Welcome to the new RK Drums website. A new, faster experience. Your account was migrated, but you will have to reset your password. Click here.

RK Drums
≡
How-To Guide

Making Crystal Malt at Home: Stewing, Caramelising and Why a Drum Helps

The malt you cannot fake with heat alone

You can get respectably close to a brown malt or a chocolate malt in a domestic oven if you are patient and you accept some unevenness. Crystal malt is a different proposition. It is not simply a darker roast of the same process. It is a different process that happens to finish with heat, and if you run it as a roast you will produce something that looks like crystal malt and behaves nothing like it.

This page is the most technical one in the malt roasting group, and it is also the one where a turning drum with a product thermocouple stops being a convenience and starts being the difference between a repeatable result and a lottery.

What is actually inside a kernel of crystal malt

Crack a finished crystal malt kernel and the centre is glassy and hard, like set caramel, rather than floury like a base malt. That glassy core is the entire point of the product. It is there because the starch inside that husk was converted to sugar, and then that sugar was cooked.

So there are two stages, and they are different in kind:

  • The stew — a mash that happens inside each individual kernel, using the grain's own enzymes, in the grain's own water.
  • The caramelise — driving the water off and cooking the sugar that the stew produced, which is what sets the final colour and flavour.

Because those sugars are partly unfermentable, crystal malt contributes sweetness, body and colour rather than alcohol. How much of it belongs in a given beer is a recipe decision and not ours to make.

Stage one: the stew

Enzymes only work in solution. Dry malt has almost no free water in it, so nothing converts until you put water back. That means wetting the malt thoroughly — steeping it until it is saturated rather than damp — and then holding it hot while keeping that water in.

The working window is 210°F to 230°F. That is a corridor about twenty degrees wide, and both walls are real:

Below it, conversion is slow, and warm wet grain sitting for an extended period is a good environment for souring organisms to get established before the enzymes finish. You can taste that in the finished malt and it is not the sourness anyone wanted.

Above it, you denature the enzymes before conversion is complete. What you get is a hard, glassy-looking kernel with unconverted starch still in the middle — the outward appearance of crystal malt without the sugar that makes it work.

Time in that corridor is commonly somewhere in the region of thirty to ninety minutes. Take that as a range to calibrate against rather than a number to obey, because it depends on how saturated the grain is, how much of it you are doing at once and what your base malt started with.

Moisture retention is the whole trick

If you remember one thing from this page, this is it. The stew fails when the grain dries out. Enzymes need liquid water to move and to act. The moment the free moisture is gone, conversion stops wherever it happened to be, and no amount of additional time recovers it.

What you have then is a lightly toasted malt with a starchy centre. It will not taste like crystal, it will not behave like crystal, and it will be a nuisance in the mash. This is by some distance the most common way home crystal malt goes wrong, and it goes wrong invisibly — the grain looks fine from the outside.

So: start wetter than feels sensible, and design the stew phase around keeping steam in rather than letting it out. The general moisture behaviour of grain during roasting applies here in reverse; every other roast wants moisture gone, and this one wants it kept.

Stage two: the caramelise

Once the core is sugar, the job changes completely. Now you raise the heat, drive the water off, and cook that sugar inside the husk. This phase is quick compared with the stew, and it is what decides which crystal malt you end up with.

Take it lightly and you get honey and toffee, pale gold in the kernel. Push further and it moves through caramel and raisin. Further still and you are into burnt sugar, dark stone fruit and a faint bitterness at the back. The same stewed grain gives you three quite different products depending only on where you stop.

Which means the finish is a temperature and colour decision made in real time, not a clock time you set at the start. Agitation and a live product temperature reading matter more in these last minutes than anywhere else in the process.

Why an oven struggles with this

Not out of loyalty to our own product — an oven is genuinely fine for plenty of light roasting work. It is a poor fit for this one specifically, for three reasons.

No agitation

Grain sits in a static bed. The layer against the pan conducts heat directly from hot metal. The layer on top sits in air. The middle steams. You are running three different stews in one tray and then mixing the results together.

The dial reads air, and it swings

An oven holds temperature by cycling an element on and off around a setpoint, and the swing in a domestic oven is frequently comparable to the entire width of your stewing corridor. Set it to 220°F and the cavity spends its time above and below that. Worse, cavity air is not grain. Even a perfectly stable oven does not tell you the grain is at 220°F, and the grain is the thing that has to be at 220°F. This is the same failure discussed on roasting grain in an oven, but a narrow window makes it much less survivable.

