Ericsson FSU 995PM Fusion splicer repair

Ericsson FSU 995PM fusion splicer repair

1–3 business days
NIST-traceable calibration
Free evaluation
Ships worldwide

Ericsson's approach to PM alignment was different from everyone else's. Full 360-degree rotation of both fibers, not one. That POL alignment method is either working correctly or it is not. There is no halfway.

The Ericsson FSU 995PM is the PM variant of the FSU 995 series, purpose-built for polarization-maintaining fiber in lab and production environments. It handles SM, MM, DS, PM, and erbium fiber, but PM is where the design decisions went. The alignment system is what sets it apart from Fujikura and Sumitomo PM splicers of the same era.

Most PM splicers rotate the fiber in one direction to find the stress rod position and stop. The FSU 995PM uses Ericsson's POL alignment, rotating both fibers a full 360 degrees independently. That full rotation gives the system a complete angular scan rather than a partial one, which produces more accurate stress rod position data and better crosstalk results on fiber types where the stress structure contrast is low. On PANDA fiber with high birefringence contrast, the difference is marginal. On elliptical clad or low-contrast PM fiber, it matters.

Beyond PM work, the FSU 995PM carries the full FSU 995 feature set: thin core alignment for EDF and DCF splicing, negative index matching for gradient-index fiber, negative altitude compensation for high-elevation field deployment, automatic arc recentering, attenuator making, and tapering capability. The platform was also designed for network deployment from the start. Splice programs can be created on a PC and pushed to a fleet of splicers. Results download directly to Excel, which is why this unit ended up in large production operations and research institutions that needed traceability across a large splice count.

The FSU 995PM combines cold image alignment with warm image processing during the arc cycle, which is where Ericsson's "total cycle time under 40 seconds" claim came from. Cold image positions the fiber before the arc. Warm image during the arc gives the system real-time feedback on what the glass is actually doing during fusion, not just where it started.

These units are old. Parts availability narrows every year. When one goes down, finding someone who has actually worked on an FSU 995PM rather than just listing it is a real problem. We have.

What Fails on the FSU 995PM

The POL alignment mechanism is the first place to look when crosstalk results are degrading. The 360-degree full-rotation system uses a rotational drive and encoder that wear over time and with cycle count. When the encoder loses accuracy or the drive mechanism develops backlash, the angular scan data becomes noisy and the system selects a worse-than-optimal theta position before the arc fires. Unlike a splicer where you can compensate with a theta offset parameter, a faulty POL drive produces inconsistent angular errors that a fixed offset cannot correct.

Arc electrode wear on the FSU 995PM is accelerated compared to a standard telecom splicer. Thin core splicing, erbium splicing, and attenuator making all involve arc parameters that are harder on the electrodes than standard 125µm SMF work. Electrode sets rated for 4,000 arcs on standard fiber reach that point faster when the arc cycle mix includes high-energy specialty programs. Elevated splice loss and irregular arc shape on the warm image are the first signs.

The V-groove system can collect debris from stripped erbium and specialty fiber coatings differently than standard fiber prep. Specialty coatings strip at different temperatures and leave different residue. We inspect V-groove condition and clean the grooves on every FSU 995PM that comes in regardless of the presenting fault.

Camera and image processing system degradation on this platform has a specific consequence beyond just image clarity. The warm image processing during arc discharge is an active measurement, not just a display function. When the camera system degrades, the warm image data quality drops and the arc recentering function loses effectiveness. The symptom is splice geometry that starts showing offset on the post-splice image even when the cold alignment looked clean.

Communication interface faults on the FSU 995PM typically hit the serial data port used for PC-based program upload and download. This does not stop the splicer from operating manually, but it breaks the workflow for any production environment that uses central program management across multiple units.

The heating oven on high-cycle units eventually shows uneven sleeve shrink or longer-than-rated heat times. On a unit doing attenuator making and tapering in addition to standard splicing, the oven cycle count is higher than a unit used only for telecom fiber.

How We Work

Full diagnostic before any repair. We run SM and PM splice cycles, verify POL alignment accuracy and crosstalk results, check arc performance and warm image behavior, test V-groove condition, inspect camera quality on both axes, and test the serial communication interface. We tell you exactly what is wrong before any repair work starts.

Component level throughout. POL drive and encoder issues get addressed at the mechanical component level. Arc calibration runs after any electrode work. Warm image processing gets verified after any camera or imaging work. NIST-traceable calibration with a certificate ships with every completed unit, along with a test report showing pre and post repair performance on SM and PM fiber.

Service Specifications

ParameterValue
Fiber TypesSM, MM, DS, PM (PANDA, Bowtie, elliptical), EDF, attenuator fiber
PM Alignment MethodPOL alignment, full 360-degree rotation of both fibers
Alignment SystemCold image + warm image processing
Splice Loss (SM)0.02dB typical
Thin Core AlignmentYes
Negative Index MatchingYes
Negative Altitude CompensationYes
Automatic Arc RecenteringYes
Additional CapabilitiesAttenuator making, tapering, erbium splicing
PC InterfaceSerial, upload/download splice programs and results
Data ExportExcel-compatible results download
Operating Temp0°C to 40°C, 0 to 95% RH

Service specifications

ParameterValue
Fiber typeSingle-mode, PM
AlignmentCore-to-core (X/Y/Z axes)
Arc systemAutomatic arc calibration

Frequently asked questions

The FSU 995PM POL alignment is different from other PM splicers. Does that make it harder to repair?
Different, not harder. The 360-degree dual-fiber rotation is a more mechanically complex alignment system than single-axis theta. It has more components that can wear and more encoder positions that need to be accurate. That said, the failure modes are well defined and the fix is the mechanical components, not a firmware patch. When the POL drive is right, the crosstalk numbers show it immediately.
Can the FSU 995PM still do attenuator making and tapering after repair?
Yes. Those functions are part of the core FSU 995 platform and get verified during our post-repair test cycle. If your repair is related to a fault in the attenuator or tapering programs specifically, tell us when you ship so we include those test cases in the diagnostic.
My crosstalk results are inconsistent, sometimes good, sometimes bad on the same fiber type. Is that the splicer or the prep?
Inconsistency specifically points to the splicer more than prep. Bad cleave or dirty fiber produces consistently poor results. Variable results on the same program and fiber type usually mean the POL alignment is finding a different angular position on each run due to encoder noise or drive backlash. That is a mechanical wear problem, not a calibration problem.

Documents