) that cuts off the charging > > >>> current when fully charged. And that BMS is probably made by Analog > > >>> Devices. Last time I looked the charge management cuts off the float > > >>> mode charge current when it has fallen to 25% or so of the initial Ipk > > >>> delivered to the uncharged battery. And there's a sensitivity of 100:1 > > >>> state of charge to the temperature dependent open terminal voltage, so > > >>> the process requires great precision. Nonetheless, it's simple to very > > >>> reliably put the lithium into the SLA charge circuit- within reason. > > >> Do you really think a UPS treats a battery as an ideal voltage source? > > >> > > >> Do you really think a product designed years ago can just tolerate a > > >> replacement battery chemistry? That the charger isn't actively monitoring > > >> charge current? And determining state of charge by noticing its > > >> relationship to open cell voltage over time? > > >> > > >> How do you think the UPS decides that the battery needs replacing? > > >> > > >> "Gee, battery stopped taking charge current. Has it become disconnected? > > >> Has it failed?" > > >> > > >> Why are lithium based solutions MULTIPLES of the price of lead-acid > > >> solutions? Surely they could just take an EXISTING lead acid design and > > >> drop in a "lead-acid emulator" for the incremental cost of the "emulator > > >> battery" over the lead-acid original. > > >> > > >> Maybe there's more to the technology than you think! > > > > > > Not really. I don't care what kind of battery charge management the UPS has, > > > the replacement battery can be made to look identical to the original. > > > Cutting off the charge current in float mode does not mean opening a relay > > > switch or something crude like that. The replacement battery can present > > > enough of a self-discharge current loading to make itself look real enough. > > And, you'll discover on one UPS the UPS won't turn on because it doesn't > > see a battery; on another, it won't run it's self checks; etc. You don't > > know what "firmware A" expects of it's battery that "firmware B" doesn't. > > > > All of the battery-backed products I've designed predated lithium. Put > > a lithium battery in them and they'll throw a fault -- because the > > battery isn't discharging at the expected rate; or is charging at > > the wrong rate (dv/dt); or, the monitoring circuitry appears defective; etc. > > > > "Hey, let's make this lithium battery look EXACTLY like a lead acid -- so > > it HAS NO ADVANTAGES over a lead acid battery! But, we can charge more > > for those non-differences!" > > > > MSRP $1,865.88 <https://www.amazon.com/dp/B08QQRFQ4W> > > MSRP $2,899.99 <https://www.amazon.com/dp/B09JR6ZCQX> > > > > $1000 premium for lithium battery pack. Same exact run-times, > > same load capacity, same networking capability, same display > > (literally the same module). Ah, but the more expensive one has > > a shiny front panel (and isn't intended to be floor-standing)! > > > > [Latter has shallower aspect ratio and weighs less -- though at the cost > > of an additional rack unit. *Might* be important in a datacenter; > > doubtful a SOHO user would bother with it as he'd likely not have > > many such devices, many batteries to replace, etc. Note the number > > of lead-acid offerings vs. lithium -- why is that?] > You're becoming hysterical. Batteries are not precision components. Any design that's looking at voltage time drift or charge rates to make any decisions is deluding itself. All you end up doing is set up the user for quirky indications they'll just learn to ignore.Charge rate is no big deal. A small 18650 can take up to 3A to 5A charging. Voltage is more important. My eTank #2 is 34x 3S batteries without BMS. Typical EV equivalence is 32x, which run the cells between 3.5V to 4.2V. I run mine at 34x with cells between 3V and 4V. I check and balance them manually. No pillows and/or firecrackers yet. My eTank #3 will have EMS.
