9-Volt Batteries - Part 1 : The Tear Down - Battery No.1 thru No. 5

The Rocketry Forum

Help Support The Rocketry Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
After reading this thread, I recently went looking for Duracell Procell 9V batteries, 6LR61. These are generally not sold at retail, they come from the industrial tool and parts shop. I was expecting to find the Procell orange packaging, instead I found two different varieties. Procell Intense and Procell Constant. Different colors, different labels. Two seemingly different flavors marketed to different applications. Marketing being the key word here. Very important step in extracting maximum revenue from the consumer is to create "value". One way to do that is to 1) Create the [appearance] of a problem, and 2) Offer the consumer the solution.

Product data sheets for the two batteries are attached. Cynicism aside, I'm sure these are both high quality products, but for all practical purposes, but the two versions of the Procell appear to be nearly identical within testing and manufacturing tolerances other than the hip new labeling. Oddly, the Intense actually has marginally more constant current capacity than the Constant...

Remember, if you cannot compete on price or quality, that leaves fraud. Which we shall call marketing....
I thought it interesting that the highest constant current in the performance charts was 75 300mA. That’s a small fraction of what is needed to fire an Estes igniter/starter (all fire 2A) and even for an electric match (1A recommended per electricmatch.com). Of course our usages aren’t constant loads, but short ones.

That leads me back to wondering how to do a capacity test on these batteries that is applicable to our usage. I’m still trying to figure it out. But —

I wonder if I can program something like a 1.5s full on every few minutes cycle into the Emeter II’s servo testing function. Then I could use a small brushed-motor electronic speed control with an appropriate power resistor in the place of the motor and do a cyclic test that would be sort of representative…. Hmmmmmm.

I’ll have to buy a quantity of resistors because I don’t see any quick way to get just the ones I’d need for less than buying a $10 assortment on Amazon…. But that way I could simulate both an Estes igniter/starter (0.7 ohms) and an e-match (1 ohm) without burning up a bunch of them.

I will take apart one of the Voniko batteries after I run it all the way down testing it and post pictures here. If I go ahead and get the lithium-based ones I will do likewise for that one.

Added: a little high-discharge-current 2s LiPoly like the one in the post just above might well be a good choice for a small launch controller, too.

Added later: Voniko lithium 9V and pack of resistors ordered. Should be here tomorrow.
 
Last edited:
The maximum current you can get out of a battery is V/Rsum, where Rsum is the sum of the resistance of your load (the igniter) and the internal resistance of the battery (which you'll probably have to look up from the datasheets). Alkaline batteries have a relatively high internal resistance compared to LiPo's, which is why a LiPo can source so much more current. The 300 mA figure that you referred to is the battery's rated maximum CONSTANT load... obviously the 9V battery isn't designed for very high current loads (such as a motor, for example). That is different that the battery's MAXIMUM load... which will certainly be much higher, but isn't sustainable.

The standard way to test battery time vs. capacity is to put a resistive load on it equivalent to the expected load, then measure the voltage at timed intervals. For an altimeter, that might be as low as 10 mA when idle, for a tracker it might be a few hundred mA. Eggfinder GPS trackers (TX & Mini) draw about 70 mA on average, the Eggtimer WiFi devices typically average about 80 mA (Quantum/Switches/ION) and the Quasar draws about 250 mA when it's in-flight and the radio is constantly transmitting. The only 9V battery that's capable of that amount of sustained current is one of those Energizer Ultimate Lithium batteries... at $10 a pop.
 
The maximum current you can get out of a battery is V/Rsum, where Rsum is the sum of the resistance of your load (the igniter) and the internal resistance of the battery (which you'll probably have to look up from the datasheets). Alkaline batteries have a relatively high internal resistance compared to LiPo's, which is why a LiPo can source so much more current. The 300 mA figure that you referred to is the battery's rated maximum CONSTANT load... obviously the 9V battery isn't designed for very high current loads (such as a motor, for example). That is different that the battery's MAXIMUM load... which will certainly be much higher, but isn't sustainable.

