Kiln Controller Programming for Annealing: Segments, Schedules, and Garage Mode

Short answer: Every programmable kiln controller, whatever the brand, builds a firing out of segments, and each segment is just three numbers: a ramp rate (degrees per hour), a target temperature, and a hold time at that target. To program an annealing cycle, you translate your glass’s annealing schedule into a chain of those segments: ramp up to the soak temperature, hold, then explicitly program the controlled cool-down with one or more down-ramp segments. The cool-down is where most programming mistakes happen, and it matters as much as the soak itself. Lampworkers add one more trick: a long “garage” hold at the front of the program so pieces can wait at temperature during the torch session.

This article covers the programming side. For the schedules themselves, see annealing schedules for glass, and for choosing the kiln in the first place, see the lampworking kiln guide.

Why annealing needs a controller, not just heat

Annealing is not “get the kiln hot.” It’s a sequence: reach a specific soak temperature, hold it long enough for the glass to equalize, then cool at a controlled rate down through the strain point before the kiln is allowed to coast. A programmable digital controller executes that whole sequence unattended, including the slow cool that happens hours after you’ve gone to bed.

The alternatives can’t do this. A kiln sitter (the cone-based mechanical shutoff on many old ceramic kilns) provides shutoff only; it cannot anneal at all. An infinite switch regulates power input, not temperature, so holding a soak or cooling at a controlled rate means a pyrometer and constant manual babysitting. More on that at the end.

Anatomy of a segment: ramp rate, target, hold

A segment is the basic building block of any kiln program, and it consists of three values:

So “60°F per hour to 200°F, hold 2 hours” is a complete one-segment firing. Chain segments together and you can describe any annealing cycle. Two conventions worth knowing on Bartlett- and Paragon-style controllers (do not assume they are universal to every brand):

• A rate of 9999 means “as fast as possible.” The controller applies full power until the target is reached. This is the usual choice for the initial climb.
• A down ramp is programmed exactly like an up ramp: you enter a rate plus a target temperature that is lower than the previous segment’s target. The lower target tells the controller this segment is a controlled cool.

That second point is the single most important thing in this article. If your program ends at the soak and the kiln free-cools, thin beads may survive, but thick work can crack. The controlled descent through the strain point must be written into the program as its own segment or segments.

Translating an annealing schedule into segments

Any written annealing schedule decomposes the same way:

1. Climb segment: as fast as possible (9999 on Bartlett/Paragon-style units) to the soak temperature. For lampworkers this is usually replaced by the garage segment described below.
2. Soak segment: hold at the annealing temperature for the time your schedule calls for, based on glass type and thickness.
3. Controlled cool segment(s): a stated rate down to a target at or below the strain point. Some schedules use one slow ramp; others use a stepped series of holds at descending temperatures.
4. End of program: once the glass is safely below the strain point, most schedules let the kiln cool naturally to room temperature.

Where does the schedule itself come from? Your glass manufacturer, first and always. The annealing schedules guide collects published numbers, but the manufacturer’s figures for your exact glass take precedence over anything here.

Worked example: COE 104 soft glass bead session

For COE 104 soda-lime glass (the Effetre/Moretti family), reported annealing soak temperatures are genuinely contested, clustering between 940°F and 968°F depending on brand and source, with the strain point around 850°F. Prefer your glass manufacturer’s published number where one exists. Community practice puts the garage temperature roughly 30 to 40°F below the anneal point, around 920 to 930°F. A typical bead cycle soaks in the 940 to 960°F range, then cools slowly at about 1 to 2°F per minute (60 to 120°F/hr) down through the strain point before free-cooling. Translated into segments:

Segment 1’s hold is set long enough to cover the whole session (or indefinite, where the controller supports it); when the last bead goes in, the program advances, or you advance it manually, into the soak. The soak length comes from your schedule and scales with the size of the work.

Worked example: COE 33 borosilicate

For COE 33 boro, ~1050°F is the standard annealing temperature and the strain point is around 950°F (industrial borosilicate 3.3 datasheets put the annealing point near 565°C/1049°F and the strain point at 515°C/959°F, consistent with the lampworking figures). A commonly cited rule of thumb is to hold at annealing temperature roughly 1 to 2 hours per inch of thickness, with thick or solid work getting the conservative end. A typical reported shop cycle: garage around 1000°F, hold about 30 minutes after the last piece goes in, ramp to 1050°F, hold about 2 hours, then a controlled ramp down to about 950°F over about 2 hours before faster cooling. As segments:

For thicker work, Northstar publishes a stepped chart instead of a single down ramp: soak at the annealing temperature, then hold for 50% of the anneal time at 125°F below it (100% for pieces over 0.25 in thick), then 25% of the anneal time at each of 200°F, 350°F, and 550°F below it. Each step becomes its own segment: a down ramp to the step temperature plus the stated hold. That turns a four-segment program into seven or eight, exactly the kind of schedule a multi-segment controller exists to run. Boro garage temperatures also vary by shop, from about 930°F up to about 1000°F, with the strain point as the logic anchor.

Programming garage mode: the all-day hold before the anneal

“Garaging” means holding finished pieces at a set temperature for the duration of a torch session, adding pieces as you finish them, then starting the true annealing program only after the last piece goes in. The critical point, made bluntly by boro lampworker Mike Aurelius: garaging alone does not anneal the work. The garage hold keeps pieces from cracking while they wait; the soak and controlled cool afterward do the actual stress relief.

In segment terms, garage mode is simply segment 1: ramp as fast as possible to the garage temperature with a long or indefinite hold, with the soak and down ramps as later segments you start or advance to after loading the last piece. Common garage ranges are about 920 to 960°F for soft glass and up to about 1000°F for boro. Purpose-built lampworking annealers (Paragon’s F-series and BlueBird lines, for example) offer bead doors so you can slip a piece in without dumping chamber heat mid-session.

