Powering Music Effects Pedals: NiMH and Lithium-ion Rechargable Batteries Compared

24-Jul-2021

Almost all compact portable effects pedals require "9 volt power." The two basic ways are via an external 9v power supply or a battery. There are various logistic, operational, ecological, and signal interference issues which favor one or the other, but for this rant we are focusing specifically on rechargeable 9v form factor batteries, i.e. not clever series-assembly of high-mAh output 1.5v batteries with a 9v connector (battery simulator or power supply barrel plug) that would have to live outside the unit.

Because as we will see: All 9v rechargeables are NOT the same.

Important facts and observations to know before we begin

"9v" often does not mean "9 volts nominal from terminals"
The term "9v battery" commonly refers to the form factor of the battery, a size-standard rectangular unit with standardized terminals on one of the short sides, not strict 9 volts output from the terminals. Rechargeable batteries come in different nominal output voltages depending on their chemistry:
1. 9.6v for nickel metal-hydride or NiMH. This is created by using 8 X 1.2v NiMH "cells" in the package.
2. 7.2v for lithium ion or Li-ion. This is created by using 2 x 3.6v cells in the package. (but see below re. topping off)
There are other chemistries and voltages available but these are clearly not in the proper range for driving a "regular 9v" device. Note that fresh disposable ("regular") alkaline batteries show 9.6 to 9.7 volts, not 9.0. Also, as we will see below, freshly recharged batteries are capable of providing significantly higher voltages.

Not all devices drain the battery at the same rate
A critical figure to know about your pedal is the current drain, typically expressed in milliamps, or mA. Besides voltage, batteries are rated in milliamp hours or mAh which is roughly -- for our purposes -- the maximum amount of current the battery can supply in one hour before it gives out, i.e. there are no more mA to give. In general, if the current drain is less, then the battery lasts proportionally longer. For example, a 250 mA battery can supply 50 mA for roughly 5 hours. Accordingly, the higher the mAh rating, the longer the battery can last. In general, modern digital pedals especially delays can consume 45-50 mA whereas analog compressors draw only 20 mA and older analog distortion boxes only 10 mA.
BUT: acceptable voltage will fall off over time, so some devices (like pedals) will stop working before the battery is completely dead.
Not all batteries maintain voltage evenly while draining
The ideal 9v battery would source 9 volts completely level until the last mA was delivered and then abruptly just die, but that is not how discharge physics works. All batteries "top off" with voltage greater than what the nominal cell voltage should be, and all will lose voltage "pressure" as they are delivering mA over time but some drop faster than others.

"Undervolting" a device will affect performance -- if it works at all
This should be obvious but a dead battery from a pedal point of view is not 0v, 0 mAh remaining. "Dead" batteries routinely show 4-6 volts but that is not enough to drive the circuitry -- especially digital pedals.

"Overvolting" a device a little bit will not hurt it -- but beware
Fresh alkaline disposable 9v batteries regularly deliver 9.6v or 9.7v and we have all used them forever without problems so clearly the input circuitry can take it. But what about 10v? 11v? Read on!

The Actual Experiments

All voltages are measured with Sperry DM-4400A multimeter and unless specified otherwise are with battery in operational unit with readings taken at the connector terminals.

First Run with Fresh NiMH Precharged Battery

I purchased 2 Powerex 230 mAh NiMH 9.6v precharged batteries and the Powerex MH-C490F Stealth 4-dock charger. I bought the precharged batteries because uncharged batteries from my usual sources were curiously out of stock. Out of box, the batteries showed 10.24v! Surprised as I was, I thought the extra 0.6v over a disposable alkaline might be a good thing and put it into my Boss PS-6 Harmonist pedal, which has a hefty 45 mA drain. This is the drain profile:

Elapsed Time	Voltage
0 (start)	10.24v
60 mins	10.08v
220 mins	8.91v
280 mins	5.93v (unit no longer functioning)

As we can see, we only got 4.7 hours out of the battery but perhaps that is OK for a jam. Besides the 10.24v starting voltage, not really a surprise. 230 mAh is not a huge amount of capacity. Also, rechargable batteries need to be drained and charged ("cycled") 5 or 6 times to get them to maximum mAh capacity.

Recharge -- Then...

The battery was charged overnight and the charger automatically settled into trickle mode. Fresh out of the charger, a whopping 11.55v was recorded! That is heading into 12v territory -- and I would rarely overvolt a 9v device with a 12v battery.
But I enthusiastically (recklessly...?) moved on and installed the battery this time in my Boss DD-3T Digital Delay, which also has a hefty 45 mA drain...

And it did not work. The check LED did not illuminate!
Horrified that I might have fried the unit, I performed series of tests:

Put a regular disposable 9v back into the device ... and it seemed to work OK. Whew!
Operating under the assumption that burned electronics do not un-burn, I put the 11.55v NiMH back in a second time and again the device did not work.
I repeated this sequence again and it seems that the circuitry was (temporarily...?) unhappy with 11.55v but could "ignore it". At least for the quick test, I did not blow up the DD-3T with the "hot" 11.55v NiMH battery.