The standard way to test battery time vs. capacity is to put a resistive load on it equivalent to the expected load, then measure the voltage at timed intervals. For an altimeter, that might be as low as 10 mA when idle, for a tracker it might be a few hundred mA. Eggfinder GPS trackers (TX & Mini) draw about 70 mA on average, the Eggtimer WiFi devices typically average about 80 mA (Quantum/Switches/ION) and the Quasar draws about 250 mA when it's in-flight and the radio is constantly transmitting. The only 9V battery that's capable of that amount of sustained current is one of those Energizer Ultimate Lithium batteries... at $10 a pop.
The maximum current you can get out of a battery is V/Rsc (resistance under short circuit). The internal resistance of a battery effectively varies depending on how much current you attempt to draw. The chemistry change rate is dependent on the battery temperature and the rate of discharge.

The maximum current you can get out of a battery in a circuit is as above. At a test temperature. I agree with what Chris is trying to say. :)
 
Last edited:
Those two recently-posted data sheets quote a 1 kHz AC impedance of 1692 milliohms, so ~1.7 ohms — or each cell at 282 millohms. I'm not sure what that means to a DC load.

Compared to a LiPoly or even nickel-based batteries that's a huge number. That explains their lousy performance at firing-igniter currents. Using that figure, and the quoted resistance of an Estes igniter (rounded to 0.7 ohm), with no other losses, gives an initial max current on a fresh battery (9.6V) of 3.5A. So that's a best-case, for a really short period of time.

I've tested a bunch of batteries (and motors and speed controls and props) for electric flight uses using tools I can't anymore (Windows 7-based computer interface/control software, company out of business so no hardware driver updates). I have some idea how I want to approach this for the launch controller use case I think, as noted in my previous post. I just need to spend some time with my eMeter II documentation to see how I can implement it. Test resistors and Voniko lithium 9V test victims should be here later today.
 
Last edited:
Here is a Lipo battery that I came across that safely meets the voltage and amp conditions
for the Stratologger, RRC2 and RRC3 flight computers.

View attachment 612701 View attachment 612702

The Blade battery has only seen modest test flights so far. I hope to get in some high G
testing in the late fall; Warp motors and the sorts. That rocket is nearing completion.
The picture shows the sled for that rocket; RRC2+ and RRC2L.

I have not found any other batteries like the Blade. So it's a risk that this will fade into
the sunset like many electrical components. But I do enjoy the "field testing".
Watch the protection circuit. Most just turn off the battery and any electronics will fail. Apogee event, no main :( They don't turn on automatically :( This battery is rated 4.8A. With a 1 ohm igniter,it can draw as much as 7A, or even worse with a shorted igniter.
 
In general, R/C batteries don't have a protection circuit, that's why you don't want to abuse them (like the R/C car guys do charging them at 10C or something ridiculous like that). Only the smaller 1S batteries have a protection circuit, and I can tell you with 100% confidence that a little 110 mAH 1S LiPo will easily fire a standard ematch. That's what I ship with the Apogee... and they work fine.
 
Cris beat me to it.

Blade is a brand of small RC helicopter. I’m sure that battery has no protection circuit. I found several similarly-sized 2s LiPolys on Amazon yesterday when I looked. They are out there.
 
I have my test rig assembled. More details later, but here is a first draft graph of what it might show. This is a test, using the same Voniko 9V battery I've been torturing up to now. A 1 ohm and 2 ohm power resistor wired in parallel are standing in for an Estes igniter in this rig. Power is applied to this 0.67 ohm resistor for 1.5s at 4 minute intervals. The current trace is shifted 10 seconds to the right so it is easier to see it.

Here is what the graph looks like:

Emeter log  038 Voniko ignitor test 3.png

The currents are higher than the graph I posted before, even though this is the same battery, mainly because there is a whole bunch less wiring in the circuit and all of it is much larger gage than that used by the 9V Estes launch controllers.

It is interesting how the voltage of the battery falls to as low as 2.5V yet still the current is over 3A, which is more than enough to fire an actual Estes igniter.

My inclination is to set the test to run until the battery falls below 2V when "firing" the "igniter". I also want to increase the time between "firings" to perhaps 10 minutes to make this test slightly more realistic for a launch controller.

