• Updated 2023-07-12: Hello, Guest! Welcome back, and be sure to check out this follow-up post about our outage a week or so ago.

I need a new battery.

PB170

Well-known member
Hi all,

I need a new battery for my PowerBook…


Short background (feel free to skip):

I built two batteries for my PowerBook 170 back in 2017 (briefly documented in this thread here and here) that I've been using regularly since then. However, all of a sudden at the end of 2020, they both went from working just fine to not being able to power the computer at all, unless all power saving features were set to the absolute minimum, and barely even then. At first I suspected something was wrong with either the PowerBook or the charger I'm using since it seemed unlikely that both batteries would fail at the same time, but after some testing that indeed appears to be the case.

When I built the batteries, I did a ton of research into NiMH battery technology to get the best use out of them, and I've been charging them at 0.1C and cycling them between 0–100% continuously. They both only had about 50 charge cycles each when they failed, so they should still have had plenty of use left. I suspect the problems might be due to internal crystallization. During the first half of 2021 I tried the best I could to revive them without taking them apart but didn't have much success and eventually gave up.

So I've been tethered to the wall ever since, which is starting to feel increasingly limiting since I often use my PowerBook when I'm out and about. It would be great to have it back running on batteries again now that summer is on its way, and avoid having to hunt for power outlets at cafés etc.


Meat of the matter:

When I was preparing to build the batteries in 2017 and considered different battery options, I quickly ruled out lithium based batteries because of the safety risks involved and went for gen 2 NiMh instead. They've served me well, giving me about 4–5 hours of use on average (more than double the capacity of the original batteries!). At this point however, it feels very tempting to leave the long charging times and the 0–100% charging schedule behind and go for the more convenient usage properties and energy density of lithium based batteries instead.

I enjoy tinkering, and it was a fun project to plan/design and build the two batteries back in 2017, but this time around I would very much just want to buy an off-the-shelf battery, if such an option existed. So I'm looking for the easiest possible path to a lithium based battery. At this point I'm contemplating taking a suitable USB power bank and build from that, should that turn out to be a realistic option. One challenge would be to find one with battery cells and circuity that will fit inside the PowerBook battery cases.

To begin with, these are my requirements:
  • At least the same capacity as my first batteries (31.5 Wh), but preferably at least 50 Wh or more (I guess LiPo pouch-style batteries would be the given choice here)
  • Able to use it safely in my PowerBook with the AC adapter connected, but with the actual charging taking place outside of the computer through a separate port (already in my first version of the batteries)
  • Unregulated power so that the the PowerBook's power management will work correctly
I've browsed through the two threads by @360alaska and @sutekh below, and as I understand it the biggest obstacle seems to be to get the battery to charge safely in the computer. If this step could be eliminated, would a lithium based battery be easier to implement you think?

Lithium Polymer battery for PB100 (and maybe portable)
PB 1[4-8][0,5]c? Li-ion Battery

I also watched this video (YouTube: How Does A Power Bank Work?) that I stumbled across, and briefly looked at a couple of other sources covering the basic circuity of USB power banks.

A couple of questions:
  1. Do you think using a power bank would be a good and safe starting point?
  2. Is it possible to safely and easily modify the circuity of USB power banks so that they output unregulated power at circa 7 V (preferably ≈6.95…5.85 V so that the PowerBook's battery management works correctly) rather than the, presumably, regulated 5 V? If not, would it be easy to bypass just the regulating circuity and add a separate boost/buck converter? Any recommendations on a suitable chip/circuit for that in that case?
  3. Unless power banks are already protected from current going the wrong way, is it possible and efficient to protect the output with a diode or some other solution? Any recommendations on a suitable part in that case?
If this turns out to be too tricky to achieve, I might just rebuild one of the batteries with new NiMH cells to have it ready before summer, but like I said, the more modern properties and power density of lithium batteries is very, very tempting this time around.

