Outside Canada? genetixbay.com shows prices in US dollars.Go to genetixbay.com
GenetixBayEST. 2009 · CANADA
Lab notes · 14 November 2026 · 11 min read

Lab notes #5: dosing by the element

You can't buy a bottle of pure nitrogen. How Jardin is designed to hit element targets anyway, with salts, a small solver, a scale under every jug and a strong preference for pH Down.

Lab notes #5JardinNutrientsGrow scienceBuild log
Professor Greenleaf

Updated 18 November 2026: the dosing tiers and the 60-litre reservoir below come from the earlier design. WaterWorks now keeps its stock bottles on a scale outside the tent, mixes in a larger batch tank, and sends each batch to the reservoirs you choose. The chemistry and the solver are unchanged. See what Jardin looks like now.

Most feed charts speak in millilitres per gallon. Plants don't. Plants care about nitrogen, potassium, calcium, magnesium and a dozen others, in parts per million. The goal for Jardin is simple to say and annoying to build: tell it the elements you want, and it works out how to get there.

Fair warning: there's maths in here. We've hidden most of it behind jokes.

You can't buy a bottle of pure nitrogen

Every fertiliser salt brings at least two things with it. Calcium nitrate brings calcium and nitrogen. Epsom salt brings magnesium and sulfur. Potassium nitrate brings potassium and more nitrogen.

So "complete control of each element" really means choosing how much of each salt to add so that all the passengers together land where you want them. Think of each salt as an arrow pointing in a few directions at once. You can mix arrows, but you can't bend them.

That has a real consequence. If your only calcium source is calcium nitrate, then every bit of calcium drags nitrogen along with it, about three-quarters as much again. A recipe that wants lots of calcium and very little nitrogen simply isn't possible with those salts, and a good system should say so, not quietly flood the tank trying.

The solver

When you set targets, Jardin's solver looks at three things:

  • What's already in the water. Zero for fresh filtered water. For tap water it's your water report, because tap water brings its own calcium, magnesium and alkalinity.
  • What's in each jug. Every stock has a known recipe, so the solver knows exactly how many milligrams of each element one millilitre adds.
  • The rules. A ceiling for every element, a ceiling for total salt strength (EC), and per-pump limits per dose and per day.

Then it finds the mix that misses your targets by the least, weighted by what matters most. Nitrogen, potassium and calcium count fully. Sulfur mostly rides along, so it gets a ceiling instead of a target. The maths is a small, well-understood optimisation that finishes in a blink on the mix station.

The best part is the report. You don't just get a dose. You get the achieved number for every element next to your target, and a plain sentence when something can't be hit: "Magnesium is 18 % high because it rides on sulfate. A magnesium nitrate channel would decouple it."

A worked example

Here's the solver's answer, in the design notes, for a fresh 60-litre reservoir of filtered water with veg targets. Numbers are rounded.

Stock Aimed at Millilitres Also brings
Calcium nitrate Calcium 140 ppm about 295 Most of the nitrogen
Epsom salt (magnesium sulfate) Magnesium 50 ppm about 152 Sulfur 66 ppm
Monopotassium phosphate Phosphorus 45 ppm about 119 Some potassium
Potassium nitrate The rest of the nitrogen about 203 Most of the potassium
Potassium sulfate Nothing 0 Would push sulfur over its ceiling
Iron and micronutrients Iron and trace elements about 64

Result: nitrogen, phosphorus, calcium and magnesium land on target. Potassium comes in about 3 % low, because the only salt that could top it up would push sulfur past its ceiling. The solver says exactly that, and tells you what a slightly higher sulfur ceiling would buy. That's the kind of honesty we want from it.

What lands in the tank · ppm · illustrative

Dashed lines are the targets, the gold one is sulfur's ceiling. Potassium lands 3 % short because the only salt that could top it up would push sulfur over its ceiling, and the solver says so instead of hiding it.

Order matters

Calcium and concentrated phosphate or sulfate are the nutrient equivalent of two exes at a wedding. Put them together at stock strength and they make rock, and that rock clogs tubes and coats probes.

So a fresh batch always goes in the same order: plain water and a reading first, then calcium on its own into clean water, then a good few minutes of mixing, then the other salts one at a time from a different inlet, then the micronutrients and iron last. Each dose trickles into the moving stream slowly, so the concentration at the inlet never gets anywhere near trouble.

