Selecting CR Onsrud vs Thermwood CNC Routers for Acrylic

For this acrylic-router purchase, the table arrangement and brand name do not decide the result; the right machine is the one that proves the required edge finish, dimensional tolerance, and production changeover on representative parts. Compare the specific configurations quoted, because a moving bridge, fixed bridge, twin tables, and vacuum zones change work flow but do not by themselves establish cut quality or accuracy.

Price, brand preference, and showroom impressions

A brand vote, a favorable service story, or a single showroom cut is not a purchase specification. Operator experience is useful for identifying questions, but individual reports describe particular machines, maintenance histories, and applications. They do not establish capability for the buyer’s acrylic, drawing tolerances, duty cycle, or current service terms.

A price-only decision can miss features that lower setup time or reduce downtime. The field comparison included an Onsrud machine described as having more features and costing roughly twice as much as a Thermwood machine, but the machines were not established as equivalent configurations and that historical figure is not a current quotation. Conversely, buying every available option without a defined job list does not prove that the extra cost improves output.

A vendor demonstration is more useful when it runs the buyer’s files and material. A polished sample from a prepared demo program cannot show how the machine handles the buyer’s tool, workholding, nesting, tool changes, or second-table setup. Ask both suppliers to demonstrate the same representative jobs, then record measured part dimensions, edge condition, cut time, setup time, and any manual intervention.

Table and bridge mechanics in the quoted configurations

The two candidate layouts differ in how work is presented to the cutting envelope. In the Thermwood configuration under consideration, two separate tables move and can be combined into one, with a fixed bridge. The CR Onsrud configuration has one table with separate vacuum zones and a moving bridge. These are configuration-specific descriptions, not claims about every machine either manufacturer sells.

Feature	Thermwood candidate	CR Onsrud candidate	Purchase implication to test
Table layout	Two moving tables; can be combined	One table with separate vacuum zones	Test both the regular production setup and the largest intended sheet or assembly.
Bridge motion	Fixed bridge	Moving bridge	Confirm usable travel and accessible work area against the actual part envelope and toolpaths.
Changeover	One table can be prepared while the other is cutting, according to an operator report	Vacuum zones allow areas of a single table to be configured separately	Time a real load, unload, locating, and vacuum setup sequence rather than estimating from layout.

Two tables can reduce idle time when the next part can be set up safely while the current part runs; they do not automatically double output. The advantage depends on setup time, cycle time, staffing, workholding, and whether the next job is ready. Combining tables may be useful for larger work, but the buyer must confirm how the combined configuration affects the usable work area and production sequence.

Vacuum zones can help configure a single table for different work areas, but the trial should prove that the intended acrylic part stays located during the cut. Ask each supplier to show the offered table, bridge, controller, spindle, and workholding options on the written quote. Do not compare a feature-rich build with a base build and treat the difference as a brand-level performance comparison.

Acrylic cutting load, heat, and edge finish

Acrylic edge quality comes from the interaction of cutter geometry and condition, feed and speed settings, chip formation, workholding, and machine motion. When the cutter shears cleanly, the edge can be clean; when the tool rubs or chips are recut, friction and heat can degrade the edge. A smooth-running machine helps, but the cutter and process settings still matter. A favorable edge report for one machine is not proof that the other configuration cannot achieve the same result.

Use the same acrylic grade and thickness, cutter, programmed toolpath, and intended finishing operation for the comparison. Have the supplier state the tool and process settings used, then run the buyer’s normal production settings as a separate trial if they differ. Inspect the cut for tool marks, smearing, burrs, dimensional error, and consistency from part to part. Record changes to the cutter or feeds and speeds; otherwise, a process adjustment can be mistaken for a machine difference.

Separate a process-quality symptom from a control or positioning symptom. A visibly poor edge with acceptable part dimensions calls for checking cutter selection and condition, feeds and speeds, chip evacuation, and workholding first. A repeatable dimension shift or lost motion on axis reversal points toward measurement, offsets, mechanical condition, or machine setup rather than edge heat. If the symptom changes with settings or tool condition, preserve those settings and compare repeat cuts before changing controller parameters.