Would be happy to hear your thoughts. To recap, my priorities here are 1) safety and 2) the easiest/quickest possible solution :)

Sorry about the long post :)
 
Last edited:

3lectr1cPPC

Well-known member
I would also be very interested in this for my 170, 145 and 180c which all use the same pack. I believe alaska360 indicated they were planning on building such a battery, albeit with internal charging just like they're pack for the PB 100. I suppose we'll find out soon when they reply :)
 

davidg5678

Well-known member
I can't remember where I saw it, but I think it is possible to rig up a power bank to connect directly to the charger port of a PowerBook and effectively act as a portable power brick. This would be easier to implement than designing a new internal battery, but it would be much less sleek.
 

4seasonphoto

Well-known member
Don't make the mistake that I did when I tried to power a portable record player from USB power bank: I got an inexpensive DC-DC converter to step up the 5 volt output to 12 volts. Which it did, but the power bank output was still limited to 5V @ 1A, and the power bank would keep cutting out until I got past the initial current inrush.

The right way to do it is with a USB-PD (Power Delivery) compatible power bank (or other similar, competing technology). Combine that with a Trigger which properly enables higher voltage / current output from the power bank. USB-PD triggers can be had pretty cheaply.
 

4seasonphoto

Well-known member
Fresh nicad batteries can still be had from sources like Digikey, and would save you the hassle of having to devise a new charge controller. Pretty sure these things are still being used in applications like electric toothbrushes and razors, which tend to be left charging continuously.
 

360alaska

Well-known member
Actually I have a design that can be used on the black and white models. I'm just very busy right now and can't mass produce them, but I'll sell you my one and only prototype. I can also release the files at some point, but I'm very busy right now so I will not be producing more at this very moment. It is a 4000 mah battery.

All of my Lipo batteries involve individually charging the cells with lipo charging ICs and those require a 5v regulator. (I use TSR 1-2450)
 

PB170

Well-known member
Don't make the mistake that I did when I tried to power a portable record player from USB power bank: I got an inexpensive DC-DC converter to step up the 5 volt output to 12 volts. Which it did, but the power bank output was still limited to 5V @ 1A, and the power bank would keep cutting out until I got past the initial current inrush.

The right way to do it is with a USB-PD (Power Delivery) compatible power bank (or other similar, competing technology). Combine that with a Trigger which properly enables higher voltage / current output from the power bank. USB-PD triggers can be had pretty cheaply.
Thanks, that’s valuable information. I actually thought about that aspect of it earlier but forgot to include it in the post. That might complicate things considerably, unless the USB output circuity can be bypassed easily and safely, or if output at higher currents can be enabled without data signalling. The PowerBook 170 needs at least 15 W. I’ll have to look into that further.

Wouldn’t want to end up with the same low current situation that I’m currently facing with my first two batteries :giggle:
 

PB170

Well-known member
Actually I have a design that can be used on the black and white models. I'm just very busy right now and can't mass produce them, but I'll sell you my one and only prototype. I can also release the files at some point, but I'm very busy right now so I will not be producing more at this very moment. It is a 4000 mah battery.

All of my Lipo batteries involve individually charging the cells with lipo charging ICs and those require a 5v regulator. (I use TSR 1-2450)
Thank you very much for your kind offer, but I would prefer a solution that’s as close to an off-the-shelf part as possible (original case (with my added port for charging), easy external charging, preferably through USB etc.). Also, I need more than 4000 mAh :) (especially if that’s at 3.7 V – my first batteries were 5200 mAh @ 6 V :)) But thanks!
 

4seasonphoto

Well-known member
Thanks, that’s valuable information. I actually thought about that aspect of it earlier but forgot to include it in the post. That might complicate things considerably, unless the USB output circuity can be bypassed easily and safely, or if output at higher currents can be enabled without data signalling. The PowerBook 170 needs at least 15 W. I’ll have to look into that further.
Should be a piece of cake for most modern power banks + the right trigger (cheap, easy to implement). Modifying the power bank itself or bypassing it's output circuitry would be counterproductive. All you really need to create is a custom USB cable with Powerbook barrel connector on one end, an in-line USB trigger, and a USB connector on the other.
 