A fresh batch, in order · animated Dose sequence: water is filled and read, then calcium nitrate, potassium nitrate, MKP with Epsom salt, and micros with iron are dosed one at a time, each jug weighed on its own scale; then the tank is mixed, allowed to settle and verified by EC. CaNKNO₃MKP·MgFe·µ 60 L reservoir fill + read60.0 kgcalcium nitrate295.4 g ✓potassium nitrate203.1 g ✓MKP + Epsom271.0 g ✓micros + iron64.2 g ✓circulate3 turnoverssettled, verifiedEC 1.71 ✓ jug scale
  1. 0 · Water + reading
  2. 1 · Calcium nitrate
  3. 2 · Potassium nitrate
  4. 3 · MKP + Epsom
  5. 4 · Micros + iron
  6. 5 · Mix
  7. 6 · Settle + verify

Calcium goes in alone, into clean water, and gets mixed before anything with phosphate or sulfate follows. Every dose is booked as the grams the scale saw leave the jug, not what the pump was told. Illustrative numbers from the worked example above.

Verified by weight

Every stock jug sits on its own scale. When the solver asks for 295 mL of calcium nitrate, the pump runs, and then the scale says how much really left the jug. The dose that gets logged is the measured one, not the hoped-for one.

That one idea does a lot of work:

  • Pump wear shows up early. Tubes slowly stretch, so each pump's real flow rate is re-learned from every dose. A 15 % drop flags the tube before it ruins a batch.
  • A pump that lies gets benched. If a dose comes in more than 10 % off, that channel locks until you look at it.
  • Wrong jug? Caught. After a refill, the first dose from that jug is a small 10 mL test. Each stock raises EC by a known amount, so if "calcium nitrate" suddenly behaves like phosphate, the channel locks and asks.
  • EC has to agree. After a batch, the measured EC is compared with what the solver predicted. More than 15 % off and the whole sequence stops.

Knowing what's in there later

EC is one number, and there are thirteen elements in the tank. Over days, plants drink unevenly, so Jardin keeps a running estimate of each element from everything it measured going in and out, with error bars that grow over time. When the estimate gets too fuzzy, or the reservoir is old, or the water has been topped up past its own volume, Jardin suggests a fresh batch. Want certainty? Send a sample to a lab every few weeks, upload the PDF, and the estimate resets to the real numbers.

pH Down only, by default

Here's something that surprises new hydro growers: in a normal nitrate-based feed, pH goes up over time. Plants drinking nitrate push the solution alkaline, often by 0.1 to 0.4 a day in a small reservoir.

So Jardin ships with pH Down only. It doses in small steps (about 2 mL in a 60-litre reservoir), waits 10 to 15 minutes for things to settle, and has a daily cap. It aims for the upper edge of your pH band rather than the middle, because pH keeps drifting while things settle and chasing the centre overshoots. pH Up exists, but it's off by default and switched on per grow only when the recipe needs it, with a smaller daily cap. Fewer ways to pour the wrong bottle in.

Do you need all this?

Probably not all of it, and we'd rather tell you that now. The design has four dosing tiers on the same firmware:

Tier Dosing channels Who it's for
Starter 3: two-part base plus pH Down First grow, soil or coco, anyone using a two-part feed
Standard 5: three-part base, pH Down, optional pH Up The default for most tents
Precision 7: single salts for calcium, magnesium and phosphorus Serious hobbyists, breeders characterising a line
Lab 10 to 12: one salt per job Research benches and commercial R&D

For a four-plant tent, the difference between Standard and twelve jugs is small, and twelve jugs is twelve things to run dry. The solver runs on every tier anyway, so you always see your element targets next to what you actually got, and which upgrade would close the gap.

Every number on this page is an engineering estimate from our design notes, and all of it will be checked with an agronomist and on a real bench before it becomes a default. The Mix Pro station behind WaterWorks has been mixing for over nine years; the dosing test report is the next proof we publish, mess included.

Professor Greenleaf: puff, puff, pass

Puff, puff, pass it on.

Founding Members get first access and a permanent badge. Share your link: when a friend who joins through it becomes an active member, you each earn 500 Pearls. Want to help us build it faster? Fund the Bay: reserve a reward now, nothing is charged until checkout opens.