Battery School

42 posts / 0 new
Last post
Battery School

So, anyway, I'm over here in my lab thinking about circuits and wonder about the inherent impedance of a 9 Volt battery.

To speak in greater detail, I'm wondering whether the impedance of the battery is low enough to consider it a useful path to signal ground in an audio circuit.

Otherwise I could add a bypass capacitor to bring signal ground up past the battery and viola.

Oh I know, let's check Wikipedia.

Checking Wikipedia for 9 Volt Battery

The Thought Process

The link gave a link for data on a 9 Volt Alkaline and here's what it says:



620 PULSE.

Um, I used the term "inherent impedance" but...

I tend to think a "LOAD" is the circuit the battery is supplying voltage to, whereas the impedance of the battery itself would be a "source" impedance, which may equate with "inherent" impedance.

But wait. We have an alternate phantom power source and... I strongly suspect that a bypass capacitor would be recommended and since such a capacitor is probably advisable also with the battery, I think we've reached a decision.

That's it for the day. No more thinking until tomorrow.

Carl Blare


Is the opposition to the flow of alternating current. How can a DC device like a battery have an impedance?

Druid Hills Radio AM-1710- Dade City, FL. Unlicensed operation authorized by the Part 15 Department of the FCC and our Resident Hobby Agent.  

What I Think Absent the Facts

In an audio circuit that is battery powered there is AC present in the form of the audio signal which is referenced to ground but the AC impedance of the battery will raise the AC component above signal ground by the internal AC resistance of the battery unless, as I'm here pre-supposing, a bypass electrolytic is used to bypass the battery and raise signal ground to the AC audio.

Everything has an internal impedance. Even our heads, which also function on DC with brain waves being AC.

DC is AC standing still.

Carl Blare

Does This Count?

My question pertains to 9 Volt batteries but this reading from the Instruction Manual for my Smart AA Battery Charger suggests that battery resistance is a thing:

"The charger will analyze the dynamic internal resistance  by applying a load current and this current reading is referred to the voltage drop detected on the battery... etc."

It seems to me both the DC and AC in a circuit need to complete a full circle (circuit) for a device to operate at optimum.

Carl Blare

DC is AC standing still

hmmmmm.... a new way to describe DC.




Classic Symptom

There is a classic symptom of the failure of the bypass capacitor across the battery in battery powered radios called motorboating.  This is a low frequency oscillation which sounds like "putt...putt...putt.." hence the nickname.

What happens is when the bypass capacitor becomes weak the internal resistance of the battery allows the supply voltage to drop when the speaker is drawing current and this drop causes a decrease in the signal to the speaker which causes the supply voltage to rise which increases the signal to the speaker which causes the supply voltage to drop and so on. (was this as much fun to read as it was to type?)  A good bypass capacitor will stabilize the supply voltage and prevent this oscillation.  For this reason it is good practice to place a bypass capacitor across a battery supply.

As energy is extracted from a battery the internal resistance increases so this effect is most pronounced with a weak battery.  A weak battery can have a normal terminal voltage under no load but still not be suitable for use.  Batteries need to be tested under load for this reason.  If, when under load, the terminal voltage is low then the battery needs replacement.



More Useful Engineering Data Shared

Thank you Neil for filling in with knowledge my technical books don't mention.

Your opening sentence in paragraph 2 reads:  "What happens is when the bypass capacitor becomes weak the internal resistance of the battery allows the supply voltage to drop..."

Did you mean to say "when the bypass capacitor becomes weak" or "when the battery becomes weak"?

If the former, what constitutes a weak bypass capacitor?

Carl Blare


I was referring to the capicator becoming "weak" meaning the capacitance is reduced.  This happens with age and temperature as the electrolyte dries out and the capacitance decreases and the ESR (internal resistance) increases so the cap is no longer effective in dampening the quick changes in voltage.



Choosing a Value

Thanks for the clarification.

I'm using 100 uF for the bypass but in other circuits in books a variety of values are used for the same purpose.

How is the value chosen?

Carl Blare

Choosing a Capacitor Value

There are equations which can be used to select a value based on the lowest freequency to suppress, the AC current expected, and the allowable voltage changes but for a one time prototype a rule of thumb I use is 1 uF for each mA of current supplied by the battery.

It would seem that bigger is better but the larger the capacitance generally the larger the ESR so it is a compromise.



Credit Payment

From the "Give Credit Where Credit Is Due" department:

Neil, you are a better educator on these topics than this huge stack of books I have here which mainly put weight on the floorboards.