360alaska

Well-known member
Thank you very much for your kind offer, but I would prefer a solution that’s as close to an off-the-shelf part as possible (original case (with my added port for charging), easy external charging, preferably through USB etc.). Also, I need more than 4000 mAh :) (especially if that’s at 3.7 V – my first batteries were 5200 mAh @ 6 V :)) But thanks!

No, When I say 4000 mah, it's two 3.7v batteries in series for running but split up for charging. One 5v regulator but two charging ICs.
 

PB170

Well-known member
Should be a piece of cake for most modern power banks + the right trigger (cheap, easy to implement). Modifying the power bank itself or bypassing it's output circuitry would be counterproductive. All you really need to create is a custom USB cable with Powerbook barrel connector on one end, an in-line USB trigger, and a USB connector on the other.
Thanks, but what I want is a lithium based battery for my PowerBook, not a separate power bank to power it from :)
 

PB170

Well-known member
After looking into the details of USB Battery Charging and USB Power Delivery, it seems like it would be a much too difficult task to be dealing with. While safe, it sort of takes the quick and easy out of the equation.

So, would it be safe to leave the power bank’s circuity unmodified and instead:
1) draw current directly from the battery terminals,
2) pass it through a power converter (booster, I’d assume) to get a voltage range suitable for the PowerBook,
3) add a diode for reverse current protection in case the AC adapter should get connected when the battery is in use, and
4) add a fuse for short circuit protection?

I’m more read up on NiMh battery technology, but as far as I’m aware, the main risks with lithium batteries is overcharge and over discharge.

Overcharge shouldn’t be a problem with the original circuity in place – if the direct connection to the cells can be safely protected with a diode. And if I set up the power converter such that the PowerBook will go to sleep when the voltage of the lithium cells are still within the safe margin, I imagine that should protect the cells (provided the power conversion is linear). And should the voltage of the cells fall below the safe limit, the power bank’s circuity would still disable the battery, I assume.

Anything crucial I’m missing here?

The downside of this, if at all realistic, is that the extra circuity will take up extra space at the expense of a more powerful battery (though maybe getting rid of the connectors on the power bank board would allow stacking the additional circuity on top), but it seems a lot more straightforward than having to deal with USB signaling and any power bank specific circuity.

Would this be a realistic way to go, you think?
 

sutekh

Well-known member
As long as you're comfortable using a separate charge port (or switch like I've implemented in my design you linked to above) this is quite simple, yes. I'll take a stab at your questions:

1) Assuming your power bank is using Li-Ion (not LiFePO4 or another chemistry) and is a 2S design (two cells in series), its nominal, unregulated output should be 7.4v. That's just about perfect for a 170. If it's a 1S design with a step-up converter from 3.7v -> 5v, that won't work obviously.

2) See #1. Not necessary if it's already outputting 7.4v. I'd find one that does and avoid stepping up from 3.7v if possible.

3) I doubt this is necessary, and will almost certainly cause you voltage problems. The voltage drop across the diode, which will increase with current draw, will almost certainly push your PB into the low voltage / battery warning range of operation.

Connecting the charger to the USB battery bank (which would be routed through some internal charge circuitry) shouldn't hurt the PB drawing directly from the cells themselves. That said, I doubt it will be able to keep up with the laptop's demands if it's powered on. These are thirsty beasts!

4) Never a bad idea.
 

PB170

Well-known member
Thank you very much for your input!

1) Assuming your power bank is using Li-Ion (not LiFePO4 or another chemistry) and is a 2S design (two cells in series), its nominal, unregulated output should be 7.4v. That's just about perfect for a 170. If it's a 1S design with a step-up converter from 3.7v -> 5v, that won't work obviously.

2) See #1. Not necessary if it's already outputting 7.4v. I'd find one that does and avoid stepping up from 3.7v if possible.
I haven't purchased any power bank yet, so I won't know for certain until I buy one and take it apart, but I'm aiming for lithium polymer for the greater energy density (and smaller volume). Also, I don't know, but my impression when looking around a bit on the web is that most power banks with two cells have them connected in paralell.

