The Versana Balance platform: internal cage, power box, panel, ports and software
Identify the Versana Balance variant, hardware generation and software line in front of you, name the field-replaceable assemblies on the signal and power paths, and read a normal start-up so that you notice when it is not normal.
What this machine is and how it differs from its relatives
The Versana Balance is a compact cart-based general imaging system built by GE in its Wuxi factory and sold in two model designations, VA and VS. Both share one Basic Service Manual (Direction 5808768-100; this course was written against Rev. 9 of 2022), one software line called R1.x.x and one set of field-replaceable units. The manual does not explain what separates VA from VS beyond the designation; treat the model letter, the rating plate and the option list as three separate facts you record at every job.
If you have done the LOGIQ courses, put some of that knowledge aside. The LOGIQ P carts are built around a nest box with a separate PC module, and the LOGIQ e is a laptop. The Versana Balance is simpler and more integrated: one small internal cage carries the processing boards, the PC is not a separate box but lives on a combined board, and there is very little to adjust. Several pages of this service manual still say 'Versana Essential', a sibling product whose documentation it was derived from. When a screen name or a diagnostic says Essential, it is the same function; when a part number says Essential, check it against the Balance parts list before you order.
In us-l1 and us-l2 you learned the physics, the probe types, the swap method and the artefact atlas. This course does not repeat them. It adds what is specific to this platform: which board owns which function, what the rear diagnostic LEDs and the service desktop can prove, what the manufacturer actually requires for maintenance, and where engineers lose data, options or time on this model.
Three blocks: top modules, internal cage, power box
The manual describes the system as three groups. On top sit the keyboard (operator panel), the speaker, the monitor on its arm and the optional peripherals: a printer powered from the system, the USB ECG module, a DVD writer and a gel warmer. The monitor exists in two sizes, 17.3 and 21.5 inch, so note which one you have before you order a panel or a cable. In the body sits the internal cage with the processing boards and the fans. At the bottom sits the power box with the mains switch, which is a circuit breaker on the rear panel, and the AC/DC module.
The AC/DC module turns mains into a single 19 V DC supply for the rest of the system, according to the theory chapter. Everything else is derived from that rail on the boards. This matters for fault-finding: the AC/DC module either gives its output or it does not, and there are no adjustments on it. The manual says plainly that the DC supplies regulate themselves, that a voltage out of range means either the supply or the load on that output is faulty, and that the system switches off when such an error happens.
A charger module and an internal battery appear in the manual only as 'if present'. The lock-out steps, the shipping note and the inspection list all mention removing or checking a battery where one is fitted, and the diagnostics contain a battery fuel-gauge test. Do not assume a battery is absent. Look for it, and if it is there, treat it as a second energy source that keeps parts of the system live after the mains plug is pulled.
Inside the cage: CWI, MST, PIB, CWD and the docking board
The internal cage holds four printed wiring assemblies. The CWI board is, by the manual's own abbreviation list, the CPU and I/O board: it is the back end processor, the PC that runs Windows and the application, with its memory, BIOS, real-time clock and network interface. When the rear LEDs or the troubleshooting table point to a CPU, BIOS self-test or CWI power problem, this is the board. The service diagnostics for it include memory, network and CPU temperature tests and an HV STOP test, which checks that each board can raise the signal that stops the transmit high voltage.
The MST, the master board, carries the receive processing and the digital beamforming and signal processing functions; its diagnostics test the swept and fixed demodulators, the interface to the front-end boards, the DSP master clock, its own memory, temperature and voltages. The PIB is the probe interface board with the three probe ports, which the parts list calls 3PP. The CWD board is optional and exists only on systems with continuous-wave Doppler. A 4D module, also optional, mounts on the cage for volume probes.
Outside the cage, on the frame, sits the IODKG board, which the troubleshooting table also calls the docking board. It is the junction where power and I/O signals enter and leave the cage, and the keyboard, monitor, speaker, DC printer and gel warmer connect through it. That is why one fault pattern in the hang-up table asks you to disconnect the top-module peripherals from the docking board and try to power on with the button on the docking board itself: it separates a shorted peripheral from a docking-board failure.
Hardware generation matters. The parts list of the manual revision used here notes that software R1.0.x is only for systems with the older MST and R1.2.x only for the newer MST; newer lines may exist, so check the current parts list and release notes. Record which line the system runs before you order an MST or a software USB, and never assume a board from another Balance will work with the software in front of you.
Storage, software and options at concept level
The system boots from a solid-state drive divided into partitions with separate jobs. The service desktop shows them: C for the operating system and application, D for service data, logs and exports, E for spoolers and the clipboard, a small F partition that holds the serial number, and Z for the package repository. You do not need to manage these by hand, and the manual warns against deleting configuration files from the user-defined settings area, but knowing that the serial number and the network identity live on the drive explains why a full software load asks you to rewrite the serial number.
From R1.0.4 the local patient data drive is encrypted by default, with no way for the user to switch that off, and removable media can be encrypted as well. The customer owns the encryption password and the recovery key, and the manual is explicit that the manufacturer has no back door. For you this means one rule: before any work that touches the drive, find out whether the customer still has the recovery key. A lost key plus a failed drive is lost patient data.
Options are software keys that appear under the system admin option page with a status such as Permanent or Disabled. Two of them need hardware: CW Doppler needs the CWD board and 4D needs the 4D assembly. A key without its board does nothing. Installing, transferring or generating option keys is the manufacturer's business, done by authorised personnel with the proper keys; your job is to record what is installed, keep that record, and check after any repair that the same options show as active.