Thoughts?

Added: to reduce the log file size for such a long test (probably hours with a lithium-based battery) I think I’ll reduce the sample rate to 4/s or maybe even 2/s. That should still catch the spikes down in voltage and up in current that are programmed to be 1.5s long. The graph above is at 8 samples per second and the resulting file was over 12,000 data points long. I should run another test with the same Voniko battery at 2 samples per second to see what the result looks like before doing a full discharge test on a fresh battery.

Also: the battery internal resistance function of the Emeter reported a value of a little over 1.7 ohms for the test plotted above. Pretty much in line with the data sheets for the two Procells posted earlier in this thread.
 
Last edited:
A few musings on this thread... opinions included.

9V Energizer Ultimate Lithium batteries are nearly $10. That's a lot for one "disposable" battery.

Not opening up your AV bay before every flight to inspect wires, terminal block screws (all you old-school guys use them, I know), battery hold-downs, etc. is like not regularly checking the oil in your car just because you bought the expensive oil. Same result... it will come back and bite you in the butt.

9V alkaline batteries aren't even being used in smoke alarms anymore... the new ones all have non-replaceable lithium batteries built into them. That should tell you something...
Locally, the Energizer 9 volt Lithium battery has really shot up in price. The best I could find for a single pack is $12.99 at Target. A lot of stores are selling the single pack for around $17 and Home Depot is selling the 2 pack for $33.87.

I wonder if using two 9 volt alkaline batteries in parallel would work and also be much cheaper. Theoretically, two in parallel should double the amps.
 
Last edited:
I'm beginning to think that I should split off my particular madness to a separate thread, but first, a couple more things, which will bring me to a teardown and so get me back to the main topic.

I ran another automated test on that same Voniko alkaline battery which did eleven simulated firings (about 10 minutes apart) of an Estes starter. At 2 samples per second, some key data — actual peak currents and minimum voltages — don't show well in the resulting plot, so when I do a full discharge test of any battery, I'll be bumping that back up to at least 4 samples/second. But after that, along with approximately another dozen simulated and real igniter firings yesterday along with the igniter firings before I created the initial graph I posted, the battery was down to the point where it couldn't deliver enough power into a Solar Starter to burn out the bridge wire. It might still be able to start an Estes motor, but it would require good installation, and patience.

So, first, here's an attempted firing of a Solar Starter where I held the launch button down for about six seconds without the bridge wire burning through and not even all the protective coating burning off. This is overlaid on and aligned with the same firing I posted earlier in the thread where the bridge wire burnt through in less than 1 1/2 seconds. That earlier one was probably the 5th or so on the battery. The later one would be more like the 35th or so.

As you can see, the current never made it to two amps on this later test.

Emeter Voniko single igniter tests.png

This also meant that the battery, being depleted, could be disassembled. No surprises inside, given the 6LR61 designation:

IMG_7356.JPEG

So, more data here, or should I start another thread?

I have two of these Voniko alkalines available, and two Voniko lithiums. Of course I have current Duracells...I suppose I should do one of those, just to show how lousy it is. Other test victims (such as the Amazon Basics alkalines or lithiums) will have to be purchased.
 
Last edited:
I'm beginning to think that I should split off my particular madness to a separate thread, but first, a couple more things, which will bring me to a teardown and so get me back to the main topic.

I ran another automated test on that same Voniko alkaline battery which did eleven simulated firings (about 10 minutes apart) of an Estes starter. At 2 samples per second, some key data — actual peak currents and minimum voltages — don't show well in the resulting plot, so when I do a full discharge test of any battery, I'll be bumping that back up to at least 4 samples/second. But after that, along with approximately another dozen simulated and real igniter firings yesterday along with the igniter firings before I created the initial graph I posted, the battery was down to the point where it couldn't deliver enough power into a Solar Starter to burn out the bridge wire. It might still be able to start an Estes motor, but it would require good installation, and patience.

