Yaesu: FT-817ND - Charging NCR18650

Status
Not open for further replies.

andrew30

Member
Joined
Sep 14, 2018
Messages
14
I'm not sure how charging circuity works on 817ND, but when connected to regulated power supply, it draws constant current of about 320mA during charging in power-off state.
This would make it suitable for charging li-ion cells without any additional controller circuity, provided the input voltage is less than 12.3V.
I don't mind 0.1C charging speed, provided it'd cut-off at 12.3V (4.1V per cell)
Or am I missing something (pulse charging?)?
 

prcguy

Member
Joined
Jun 30, 2006
Messages
15,228
Location
So Cal - Richardson, TX - Tewksbury, MA
The original FT-817 and ND series had an internal charger designed to charge Ni-Cad and Ni-MH batteries, which would not be suitable for any Lithium Ion types.

I'm not sure how charging circuity works on 817ND, but when connected to regulated power supply, it draws constant current of about 320mA during charging in power-off state.
This would make it suitable for charging li-ion cells without any additional controller circuity, provided the input voltage is less than 12.3V.
I don't mind 0.1C charging speed, provided it'd cut-off at 12.3V (4.1V per cell)
Or am I missing something (pulse charging?)?
 

andrew30

Member
Joined
Sep 14, 2018
Messages
14
The original FT-817 and ND series had an internal charger designed to charge Ni-Cad and Ni-MH batteries, which would not be suitable for any Lithium Ion types.

Please elaborate on this (rather generic) statement.

Li-Ion cells should be charged with a combination of CC/CV or just CC without reaching saturation phase (which starts around 4.1V).
Full charge to 4.2V must have cut-off current (usually 85mA), otherwise the cell might be damaged. However in this scenario, I am aiming for partial charge, without saturating the cell.
Powering the 817 from exactly 12.3V, one should achieve maximum of 4.1V per cell, just below CV phase of charging process (85%-90% capacity).

From what 817 exhibits when charging is enabled, it would seem the internal NiMH battery is charged with constant current of 320mA for a set period of time (6h,8h or 10h). This seems to be limited by input voltage.
Therefore by using 12.3V instead of 13.5V, one should achieve optimal charging conditions for 3S Li-Ion pack, provided the charging circuit does not implement pulse charging or deltaV detection (which doesn't seem likely).
 

prcguy

Member
Joined
Jun 30, 2006
Messages
15,228
Location
So Cal - Richardson, TX - Tewksbury, MA
All I can say is the FT-817 is designed to charge their specific battery and also aftermarket individual Ni-Cad or Ni-MH AA cells. The radio and its internal charger design is nearly 20yrs old by now and it was not intended to charge anything else.

In my opinion, if you have a suitable 12V range pack with built in cell balancing and overcharge protection and if 320ma is a safe charge range for that, then you might be on to something. If you have to carefully watch the input voltage to the radio during charge, and there is no cell balancing between the 18650s, then you are on your own and good luck with that.

I build 12V packs with 18650s using $3 boards that will handle from two to four cells, have cell balancing, overcharge protection, over discharge protection, short circuit protection, etc. However I would use something else to power an FT-817 in the field.





Please elaborate on this (rather generic) statement.

Li-Ion cells should be charged with a combination of CC/CV or just CC without reaching saturation phase (which starts around 4.1V).
Full charge to 4.2V must have cut-off current (usually 85mA), otherwise the cell might be damaged. However in this scenario, I am aiming for partial charge, without saturating the cell.
Powering the 817 from exactly 12.3V, one should achieve maximum of 4.1V per cell, just below CV phase of charging process (85%-90% capacity).

From what 817 exhibits when charging is enabled, it would seem the internal NiMH battery is charged with constant current of 320mA for a set period of time (6h,8h or 10h). This seems to be limited by input voltage.
Therefore by using 12.3V instead of 13.5V, one should achieve optimal charging conditions for 3S Li-Ion pack, provided the charging circuit does not implement pulse charging or deltaV detection (which doesn't seem likely).
 

andrew30

Member
Joined
Sep 14, 2018
Messages
14
All I can say is the FT-817 is designed to charge their specific battery and also aftermarket individual Ni-Cad or Ni-MH AA cells. The radio and its internal charger design is nearly 20yrs old by now and it was not intended to charge anything else.
Upon further investigation there no "internal charger" to speak of. The 817 dumps approx. 250mA into the battery connector for a set period of time. (regardless if the battery is present or not, 70mA is used for the display and LED in off state). NiCd and NiMH can 'cope' with it, as when fully charged, they dissipate the extra energy in heat, and coincidentally, there's plenty of heat conductors in the battery area.
Of course this process damages the cells slightly (esp. NiCd), as it's not IEC specified 0.1C for 16hours.
The only measure that's keeping 817 from charging Alkaline cells is that green wire.