Reply by ●October 22, 20222022-10-22
On 10/22/2022 2:05 PM, Fred Bloggs wrote:> On Saturday, October 22, 2022 at 11:16:46 AM UTC-4, Don Y wrote: >> On 10/22/2022 6:34 AM, Fred Bloggs wrote: >>> On Friday, October 21, 2022 at 9:39:16 PM UTC-4, Don Y wrote: >>>> On 10/21/2022 5:51 PM, Fred Bloggs wrote: >>>>> On Tuesday, October 18, 2022 at 5:10:35 PM UTC-4, Don Y wrote: >>>>>> On 10/18/2022 8:43 AM, Bert Hickman wrote: >>>>>>> Does anyone have any experience swapping out old lead-acid >>>>>>> batteries for Li-ion equivalents in a UPS? >>>>>>> >>>>>>> I see these batteries advertised as replacements, but I'm >>>>>>> concerned about their safety in this application. New Li-ion UPS >>>>>>> systems seem to be extremely expensive. >>>>>> I wouldn't trust them -- unless the UPS manufacturer had blessed >>>>>> them (and dubious, even then). >>>>>> >>>>>> If you are having problems with battery life in UPS, look to see >>>>>> how it is being charged. Many "cook" their batteries (poor >>>>>> circuit tolerances). Also, they seem intent on recharging QUICKLY >>>>>> (in case another outage??) instead of trying to prolong the >>>>>> batteries' service life. >>>>>> >>>>>> Finally, note that many places will pay you for the lead in your >>>>>> old batteries. If you're just using 12V7.2AHr batteries, then >>>>>> there's not much money there (~$0.20/pound). OTOH, if you have a >>>>>> bigger UPS, you can get a fair bit back (I got over $100 for the >>>>>> dead batteries in my biggest UPS) >>>>> >>>>> Do you really think the modern lithium battery is just a dumb >>>>> collection of cells. The products going after the SLA market almost >>>>> certainly have built-in BMS ( battery management system) that cuts >>>>> off the charging current when fully charged. And that BMS is >>>>> probably made by Analog Devices. Last time I looked the charge >>>>> management cuts off the float mode charge current when it has fallen >>>>> to 25% or so of the initial Ipk delivered to the uncharged battery. >>>>> And there's a sensitivity of 100:1 state of charge to the >>>>> temperature dependent open terminal voltage, so the process requires >>>>> great precision. Nonetheless, it's simple to very reliably put the >>>>> lithium into the SLA charge circuit- within reason. >>>> Do you really think a UPS treats a battery as an ideal voltage >>>> source? >>>> >>>> Do you really think a product designed years ago can just tolerate a >>>> replacement battery chemistry? That the charger isn't actively >>>> monitoring charge current? And determining state of charge by noticing >>>> its relationship to open cell voltage over time? >>>> >>>> How do you think the UPS decides that the battery needs replacing? >>>> >>>> "Gee, battery stopped taking charge current. Has it become >>>> disconnected? Has it failed?" >>>> >>>> Why are lithium based solutions MULTIPLES of the price of lead-acid >>>> solutions? Surely they could just take an EXISTING lead acid design >>>> and drop in a "lead-acid emulator" for the incremental cost of the >>>> "emulator battery" over the lead-acid original. >>>> >>>> Maybe there's more to the technology than you think! >>> >>> Not really. I don't care what kind of battery charge management the UPS >>> has, the replacement battery can be made to look identical to the >>> original. Cutting off the charge current in float mode does not mean >>> opening a relay switch or something crude like that. The replacement >>> battery can present enough of a self-discharge current loading to make >>> itself look real enough. >> And, you'll discover on one UPS the UPS won't turn on because it doesn't >> see a battery; on another, it won't run it's self checks; etc. You don't >> know what "firmware A" expects of it's battery that "firmware B" doesn't. >> >> All of the battery-backed products I've designed predated lithium. Put a >> lithium battery in them and they'll throw a fault -- because the battery >> isn't discharging at the expected rate; or is charging at the wrong rate >> (dv/dt); or, the monitoring circuitry appears defective; etc. >> >> "Hey, let's make this lithium battery look EXACTLY like a lead acid -- so >> it HAS NO ADVANTAGES over a lead acid battery! But, we can charge more for >> those non-differences!" >> >> MSRP $1,865.88 <https://www.amazon.com/dp/B08QQRFQ4W> MSRP $2,899.99 >> <https://www.amazon.com/dp/B09JR6ZCQX> >> >> $1000 premium for lithium battery pack. Same exact run-times, same load >> capacity, same networking capability, same display (literally the same >> module). Ah, but the more expensive one has a shiny front panel (and isn't >> intended to be floor-standing)! >> >> [Latter has shallower aspect ratio and weighs less -- though at the cost >> of an additional rack unit. *Might* be important in a datacenter; doubtful >> a SOHO user would bother with it as he'd likely not have many such >> devices, many batteries to replace, etc. Note the number of lead-acid >> offerings vs. lithium -- why is that?] > > You're becoming hysterical.Hysterical? Wow, you've sure got a perverse sense of reality!> Batteries are not precision components. Any > design that's looking at voltage time drift or charge rates to make any > decisions is deluding itself. All you end up doing is set up the user for > quirky indications they'll just learn to ignore.I guess products that have done so and are in the field, WITHOUT COMPLAINT, must not exist in your perverse reality! Designs that just look at voltage trip points are the ones that annoy users with erroneous data. If I dump C/10 into a battery pack FOR TWELVE HOURS and see no change in pack voltage, is that just a quirk? What about three hours? One hour? Maybe I should just wait a week and, at the end of the week, the cell voltage will spontaneously indicate a full charge? Despite the fact that, HISTORICALLY (remember, I can watch the battery FOREVER; it's built into the circuit!) it has registered a specific change in each of those intervals? If I'm a constant load and, today, see the pack voltage dropping at rate R1 and, historically, it has always dropped at ~R0 (R0 << R1), should I just assume that's a quirk and NOT act on it? Has my implementation ACTUALLY changed so that the components are dissipating more power -- yet NOTHING IS WRONG? We EXPECT components to behave in largely predictable ways. Otherwise, we couldn't design with them.
