Dental Office Network Support in Philadelphia: Operatory Drops, Slow Imaging and PECO Outages, Room by Room

Wednesday, 9:15, South Philadelphia. The back operatory, the one on the second floor of the rowhouse the practice has occupied since the previous dentist, has dropped its sensor for the third time this week. The front two rooms are fine. The hygienist has moved the patient to room 1 to finish the bitewings, the schedule is now twenty minutes behind, and the person who "does the computers" has not called back.

Below that sits a symptom table mapping what staff see to a probable cause and a remedy, an explanation of why rowhouse and converted-office practices here behave unlike a new build, the vendors' own requirement sheets used to put imaging traffic in proportion, a plan for the morning the internet fails and the afternoon PECO does, and finally a set of questions that show whether your present provider grasps any of this.

What does "network support" mean for a dental office, and who owns it?

For a dental office, network support means keeping four things reliable: the cable and switch path from every operatory to the practice-management and imaging servers, the wireless the hygiene tablets and intraoral camera carts use, the firewall and internet circuit that carry cloud imaging, e-claims and phones, and the closet that all of it lives in.

That is a narrower and more specific job than "IT support" in general, and it is worth being precise about it because the symptoms you see in a clinic are different from the symptoms an accounting office sees. An accounting office notices a slow website. A dental office notices a sensor that captures and then sits on "sending", a panoramic image that takes a minute to reach the viewing workstation, a hygiene tablet that loses the chart when the door to the back room closes, and phones that crackle while a CBCT export is running.

The parts, named the way your provider will name them:

• • Demarc: where the carrier's line ends and your equipment starts. In a rowhouse practice it is often in the basement; in a converted office suite it may be in a shared room on another floor.
• • Firewall: the device just inside the demarc, owned by the practice rather than the carrier. All traffic between an operatory and the internet crosses it.
• • Core switch: the hub in the closet where each operatory cable, the server and the desk phones terminate. A practice of four to ten operatories is better served by a single managed switch than by a string of small unmanaged boxes added one renovation at a time.
• • Drop: one cable, closet to wall jack, serving one spot: a chair, the reception counter, the wall behind the pan.
• • Patch panel: where the drops end in the closet, each labelled with the room it serves. An unlabelled panel is the most common reason a one-room fault takes an hour to find.
• • PoE (Power over Ethernet): the switch sends electricity along the same cable that carries data, which is how ceiling access points and desk phones run without a nearby socket.
• • VLAN: a logical partition on the switch. One partition keeps patients' phones on guest Wi-Fi out of the clinical network; another can let calls jump the queue ahead of image transfers.
• • Failover: a backup route to the internet, most often a cellular modem, that the firewall falls back on if the primary line goes dead.

Who owns it: the national owner page for this topic is Legend's dental office network design service, and the broader Philadelphia picture, including who to call for the rest of the practice's technology, is on the IT support for Philadelphia dental offices page. This article covers the part those pages do not: what goes wrong in a Philadelphia operatory network, why, and what fixes it.

Which part of the network is behind the symptom you are seeing?

What staff see narrows the fault to a layer before anyone opens a wall. A single failing room implicates its run, jack or switch port. All rooms failing at once implicates a scheduled task or the server. Cloud tools down while local charts still open implicates the circuit or firewall. A fully dark office implicates the closet or its power.

Settle three things before you pick up the phone; any engineer worth the call will open with them:

1. 1. Is it one room or the whole office? A single room means its drop, jack, patch cord or switch port. The whole office means the core switch, the server or the internet line.
2. 1. Can a wired desk still reach the practice-management server in the closet? When Dentrix, Eaglesoft or Open Dental still loads the day's schedule at reception while e-claims, cloud imaging, phones and email have all gone quiet, the trouble lies with the internet line or the firewall, and rebooting workstations will not reach it.
3. 1. Are only the wireless devices affected? If the tablets, the hygiene laptop and the camera cart are struggling while every cabled PC is fine, look at the access points, or at a clinical device that belonged on a cable all along.

Then find your row.

Two rows need an honest note. The "second-floor back operatory" row and the "converted-office closet" row describe conditions that are common in Philadelphia's older buildings and are stated as things to survey, not as a statement about any particular practice. The "nothing reaches the internet after a flicker" row describes a common pattern after any power event, not only a PECO one, and the power-up order it points to is general practice, not a procedure specific to one provider.