In my opinion, if you have a suitable 12V range pack with built in cell balancing and overcharge protection and if 320ma is a safe charge range for that, then you might be on to something. If you have to carefully watch the input voltage to the radio during charge, and there is no cell balancing between the 18650s, then you are on your own and good luck with that.
Cell balancing is not necessary provided the cells are capacity, voltage and IR matched from the start (100mAh tolerance is generally accepted). Laptop batteries rarely implement balancing and do not usually exhibit explosive tendencies. :)

I build 12V packs with 18650s using $3 boards that will handle from two to four cells, have cell balancing, overcharge protection, over discharge protection, short circuit protection, etc. However I would use something else to power an FT-817 in the field.
No reason not to opt in for li-ion cells IMO.
Gone are the days where I'd willingly pack 40Wh/kg lead-acid for QRP backpacking. AAs don't fare much better.
Li-Ion is a way to go, NCR specifically both due to lower temperature tolerances, higher currents and higher densities.

There are a few aftermarket solutions:
3D printed holder
https://www.thingiverse.com/thing:2527434

and

3000mAh battery with back cover and adapter:
https://www.aliexpress.com/item/NEW...-Hatch-Cover-Charger-Bracket/32798891704.html

Neither of those solutions I'd consider optimal.
The 3D holder seems good solution at start, but requires protected cells and springed contacts increase resistance.
3Ah battery pack looks great, but it's expensive ($60) and replacement would mean buying proprietary battery, likely another $40 value.

Best solution IMO would be to 3D model the battery door to allow accommodating 18650 with their 18mm height + some extra room for heat-shrinked taping, 5.5x2.1mm jack and 5A+ switch.
This would enable swapping 3S 18650 packs on the go, low resistance across contacts and retain all the benefits of above battery pack, also by being non-proprietary (3S packs are readily DIY-able for $5-10 using NCR18650B or INR18650-35E).

With that, a question follows - has anyone managed to 3D model the battery door?
 

prcguy

Member
Joined
Jun 30, 2006
Messages
15,228
Location
So Cal - Richardson, TX - Tewksbury, MA
At one time I considered marketing a set of aluminum 817 replacement covers to save weight and also a bottom cover with more space to accommodate an auto antenna tuner or batteries. That and a set of A123 (now NEC Energy Solutions) 26650 cells with cell balancing and overcharge/discharge protection would be an ideal solution.

This type of battery will give upwards of 7000 charge/discharge cycles and will last 10yrs or more. You only get that with cell balancing during charging and protecting the cells from being charged or discharged beyond a critical point.


Upon further investigation there no "internal charger" to speak of. The 817 dumps approx. 250mA into the battery connector for a set period of time. (regardless if the battery is present or not, 70mA is used for the display and LED in off state). NiCd and NiMH can 'cope' with it, as when fully charged, they dissipate the extra energy in heat, and coincidentally, there's plenty of heat conductors in the battery area.
Of course this process damages the cells slightly (esp. NiCd), as it's not IEC specified 0.1C for 16hours.
The only measure that's keeping 817 from charging Alkaline cells is that green wire.


Cell balancing is not necessary provided the cells are capacity, voltage and IR matched from the start (100mAh tolerance is generally accepted). Laptop batteries rarely implement balancing and do not usually exhibit explosive tendencies. :)


No reason not to opt in for li-ion cells IMO.
Gone are the days where I'd willingly pack 40Wh/kg lead-acid for QRP backpacking. AAs don't fare much better.
Li-Ion is a way to go, NCR specifically both due to lower temperature tolerances, higher currents and higher densities.

There are a few aftermarket solutions:
3D printed holder
https://www.thingiverse.com/thing:2527434

and

3000mAh battery with back cover and adapter:
https://www.aliexpress.com/item/NEW...-Hatch-Cover-Charger-Bracket/32798891704.html

Neither of those solutions I'd consider optimal.
The 3D holder seems good solution at start, but requires protected cells and springed contacts increase resistance.
3Ah battery pack looks great, but it's expensive ($60) and replacement would mean buying proprietary battery, likely another $40 value.