The controller families you’ll meet

Menus and keystrokes differ by brand, model, and even firmware version, so this article deliberately avoids model-specific button sequences. Your controller’s manual takes precedence over everything here. That said, almost everything on the market falls into three families, and all of them speak the same ramp/target/hold language:

• Bartlett-style digital controllers. Bartlett Instrument makes the controllers fitted to many kiln brands. The current touchscreen Genesis stores up to 30 custom programs of up to 32 segments each; older 12-key Bartlett boards use the same segment model with fewer stored programs. The 9999 maximum-rate and lower-target down-ramp conventions are documented in the Genesis manual and Bartlett’s newsletters.
• Paragon Sentry-style 3-key controllers. Paragon’s Sentry Xpress, common on small bead annealers, uses Ramp-Hold programming: at each segment prompt you enter the firing rate (1° to 9999° per hour), the target, and the hold. Capacity depends on firmware generation: earlier Xpress versions store 4 programs of 8 segments, the Xpress 5.0 up to 25 programs of 20 segments. Paragon controllers on lampworking kilns have one firing mode, ramp-hold, rather than the ceramic cone-fire modes.
• Retrofit controller boxes. Standalone controllers from manufacturers such as SDS Industries exist specifically to convert manual and infinite-switch kilns to programmable operation, built around the same segment model. If you’re weighing a retrofit against a purpose-built annealer, see DIY vs manufactured kilns.

One caution that applies to every family: the displayed temperature is the thermocouple temperature, not the glass temperature. Thermocouple placement, drift, and offsets mean a small-chamber bead kiln can read differently from the work inside it, so verify your kiln’s actual behavior rather than trusting the display blindly.

Infinite-switch kilns: what they can and can’t do

An infinite switch is a duty-cycle power control, like a stove dial: set to 50%, it sends full power half the time and no power the rest. It controls power input, not temperature. It cannot hold a setpoint, cannot execute a ramp rate, and has no idea what the chamber temperature is. Annealing with one requires, at minimum, a pyrometer plus constant manual attention to approximate the holds and ramps by nudging the dial, for the entire multi-hour cycle.

That is the practical case for a controller: annealing cycles run for hours, often overnight, and a programmable controller executes the whole schedule, including the cool-down, unattended. Two safety notes belong here. Never leave an infinite-switch kiln running unattended; without a controller there is nothing regulating it. And any kiln that runs for hours needs adequate wiring capacity and proper clearance from combustibles, controller or not. The kiln and controller manufacturers’ manuals govern installation and operation and take precedence over anything here.

Common programming mistakes

• No programmed cool-down. Ending the program at the soak and letting the kiln free-cool is the most common translation error; thick work needs the down ramp written in.
• Forgetting a down ramp needs both numbers: a rate and a lower target temperature.
• Assuming 9999 is universal. The maximum-rate convention is Bartlett/Paragon-style; check your manual before assuming another brand reads it the same way.
• Treating the garage hold as the anneal. It is not a substitute for the soak and controlled cool.
• Copying segment counts across models. Capacity changes between firmware generations even within one product line; verify what your unit actually stores.

Key takeaways

• Every kiln program is a chain of segments: ramp rate, target temperature, hold time.
• Translate your schedule directly: climb (or garage), soak, programmed controlled cool through the strain point, then free-cool.
• COE 104: soak around 940 to 968°F (sources vary), cool 60 to 120°F/hr to about 850°F. COE 33: soak ~1050°F for 1 to 2 hours per inch, controlled cool to about 950°F.
• Garage mode is just segment 1 with a long hold. Garaging keeps pieces safe during the session; it does not anneal them.
• All the common controller families speak ramp/target/hold, but menus, conventions, and capacities differ: the manufacturer’s manual wins.
• An infinite switch controls power, not temperature. It cannot anneal unattended, and a kiln sitter cannot anneal at all.

Sources

• Bartlett Genesis Operation Manual (hosted by L&L Kilns) - https://hotkilns.com/sites/default/files/2024-04/Genesis%20Operation%20Manual%204-10-2024-web.pdf
• Bartlett Instrument, “Slow That Kiln Down! Program a Down Ramp” - https://binst.s3.amazonaws.com/uploads/newsletter/picture/88/Kiln_-_47_-_Program_A_Cool_Down.pdf
• Paragon, “Sentry Xpress 4.0 3-key controller: single segment and ramp-hold” - https://corp.paragonweb.com/documents/sentry-xpress-4-0-3-key-controller-single-segment-and-ramp-hold/
• Northstar Glassworks annealing chart - https://northstarglass.com/annealing/
• Mike Aurelius, “Garaging and annealing, two different aspects of the same beast” - https://mikeaurelius.wordpress.com/2008/01/30/garaging-and-annealing-two-different-aspects-of-the-same-beast/
• SDS Industries, “Kiln controller upgrade” - https://www.kilncontrol.com/blog/kiln-controller-upgrade/
• SIMAX Physical and Chemical Properties datasheet - https://www.vitrocom.com/wp-content/uploads/2023/04/SIMAX%C2%AE-Physical-and-Chemical-Properties.pdf
• Chatter Glass, “The mystery of softening, annealing and strain points” - https://chatterglass.wordpress.com/2014/03/23/the-mystery-of-softening-annealing-and-strain-points/

Editor’s note: annealing temperatures, garage temperatures, and hold-time rules of thumb vary by glass brand, source, and shop practice; where they conflict, this article gives the range. Your glass manufacturer’s published schedule and your kiln and controller manufacturers’ manuals take precedence over everything here.