Why a Batteryless Soil pH Meter Measures Salt

This page contains affiliate links. If you buy through one we may earn a commission, at no cost to you.

The question skips a step, and the skipped step is the whole problem. Before you can lower a soil pH you have to know what it is, and the instrument almost everybody reaches for cannot tell you. Not roughly. Not as a trend. It is not measuring pH at all, and the clue is printed on it as a selling point: no batteries required.

That sounds like an accusation. It is arithmetic, and it comes out of two numbers that have been known for a century.

What a pH measurement is, and why it cannot be passive

Measuring pH is a potentiometric measurement. A thin glass membrane develops an electrical potential across itself that depends on the hydrogen ion activity on each side, and that potential follows the Nernst relation: 2.303 times the gas constant times absolute temperature, divided by the Faraday constant. At 25 degrees Celsius that is 59.17 millivolts per pH unit.

Soil sits in a narrow band. From pH 5 to pH 8 covers nearly everything a gardener will meet, and three pH units at 59.17 millivolts each is 177.5 millivolts in total. That is the entire signal, end to end.

Two things are needed to read it, and both are the obstacle.

First, a reference half-cell. A single electrode has no voltage; voltage is a difference. So a pH meter carries a second electrode of known, fixed potential to measure against, usually a silver chloride cell in a salt bridge. Two bare metal prongs pushed into soil are not that. They are two unknowns.

Second, an amplifier. A glass membrane thin enough to respond to hydrogen ions has an internal resistance somewhere between ten million and a billion ohms. To read a voltage across a source that stiff without collapsing it, the meter needs an input impedance a hundred to a thousand times higher again, which is a billion to a trillion ohms. There is no passive component that does that. It is an active circuit, and an active circuit needs a power source.

So a meter whose headline feature is that it has no battery has told you, on the front of the packet, that it contains no pH measurement.

What the needle is actually responding to

A three in one houseplant health meter for moisture, light and pH publishes a single elongated probe, a three position toggle, no batteries to buy or replace, indoor or outdoor use, and a note that it performs best in loose aerated potting mixes. It also publishes, in the same paragraph as the three readings, the sentence that settles the whole question: "A dual-sensor design drives the readings."

Two sensors. Three dials. One of the three is reading a sensor that belongs to another one.

Here is the mechanism. Two dissimilar metals in a moist electrolyte are a galvanic cell. The device is the battery, which is precisely why it does not need one. Take a copper and zinc pair as an illustration: copper's standard potential is plus 0.3419 volts and zinc's is minus 0.7618, so the cell delivers 1.104 volts. That is the needle's driving voltage, and it is a property of the metals, not of the soil.

What the soil contributes is resistance. The current through the galvanometer is 1.104 volts divided by the soil resistance plus the meter's own, and soil resistivity runs from roughly 10 ohm metres when wet and salty to around 1,000 when dry and sandy. That is a factor of 100, and it is the full sweep of the needle.

Now set the two sensitivities against each other. Moisture and dissolved salt move the reading across two orders of magnitude. The entire pH range, if a pH sensitive junction existed in there at all, could move the driving voltage by 177.5 millivolts out of 1,104, which is 16.1 percent. Divide: the instrument is roughly 620 times more responsive to how wet and salty the soil is than to the quantity a glass electrode would be reading.

Which is exactly why the same two prongs are also sold as the moisture meter. They are not two instruments sharing a body. The pH dial is the moisture measurement on a different scale.

And the light reading, to be fair to the product, is the one that legitimately works with no battery, because a photovoltaic cell generates its own current from the light falling on it. So the "dual sensor design" is a galvanic prong pair and a photocell, both honest, and the third scale is the first one relabelled.

The record's own caveats are the mechanism confessing

That listing is more candid than most, and once you know what is going on, its drawbacks read like a description of a conductivity meter.

It says readings "show trends, not exact lab values", and that "basic probe meters give useful trends rather than lab-grade numbers". On moisture, that is simply true and useful. On pH, the trend it shows is a trend in wetness.

It says "Dense or compacted mixes can throw readings off", and that the probe "performs best in the loose, aerated soils typical of potted houseplants". That is the conductivity mechanism stating itself. Compaction changes how much soil is pressed against the metal, which changes the contact resistance, which moves the needle. A glass electrode does not care how tightly anything is packed around it, because it is reading a potential and not passing a current.

What the probe is genuinely good for is the thing the question underneath this one is usually really asking: whether to water yet. On that it is a real instrument, and far better than a finger in the top inch, because it reads at the depth the roots are at. That is the subject of how to know when to water your houseplants, and the full record is in our review of the three in one houseplant meter.

The trap: feeding the plant moves the pH needle

This is where the arithmetic becomes practical rather than pedantic, and it bites exactly the person asking how to change a pH.

A bottle of fiddle leaf fig plant food, an 8 ounce 3-1-2 liquid fertilizer publishes a 3-1-2 nitrogen, phosphorus and potassium ratio, three forms of nitrogen, and a dilution: one teaspoon per two cups of water, usable at every watering.