Best solution IMO would be to 3D model the battery door to allow accommodating 18650 with their 18mm height + some extra room for heat-shrinked taping, 5.5x2.1mm jack and 5A+ switch.
This would enable swapping 3S 18650 packs on the go, low resistance across contacts and retain all the benefits of above battery pack, also by being non-proprietary (3S packs are readily DIY-able for $5-10 using NCR18650B or INR18650-35E).

With that, a question follows - has anyone managed to 3D model the battery door?
 

andrew30

Member
Joined
Sep 14, 2018
Messages
14
At one time I considered marketing a set of aluminum 817 replacement covers to save weight and also a bottom cover with more space to accommodate an auto antenna tuner or batteries. That and a set of A123 (now NEC Energy Solutions) 26650 cells with cell balancing and overcharge/discharge protection would be an ideal solution.


This type of battery will give upwards of 7000 charge/discharge cycles and will last 10yrs or more. You only get that with cell balancing during charging and protecting the cells from being charged or discharged beyond a critical point.
26650 constitutes only battery size - what chemistry would you use to get 7000 cycles?
In LiNiCoALO2, 5500mAh would be one of the highest capacities. More economical solution, which wouldn't require considerable level of modding would be to add 6 more 18650 cells on the bottom.

NCRs might give you approx 2500 cycles with 50% SoC, but 7000 cycles? Other than Lithium Titanate cells (which are super expensive), I can't think of any other chemistry rated for such cycle life.

So you managed to redraw the batt cover shape and dimensions?
 

prcguy

Member
Joined
Jun 30, 2006
Messages
15,228
Location
So Cal - Richardson, TX - Tewksbury, MA
Lithium Iron Phosphate is the chemistry developed by A123 systems and some of their line was recently sold to NEC. The 12V7 pack originally sold by A123 and now NEC has a 4S2P configuration for 5AH in an impossible to screw up package.

Each 2.5AH 26650 cell is rated at 70 amp continuous discharge during everyday use and they can be 100% charged in 15 minutes. I have an old Buddipole 4S2P 26650 pack with A123 cells that is about 10yrs old and it will still start my car.A single string of these 26650 cells with balancing would be a good match for a QRP rig or if space permits a 4S2P version.

I had an AutoCad dwg started many years ago for the 817 covers but its probably long lost by now.

26650 constitutes only battery size - what chemistry would you use to get 7000 cycles?
In LiNiCoALO2, 5500mAh would be one of the highest capacities. More economical solution, which wouldn't require considerable level of modding would be to add 6 more 18650 cells on the bottom.

NCRs might give you approx 2500 cycles with 50% SoC, but 7000 cycles? Other than Lithium Titanate cells (which are super expensive), I can't think of any other chemistry rated for such cycle life.

So you managed to redraw the batt cover shape and dimensions?
 

andrew30

Member
Joined
Sep 14, 2018
Messages
14
Even LiFePO4 cells don't achieve 7000 cycles. For QRP rig those parameters seems quite excessive (max. 2.2A), and a noticeable tradeoff for parameters not required out of weight/size constrained rig.

Even if 26650 LiFePO4 could pull off 7000 cycles, at 2.5Ah and 90Wh/kg (Titanate can pull off up to 20000 cycles at 60Wh/kg), this is wholly unnecessary for QRP use. All the buttons and knobs would have long stopped working before the battery started to age. Three times lower Wh/kg does make a difference more than theoretical cycle life. At 2000 cycles for 60% SoC, that's more than enough anyway.
Needless to say, NCR18650 are 1/6 to 1/30 of the price of LiFePO4 and usually have better low temperature performance.
Also, FT-817 specifically runs more efficiently at 11.5V compared to 13.5V.

When hiking with 817, every gram counts, lead-acid and AAs are naturally first to go. Neither LiFePO4 nor Li2TiO3 are good match there. Even with using 50% SoC Li-Ion, that's 135Wh/kg, far better than aforementioned chemistries. As far as safety, I have never seen NCR cell explode. When shorted, they either get hot and vent, or just get hot (ICRs are the most violent out of cylindrical cells). Compare that to every other Lipo, of which 50% managed to get swollen within first two years of use. And inside crammed space of 817, that wouldn't be good.
 

prcguy

Member
Joined
Jun 30, 2006
Messages
15,228
Location
So Cal - Richardson, TX - Tewksbury, MA
For a number of years the LifePO4 cells have been the go too battery for portable radio operation, electric cars and a number of other uses due to their specs and safety. I know a number of people who have experimented with other chemistry for their remote control airplanes and helicopters which ended in explosions and fire.