Why do Wi-Fi and sensors fail in older Philadelphia buildings?

Older Philadelphia buildings fail dental networks in predictable ways: plaster-and-lath and brick party walls absorb Wi-Fi, runs through three floors of a rowhouse are long and pass close to old electrical, closets were built for coats rather than switches, and reused telephone wiring still carries data in some suites. Each is a survey item, not a mystery.

One figure, with its caveat stated plainly: the regional planning commission's July 2025 housing profile reports that just over 40 percent of Philadelphia's homes were built before 1940, against 22 percent across the region. That is a residential figure; nobody has published an equivalent one for the commercial and converted stock most practices occupy. It is cited here only to say what every Philadelphia dentist already knows: the buildings are old, and a network that would be routine in a suburban new build becomes a project in a 1920s rowhouse or a converted Center City office floor.

What a competent survey of an older Philadelphia practice should look at before anyone quotes you for new cable:

• • Wall construction, room by room. Plaster-and-lath, brick party walls and the thick masonry of older commercial buildings are the reason one access point in reception does not reach the back hygiene room. The fix is placement and a second access point, not a bigger antenna.
• • Cable paths and run length. In a rowhouse practice the closet is often in the basement or on the ground floor and the back operatories are upstairs. The survey should trace where each run can physically go, whether existing chases or the old telephone riser can be reused as a path (not as cable), and whether any run will exceed the length at which copper no longer links reliably.
• • What the existing cable actually is. Some suites still carry data over the flat or four-pair telephone wiring a previous tenant installed. It links, badly, until it does not. The survey should identify every run that is not Cat5e at minimum and mark it for replacement with Cat6 or Cat6A.
• • The closet. Airflow, temperature on a July afternoon, a dedicated circuit, space for the UPS, and whether the door can be closed without the equipment cooking. A converted coat closet with the door shut is where switches go to die in summer.
• • Electrical sharing. A pan or CBCT unit that shares a circuit with the closet, or with the acquisition PC, produces exactly the kind of intermittent fault that gets blamed on "the network". The survey should note which outlets share a breaker with imaging equipment.
• • The demarc and the basement. Where the carrier's line enters, whether the modem or optical terminal sits on a bare outlet, and whether that spot is one that has ever been damp. Philadelphia basements and heavy rain are not a theoretical combination.
• • Drop ceilings and the lack of them. A converted office floor with a drop ceiling is straightforward to recable. A rowhouse with finished plaster ceilings is not, and the quote should say so.

Two practical rules follow from all of this. First, anything clinical that can take a cable gets one: sensors, acquisition PCs, the server, the pan. Wireless, in the Dentrix requirements document, is "to be used at your own risk due to potential interference from X-ray machines and other devices", and it is said to "require professional installation". That is the vendor's own sentence, and it is the right default for every clinical device that can take a cable. Second, Wi-Fi is for the things that move: tablets, the camera cart, the laptop the doctor carries between rooms, and patient devices on their own fenced-off network.

The build-order detail for a new or renovated network, including what to decide before rough-in, lives on the Dental Office Network Setup Service page; this article does not repeat it. If the practice is a new build or a full renovation rather than a fix, the planned dental startup IT guide for Philadelphia covers the sequence from lease to opening day.

How much network does imaging actually need?

Imaging needs a gigabit wired path far more than it needs a faster internet circuit. The vendors' own requirement documents ask for gigabit Ethernet between acquisition PCs, the imaging server and viewing workstations, and they ask for very little of the internet unless the images live in a cloud. The table below uses only what those documents say.

Three things to take from the table rather than from anyone's opinion.

The ratio is the point. The one vendor that publishes internet speeds for cloud imaging asks for ten times the connection for 3D than for 2D. Even for a practice that keeps every image on its own server, the ratio shows where the weight sits: CBCT volumes are what find the weak link, and one 100 Mb port anywhere between acquisition PC and viewing station is all it takes to stretch a seconds-long open into minutes.

The gigabit requirement is on the wire, not on the internet. Every local requirement above is a gigabit Ethernet card and a gigabit switched network. None of them is "a faster internet plan". A practice that upgrades its circuit to fix slow imaging has usually paid for the wrong thing.