Either way, I want to be able to calibrate the output voltage precisely so that the PowerBook's battery management will work correctly. (Are voltage converters generally able to give a linear conversion, such that the input voltage equals the output voltage plus or minus a given amount?)

As a side note though, I actually got hopeful that a lithium battery might be doable when I opened and restored a broken chromebook some time ago and realized that its battery voltage matched the PowerBook's requirements quite well (as well as the dimensions of the battery cells – turned out that they were only a couple of millimeters too wide to fit in the PowerBook's battery bay…)

3) I doubt this is necessary,
Uh, how do you mean? As far as I'm aware, one of the main issues with lithium batteries is their precise charging requirements, so I wouldn't want the battery to get charged by the PowerBook should the AC adapter be connected. Right?

and will almost certainly cause you voltage problems. The voltage drop across the diode, which will increase with current draw, will almost certainly push your PB into the low voltage / battery warning range of operation.
Thanks, I didn't think about that. After looking into reverse voltage protection a bit more, it sounds like a P-channel MOSFET would be a better solution.

Connecting the charger to the USB battery bank (which would be routed through some internal charge circuitry) shouldn't hurt the PB drawing directly from the cells themselves. That said, I doubt it will be able to keep up with the laptop's demands if it's powered on. These are thirsty beasts!
Ah, I won't be charging the battery through the USB port when the PowerBook is in use, if that's how you mean. I was referring to (unintentional) charging by the PowerBook's AC adapter.


By the way, any reason to be concerned that drawing power directly from the battery terminals might interfere with or confuse the charging circuity? I assume it would be okay since the battery voltage could drop by just long storage, but it would be good to know for certain.
 

sutekh

Well-known member
Thank you very much for your input!


I haven't purchased any power bank yet, so I won't know for certain until I buy one and take it apart, but I'm aiming for lithium polymer for the greater energy density (and smaller volume). Also, I don't know, but my impression when looking around a bit on the web is that most power banks with two cells have them connected in paralell.

If that's the case and you can't find one designed around a 2S cell arrangement (would make sense now that I think about it, since 3.7->5v is only a 1.3v gap while 7.4->5 is 2.4...) then I'd bag that idea TBH. Just get any of the many available 2s li-ion / lipo charge boards that are readily available (that's what the battery bank would have in it anyway) and connect its output to the 7.4v / 2S cell network, and its input to whatever charge source you prefer (recognizing it'd need to be at least 9v).

Either way, I want to be able to calibrate the output voltage precisely so that the PowerBook's battery management will work correctly. (Are voltage converters generally able to give a linear conversion, such that the input voltage equals the output voltage plus or minus a given amount?)

No, they intentionally provide a consistent output regardless of the input voltage. That's desirable in 99% of use cases, just not this one :) It'd be much nicer if the output +/- the step down or step up voltage scaled linearly relative to the input for battery monitoring purposes.

As a side note though, I actually got hopeful that a lithium battery might be doable when I opened and restored a broken chromebook some time ago and realized that its battery voltage matched the PowerBook's requirements quite well (as well as the dimensions of the battery cells – turned out that they were only a couple of millimeters too wide to fit in the PowerBook's battery bay…)


Uh, how do you mean? As far as I'm aware, one of the main issues with lithium batteries is their precise charging requirements, so I wouldn't want the battery to get charged by the PowerBook should the AC adapter be connected. Right?

Didn't realize you intended to connect the PB's AC adapter since you were talking about a separate battery charger port / circuit. Yes, if you want this to work in concert with the factory AC adapter, then you'll need a diode (recognizing their caveats) or a current monitor / disconnect circuit like @360alaska has implemented in his design.

Thanks, I didn't think about that. After looking into reverse voltage protection a bit more, it sounds like a P-channel MOSFET would be a better solution

A few ways to do this, yes. A Schottky, P-channel, also look up "ideal diodes".