Reply by ●October 22, 20222022-10-22
On Saturday, October 22, 2022 at 5:29:19 PM UTC-4, Don Y wrote:> On 10/22/2022 2:05 PM, Fred Bloggs wrote: > > On Saturday, October 22, 2022 at 11:16:46 AM UTC-4, Don Y wrote: > >> On 10/22/2022 6:34 AM, Fred Bloggs wrote: > >>> On Friday, October 21, 2022 at 9:39:16 PM UTC-4, Don Y wrote: > >>>> On 10/21/2022 5:51 PM, Fred Bloggs wrote: > >>>>> On Tuesday, October 18, 2022 at 5:10:35 PM UTC-4, Don Y wrote: > >>>>>> On 10/18/2022 8:43 AM, Bert Hickman wrote: > >>>>>>> Does anyone have any experience swapping out old lead-acid > >>>>>>> batteries for Li-ion equivalents in a UPS? > >>>>>>> > >>>>>>> I see these batteries advertised as replacements, but I'm > >>>>>>> concerned about their safety in this application. New Li-ion UPS > >>>>>>> systems seem to be extremely expensive. > >>>>>> I wouldn't trust them -- unless the UPS manufacturer had blessed > >>>>>> them (and dubious, even then). > >>>>>> > >>>>>> If you are having problems with battery life in UPS, look to see > >>>>>> how it is being charged. Many "cook" their batteries (poor > >>>>>> circuit tolerances). Also, they seem intent on recharging QUICKLY > >>>>>> (in case another outage??) instead of trying to prolong the > >>>>>> batteries' service life. > >>>>>> > >>>>>> Finally, note that many places will pay you for the lead in your > >>>>>> old batteries. If you're just using 12V7.2AHr batteries, then > >>>>>> there's not much money there (~$0.20/pound). OTOH, if you have a > >>>>>> bigger UPS, you can get a fair bit back (I got over $100 for the > >>>>>> dead batteries in my biggest UPS) > >>>>> > >>>>> Do you really think the modern lithium battery is just a dumb > >>>>> collection of cells. The products going after the SLA market almost > >>>>> certainly have built-in BMS ( battery management system) that cuts > >>>>> off the charging current when fully charged. And that BMS is > >>>>> probably made by Analog Devices. Last time I looked the charge > >>>>> management cuts off the float mode charge current when it has fallen > >>>>> to 25% or so of the initial Ipk delivered to the uncharged battery. > >>>>> And there's a sensitivity of 100:1 state of charge to the > >>>>> temperature dependent open terminal voltage, so the process requires > >>>>> great precision. Nonetheless, it's simple to very reliably put the > >>>>> lithium into the SLA charge circuit- within reason. > >>>> Do you really think a UPS treats a battery as an ideal voltage > >>>> source? > >>>> > >>>> Do you really think a product designed years ago can just tolerate a > >>>> replacement battery chemistry? That the charger isn't actively > >>>> monitoring charge current? And determining state of charge by noticing > >>>> its relationship to open cell voltage over time? > >>>> > >>>> How do you think the UPS decides that the battery needs replacing? > >>>> > >>>> "Gee, battery stopped taking charge current. Has it become > >>>> disconnected? Has it failed?" > >>>> > >>>> Why are lithium based solutions MULTIPLES of the price of lead-acid > >>>> solutions? Surely they could just take an EXISTING lead acid design > >>>> and drop in a "lead-acid emulator" for the incremental cost of the > >>>> "emulator battery" over the lead-acid original. > >>>> > >>>> Maybe there's more to the technology than you think! > >>> > >>> Not really. I don't care what kind of battery charge management the UPS > >>> has, the replacement battery can be made to look identical to the > >>> original. Cutting off the charge current in float mode does not mean > >>> opening a relay switch or something crude like that. The replacement > >>> battery