The documents age. The Carestream CS Imaging 8 sheet cited here is a 2019 revision that still lists Windows 7; the CS 3D Imaging sheet is from March 2024. Use the current sheet for your exact unit, and note that intraoral scanner vendors and the DEXIS, Vatech and Planmeca CBCT lines each publish their own; this page cites only the documents it opened. Who looks after the acquisition PC and the imaging database, and the rest of the imaging-software question for practices here, is the subject of the planned dental software and imaging support in Philadelphia article, and the server itself is covered by the Dental Office Server Setup Services page.

What should a Philadelphia practice do when the internet goes down?

Check scope first, then the carrier, then the firewall, then single devices. If the practice-management server still opens the schedule from a wired desk, nothing inside the building needs restarting; the fault is the circuit or the firewall. A practice with cellular failover should already be on it, and the front desk should be working the patient script.

The order, written for the office manager who has to run it at 8 AM with a full waiting room:

1. 1. Scope. Open the schedule on a wired front-desk PC. If it opens, the server and switch are fine. If a phone on the office Wi-Fi shows the network but no internet, the wireless is fine too. So the fault lies where the line leaves the building, or somewhere past it.
2. 1. The carrier. Check the carrier's outage status from a phone on cellular data, not from the office network. After a thunderstorm a carrier outage is a likely answer, and it is one your IT provider cannot repair.
3. 1. The firewall. Look at the outside (WAN) light on the firewall and at the carrier's modem or optical terminal at the demarc. If the carrier reports no outage and the modem's lights are wrong, power the modem down, bring it up alone, and wait for the lights to settle before anything else is touched.
4. 1. The failover. If the practice has a cellular backup on the firewall, confirm it has taken over; cloud practice-management systems such as Curve and Denticon, cloud imaging, e-claims and email keep working on it, slower. If there is no failover, a phone hotspot will carry a cloud practice-management login for the front desk, and nothing else.
5. 1. The phones. Modern VoIP systems ring to a mobile app or forward to a cell number when the office loses its circuit; the setting has to exist before the outage. If the office still has a copper line, this is the morning it earns its keep. How phones should be arranged so this works is on the VoIP for dental offices page.
6. 1. One device. Only after the above do you restart a workstation, and only the one that is misbehaving.

The patient script matters as much as the sequence. While the circuit is down the practice can still see patients whose charts are on a local server, can still take images to a local imaging server, and can still process cards on a terminal with its own cellular connection. What it cannot do is e-claims, cloud imaging or anything that lives in Microsoft 365, until the failover or the carrier comes back. Decide in advance which of those the front desk should reschedule and which it should simply defer, and write it on the same sheet as the sequence.

A note on the national guide: the general version of the outage sequence, and the setup decisions that make the failover step possible, are on the Dental Office Network Setup Service page. This page adds the Philadelphia layer: the power events below.

What happens to the network when PECO goes down, and how do you power back up?

A power outage takes down the closet first and the carrier's equipment second, and the order in which things come back decides whether the practice is working ten minutes after the lights return or still "down" at lunch. The answer is a UPS on the right devices, a clean shutdown before the battery dies, and a written power-up order.

Philadelphia's pattern is worth naming because it drives the plan. On 19 June 2025 a line of thunderstorms moved through the region and PECO reported 327,000 customers had lost service (The Philadelphia Inquirer, 20 June 2025). In the first days of February 2021 a nor'easter put the City of Philadelphia under a declared snow emergency (WHYY, 1 February 2021). Summer thunderstorms, winter nor'easters and summer heat waves are the three events the network plan has to survive, and each produces a different failure: a hard power cut, a multi-day closure with the equipment running unattended, and a hot closet.

The power-up order in the table is general practice for any small network: bring up the device nearest the carrier first and work inward, waiting for each to finish before starting the next. It is not a procedure specific to any provider, and your own IT provider should put their version in writing for your closet, with the exact devices named. If they cannot, that is one of the questions below.

Should the firewall and switches be managed, and by whom?

Yes. A managed firewall and a managed switch mean someone is watching their logs, applying their updates, holding a copy of their configuration and getting an alert when a port goes down or the closet heats up. An unmanaged firewall is a box that was configured once; its failure is discovered by the first patient of the day.

"Managed" is a plain word with a plain test. Ask whether anyone would know, without being told, that the switch port serving operatory 4 had dropped to 100 Mb last Tuesday, or that the closet reached an unsafe temperature on a July afternoon. If the answer is that staff would notice when something broke, the firewall and switch are not managed; they are installed.