So, first, here's an attempted firing of a Solar Starter where I held the launch button down for about six seconds without the bridge wire burning through and not even all the protective coating burning off. This is overlaid on and aligned with the same firing I posted earlier in the thread where the bridge wire burnt through in less than 1 1/2 seconds. That earlier one was probably the 5th or so on the battery. The later one would be more like the 35th or so.

As you can see, the current never made it to two amps on this later test.

View attachment 613122

This also meant that the battery, being depleted, could be disassembled. No surprises inside, given the 6LR61 designation:

View attachment 613123

So, more data here, or should I start another thread?

I have two of these Voniko alkalines available, and two Voniko lithiums. Of course I have current Duracells...I suppose I should do one of those, just to show how lousy it is. Other test victims (such as the Amazon Basics alkalines or lithiums) will have to be purchased.

I think it would be a good idea if you split this out to another thread.

It would keep the original thread in a manageable state for gleaming info.

And your new thread would highlight your work and findings much better.

Your approach and findings have been pretty interesting, and they deserve their own thread.
 
Thanks for the thoughts.

I just went back to your first post and it occurs to me that I could EASILY put in my test two simulated firings 15 seconds apart, then let things rest for some minutes before repeating the test. That might be a way for the proposed testing to be useful both for launch controllers AND dual deploy flight computers.

I suppose, too, if I had enough battery test victims to hand I could run my launch controller test (one 1.5s “firing” through the 0.67 ohm load every 10 minutes) and another that’s more for dual deploy (two 1s “firings” through a 1 ohm resistor separated by 15s, then a longer period — maybe 20 or 30 minutes between test cycles) for each battery type.

But it seems my launch controller test might well give enough information if done on different battery brands/types without repeating the test in a more flight computer-like fashion.

Based on what you’ve already done manually, what do you think?

Since I’ve never flown dual deploy (and probably would use LiPoly batteries if I did) I hadn’t really designed what I want to do with the dual deployment application in mind….
 
I guess if I get to doing these discharge tests I’ll have to break down and buy some of these. I don’t have any use where I need that many though. :)
You can get them in smaller quantities, just not free shipping unless you have other items in the order.
 
You can get them in smaller quantities, just not free shipping unless you have other items in the order.
Yeah, I'm not worried about that too much. I'm one of the suckers who has Prime.


I will likely run the first structured test series on another of those Voniko alkaline today.
 
Thanks for the thoughts.

I just went back to your first post and it occurs to me that I could EASILY put in my test two simulated firings 15 seconds apart, then let things rest for some minutes before repeating the test. That might be a way for the proposed testing to be useful both for launch controllers AND dual deploy flight computers.

I suppose, too, if I had enough battery test victims to hand I could run my launch controller test (one 1.5s “firing” through the 0.67 ohm load every 10 minutes) and another that’s more for dual deploy (two 1s “firings” through a 1 ohm resistor separated by 15s, then a longer period — maybe 20 or 30 minutes between test cycles) for each battery type.

But it seems my launch controller test might well give enough information if done on different battery brands/types without repeating the test in a more flight computer-like fashion.

Based on what you’ve already done manually, what do you think?

Since I’ve never flown dual deploy (and probably would use LiPoly batteries if I did) I hadn’t really designed what I want to do with the dual deployment application in mind….

Sorry about the long hiatus.

I read your post a couple of times, so I have to admit I'm not sure what you're asking,
but I'll take a stab at a reply.

I think you're onto a good project for launch controller testing. What would be nice to see
is how well the batteries hold up under a "typical" launch cycle at a club's launch day.

I wouldn't worry about the dual deployment application. I think it's covered fairly well (but not perfect)
in what I have posted to date. An important part of my experimenting was to highlight the different
battery constructions, and how that related to amps & volts. And to also help people easily identify
which batteries had which type of cell construction.
 
Sorry about the long hiatus.

I read your post a couple of times, so I have to admit I'm not sure what you're asking,
but I'll take a stab at a reply.

I think you're onto a good project for launch controller testing. What would be nice to see
is how well the batteries hold up under a "typical" launch cycle at a club's launch day.