One teaspoon is 4.93 millilitres and two cups is 473.2, so that is a 1 in 96 dilution, 1.042 percent by volume. And a liquid fertiliser concentrate is a salt solution: nitrates, phosphates and potassium salts, all dissolved, all conductive.

So every feeding raises the medium's electrical conductivity. Which raises the current through the prong meter. Which moves the pH needle, with no change in acidity required at all. Feed the plant and re-test, and the dial will have shifted, and you will conclude that the feed changed the pH.

How far it shifts cannot be computed, and that is worth being explicit about: the record publishes an NPK ratio and a dilution, and no concentration. A 3-1-2 is three parts to one part to two parts of something, and the listing does not say three percent of what. The direction is certain and the magnitude is not. Details in our review of the 3-1-2 liquid fertilizer.

The instrument in this catalogue that honestly measures what the prongs measure

There is one, and it is in a completely different aisle. A ZeroWater 10 cup pitcher with a five stage ion exchange filter publishes a built in TDS meter in the lid, IAPMO certification for lead, chromium, mercury and PFOA and PFOS, BPA free plastic, and the honest admission in its own drawbacks that the meter "needs a manual check" and "will not alert you automatically when the filter is spent".

TDS is total dissolved solids, and it is measured as electrical conductivity. Which is precisely the quantity the soil prongs respond to. So that pitcher lid contains a properly scaled version of the measurement the garden probe is actually making, and it is sold for what it is.

And now the orthogonality, because this is the cleanest way to show that pH and conductivity are nearly unrelated quantities. At pH 4, the hydrogen ion concentration is a ten thousandth of a mole per litre, which at 1.008 grams a mole is 0.1008 milligrams a litre. At pH 7 it is 0.0001. The difference across three whole pH units is 0.1007 milligrams a litre.

Against a typical tap reading of 150 to 400 milligrams a litre, that is between one part in 4,000 and one part in 1,500. Moving the acidity by a factor of a thousand changes a conductivity reading by about a twentieth of one percent. The two quantities are close to independent, which is why a conductivity instrument cannot be rescaled into a pH one by any amount of cleverness, and why the hydrogen ion is not what the needle is counting.

The filter record's other point is also conductivity doing its job properly: high TDS water exhausts an ion exchange bed faster, because the bed is capturing ions and there are more of them. More in our review of the ZeroWater pitcher and the category in best water filter pitchers.

What this leaves the question with

Our page on why an aquarium's pH falls while you change the water works the chemistry side of this problem properly: pH is the symptom and the buffer capacity is the thing that moves, and adding or removing acid without accounting for the carbonate reserve produces exactly the confusing result the aquarium question describes. A raised bed has the same structure, with the carbonate reserve living in the mineral fraction and the organic matter instead of in the water.

What the arithmetic here adds is narrower and more useful: the prong meter is not the instrument for the job, and nothing you do to it makes it one. A measurement with a reference half-cell and an amplifier is a different device; a dilute soil slurry read with pH paper or test drops at least has a reference reaction and no conductivity term at all; and either of those is a starting point for the question as asked.

What the prong meter is genuinely for is moisture, where it is a good tool, and relative light, where it is a rough one. Those are two real measurements and they are both on the same instrument. Soil preparation, which is what actually shifts a bed's buffer, is covered in preparing garden soil, and the water-holding side of the same physics is in why raising a drain hole five centimetres halves a pot's air.

The honest limit, and three things not computed

What is exact: the 59.17 millivolts per pH unit from the Nernst relation at 25 degrees, the 177.5 millivolts across three pH units, the hydrogen ion mass arithmetic giving 0.1008 and 0.0001 milligrams a litre, and the 1 in 96 dilution from two published volumes.

What is representative rather than published, and the first item matters most. The copper and zinc couple at 1.104 volts is an illustration: the meter publishes no probe metal at all, so that is a plausible galvanic pair and not its pair. A different couple shifts the baseline and therefore the 16.1 percent, but not the structure of the argument, because any two dissimilar metals give a baseline of the same order and the resistivity term still spans a factor of 100. The 10 to 1,000 ohm metre soil resistivity range, the 150 to 400 milligrams a litre tap TDS, and the ten million to a billion ohms of glass electrode resistance are all standard values.

Three things deliberately not computed. The actual needle deflection, because the record publishes no internal resistance, no galvanometer full scale current and no probe spacing or surface area, and all three are needed; the shape of the argument survives their absence and a number would not. The added conductivity from one feeding, because the fertiliser record gives a ratio and a dilution and no concentration, so the magnitude is genuinely unavailable rather than merely tedious. And whether any particular bed needs its pH changed, which is an agronomic question about a specific soil and a specific crop, and is not something a page about instruments should answer.

Three numbers to carry. A pH measurement is 59.17 millivolts per unit read across a source of a billion ohms, so it requires a reference electrode and an amplifier and therefore a battery. A batteryless prong meter is a galvanic cell whose reading swings by a factor of 100 with moisture while all of pH 5 to 8 could move it by 16 percent, a sensitivity ratio of about 620 to 1. And three whole pH units change a conductivity reading by one part in a couple of thousand, which is why one instrument cannot be the other.