Ah, I won't be charging the battery through the USB port when the PowerBook is in use, if that's how you mean. I was referring to (unintentional) charging by the PowerBook's AC adapter.


By the way, any reason to be concerned that drawing power directly from the battery terminals might interfere with or confuse the charging circuity? I assume it would be okay since the battery voltage could drop by just long storage, but it would be good to know for certain.

Should be just fine. That's exactly what happens when you plug most laptops in. The load is draining off the battery, and the charge circuit is replenishing it (the latter hopefully faster than the former!)
 

PB170

Well-known member
Just get any of the many available 2s li-ion / lipo charge boards that are readily available (that's what the battery bank would have in it anyway) and connect its output to the 7.4v / 2S cell network, and its input to whatever charge source you prefer (recognizing it'd need to be at least 9v).
My plan is to keep the power bank's circuity intact, to be able to conveniently charge the pack with standard USB chargers, and add a separate power converter for drawing power from the battery.

No, they intentionally provide a consistent output regardless of the input voltage. That's desirable in 99% of use cases, just not this one :)It'd be much nicer if the output +/- the step down or step up voltage scaled linearly relative to the input for battery monitoring purposes.
I guess what I meant was unregulated power converters. Wikipedia's article on boost converters mentions that "Boost converters are highly nonlinear systems and a wide variety of linear and nonlinear control techniques for achieving good voltage regulation with large load variations have been explored."

Are boost/buck converters that provide a stable, linear conversion available?

Any way around it, if not? I really need the output voltage to reflect the battery's state of charge in order for the PowerBook's power management to work correctly (don't want the PowerBook to turn off without warning when the battery is empty).

Should be just fine. That's exactly what happens when you plug most laptops in. The load is draining off the battery, and the charge circuit is replenishing it (the latter hopefully faster than the former!)
I think you misunderstood me. Like I mentioned, I plan on charging the battery using the power bank's circuity, but discharge the battery directly from the battery terminals (and not at the same time). So my question was if this might confuse or interfere with the charging circuity.
 

PB170

Well-known member
If voltage converters that provide a stable, linear conversion (where the output voltage equals the input voltage plus/minus a set amount) aren’t available, would it be possible electrically to somehow combine/offset the converted output voltage with the unconverted voltage?

Anyone with knowledge in electronics who can help me out a bit here?

Summer is approaching and I’d really like to have my PowerBook up and running on batteries again :)
 

Chopsticks

Well-known member
going liPo is probably the best option for power density vs physical size. not to mention that a LiPo can handle much higher initial current draw.

modifying a existing power banks output voltage probably isn't the way to go, off the shelf power banks generally only follow specific voltages mainly 5v or the 5/9/12/15/20 in the case of USB PD. it might be worth looking into a power bank that supports Qualcomm QC 3.0/4.0. as this can output any voltage between 3.6-20v (QC3.0 in increments of 200mV), and QC 4.0 I believe adds USB PB and higher output currents among other things.

this would allow you to get the exact output voltage you want. im not sure what battery contact setup is used but id imagine it to be pretty easy to prevent the charging of the cells while the AC is connected, though at the same time a lot of power banks have a wide input voltage tolerance and a good quality one will have all the overcharge/discharge protection built in. otherwise if doing a total DIY build then you can get protection circuit PCB's (been awhile but I used something like an 8205A? IC) for a LiPo very cheaply or a QC3.0/USC PD charging module will have those components build in anyway.
 

PB170

Well-known member
Thank you for your thoughts.

Since I wrote the post I have abandoned the idea of modifying the power bank's output, and instead plan on drawing power directly from the battery terminals (while still, hopefully, being able to rely on the BMS for safely charging the battery via USB).

The main problem I see before I can get started with this, is that I need the output voltage to reflect the battery’s state of charge (such that the output voltage equals the input voltage plus/minus a set amount).

Is this possible and easy to achieve with a voltage converter connected directly to the battery terminals?
 
Top