Tolerance, backlash, and material sensitivity

The deciding quantity is the buyer’s required part tolerance and the measured error across the actual working envelope. A machine’s nominal model description, a single tolerance anecdote, or a backlash number from another installation cannot replace an acceptance test against the drawing. The reports available include useful examples, but they are not guaranteed specifications:

Reported measurement	Context	How to use it
Within 0.005 in consistently	A Thermwood operator reported cutting plastics on machines with 5 ft x 10 ft and 5 ft x 20 ft tables.	Ask for a comparable test on the buyer’s part, material, and measurement method.
+/- 0.007 in	A reported 5-axis Thermwood installation produced long mold patterns checked with a laser tracker.	Use as an example of a measured application, not as an acceptance limit for another configuration.
0.003-0.004 in backlash	An operator reported this on an Onsrud rack-and-pinion gantry and compared it with a ball-screw Thermwood machine.	Measure reversal error on the quoted machine; the operator noted greater visibility of the error in aluminum than in wood.

Backlash is lost motion when an axis reverses direction. It can matter more on a toolpath with repeated reversals or when the material and feature make dimensional error conspicuous. The reported rack-and-pinion and ball-screw comparison describes two machines, not a universal drive-system ranking. Request the drive arrangement for the exact build and test reversal behavior at the locations and directions that matter to the part.

Agree on a written acceptance method before the trial: identify the part dimensions, the allowed deviation from the drawing, the measurement locations, the measuring equipment, and the number of repeat parts. Measure across more than one location in the working area if the job uses the whole table. Keep the programmed file, offsets, tools, material, settings, and recorded results together so both parties can reproduce the result.

Bearing inspection and tolerance drift

Lubrication and bearing condition are related but separate maintenance tasks. One reported Thermwood machine began to lose tolerance even though bearings were lubricated on schedule; the missed task was their recommended removal, cleaning, and inspection. Replacing worn bearings restored tolerance on that machine. The practical lesson is to investigate maintenance condition before treating tolerance drift as an inherent machine limitation.

Use the manufacturer’s maintenance plan for the exact machine and record both lubrication and inspection work. An automatic lubrication system can perform its specified lubrication function; it does not replace a separate inspection or cleaning task if the maintenance procedure calls for one. When measured error changes over time, compare the current measurements with the baseline acceptance results, review maintenance records, and have the machine inspected for wear or adjustment conditions before changing the process to conceal the drift.

Keep the failure categories distinct. A mechanical error that grows with axis reversal or travel suggests a different investigation from a cut edge that degrades as the tool dulls or the settings change. A machine that repeats a wrong programmed position may also involve offsets or control settings. Capture the exact test conditions and error pattern before requesting service so the technician can direct the next check rather than rely on a general complaint that the machine is inaccurate.

Automatic tool measurement and locating features

Tool measurement and pop-up locators affect setup repeatability and handling time, not just headline cutting accuracy. An operator reported that Thermwood’s automatic tool measurement feature was accurate within a couple of thousandths when checked manually, and that pop-up locators made part changes faster. Treat those observations as a reason to test the feature on the quoted configuration, not as a guarantee that every machine includes it or performs identically.

During a demonstration, measure a tool manually and with the offered automatic measurement system, then compare the resulting cut or reference dimension. Ask which tools and workflows the option covers, what the operator must enter or verify, and how the system handles a tool change. Test locators with the buyer’s actual part-loading sequence and confirm that the location method is compatible with the intended vacuum or fixture arrangement.

Include the option names and operating procedures in the quote and acceptance record. If the feature is absent from the build, account for the operator time and repeatability of the alternative process instead of assuming a later retrofit is straightforward.

Production throughput, software, and job expansion

Estimate capacity from the full production sequence, not spindle-on time alone. Twin tables may reduce waiting when one table can be prepared during machining on the other. Measure the current and demonstrated time for unloading, cleaning, locating, vacuum setup, tool changes, and first-piece checks. A shorter cut cycle may not improve completed parts per shift if setup, loading, or programming remains the constraint.