Two separate statements apply to Legend here, and they are deliberately stated apart. Legend's published facts are 24/7 monitoring, and, as a separate matter, a help desk with the staffed hours given in the next section. In general terms, monitoring is an automated watch on devices such as the firewall, the switch and the server; a help desk is the people you reach. Legend also states: in-house engineers; no outsourced first-level triage. What that means for your Philadelphia practice is a question to put to any provider you are considering, in writing: who sees the alert, and who answers the phone.

The firewall also carries the practice's security posture, which is a larger subject than a network article should try to cover. The dental cybersecurity services page owns that: segmentation, endpoint protection, what a readiness review looks at. This article's point is narrower: a firewall that is not managed is a network fault waiting for a date.

When can a Philadelphia office reach network support?

Legend's published hours are: help desk Mon–Fri 7 AM–7 PM CT, Saturday 8 AM–2 PM CT; 24/7 critical/emergency line. For a Philadelphia office that is Mon–Fri 8 AM–8 PM ET and Saturday 9 AM–3 PM ET; a 7 AM first patient arrives an hour before the staffed day, and the emergency line covers the hours outside it.

The time-zone conversion is worth stating because a Central-time provider working with an Eastern-time practice is a real consideration, and it cuts both ways: the staffed day starts at 8 AM on the Eastern clock, an hour after a 7 AM opening, and runs to 8 PM, an hour past a 7 PM close. Legend Networking's headquarters are in Dallas, Texas, and the company's published position on where people go is: onsite dispatch by appointment across served dental markets; remote-first nationwide. Before signing, have any provider spell out in writing what that means for a Philadelphia building, because most of the symptom table above is solved by remote diagnosis and then either a cabling contractor or an engineer visit, and you should know which one your building will get.

How to reach the help desk and what the ticket process looks like is on the dental computer support and help desk page. Keep the general Philadelphia page for everything that is not a network question: dental IT support in Philadelphia covers the rest of the practice's technology and the nearby markets.

What does network work cost inside a plan?

Hardware procurement, major cabling projects and software licensing are quoted separately; what a given plan covers for an existing network beyond that is a question to settle with the provider in writing. The rest of Legend's published figures are: flat monthly plans $399.99–$1,200 per practice per month; typical single-location practice $600–$900 per month all-in.

Applied to the table above, that split is simple. Diagnosing the dropped sensor, disabling power management on a workstation's network card, moving the backup window, reconfiguring the guest VLAN and confirming the power-up order are support. Re-pulling six operatories to Cat6A through plaster, replacing the switch that died in the hot closet and building a dedicated server are a cabling project and hardware procurement, and Legend states those are quoted separately. Whether a cellular failover device is part of a given plan or a separate line is a question to ask any provider directly; this page does not assume the answer.

The cost that does not appear on any invoice is the one the opening of this page described: the FMX redone in another room, the hygiene column twenty minutes behind, the doctor waiting for a CBCT volume to open. Nobody can put an honest dollar figure on that for your practice without your schedule in front of them, and this page does not try. It is the reason the table exists.

Which Philadelphia-area work has Legend documented?

Legend's documented Philadelphia-area project is Maple Glen Modern Dentistry in Ambler, Montgomery County, in the Philadelphia suburbs: a dental startup where Legend worked alongside Dr. Adam Rudin from early concept and planning through the opening of the practice, with a four-photo gallery of the completed office on the project page.

The project page does not list a network scope, and this article does not add one. What it does show is the stage at which Legend's involvement began, early concept; in any build, that is the point at which cable paths, the closet and the demarc are still choices rather than constraints, the opposite of the situation most of the symptom table describes.

Legend's published facts: 22+ years in dental IT (since 2004), 3,500+ practices supported, and a 4.8 rating · 173 Google reviews. Ian Lynch — 22 years in dental IT; oversees the NOC, onsite dispatch and client technology roadmaps. On a network fault, the question that matters is whether the person reading the switch alert understands what a dropped sensor in a second-floor operatory looks like on a dashboard; the roadmap conversation is where the re-cabling of an older building gets planned rather than patched.

What to ask your IT provider about your network

Take these to your current provider before you take them to a new one. Written answers, with the device names for your closet.

A provider who answers these in writing understands dental networks in older buildings. A provider who answers "we'll take care of it" does not yet.