can present enough of a self-discharge current loading to make > >>> itself look real enough. > >> And, you'll discover on one UPS the UPS won't turn on because it doesn't > >> see a battery; on another, it won't run it's self checks; etc. You don't > >> know what "firmware A" expects of it's battery that "firmware B" doesn't. > >> > >> All of the battery-backed products I've designed predated lithium. Put a > >> lithium battery in them and they'll throw a fault -- because the battery > >> isn't discharging at the expected rate; or is charging at the wrong rate > >> (dv/dt); or, the monitoring circuitry appears defective; etc. > >> > >> "Hey, let's make this lithium battery look EXACTLY like a lead acid -- so > >> it HAS NO ADVANTAGES over a lead acid battery! But, we can charge more for > >> those non-differences!" > >> > >> MSRP $1,865.88 <https://www.amazon.com/dp/B08QQRFQ4W> MSRP $2,899.99 > >> <https://www.amazon.com/dp/B09JR6ZCQX> > >> > >> $1000 premium for lithium battery pack. Same exact run-times, same load > >> capacity, same networking capability, same display (literally the same > >> module). Ah, but the more expensive one has a shiny front panel (and isn't > >> intended to be floor-standing)! > >> > >> [Latter has shallower aspect ratio and weighs less -- though at the cost > >> of an additional rack unit. *Might* be important in a datacenter; doubtful > >> a SOHO user would bother with it as he'd likely not have many such > >> devices, many batteries to replace, etc. Note the number of lead-acid > >> offerings vs. lithium -- why is that?] > > > > You're becoming hysterical. > Hysterical? Wow, you've sure got a perverse sense of reality! > > Batteries are not precision components. Any > > design that's looking at voltage time drift or charge rates to make any > > decisions is deluding itself. All you end up doing is set up the user for > > quirky indications they'll just learn to ignore. > I guess products that have done so and are in the field, WITHOUT COMPLAINT, > must not exist in your perverse reality! > > Designs that just look at voltage trip points are the ones that > annoy users with erroneous data. > > If I dump C/10 into a battery pack FOR TWELVE HOURS and see no change in > pack voltage, is that just a quirk? What about three hours? One hour? > Maybe I should just wait a week and, at the end of the week, the cell > voltage will spontaneously indicate a full charge? > > Despite the fact that, HISTORICALLY (remember, I can watch the battery > FOREVER; it's built into the circuit!) it has registered a specific change > in each of those intervals? > > If I'm a constant load and, today, see the pack voltage dropping at rate R1 > and, historically, it has always dropped at ~R0 (R0 << R1), should I just > assume that's a quirk and NOT act on it? Has my implementation ACTUALLY > changed so that the components are dissipating more power -- yet NOTHING > IS WRONG? > > We EXPECT components to behave in largely predictable ways. Otherwise, > we couldn't design with them.Sounds like you're confusing predictability with eternality. Talk about oversimplification. The measurements you touched on are largely a big unknown, not only in principle but in practice. Any instance of a particular battery is more than just the battery itself but also it's entire life history of charge and discharge cycles, storage temperature, maybe a bunch of other stuff known to have significant impact on battery life. Did you do any pulsed impedance measurements with varying test loading while you were at it. IIRC the terminal impedance tells some people a lot about the battery condition.