I'll stick with the LiFePO4 cells and preferably from A123 Systems or NEC Energy Solutions for my radio needs. Bioenno Power also puts out a good product for radio use, although A123 Systems cells are better. I've also moved on from the FT-817 series to the Elecraft KX3 and KX2, which has a really good internal factory battery option.


Even LiFePO4 cells don't achieve 7000 cycles. For QRP rig those parameters seems quite excessive (max. 2.2A), and a noticeable tradeoff for parameters not required out of weight/size constrained rig.

Even if 26650 LiFePO4 could pull off 7000 cycles, at 2.5Ah and 90Wh/kg (Titanate can pull off up to 20000 cycles at 60Wh/kg), this is wholly unnecessary for QRP use. All the buttons and knobs would have long stopped working before the battery started to age. Three times lower Wh/kg does make a difference more than theoretical cycle life. At 2000 cycles for 60% SoC, that's more than enough anyway.
Needless to say, NCR18650 are 1/6 to 1/30 of the price of LiFePO4 and usually have better low temperature performance.
Also, FT-817 specifically runs more efficiently at 11.5V compared to 13.5V.

When hiking with 817, every gram counts, lead-acid and AAs are naturally first to go. Neither LiFePO4 nor Li2TiO3 are good match there. Even with using 50% SoC Li-Ion, that's 135Wh/kg, far better than aforementioned chemistries. As far as safety, I have never seen NCR cell explode. When shorted, they either get hot and vent, or just get hot (ICRs are the most violent out of cylindrical cells). Compare that to every other Lipo, of which 50% managed to get swollen within first two years of use. And inside crammed space of 817, that wouldn't be good.
 

RFI-EMI-GUY

Member
Joined
Dec 22, 2013
Messages
6,859
I read an article about using avalanche zener diodes to balance LifePO4 cells. One diode, two cells. Not sure what you would do for odd number of series cells. Apparently the low voltage avalanche zeners have a very abrupt knee.

Sent from my SM-T350 using Tapatalk
 

prcguy

Member
Joined
Jun 30, 2006
Messages
15,228
Location
So Cal - Richardson, TX - Tewksbury, MA
Elaborate cell balancing and protection boards are really cheap, I use these rather than just Zeners and they stick right on the end of the cells: https://www.ebay.com/itm/3S-20A-11-...m=122183531720&_trksid=p2047675.c100005.m1851

I read an article about using avalanche zener diodes to balance LifePO4 cells. One diode, two cells. Not sure what you would do for odd number of series cells. Apparently the low voltage avalanche zeners have a very abrupt knee.

Sent from my SM-T350 using Tapatalk
 

RFI-EMI-GUY

Member
Joined
Dec 22, 2013
Messages
6,859
Elaborate cell balancing and protection boards are really cheap, I use these rather than just Zeners and they stick right on the end of the cells: https://www.ebay.com/itm/3S-20A-11-...m=122183531720&_trksid=p2047675.c100005.m1851
I have repacked some Saber batteries with 2200 mAh Tenergy 2 cell lithium Ion packs. They are protected internally. I made up a charger using a 2 cell lithium Ion development board from Microchip. They work well.


I am now looking for a solution for the Alkaline D cell 11S2P 16.5V battery pack on my UHF PX300-S project. 22 D cells heft some weight. No doubt it will run a week on that, but Lithium Ion seems to be the way to go. I figure I need a 4S2P string to replace it if I go lithium Ion. LIFEPO would require 5S2P if I have my numbers correct. Not sure how I would protect those.

If you construct a 4S2P do you use one protection board and pair up 2 cells in parallel for each of the 4 series pairs? Or are their two series branches each with a protection board?

Sent from my SM-T350 using Tapatalk
 

andrew30

Member
Joined
Sep 14, 2018
Messages
14
If you construct a 4S2P do you use one protection board and pair up 2 cells in parallel for each of the 4 series pairs? Or are their two series branches each with a protection board?