Test the software path with representative files. An operator described a Thermwood workflow accepting rectangles from a spreadsheet, cabinet-design output, and direct DXF import. That experience does not confirm compatibility for another software version or a different machine configuration. Provide the supplier with actual files, including the formats used today and the formats planned for near-term work, and confirm how toolpaths, offsets, and revisions reach the controller.

Build the option list around a defined job mix and likely expansion. Ask the supplier to explain alternate ways to make the same part and compare efficiency, operator work, and required options. Specify future needs such as larger work, multiple heads, rotary work, or additional tools only if they are relevant to the production plan, then price and demonstrate those functions as part of the configured offer.

Service, spare parts, and downtime exposure

Service quality should be evaluated as a recovery process: fault isolation, access to the needed part, diagnosis, and return to production. Individual reports for both manufacturers ranged from positive support experiences to costly delays. For example, one owner of an older Onsrud machine reported a wait exceeding a month for an interface cable quoted at $900, while another Onsrud operator described phone diagnosis of a spindle inverter fault, overnight shipment of a replacement, and a roughly three-hour installation. These historical cases are not current prices or lead-time commitments; ask for written current terms for the specific build and installed options.

One reported Onsrud spindle problem took more than two days of production loss after a cable, motor, and controller were replaced one at a time; a later online diagnostic session adjusted parameters and restored operation within an hour. That sequence shows why symptom-based parts swapping can increase downtime. Preserve fault messages, test results, and changes made, and follow a diagnostic sequence directed by the manufacturer before replacing multiple components.

Ask the supplier to identify the official service contact route, response terms, remote-diagnostic capabilities, field-service availability, recommended critical spares, and parts lead times for the exact configuration. Confirm whether the controller, interface cables, spindle drive, and other configuration-specific parts remain supported, how the buyer can obtain service documentation, and which repairs the supplier expects the customer to perform. Compare the cost of planned spares and downtime exposure with the purchase price.

Acceptance trial and decision record

Use a controlled trial to make the two bids comparable. Bring production files and representative acrylic stock; tell each supplier what the part must do, which dimensions govern acceptance, what cut quality matters, and what throughput the shop needs. Request a demonstration of both the normal table arrangement and the larger or alternate setup that influenced the purchase.

1. Freeze the trial conditions. Record material, cutter, toolpath, workholding, vacuum setup, program settings, and machine options. Run the same part and process on each quoted configuration where practical.
2. Measure the result. Compare drawing dimensions, repeat-part variation, edge appearance, setup time, changeover time, and complete cycle time. Use the written drawing limits and agreed measurement method rather than a verbal claim of accuracy.
3. Test the production workflow. Load the buyer’s actual files, check tool measurement and locator functions if quoted, and time the sequence from unloading through the next job’s first cut.
4. Close the commercial and service gaps. Match the tested machine to the final quote by model/configuration, options, software path, maintenance tasks, warranty and service terms, parts support, and acceptance results.

Quantity	Acceptance limit	Where to read or establish it
Part dimensions and repeatability	Drawing tolerance and buyer-defined repeat requirement	Part drawing, quality plan, and recorded trial measurements
Edge condition	Buyer’s sample or stated inspection requirement	Representative cut sample under recorded tool and process settings
Cycle and changeover time	Production target for the job mix	Timed demonstration of load, setup, cutting, unload, and next-part preparation
Backlash or lost motion	Limit derived from the part’s tolerance budget	Measured reversal test on the quoted machine, following the supplier’s procedure
Service and spares	Shop’s allowed recovery time and risk threshold	Current written service terms, parts list, and lead-time confirmation

Choose the configuration that meets the drawing and throughput requirements with documented service coverage, not the machine with the strongest anecdote. Preserve the test files, settings, measurements, and final option list in the purchase record; they provide a baseline for commissioning and later tolerance checks.

Frequently asked questions