Reply by ●October 22, 20222022-10-22
On 10/22/2022 11:19 AM, Ed Lee wrote:>>>> Maybe there's more to the technology than you think! >>> >>> Not really. I don't care what kind of battery charge management the UPS has, >>> the replacement battery can be made to look identical to the original. >>> Cutting off the charge current in float mode does not mean opening a relay >>> switch or something crude like that. The replacement battery can present >>> enough of a self-discharge current loading to make itself look real enough. >> And, you'll discover on one UPS the UPS won't turn on because it doesn't >> see a battery; on another, it won't run it's self checks; etc. You don't >> know what "firmware A" expects of it's battery that "firmware B" doesn't. >> >> All of the battery-backed products I've designed predated lithium. Put >> a lithium battery in them and they'll throw a fault -- because the >> battery isn't discharging at the expected rate; or is charging at >> the wrong rate (dv/dt); or, the monitoring circuitry appears defective; etc. >> >> "Hey, let's make this lithium battery look EXACTLY like a lead acid -- so >> it HAS NO ADVANTAGES over a lead acid battery! But, we can charge more >> for those non-differences!" >> >> MSRP $1,865.88 <https://www.amazon.com/dp/B08QQRFQ4W> >> MSRP $2,899.99 <https://www.amazon.com/dp/B09JR6ZCQX> > > 1500VA or 100Wh? See first comment. That's certainly expensive. I can get 100Wh for $28.99991500VA load capacity. Run time (for each) is 5 minutes at full load (about 13 minutes at half load -- though the lead acid unit will go almost 15 minutes). I have several of the first unit that I use to power my "bench tops" (anything located *on* my bench vs. UNDER). My point is, they could have taken the existing LA design and just put lithium batteries in it, if there was no other consequence to changing the battery chemistry. After all, they allegedly "look just like lead acid", right? :> And, other than longevity and power density, they offer no better *performance* in this example (in practical terms: run time!) My experience (processing LITERALLY *thousands* of discarded UPSs that are still operational -- except for their batteries) has been that folks think they need a UPS. And, later, decide they aren't worth the trouble. Mains power, in most places, is pretty reliable. Outages infrequent. And, systems (except life support) tend to be reasonably robust in limiting loss due to abrupt outages. [Think about it; do you really do any work during an outage? This machine -- and the modem -- have an oversized backup that will keep them operational for a couple of hours. But, every other machine just needs to stay "up" until I can get around to noting what they were doing, at the time of the outage, and cleanly shutting them down -- before the UPS daemon does that unilaterally] Most of the 12 (15?) UPSs that I have here were rescued NIB; many with functional battery packs (some happened to have been warehoused too long before being discarded). Anything with bad batteries has a (recycle) value of about $5 -- largely for the weight of the switching transformer (plastic has NO value and the tin/steel case is $0.01/pound; circuit boards have very little precious metals so pennies, there). [An unattended workstation is typically idle and the disk cache has already been written out so you're only concerned with outages WHILE you are actively working on something.] Hence my points to the OP. Think about WHY you are looking at lithium pack replacements (they are roughly double the price of lead acid; is the inconvenience of replacing them every three or four years really worth the price? Are you sure you won't be replacing (or discarding!) the UPS in that same interval?? [I see a market for oversized "super caps"; temporary energy reservoirs to let UPSs ride out very short outages -- a second or two -- without the need for replacement]
Reply by ●October 22, 20222022-10-22
On 10/22/2022 7:18 AM, Ed Lee wrote:> On Saturday, October 22, 2022 at 1:06:33 AM UTC-7, Don Y wrote: >> On 10/22/2022 12:52 AM, John Walliker wrote: >>> Not in mine. It uses lithium cells, is easy to hold in one hand and has >>> started a variety of large engines including diesel vans and a Ferrari >>> FF. One of the nice things about Li cells is that the starter pack can >>> sit on the shelf for a year without any attention and then "just work". >>> (I do leave it slightly discharged to help preserve battery life.) It >>> was very reasonably priced about 3 years ago, mostly because I bought it >>> on an Amazon "flash" sale. I had been browsing for such things but >>> considered all the good ones to be too expensive. Then an offer of a 40% >>> discount, valid for 1 hour, popped up and I took it. >> But is *it* really starting the car *or* simply quickly replenishing >> (topping off) the charge in the existing lead acid battery? I.e., >> converting "unusable" coulombs to useful ones? . > > My Leaf won't wakeup after 2 or 3 days sleeping. It's mostly because of the > 100W power relay.Are you saying that the relay actuator dissipates 100W WHILE SLEEPING? Why isn't the resting state of the relay the one that doesn't draw power? Does an additional 100W "driving load" hurt your efficiency that much that it has to be moved to the "charging" state?> I could just jumper short the relay output, when i open > up the battery next time. For now, i just added a 200Wh (3S9P) Li power > brick (around 5 pounds) to the LA battery. The Li BMS cutoff at 11.5V when > charging. It's fine in this case. The Li battery mostly keep the LA > battery from dropping too low.So, you're just using the lithium battery to keep the leaky lead acid battery "topped off". Could that brick achieve the same purpose if connected to the battery via a pushbutton that you momentarily activate prior to starting the vehicle? (Or, is the LA battery your main propulsion battery pack)>> Most of the units I've seen have ~16AWG leads -- hardly enough to carry >> the current required to turn over the engine. >> >> Said another way, if the lead-acid battery was REMOVED from the car, would >> you expect to be able to turn it over? > > Yes it does. But better to keep the LA when close to fully charge, and less > cycling on the Li.