I think I can do all the things you suggested!

Feeling smarter isn't that frequent a thing.

Carl Blare

Making More Sense Every Minute

Neil advises:  "For a one time prototype a rule of thumb I use is 1 uF for each mA of current supplied by the battery."

My (2) electret microphone capsules are each rated at 1.0 mA, which easily works out to 2 uF for the bypass capacitor.

We have been talking about bypassing the battery, but there is also the matter of placing a bypass cap at the source of a FET (in some cases)... unless you say otherwise I will suppose that the same rule applies.

Carl Blare


Ever notice some circuits will parallel a large electrolytic with a small value like .1 uf or less?  

Electrolytics have an inductance which at higher frequencies raises the impedance.  Think about how they're made.  Two long sheets of foil rolled into coils separated by insulation.

As such the reactance of the coil offsets the capacitive reactance.

Suppose an.electrolytic can have a resonant frequency like a coil?

by MRAM 1500 

Charter Member - Association of Low Power Broadcasters

Chairman - ALPB

Reply to #13 Source Bypass

I am not sure of your circuit but generally a source bypass cap is used in a common source amplifier where the signal is taken from the drain.  In this case the cap functions to raise the AC gain of the amplifier and the value is based on the lower cutoff frequency desired.  For a FET this is C = 1/(2*pi*R*f) where R is the source resistance.  For a bipolar transistor it is a bit more complicated since the resistances seen by the base appear at the emitter and have to be accounted.

If your signal is taken from the source (a source follower) then no bypass cap should be used.



I'm enjoying this TOO, and also -

Carl - you should tell people more about

your "Smart AA Charger" if you haven't

done so already.

Brooce, Hartford


Will Do

Hi Brooce... I have posted about the Smart Charger and talked about it on the show, but no longer remember where or what.

I will talk about it again soon... it's very interesting stuff.

It has made charging batteries fun and I keep it running so my portables always have fully charged power.

Carl Blare


It's a unique charger and it's

much better for the batteries.



Questionable Result

Neil, in you Reply # 15 you give the formula for determining the capacitor value to bypass the source resistor.

The numbers I used in the equation include:

R = 6.8k

lowest frequency = 80 Hz

The result I get is .0000002

What value do you come up with running the same numbers?

Carl Blare


I figured 0.29 uF but be aware that the equation was for bypassing a source resistor in a common source amplifier where the signal is taken from the drain.  From what you said dirung the ALPB TeamSpeak meeting this is not the case for your circuit.

To calculate the needed C the resistance "seen" by the capacitor needs to be used.  For a source follower used as I think you have this would include the source resistor, the transformer resistance, the a.c. resistance seen looking into the source of the FET, and the transformed a.c. resistance ol the load.  This involves drawing the a.c. equivalent circuit of your circuit and calculating the R to use.

It is not as complicated as it appears and if you will provide your circuit diagram I can explain this.




Oh, I follow that.

Amazing that so much is going on in such an otherwise small circuit.

Your recollection from the meeting description of the diagram is correct.

Today, after I make my rounds and tend to a few things I'll scan and post the circuit for Alpmic 16.3c.

Carl Blare

As the Sun Shines

Hey Hi Everybody!

PREFACE - The last hour was spent poking at the wierd software supplied with my scanner.

It is documented in a way that causes me to prefer a coma to being conscious.

It tells me I'm building a PDF file when all I want is a JPEG.

If I am persistent I end up with a JPEG but I can't grasp the logic so it never gets easier.

By this time you should admire me for the effort.

ON TO BUSINESS - We are building a microphone based around no-longer-available Radio Shack discontinued electret mic capsules. It's called the Alpmic and an earlier version of it is documented at the ALPB Website.

Neil Radio8Z has been educating me on aspects of circuit design and this is a higher learning experience.


FRUSTRATION ALERT: Repeated attempts to link the Circuit Diagram have belligerently failed. I won't put up with it and am driving away at a high rate of speed to show my contempt for the internet's failure to show even a hint of artificial intelligence. So long suckers!!

Carl Blare

Recycled Attempt

Posting JPEG Not Be Easy


Carl Blare

Can You See It?