Just one PCB will do for 2P configuration. Two PCBs might interact with each other's short circuit protection.
 

andrew30

Member
Joined
Sep 14, 2018
Messages
14
When left undecided, consider the following numbers:
LiFePO4:

Typical cycle life: 2000
Density: 90-110Wh/kg
Availability: scarce (in cell form)
Price: approx. 1.2Wh/$

Li-Ion (LiNiCoAlO2):

Typical cycle life: 500
70% SoC cycle life: 1200
50% SoC cycle life: 2000-2500
Density: 200-270Wh/kg
Availability: plenty of options
Price: from 3Wh/$ to 13Wh/$
 

nanZor

Active Member
Joined
May 28, 2009
Messages
2,807
So Andrew, (as an LFP user myself), ask yourself: What happens when my protection / balancing board fails? Are my sources of cells coming from used laptops, or old vape-shop smoke rejects, or grey market counterfeits?

Vent with flame is the possible answer.

LiFeP04, on the other hand, is one greedy molecule and hangs on to oxygen like no tomorrow. So, under abuse / failure, they may vent, but not with flame.

Overvoltage example : take a NON-lifepo4 cell to well above 4.2 volts, say 4.5v and beyond. Hello incident with flame.

Take a LiFeP04 cell to about 10v, and although the cell is ruined, you may have a mere venting incident, but without flame. The point here is that LFP is not as knife-edge critical to start a major event.

So while LFP is not as energy dense as other cells, there is NO WAY I would use anything but LiFeP04 for the average DIY project that relies upon unqualified cheap little boards.

Sounds like you'll forge on ahead anyway, but had to throw the question of how well you trust your balancing / safety boards themselves?
 

andrew30

Member
Joined
Sep 14, 2018
Messages
14
So Andrew, (as an LFP user myself), ask yourself: What happens when my protection / balancing board fails? Are my sources of cells coming from used laptops, or old vape-shop smoke rejects, or grey market counterfeits?

Vent with flame is the possible answer.
It seems you disregarded my previous messages on the subject. Oh well.
I encourage anyone to find me one, ONE NCR18650 cell that would fail explosively when stressed with short circuit, overcharge or mechanical deformation. There's a good reason why Tesla use them in their EVs.

Li-Ion cells are not all the same (ICR vs. IMR vs. NCR...) - and every recognizable manufacturer performs extensive testing before marketing new products. Check any PDF for tests done in the manufacturing plants.

There's lot of media fear-mongering about li-ion cells, but it's important to stay objective when assessing benefits vs. drawbacks.

LiFePO4, on the other hand, is one greedy molecule and hangs on to oxygen like no tomorrow. So, under abuse / failure, they may vent, but not with flame.

Overvoltage example : take a NON-lifepo4 cell to well above 4.2 volts, say 4.5v and beyond. Hello incident with flame.
See point above.

Take a LiFePO4 cell to about 10v, and although the cell is ruined, you may have a mere venting incident, but without flame. The point here is that LFP is not as knife-edge critical to start a major event.

I am kind of startled that we're discussing li-ion cell safety, when manufacturers are pouring hundreds of thousands of cells every month. Those that exhibit catastrophic failures are either fake, or the failures can be traced to electronics and/or improper installation. When it's the cell fault, and quite possibly it can happen, statistically how are those isolated cases relevant?

So while LFP is not as energy dense as other cells, there is NO WAY I would use anything but LiFePO4 for the average DIY project that relies upon unqualified cheap little boards.

To each their own. I'm not planning to charge a 12V pack with 30V and rely on protection PCB to do it's job and disconnect the pack. The boards are good for short circuit protection at least, and all those I tested performed well within specs. You can of course buy more expensive BMS', but they'll likely contain the same components, only have different, more recognizable brand printed on the PCB.

Sounds like you'll forge on ahead anyway, but had to throw the question of how well you trust your balancing / safety boards themselves?
I have build enough battery packs over the last 4 years to understand the risks.
Protection circuits are there to protect against human errors. Avoid those, and they'll never have to trigger.
Short circuit protection, thermal sensor and voltage monitoring is all you really need.
Going over 4P, balancing is recommended.

Avoid charging to 4.2V, choose well-tested original cells, balance capacity, voltage and internal resistance. Use contact welder, not soldering iron. Choose pure nickel strips based on peak current (as a last resort nickel acts like a fuse). Protect the cells from external damage, provide enough spacing to prevent cascade thermal runaway and use flame retardant ABS for cases.
With that, it's possible to enjoy all the benefits of higher density cells safely.

PS: I corrected PO4 in your quotation. Hope it's fine.
 
Status
Not open for further replies.
Top