Reply by ●October 22, 20222022-10-22
On 10/22/2022 2:46 PM, Fred Bloggs wrote:> On Saturday, October 22, 2022 at 5:29:19 PM UTC-4, Don Y wrote: >> On 10/22/2022 2:05 PM, Fred Bloggs wrote: >>> On Saturday, October 22, 2022 at 11:16:46 AM UTC-4, Don Y wrote: >>>> On 10/22/2022 6:34 AM, Fred Bloggs wrote: >>>>> On Friday, October 21, 2022 at 9:39:16 PM UTC-4, Don Y wrote: >>>>>> On 10/21/2022 5:51 PM, Fred Bloggs wrote: >>>>>>> On Tuesday, October 18, 2022 at 5:10:35 PM UTC-4, Don Y wrote: >>>>>>>> On 10/18/2022 8:43 AM, Bert Hickman wrote: >>>>>>>>> Does anyone have any experience swapping out old lead-acid >>>>>>>>> batteries for Li-ion equivalents in a UPS? >>>>>>>>> >>>>>>>>> I see these batteries advertised as replacements, but I'm >>>>>>>>> concerned about their safety in this application. New Li-ion >>>>>>>>> UPS systems seem to be extremely expensive. >>>>>>>> I wouldn't trust them -- unless the UPS manufacturer had >>>>>>>> blessed them (and dubious, even then). >>>>>>>> >>>>>>>> If you are having problems with battery life in UPS, look to >>>>>>>> see how it is being charged. Many "cook" their batteries >>>>>>>> (poor circuit tolerances). Also, they seem intent on >>>>>>>> recharging QUICKLY (in case another outage??) instead of >>>>>>>> trying to prolong the batteries' service life. >>>>>>>> >>>>>>>> Finally, note that many places will pay you for the lead in >>>>>>>> your old batteries. If you're just using 12V7.2AHr batteries, >>>>>>>> then there's not much money there (~$0.20/pound). OTOH, if you >>>>>>>> have a bigger UPS, you can get a fair bit back (I got over >>>>>>>> $100 for the dead batteries in my biggest UPS) >>>>>>> >>>>>>> Do you really think the modern lithium battery is just a dumb >>>>>>> collection of cells. The products going after the SLA market >>>>>>> almost certainly have built-in BMS ( battery management system) >>>>>>> that cuts off the charging current when fully charged. And that >>>>>>> BMS is probably made by Analog Devices. Last time I looked the >>>>>>> charge management cuts off the float mode charge current when it >>>>>>> has fallen to 25% or so of the initial Ipk delivered to the >>>>>>> uncharged battery. And there's a sensitivity of 100:1 state of >>>>>>> charge to the temperature dependent open terminal voltage, so >>>>>>> the process requires great precision. Nonetheless, it's simple >>>>>>> to very reliably put the lithium into the SLA charge circuit- >>>>>>> within reason. >>>>>> Do you really think a UPS treats a battery as an ideal voltage >>>>>> source? >>>>>> >>>>>> Do you really think a product designed years ago can just tolerate >>>>>> a replacement battery chemistry? That the charger isn't actively >>>>>> monitoring charge current? And determining state of charge by >>>>>> noticing its relationship to open cell voltage over time? >>>>>> >>>>>> How do you think the UPS decides that the battery needs >>>>>> replacing? >>>>>> >>>>>> "Gee, battery stopped taking charge current. Has it become >>>>>> disconnected? Has it failed?" >>>>>> >>>>>> Why are lithium based solutions MULTIPLES of the price of >>>>>> lead-acid solutions? Surely they could just take an EXISTING lead >>>>>> acid design and drop in a "lead-acid emulator" for the incremental >>>>>> cost of the "emulator battery" over the lead-acid original. >>>>>> >>>>>> Maybe there's more to the technology than you think! >>>>> >>>>> Not really. I don't care what kind of battery charge management the >>>>> UPS has, the replacement battery can be made to look identical to >>>>> the original. Cutting off the charge current in float mode does not >>>>> mean opening a relay switch or something crude like that. The >>>>> replacement battery can present enough of a self-discharge current >>>>> loading to make itself look real enough. >>>> And, you'll discover on one UPS the UPS won't turn on because it >>>> doesn't see a battery; on another, it won't run it's self checks; etc. >>>> You don't know what "firmware A" expects of it's battery that >>>> "firmware B" doesn't. >>>> >>>> All of the battery-backed products I've designed predated lithium. Put >>>> a lithium battery in them and they'll throw a fault -- because the >>>> battery isn't discharging at the expected rate; or is charging at the >>>> wrong rate (dv/dt); or, the monitoring circuitry appears defective; >>>> etc. >>>> >>>> "Hey, let's make this lithium battery look EXACTLY like a lead acid -- >>>> so it HAS NO ADVANTAGES over a lead acid battery! But, we can charge >>>> more for those non-differences!" >>>> >>>> MSRP $1,865.88 <https://www.amazon.com/dp/B08QQRFQ4W> MSRP $2,899.99 >>>> <https://www.amazon.com/dp/B09JR6ZCQX> >>>> >>>> $1000 premium for lithium battery pack. Same exact run-times, same >>>> load capacity, same networking capability, same display (literally the >>>> same module). Ah, but the more expensive one has a shiny front panel >>>> (and isn't intended to be floor-standing)! >>>> >>>> [Latter has shallower aspect ratio and weighs less -- though at the >>>> cost of an additional rack unit. *Might* be important in a datacenter; >>>> doubtful a SOHO user would bother with it as he'd likely not have many >>>> such devices, many batteries to replace, etc. Note the number of >>>> lead-acid offerings vs. lithium -- why is that?] >>> >>> You're becoming hysterical. >> Hysterical? Wow, you've sure got a perverse sense of reality! >>> Batteries are not precision components. Any design that's looking at >>> voltage time drift or charge rates to make any decisions is deluding >>> itself. All you end up doing is set up the user for quirky indications >>> they'll just learn to ignore. >> I guess products that have done so and are in the field, WITHOUT >> COMPLAINT, must not exist in your perverse reality! >> >> Designs that just look at voltage trip points are the ones that annoy >> users with erroneous data. >> >> If I dump C/10 into a battery pack FOR TWELVE HOURS and see no change in >> pack voltage, is that just a quirk? What about three hours? One hour? >> Maybe I should just wait a week and, at the end of the week, the cell >> voltage will spontaneously indicate a full charge? >> >> Despite the fact that, HISTORICALLY (remember, I can watch the battery >> FOREVER; it's built into the circuit!) it has registered a specific >> change in each of those intervals? >> >> If I'm a constant load and, today, see the pack voltage dropping at rate >> R1 and, historically, it has always dropped at ~R0 (R0 << R1), should I >> just assume that's a quirk and NOT act on it? Has my implementation >> ACTUALLY changed so that the components are dissipating more power -- yet >> NOTHING IS WRONG? >> >> We EXPECT components to behave in largely predictable ways. Otherwise, we >> couldn't design with them. > > Sounds like you're confusing predictability with eternality. Talk about > oversimplification. The measurements you touched on are largely a big > unknown, not only in principle but in practice. Any instance of a particular > battery is more than just the battery itself but also it's entire life > history of charge and discharge cycles, storage temperature, maybe a bunch > of other stuff known to have significant impact on battery life. Did you do > any pulsed impedance measurements with varying test loading while you were > at it. IIRC the terminal impedance tells some people a lot about the battery > condition.No. You're overcomplicating things! You want dimensioned values for observations; I just want *trends*. I don't care what the rate of current flow out of the battery is FOR THIS DEVICE. I know the nominal design rate and have selected the battery with that in mind. Nor do I care how "robust" this particular battery happens to be -- it may have been sitting on a shelf in the stockroom a few months longer than some other battery. Or, the specific gravity of the electrolyte may be at the other end of the manufacturer's tolerance. The *user* only cares about: - how long will the unit remain operational in the absence of power (will I be able to finish the uninterruptible process that I have begun before it unilaterally shuts down?) - how long will the battery continue to provide this level of performance before requiring replacement As the battery is continuously observable, I can make those predictions FOR THIS BATTERY IN THIS DEVICE. You WATCH THE BATTERY. FOREVER. You TAKE NOTES. You don't care if it performs identically to every other battery. Rather, you notice how THIS battery performs in THIS circuit with THESE tolerances in THESE environmental conditions. You expect it to perform roughly the same, tomorrow. And, you expect its performance to gradually degrade up to a point where it fails to meet the design goals. BEFORE it gets to that point (and pisses off the user), you alert the user to that fact. When the battery is replaced, you restart the observation process anew. And, you didn't have to incur additional manufacturing test or calibration steps to provide this information!
