We connect remotely within 24 hours of a fault being logged, and put an engineer on site in 12 cities.
01 — COMMITMENTS
The six items below describe how service is defined in time, place and scope. Each one is tied to a measure.
The clock starts the minute your fault is logged. We connect remotely, read the nozzle check and the error log with you, and tell you whether the problem is the head, the board or the software. Weekends and public holidays are inside that window.
Phone, WhatsApp and screen sharing straight into the RIP and the machine service menu. We do not bill remote support, and it does not stop when the warranty does.
Same working day inside Istanbul, within 48 hours in the other cities listed. For provinces outside the list we send a plan within 72 hours with travel cost itemised separately.
A one-line "one year warranty" hides which part is covered. A printhead is a consumable and is not covered by any manufacturer; we put that in the quote, not in a conversation after the sale.
Held in Istanbul: printheads, dampers, ink pumps, encoder strips, belts, filter sets, UV LED modules and driver boards. For anything not in stock we give you a date, not an estimate.
Serial number, installation date, every replaced part and the last service visit are on file for each machine. The call does not start with us asking which machine you have.
02 — SERVICES
We measure the site before the machine ships: door width, lift and floor load capacity, panel capacity, compressor flow, ventilation and extraction. That list is an annex to the order form. A machine that will not fit through your door is our mistake, not yours.
Training happens at the machine, not in a slide deck: RIP setup, building an ICC profile, the daily and weekly maintenance routine, reading a nozzle check, and the warning signs that precede a fault. It continues until your team runs the machine unaided.
On a calendar rather than on a breakdown. Every six months or 1,500 running hours — whichever comes first — we work a 32-item checklist and hand it over as a signed report with the measured values on it.
The items that stop production are held in Istanbul; a machine standing still while a part is in transit is not something we treat as normal. Stocked parts ship the same working day, and for anything on order you get a written delivery date.
03 — COVERAGE
In the cities below our own engineer travels to you; we do not subcontract service. We also cover provinces outside the list, but the lead time and the travel cost are stated in writing first.
04 — TROUBLESHOOTING
These are the eight faults we actually log. They are written as steps: if they fix it you never need us, and if they do not, you can tell us which step failed and the diagnosis gets ten minutes shorter.
First separate the two failure types: if the missing lines sit in one block it is a damper or an air leak; if they are scattered and random it is drying. In order: (1) three normal cleans with two minutes between them — ten cleans back to back will not recover a head, it just wastes ink and stresses the pump. (2) If that fails, wipe the nozzle plate with a lint-free cloth and the correct flush. (3) Still missing, park a flush-soaked cloth under the head for 30–60 minutes; never use water, alcohol or acetone on a UV head. (4) If a channel still will not return, the fault is usually not the head but the damper or the negative-pressure setting. Below 40% relative humidity the blockage comes back — a humidifier is cheaper than repeat cleaning.
Banding has three distinct causes and three distinct tests. Evenly spaced bands whose pitch equals the pass height are a media-feed (step) calibration problem: print a one-metre reference, measure the real advance and correct the step value. Fine white or pale lines are missing nozzles; confirm with a nozzle check. Irregular bands that darken in places are head height or media flutter; check media tension, vacuum and roll-tension drift. On UV machines, a lamp that has lost output produces wet-on-wet build-up that also reads as banding. Raising the pass count hides banding without fixing it, and halves your output.
White ink carries titanium dioxide pigment at roughly twice the density of the other colours, and it settles within 24–48 hours if it stands still. On machines with white circulation, do not switch the machine off — leave it in circulation mode. Without circulation, run the white channel for ten minutes every morning and mix the bottle by inverting it; shaking makes foam, inverting does not. Once settling has started, change the filter and damper and flush the line — mixing the bottle alone will not bring pigment back out of the line. White has a shorter shelf life than the other colours; do not use white older than six months.
The encoder strip is an optical ruler, and every speck of ink on it reports a false position. In order: (1) wipe the strip end to end with a dry lint-free cloth; do not spray cleaner onto it, which leaves residue and scratches. (2) Check that the strip is taut and untwisted — a kinked strip always fails at the same point. (3) Blow out the reader slot with low-pressure air. (4) If the error repeats at the same X coordinate the strip is physically damaged and must be replaced; if it appears at random points, look at the cable, the connector and the driver board. Never handle the strip bare-handed: the oil from a fingerprint produces the same fault.
On UV, do not confuse two different faults: if a fingernail marks the print, cure is short; if it does not mark but lifts in a tape test, it is a surface-energy problem and needs primer. For low cure, clean the LED glass of ink mist first, then measure output — LED modules weaken quietly over time. The second cause is excessive head-to-material distance. On DTF, measure oven temperature at the belt rather than in the air: 110–130 °C with 2–3 minutes dwell. If the powder melts glossy but the surface is rough the dwell is short; if the film yellows or curls the temperature is too high.
Colour drift is almost never a broken machine. Eliminate in order. (1) Same ICC profile and same pass mode? Change the profile, the pass count or the resolution and the colour changes with it. (2) Did the material batch change? A different lot of the same branded banner prints 5–8% off on the same profile. (3) Is the nozzle check clean? A missing nozzle in one channel shows up as a colour-balance shift long before it shows up as a line. (4) Did the room drop below 18 °C? Viscosity rises, drop volume falls and the print goes light. The permanent fix is a profile per material group plus a monthly verification print.
The first cause of misting is head-to-material distance: running at 3 mm instead of 1.5 mm is enough for the droplet to stray. The second is static — in a dry room film and vinyl charge up and pull airborne ink onto themselves; an ionising bar or proper earthing ends it. The third is a low negative-pressure (vacuum) setting, which destabilises the meniscus and starts build-up under the head. Drop the head to the lowest height the material allows and the contamination usually stops the same day.
Heads are expensive and boards are not, so test order matters. On a machine with two identical heads, swap them: if the fault moves with the head, it is the head; if it stays on the same channel, it is the driver board or the cable. On a single-head machine you run the same test by swapping the data cable and the channel on the board. Read head voltage and temperature from the service menu — if temperature reads nothing at all the problem is most likely the cable or the board, because even a dead head still reports temperature. We run these tests with you over a remote connection, and most of the time it turns out the same day that no head is needed.
05 — DOCUMENTS
The machine model, the serial number and one sentence describing the fault are enough. Describe the nozzle check as well and the first diagnosis usually finishes on the first call.