The Circuit Diagram May Be Seeable

The following is circuit data:

B! = standard 9 Volt battery

C1 = 2 uF based on the 2-mic capsules pulling 1 mA apiece

C2 = As yet undecided

M1, M2 = Discontinued Radio Shack off-the-rack electret capsules

R1, R2 = 909 Ohms

R3 = the resistor that sets the voltage seen by the mic capsules... the data sheet tells us it will accept 1 to 10 VDC and the Optimum is 4.5 VDC. Our circuit provides 3.69 VDC;

___ discussion: The output destination, the "load", is 2K (RDL RU-MX4 Mixer); According to Neumann Microphone Corporation White Paper source impedance should be 1/5 of load impedance in condenser mic circuits. This circuit does not achieve that ideal but uses available parts;

T1 = is a transformer bought for a few cents at a surplus store and is otherwise unknown. It has worked in previous tests as a viable mic transformer;

Present subjective impression of how it sounds... not too bad, acceptable, maybe pretty good.

Any questions should be addressed to me.

Carl Blare

Smarts Fade Away

In the 1990s my wife and I knew everything about print and video, including the fact that JPEG was a color format and some other thing was better for black and white, possibly BMP.

Now I only know that I don't know.

Carl Blare

Reply to #23 and #24

Without the value of R3 and the d.c. resistances of the transformer windings and the turns or impedance ratio the calculation for C2 given the cutoff frequency cannot be done.

Probably the best way to proceed is to try different values for C2 until you get the sound you want.  Bear in mind that a change in the load impedance will affect the cutoff frequency for this circuit.



The Finishing Stretch

Many thanks Neil for contributing to the microhone project.

Given the unknown status of the transformer we move to the other option you recommend... comparing the result of various capacitors for C2 until we hear the "best" result.

If curiosity arises we have a program utilizing this version (v16.3c) of the Alpmic so give it a listen unless you don't.

Blare OnAir July 4th in the TeamSpeak Open Room

Carl Blare

Today's Progress

Changed R3 to 6.6k to bring up the voltage seen by the mic capsules to 4.169 V.

Changed C2 (by the way, it was initially 1 uF as a starting point in the previous version) to 10 uF, the next closest value available on hand.

Made a recording

Alpmic V 16.3d Blare OnAir Lite

Carl Blare

Further Question

Neil said:  "Probably the best way to proceed is to try different values for C2 until you get the sound you want.  Bear in mind that a change in the load impedance will affect the cutoff frequency for this circuit."

To make sure we are talking about the same thing...

The "cutoff frequency" is the low end of the frequency spectrum, right?

Since in any case the cutoff frequency will be below voice range I am wondering how to "listen" for a difference without the presence of very low frequency audio.

I do not have any musical instruments or other low frequency devices.

In fact I'm thinking that whether the frequency rolls off at 80 Hz or 40 Hz doesn't make any difference for a voice mic.

Carl Blare

Cutoff Frequency

Yes, I was referring to the the low frequency cutoff.

Sorry to gum up the works further but it should be understood what is meant by cutoff frequency for a network. By definition, a cutoff frequency is one which results in a drop of 3 dB from the peak or "mid-band" gain.  At frequencies where C2 has a low impedance compared to R3 and the other circuit components the gain will be maximum.  As the frequency is lowered, the impedance of C2 rises and the gain is now mostly affected by the other circuit components.  Unlike conventional low pass filters where the gain continues to reduce with lowering frequency, this circuit, excepting the transformer, will not do so.  A point will be reached where as the frequency continues to be lowered the gain will remain fixed.  This is further complicated by the unknown characteristics of the transformer.  Transformers have a lower frequency limit where predictable operation ceases which in this case means that there will be a frequency below which little signal gets through the transformer.  It could be that the low frequency rolloff of the transformer dominates the effect of C2 in which case changing C2 may have no noticable effect.

If the system sounds good to you then perhaps the best thing is to leave it be and use it.



Onward Toward Future Victory

Neil:  "If the system sounds good to you then perhaps the best thing is to leave it be and use it."

That would have been possible with past versions, which today sound like what I want, but which (for now) are lost in the maze of notes packed into an over-stuffed folder never collated.

These "past versions" are audible within hundreds of Blare OnAir episodes in which I too seldom made note of the microphone in use. Forensics will weed out some clues.

Of these two examples recently linked, with C2 = 1uF compared to C2= 10 uF, the first one is better by comparison... the 10 uF version seems too "thick" or badly hollow to use imprecise language.

This circuit arrangement is very much adored by my artist's sense, but the transformer isn't coupling properly... speaking from intuition.

It's a joy god damn it working on these projects and I may do this until Trump sounds the Trumpets of the End Times.

I'm pre-raptured so it's no big deal.

Carl Blare


Log in or register to post comments