Nothing holds the moisture unless you fight for it

The usual workaround is a sealed foil pan with added water. It works, after a fashion, but you have now built a closed box you cannot see into, cannot measure and cannot adjust — and every time you open the door to check, you dump heat and steam.

What a turning drum changes, and what it does not

Rotation removes the static bed. Every kernel enters the hot zone and leaves it continuously, so the temperature spread across the batch narrows to something you can reason about. That matters more here than in any other roast, because the corridor is narrow and a twenty-degree spread across the batch means part of the charge is outside the window at all times.

The RoastPilot controller reads product temperature from a thermocouple in the rotisserie handle and holds your target by working the gas automatically. Sitting inside a twenty-degree band for an hour by hand, on a gas grill, on a windy day, is genuinely tedious work and it is exactly the sort of work a controller should be doing. The saved curve then means a crystal malt you liked is a crystal malt you can make again.

Now the honest part. A drum vents. That is a virtue everywhere else on this site — chaff, smoke and moisture all need somewhere to go — and it is a problem for the stew, the one phase where you want moisture to stay put. Three ways to deal with it:

  • Stew in a covered vessel, caramelise in the drum. Simplest and most repeatable, and it puts the drum on the phase where agitation and product temperature matter most.
  • Stew in the drum with a heavily saturated charge. The Solid drum option, listed for ultra-fine grains and powders, is also the one that vents least. Expect to lose moisture anyway and start correspondingly wetter.
  • Accept a partial conversion. Some home crystal is better than none, and a batch that is two thirds properly converted still does useful work.

For the dry roasts, barley belongs in the 1/16" mesh tier for small grains. That mesh is right for everything else you will do, and it will shed moisture faster than you want during a stew.

Expect the first few to be uneven

They will be. This is the advanced roast in the cluster and there is no honest way to present it otherwise. Your first batches will contain a mixture of properly converted glassy kernels and pale starchy ones, and the ratio is your feedback.

Cut ten kernels open and count. Mostly starchy centres means the stew was too dry, too short, or ran too hot and killed the enzymes early. Glassy but pale means the caramelise was cut short. Dark, hard and bitter means you overran the finish. Each of those is a distinct fault with a distinct cause, and the same diagnostic habit applies across roast defects generally.

Write down every one of those attempts. Three logged failures will get you to a good crystal malt faster than a dozen unlogged ones.

If you want the machine side of this handled properly, our drum roasters come in 2 through 12 LB with the mesh and controller options to match the work. Next: sizing, resting and recording roasted malt for the brewhouse.

Common questions

Why does my home crystal malt come out starchy in the middle?

The grain dried out before conversion finished, or the temperature climbed past roughly 230°F and denatured the enzymes early. Both leave unconverted starch behind a husk that looks correct from outside. Start with more thoroughly saturated grain, keep steam contained during the stew, and verify the hold temperature with a probe in the grain rather than an air reading.

Do I need to steep the malt before making crystal malt?

Yes. Finished malt is dry and enzymes cannot act without liquid water. The grain has to be soaked until it is saturated through, not merely damp on the surface, before you begin the hold. Commercial crystal production uses green malt that is already wet from germination, which is the condition you are recreating.

Can I make crystal malt in the same session as a dark roast?

It is easier to keep them separate. Crystal is a long hold in a narrow band followed by a short finish, while a dark roast is a steady climb with a critical last minute. They want different heat settings, different attention and, ideally, different moisture handling. Run them as separate sessions and log them separately.

How dark can I take crystal malt before it becomes something else?

Once the sugar is cooked far enough it stops reading as caramel and starts reading as burnt, with a hard bitterness rather than the raisin and dark fruit of a dark crystal. There is no single crossover number. Sample and cut kernels through the caramelise phase and stop at the point you can still taste sweetness under the roast.

Keep reading

Where the roast ends

We build roasters. What you do with the grain afterwards — mashing, brewing, distilling, baking, blending a grain bill — is its own craft and a deep one, and it is not ours. Unlike coffee, there is no single authority we can hand you to: brewing knowledge lives in forums and clubs rather than with one gatekeeper.

That gap is why we wrote these pages. Ask us about drums, mesh sizes, grills, burners, motors, thermocouples, batch sizes, or why half your barley came out two shades darker than the rest, and you have our full attention.

Shop the roasters Get started
Keep exploring