Solving The Mercury Battery Dilemma: A Quick Guide

What is the practical difference between 1.5v silver oxide and 1.35 mercury? A quarter of a stop in meter reading? As the silver oxide wear out the voltage drops anyway doesn’t it? Can’t we manually set the iso to compensate?

Silver oxide batteries tend to remain at, or close to, their rated voltage until they are about ready to expire. One moment the camera is working, the next moment it's dead. Little or no in-between.

I let Sherry Krauter recalibrate my Leica Meter MR4 to use alkaline batteries. It's a convenient expedient, but the downside is that alkalines taper off gradually in voltage output. A cris adapter would have been better.

Manually setting the ISO could be helpful, but error could creep in because the amount of correction required probably isn't linear over the range of exposures. It might be different in low light than in bright light. Still, it could be worth doing. If you always shoot outdoors and the correct exposure only varies plus or minus one f-stop, for example, I imagine that would probably work just fine.
 
Unfortunately, that's over-simplistic. All diode types have a forward voltage drop that depends on the example and on the current flow, which in a camera metering circuit alone is generally quite low. Shottky diodes, for instance can vary from about 0.15V to over 1V. P-N diodes vary less, typically 0.6-0.7V at lower currents. Germanium also varies. Net result, no diode provides a predictable voltage-drop without experiment. There is also the problem of the reduced voltage also varying because the chosen cell ages or becomes depleted.

I'm not saying variations couldn't be detected under laboratory conditions. For practical purposes, for the matter at hand, what I said is correct.

When I design a circuit, allowing 0.3 volts junction drop for a germanium diode is adequate. Yes, older silicon transistor types used to drop 0.7V; for newer silicon devices, I use 0.6V as a better approximation.
 
I've always set the film speed at half the rated value and have gotten fine results. Thus, for 400 speed film, set the meter to 200 to compensate, for example.

I would agree with this for the majority of the cases but I found out that if the light is getting more and more dim (or the opposite very bright ) I get an exposure error wider than a stop (I.e muddy shadows or washed out highlights). I think there is something non-linear in there.
 
Silver oxide batteries tend to remain at, or close to, their rated voltage until they are about ready to expire. One moment the camera is working, the next moment it's dead. Little or no in-between.

I let Sherry Krauter recalibrate my Leica Meter MR4 to use alkaline batteries. It's a convenient expedient, but the downside is that alkalines taper off gradually in voltage output. A cris adapter would have been better.
Why not use silver oxide cells in the meter? They have the same voltage as the alkalines, don't they? Or does internal resistance or some such come into play?
 
I'm not saying variations couldn't be detected under laboratory conditions. For practical purposes, for the matter at hand, what I said is correct.

When I design a circuit, allowing 0.3 volts junction drop for a germanium diode is adequate. Yes, older silicon transistor types used to drop 0.7V; for newer silicon devices, I use 0.6V as a better approximation.
Respectfully, I have to disagree. I'm not talking about laboratory conditions either. If you take the trouble to look up the If/Vf curves for examples of all three types of diodes, you'll see how much the voltage drop depends on the current. Since we're talking about trying to drop 0.15-0.2V here, this is hardly an insignificant error. It matters not for "typical" electronic circuitry but in this application it can. You can't just assume 0.3V for a germanium diode (or any other) without context.


EDIT: for an example, a BAT85 shottky diode shows about 0.15V drop at about 1mA, rising to about 0.55V at 100mA, at 25C junction temperature. This is a lot when you're worrying over 0.1V errors.
 
I haven't been able to find silver oxide batteries recently. Are they still around and if so where?

The last batch I bought from Amazon, (Duracell 303/357 or Everready 357 silver oxide) they currently are running about $1.50 each when ordered in lots of 5 or 10 batteries, from third party supplier shipped through Amazon.
 
303/357 batteries are readily available in the chain drug stores (Walgreens, CVS) if you dare to venture out. I think they're around $5-$6 for three.
 
Respectfully, I have to disagree. I'm not talking about laboratory conditions either. If you take the trouble to look up the If/Vf curves for examples of all three types of diodes, you'll see how much the voltage drop depends on the current. Since we're talking about trying to drop 0.15-0.2V here, this is hardly an insignificant error. It matters not for "typical" electronic circuitry but in this application it can. You can't just assume 0.3V for a germanium diode (or any other) without context.


EDIT: for an example, a BAT85 shottky diode shows about 0.15V drop at about 1mA, rising to about 0.55V at 100mA, at 25C junction temperature. This is a lot when you're worrying over 0.1V errors.

I am no electrical engineer, but I assume this correct.
The battery adapter I do have experience with and it works with some cameras and not others. I think it does not pass enough current. Also Leica M5 does not fit the combo.

Cheapest solution are the hearing aid cells and tape over the air holes when not in use.

Buy a nice hand meter and you will not worry.
 
those cheap "no-voltage-change" adapters have a use, you can use them to fit a 675 hearing aid battery which is dirt cheap. I've found them more physically stable than the o-ring or ring of cardboard. The 675 batteries are cheap enough that I don't mind if it only lasts a couple of weeks. I bought a couple from the UK that are stainless steel and have been happy.
 
those cheap "no-voltage-change" adapters have a use, you can use them to fit a 675 hearing aid battery which is dirt cheap. I've found them more physically stable than the o-ring or ring of cardboard. The 675 batteries are cheap enough that I don't mind if it only lasts a couple of weeks. I bought a couple from the UK that are stainless steel and have been happy.


These no-voltage-change adapters are also good for a silver oxide battery if the meter has been modified to take a 1.5v battery. This is a better solution than using one of the 625/675 alkaline batteries.

- Murray
 
yes, I was just going to add what Murray did. The battery compartment of my M5 (has been modified) fits the adapter + Silver better than it fits a 625A.
 
Another vote for the Kanto MR-9 style battery adapter - which allows the use for a modern Energizer 386/301 silver oxide. As mentioned, if the meter has been modified to take a 1.5 volt battery, then the simpler "no voltage changed" adapter works a treat.
 
Why not use silver oxide cells in the meter? They have the same voltage as the alkalines, don't they? Or does internal resistance or some such come into play?

Well, I think that using the silver oxide cells certainly is the best thing to do. They won't directly fit in the battery holders meant for the mercury batteries, which were larger. But you can just pop one into a cris adapter, it will fit just right, and the voltage will be just right, too!
 
Respectfully, I have to disagree. I'm not talking about laboratory conditions either. If you take the trouble to look up the If/Vf curves for examples of all three types of diodes, you'll see how much the voltage drop depends on the current. Since we're talking about trying to drop 0.15-0.2V here, this is hardly an insignificant error. It matters not for "typical" electronic circuitry but in this application it can. You can't just assume 0.3V for a germanium diode (or any other) without context.


EDIT: for an example, a BAT85 shottky diode shows about 0.15V drop at about 1mA, rising to about 0.55V at 100mA, at 25C junction temperature. This is a lot when you're worrying over 0.1V errors.

The problem with getting involved in voltage drop vs current curves is that for the purpose of selecting the best battery for a meter or camera, no matter what calculations we do, we are limited by what's available. You can get an alkaline battery that fits the old mercury battery holder. That's one option, but the voltage is wrong, so we could get the camera or meter recalibrated. I took that approach with my MR-4 meter. You could also use a Wein cell. Great solution: it fits and the voltage is right. Limited service life, though. Or, you can put a silver oxide cell in a cris adapter. Now you have the right voltage, it fits, and the service life is decent. As far as I know, these are the options. As a former radar field engineer, I learned to work with what was available!
 
Back
Top