Service access has levels. Basic access is available to a local user with service rights; the purchased classes (Service Advanced, Expert and Pro) add diagnostics, utilities and the probe assessment tool; full manufacturer access needs a service key that expires and a trained, authorised holder. This course works at the level the hospital is entitled to and tells you where a step belongs to the manufacturer.
Rear panel and what a normal start-up looks like
The rear connector panel carries the AC inlet, the circuit breaker, an equipotential ground terminal, a DC output for the system-powered printer, USB ports, Ethernet, HDMI, VGA, S-video, composite video and audio out, and three diagnostic LEDs. The LEDs are the cheapest instrument you have on this machine. Learn the normal pattern on a healthy system and write it in your notes, because when the screen stays black, the LED states and the colour of the On/Off button are what the troubleshooting table reads.
A healthy start goes like this. You close the breaker and hear the power relays click. You press On/Off once. The fans run at full speed and then slow down, power reaches the panel, monitor, front end and back end, the back end loads software behind a start screen with a progress bar, the software initialises the front end, the keyboard backlight comes on, and you land either in 2D with the connected probe or in a no-mode screen if no probe is connected. The manual gives about one and a half minutes for a normal start and up to four minutes when it is recovering from a lock-up or a loss of power. The operator panel is checked by start-up scripts every time.
Two habits protect the drive. Do not cycle the breaker off and on faster than about five seconds; let the system de-energise first. And shut down through the system exit menu, waiting for the sequence to finish before unplugging, because a forced power-off can corrupt system and patient files. The manual recommends a full shutdown at least once a week rather than living in standby.
- Record: model (VA or VS), serial number from the rating plate, software version and part number from the About page, base image version, MST generation (R1.0.x or R1.2.x line).
- Record: monitor size, probes and their serial numbers, installed options and their status, peripherals and their models.
- Ask: is disk encryption on, and does the customer hold the recovery key and the administrator password?
- Observe: breaker click, fan run-up and slow-down, rear LED pattern, On/Off button colour, start time, landing screen.
| Part | What it does | How it fails | |
|---|---|---|---|
| 01 | CWI board (CPU / I/O) | Back end PC: Windows, application, BIOS, clock, memory, network, HV STOP logic | Black screen with CPU or BIOS LED patterns, no network, boot loops |
| 02 | MST board (master) | Receive processing, beamforming, demodulators, DSP clock, front-end interface | Hardware configuration error, no image, image artefacts, failed MST diagnostics |
| 03 | PIB (probe interface, three ports) | Probe connectors, probe identification and switching | Probe not recognised on one port, element-like dropout only on one port |
| 04 | CWD board (option) | Continuous-wave Doppler | CW option present but CW mode missing or noisy |
| 05 | IODKG / docking board | Power and I/O junction between cage and top modules | No start with keyboard LED patterns; one shorted peripheral pulls power down |
| 06 | Power box: breaker and AC/DC module (19 V) | Mains isolation and the single DC supply | No light at all with mains present; system switches off on overload |
| 07 | Keyboard (operator panel) and trackball | Controls, On/Off button, start-up panel test | Dead keys, stuck trackball axis, On/Off button patterns in the LED table |
| 08 | Monitor 17.3 or 21.5 inch on arm | Display, brightness and LED light | Blank or flickering image, bad pixels, loose arm |
| 09 | SSD with C/D/E/F/Z partitions | System, logs, spoolers, serial number, package repository | Boot failure, slow start, read errors; data loss without backup |
| 10 | Fans and side air filters | Cooling of cage and CPU | Temperature warning and forced shutdown when filters clog |
A sibling with a lost identity: GE Vivid S6
A GE Vivid S6, a cardiology system from another GE family, stopped seeing its probes and its board test aborted during a power-supply check; after a restart it reported a software configuration error. Engineers found that the identity data stored in a flash memory on the DC/DC power block had been lost, so the system no longer matched its own configuration. The fix in the discussion involved board-level work that belongs to the manufacturer. The point for the Versana Balance is the pattern: modern GE systems keep identity and configuration data on boards and on the drive, and a mismatch shows up as a configuration error, not as a dead board.
On the Versana Balance a hardware configuration error with a booting PC points first to identity and compatibility: MST generation versus software line, serial number and configuration after a drive or board change. Check those before you condemn a board.
Liquid in the console, then a login prompt: GE Versana Active
A GE Versana Active, a close sibling of the Balance, had liquid spilled into it. After drying it booted, but stopped at an operating-system login that nobody could pass. In an engineers' chat someone proposed booting from a third-party rescue stick to strip the Windows passwords, and reported that this usually works; the outcome on this unit was never confirmed. Stripping the passwords does nothing for the liquid damage. It also defeats the security the manufacturer built in, and on a system with an encrypted patient data drive it can leave data unreadable or exposed.
Treat liquid ingress as a hardware event first: power off, dry, inspect the boards and the keyboard, and check leakage before clinical use. Passwords and drive security on this family are recovered only through the customer's own recovery data or the manufacturer's procedure, never through password-removal tools.
On-site checklist
- Model VA or VS, rating-plate serial number and software version recorded
- MST generation and software line (R1.0.x or R1.2.x) identified
- Monitor size, probes, options and peripherals listed
- Battery presence checked and noted
- Disk encryption status and recovery-key custody confirmed with the customer
- Normal rear LED pattern and start time observed and noted for this system