I then regrouped and assumed that borderline overvolting of analog circuits could be tolerated and tried the 11.55v on my old analog Boss DF-2 Distortion and that did work. I also tried it on an analog Boss CS-3 Compressor/Sustainer and that worked too.
Feeling bold, I tried it on the digital PS-6 and that also worked. Bottom line: It's not just about digital or analog or even current drain. I would have thought the digital delay and the harmonizer to be "equivalent" from a battery perspective but this experiment proved otherwise.

Deliberate Drain To Find Overvolt Threshold

So 10.24v (the voltage of the precharged battery fresh out of the box) seemed to work for the DD-3T but the fully recharged 11.55v was too hot. What was the threshold? To determine this, I let the PS-6 with its 45 mA drain eat up the battery and periodically swapped the battery into the DD-3T to assess proper function. Remarkably, after only 8 minutes in the PS-6, the voltage was already down to 10.80v and this worked! In the 10-something volt range, the voltage-time spent at 45 mA drain started to nicely flatten out.

Adding Lithium-ion to the mix

I started with NiMH because of their higher voltage than Li-ion which I believed would be required to properly drive digital circuits. After doing the casual NiMH run on the digital pedals, I observed that they continued to work well to 6.8v, perhaps even a little lower. Perhaps well-charged Li-ion batteries could stay above that threshold for some time? And some of the Li-ion batteries have much more mAh capacity. I purchased 2 EBL 600 mAh Li-ion 8.4v (full charge) batteries and systematically tested both batteries in a set of pedals.

Results of The Detailed Experiments

The Pedals and Summary Information:

					From manufacturer spec sheets			Test Results
Maker	Model	Description	Manu. Year	A/D	Current Drain	Carbon TTL	Alkaline TTL	NiMH TTL	Li-ion TTL
Boss	DD-3T	Digital Delay	2021	D	45 mA	3 hr	11.5 hr	5.75 hr	16 hr
Boss	PS-6	Harmonist (vari-pitch octavizer)	2021	D	45 mA	2 hr	8 hr	6 hr	16 hr
Boss	CS-3	Compressor/Sustainer	~1990	A	20 mA	10.5 hr	26 hr	>18 hr	>24 hr
DOD	FX50	Overdrive Preamp	~1984	A	<10 mA	n/a	n/a	>24 hr	>24 hr

Let us get the easy stuff out of the way:

Even for medium-high drain digital pedals, the 600 mAh Li-ion battery is the clear "winner", supplying current at enough voltage (over 6.5v) approx. 2.5X longer than the 230 mAh NiMH cell. Although 600/230 is about 2.5X, the mAh capacity alone is not enough to determine the time-to-live. The Li-ion cell shows excellent (meaning shallow) drop-off in voltage over time.
Check with your manufacturer on the maximum acceptable input voltage! I am lucky I did not damage my pedals but I suspect I came close. Conduct your own tests at your own risk.

Notes and Observations:

The line color is meant to approximate the color of the effect pedal. Apols but the DD-3T is pearl white and the CS-3 and the DS-6 are both blue. :-)
The gaps in the lines are there to draw attention to the fact that the x-axis intervals change. The first four observations are done very close together to gauge the amount of initial "overcharge draw-down", i.e. how fast the topped-off battery will settle into a more steady voltage delivering the current draw demanded. The last line grouping is measured only once every hour.
The almost 40 year old, low-drain analog FX50 distortion pedal, with a super-hot 11.87v non-settled fresh recharged NiMH, basically ran until I got bored after 24 hours and stopped the measurements.
I do not include a Li-ion test on the analog pedals because at 600 mAh they will clearly last longer than 24 hours.
Bonus: The Li-ion cell at 1oz weighs half that of the NiMH cell (2oz). Also, the EBL Li-ion cells are just a wee bit trimmer than the Powerex NiMH cells and get less wedged into the battery compartment.
The CS-3 and PS-6 NiMH tests were launched with cells settled overnight to 11.24v.
Due to the overvolting issue described earlier, the DD-3T NiMH test had to be launched at 10.8v.
All units were activated (not in bypass mode) and left "idle" with no input signal. It is unclear how bypass toggling and actual live signal processing might effect power consumption.
The CS-3 claims 26 hr life with akaline and although measurements stopped at 18 hrs, it probably could run another 4 at least.
The digital pedals seem to cease operation at about 6.5 volts. I need a variable power supply to easily determine minimum operational voltage (wink: stay tuned).
A Li-ion test on the PS-6 was performed that yielded results almost identical to the DD-3T so rather than clutter the graph it was not included.

Like this? Dislike this? Let me know