Reply by ●October 22, 20222022-10-22
On Saturday, October 22, 2022 at 2:53:30 PM UTC-7, Don Y wrote:> On 10/22/2022 7:18 AM, Ed Lee wrote: > > On Saturday, October 22, 2022 at 1:06:33 AM UTC-7, Don Y wrote: > >> On 10/22/2022 12:52 AM, John Walliker wrote: > >>> Not in mine. It uses lithium cells, is easy to hold in one hand and has > >>> started a variety of large engines including diesel vans and a Ferrari > >>> FF. One of the nice things about Li cells is that the starter pack can > >>> sit on the shelf for a year without any attention and then "just work". > >>> (I do leave it slightly discharged to help preserve battery life.) It > >>> was very reasonably priced about 3 years ago, mostly because I bought it > >>> on an Amazon "flash" sale. I had been browsing for such things but > >>> considered all the good ones to be too expensive. Then an offer of a 40% > >>> discount, valid for 1 hour, popped up and I took it. > >> But is *it* really starting the car *or* simply quickly replenishing > >> (topping off) the charge in the existing lead acid battery? I.e., > >> converting "unusable" coulombs to useful ones? . > > > > My Leaf won't wakeup after 2 or 3 days sleeping. It's mostly because of the > > 100W power relay. > Are you saying that the relay actuator dissipates 100W WHILE SLEEPING?No, 100W when the car is on. After a few days, the 12V battery is too weak to pull the 100W power relay (switching 400V on).> Why isn't the resting state of the relay the one that doesn't draw > power? Does an additional 100W "driving load" hurt your efficiency > that much that it has to be moved to the "charging" state? > > I could just jumper short the relay output, when i open > > up the battery next time. For now, i just added a 200Wh (3S9P) Li power > > brick (around 5 pounds) to the LA battery. The Li BMS cutoff at 11.5V when > > charging. It's fine in this case. The Li battery mostly keep the LA > > battery from dropping too low. > So, you're just using the lithium battery to keep the leaky lead acid > battery "topped off". Could that brick achieve the same purpose if > connected to the battery via a pushbutton that you momentarily activate > prior to starting the vehicle? (Or, is the LA battery your main propulsion > battery pack)Yes, i also did an emergency jump charging from 400V to 12V with 5S5P 200K resistor. Not very efficient, but works and i happened to have the resistors in the car, and 400V tap in the backseat. The traction/propulsion battery is 24kWh 96S Li.> >> Most of the units I've seen have ~16AWG leads -- hardly enough to carry > >> the current required to turn over the engine. > >> > >> Said another way, if the lead-acid battery was REMOVED from the car, would > >> you expect to be able to turn it over? > > > > Yes it does. But better to keep the LA when close to fully charge, and less > > cycling on the Li.
Reply by ●October 22, 20222022-10-22
On 10/22/2022 3:05 PM, Ed Lee wrote:>>> My Leaf won't wakeup after 2 or 3 days sleeping. It's mostly because of >>> the 100W power relay. >> Are you saying that the relay actuator dissipates 100W WHILE SLEEPING? > > No, 100W when the car is on. After a few days, the 12V battery is too weak > to pull the 100W power relay (switching 400V on).Is there a parasitic drain on that battery? Or, is it just undersized for the task? (one would expect a car sitting for 2 or 3 days to be somewhat "normal use")>> Why isn't the resting state of the relay the one that doesn't draw power? >> Does an additional 100W "driving load" hurt your efficiency that much that >> it has to be moved to the "charging" state? >>> I could just jumper short the relay output, when i open up the battery >>> next time. For now, i just added a 200Wh (3S9P) Li power brick (around 5 >>> pounds) to the LA battery. The Li BMS cutoff at 11.5V when charging. >>> It's fine in this case. The Li battery mostly keep the LA battery from >>> dropping too low. >> So, you're just using the lithium battery to keep the leaky lead acid >> battery "topped off". Could that brick achieve the same purpose if >> connected to the battery via a pushbutton that you momentarily activate >> prior to starting the vehicle? (Or, is the LA battery your main >> propulsion battery pack) > > Yes, i also did an emergency jump charging from 400V to 12V with 5S5P 200K > resistor. Not very efficient, but works and i happened to have the > resistors in the car, and 400V tap in the backseat. The traction/propulsion > battery is 24kWh 96S Li.This suggests either replacing the battery with something with less leakage/more reserve (assuming there is no unintentional leak) or just installing the lithium battery as a built-in jump starter. Though your approach suggests you don't have to worry about keeping the lithium charged if permanently installed (?)
