The six-head ring, sold as a module.
One welded frame, five area heads, one line scan bottom behind a mirror, and a trigger box that keeps six exposures inside a microsecond.
Status: pre production. First build targeted for the current quarter. Everything on this page marked target is not yet a shipped figure.
Specification
What the module is, and what it is not
The class figures come from published vendor datasheets. The target column is our first build commitment. We do not merge the two.
| Property | Class value | H6 target |
|---|---|---|
| Heads per ring | 5 + 1 | 5 + 1 |
| Bottom face method | line scan plus mirror | line scan plus mirror |
| Exposure sync error, six heads | not published | < 1 µs |
| Strobe pulse, set in hardware | not published | 50 to 100 µs |
| Smallest 2D code, six sided at 500 mm | 1.27 mm | 1.27 mm |
| Smallest 2D code, top face only at 500 mm | 0.38 mm | 0.38 mm |
| Belt speed, multi-face layout | < 3 m/s | 3 m/s |
| Camera interface | GigE, 2 × GbE with TSN | GigE, one port per head |
| Result output | vendor specific | JSON over TCP, one record per parcel |
| Ingress protection, frame and box | IP65 to IP67 on heads | IP54 cabinet, IP67 heads |
| Working temperature | 0 to 40 °C typical | 0 to 40 °C |
| Indicative module price, heads supplied by you | no part number exists | from €3,950 |
Control layer
Three jobs, one small FPGA
The ring is not a camera bracket. It is a timing device that happens to hold cameras.
01
Distribute the encoder trigger
One encoder input, six isolated outputs. Every head sees the same edge at the same tick, so no head lags behind the parcel it is supposed to be reading.
02
Fire the strobes with a fixed pulse width
The pulse is set in hardware, not by an operating system scheduler. This is the difference between a sharp code and a code with two millimetres of smear.
03
Merge six results into one parcel record
Six heads see parts of the same box, and the bottom head sees the same box again in the mirror. De-duplication happens against the parcel, not against the string.
What a timing error actually costs
| Skew between heads | Parcel moves | Consequence |
|---|---|---|
| 1 µs | 0.002 mm | irrelevant |
| 100 µs | 0.2 mm | below one module at 1.27 mm, tolerable |
| 300 µs | 0.6 mm | half a module, the side faces start losing digits |
| 1 ms | 2 mm | more than a module, small codes fail outright |
Boundaries, stated before the order
A ring seller who does not publish limits is selling you a return shipment. These are ours.
1.27 mm is a geometry floor, not a tuning problem
At 500 mm working distance the side heads look along the box face at a shallow angle, so the effective code resolution drops. Below 1.27 mm on the sides, no amount of extra megapixels recovers it. If your side codes are 10 mil, we will quote a top only station and say why.
Direct part marks need a look at the surface, not the datasheet
Low contrast laser marks on machined aluminium behave nothing like printed labels on corrugate. Send a photograph with a scale and we will tell you whether to budget for dome light or for a change of marking process.
The ring does not include a sorting decision
We return a code and a parcel record. What the line does with it is your PLC or your WCS. We interface to it, we do not replace it.
Typical reply: three working days, with a geometry sheet, a price band, and the list of facts we still need from you.