I wouldn't worry about the dual deployment application. I think it's covered fairly well (but not perfect)
in what I have posted to date. An important part of my experimenting was to highlight the different
battery constructions, and how that related to amps & volts. And to also help people easily identify
which batteries had which type of cell construction.
Yeah, I tend to use too many words.

I will proceed on the launch controller-focused test. I had something go wonky on my second automated run of the fresh Voniko battery so I think I'm going to try to run another and pay a little more attention while the test is running. I'm also transitioning to a new Mac, which is where I use MagicPlot to make the graphs. Otherwise I'd have started my new thread.

I also should have an actual Aerotech Phaser soon, as it's on the way, and I can then use it rather than the Astron II (or R2D2) controller for the actual igniter firing parts of the tests.

After I posted the above I figured that using a 0.67 ohm load would work as a proxy for e-matches as well, so hopefully what I do will be complementary to what you've done and not just duplicative.

Thanks.
 
Dollar Tree
Not much holding the cells together..

Tony
'Super heavy duty' batteries are the old carbon zinc. Even worse performance than the stacked alkaline batteries. They will work OK in very low drain devices like smoke detectors.
 
'Super heavy duty' batteries are the old carbon zinc. Even worse performance than the stacked alkaline batteries. They will work OK in very low drain devices like smoke detectors.
PLEASE DO NOT USE THESE BATTERIES IN SMOKE DETECTORS. USE THE RECOMMENDED ONE OR A BETTER EQUIVALENT.
YOUR LIFE DEPENDS ON IT SITTING THERE DOING NOTHING, BUT WORKING ON THE DAY YOU NEED IT. A LOW RELIABILITY BATTERY IS NOT THE ONE FOR THAT TASK.
 
PLEASE DO NOT USE THESE BATTERIES IN SMOKE DETECTORS. USE THE RECOMMENDED ONE OR A BETTER EQUIVALENT.
YOUR LIFE DEPENDS ON IT SITTING THERE DOING NOTHING, BUT WORKING ON THE DAY YOU NEED IT. A LOW RELIABILITY BATTERY IS NOT THE ONE FOR THAT TASK.

We get to test your statement.

It just so happens I bought a 3-pack of new detectors just last week. I'm in the process
of replacing and adding some in my home. The detectors included the 9-volt batteries.
So here are pictures of the detector and the battery. The battery is a 6F22.

SmokeDetectorBattery-01.jpg SmokeDetectorBattery-02.jpg SmokeDetectorBattery-03.jpg


Here are pictures of the 6F22 batteries I tested and tore apart.

IMG_6922.JPG IMG_6975.JPG

The 6F22 batteries maintained higher voltages than all the other batteries (of better construction)
throughout my testing. But the 6F22's had the poorest discharge Amp levels, barely 1.0 amp.

Not defending the Smoke Detector manufacturer, but it would seem if the batteries were inadequate
to start - they would have been open to a whole lot of litigations by now. And probably not met
code compliances.


SMI100_c7-6.jpg

The above is a picture from the First Alert website. You can buy their detectors at all the big box stores.
The picture is not the best, but on the website you can zoom in on the battery and its the exact same
battery in my units.
 
Last edited:
PLEASE DO NOT USE THESE BATTERIES IN SMOKE DETECTORS. USE THE RECOMMENDED ONE OR A BETTER EQUIVALENT.
YOUR LIFE DEPENDS ON IT SITTING THERE DOING NOTHING, BUT WORKING ON THE DAY YOU NEED IT. A LOW RELIABILITY BATTERY IS NOT THE ONE FOR THAT TASK.
Are they really unreliable in that use case, though, when they're just sitting there never being touched for years?
 
Are they really unreliable in that use case, though, when they're just sitting there never being touched for years?
The reliability has to do with working as expected when needed, not just sitting around doing nothing. So the reliability relates to working when there's smoke, which may not happen for many months...or at all. But you still want to know that if there was smoke, the battery would work.

All else being equal, heavy duty batteries doesn't have the performance of alkaline. Therefore, for mission/life critical applications, it's a better idea to use alkaline batteries and not heavy duty. Not sure why there are some people here trying to defend the use of heavy duty batteries over alkaline...
 
The reliability has to do with working as expected when needed, not just sitting around doing nothing. So the reliability relates to working when there's smoke, which may not happen for many months...or at all. But you still want to know that if there was smoke, the battery would work.
What about the construction of heavy duty batteries makes them less reliable in the smoke detector use case? @QFactor's testing showed particular issues with rocketry use cases, but those obviously aren't the same. The construction that could fail in the rocketry use case seems much less likely to fail in the smoke detector use case. Perhaps there are other flaws that you could explain.
All else being equal, heavy duty batteries doesn't have the performance of alkaline.
In what way do they have worse performance? Does that matter to a smoke detector?

I'm not saying you're wrong, but you're not explaining why "better" is necessary for the smoke detector use case.
 
What about the construction of heavy duty batteries makes them less reliable in the smoke detector use case? @QFactor's testing showed particular issues with rocketry use cases, but those obviously aren't the same. The construction that could fail in the rocketry use case seems much less likely to fail in the smoke detector use case. Perhaps there are other flaws that you could explain.

In what way do they have worse performance? Does that matter to a smoke detector?

I'm not saying you're wrong, but you're not explaining why "better" is necessary for the smoke detector use case.

Alkalines are better than heavy duty (or super heavy duty) for several reasons. First, they have more power (by power, I mean more capacity and the ability to delivery higher currents), especially in high drain devices. I know a smoke detector isn't a high-drain device, although I'd still take a battery with more capacity over one with less.

Second, alkalines have a longer shelf life by several years. This shouldn't matter with smoke detectors, but knowing how practically all consumer batteries start to lose performance the moment they leave the production factory, it's better to use a battery with more of a cushion in terms of how large of a window yo have to get full life out of the battery.

Yes, I acknowledge that heavy duty batteries may be more cost effective that alkalines depending on w/e metric you want to use. But for important applications like smoke detectors (where you're in the "business" of installing them in your home, not saving a few cents producing them), I don't see any reason to use a super heavy duty over an alkaline. But you do you do you.

I've also heard that heavy duty chemistry are less likely to leak. Not sure if it's true, but for the 9V heavy duty lovers out there, even if it were true, you should be at least checking your smoke detector's battery every 6-12 months. Any leakage that occurs in that time frame should be small or easily cleaned up. I'll take the off chance of a small battery leak in a smoke detector in return for longer shelf life, higher capacity and better high-drain performance.
 
What about the construction of heavy duty batteries makes them less reliable in the smoke detector use case? @QFactor's testing showed particular issues with rocketry use cases, but those obviously aren't the same. The construction that could fail in the rocketry use case seems much less likely to fail in the smoke detector use case. Perhaps there are other flaws that you could explain.

In what way do they have worse performance? Does that matter to a smoke detector?

I'm not saying you're wrong, but you're not explaining why "better" is necessary for the smoke detector use case.
Over the years the current drain for smoke detectors has become better. If the manufacturer has supplied the battery with it and it doesn't say in the instructions that the battery supplied is for testing only and should be replaced with a fresh one on installation, then fine for that one. But check what YOUR smoke detector says IS the correct battery.ó
You can now get ones with a 10 year sealed for life of detector battery. So you never have to replace the battery for the life of the detector. 3M make them. Anyway, use the recommended battery or better. Don't use heavy-duty zinc carbon unless it says you can.
But to answer your other question. Zinc carbon batteries have a shorter shelf life than Alkaline so make sure it's fresh. That's trickier

Or just buy the 10-year, sealed for life detectors, and stop worrying for 10 years
.
 
Last edited:
Zinc-carbon batteries' self-discharge rate would seem to me to be a contra-indication for their use in anything over a period of months or longer, even if they can provide the necessary current (certainly no more than a milliamp or two in a smoke detector EXCEPT when it actually has to sound a warning).

But a 6F22/6LP3146 type alkaline should be OK — but as @OzHybrid notes, use what the smoke detector maker says to use.

My report, inspired by this thread, but which also includes tests of two different types of lithium-based batteries, will be presented at vNARCON at approximately 2:30 PST tomorrow.
 
Back
Top