Small-batch membrane-electrode work has a shape that a specification table does not describe. A few dozen pieces of one frame format, three formats in the same week, catalyst layer arriving in a new lot every month, and a station that is expected to answer questions rather than to run flat out. A hot-press bonding machine in that setting is not a production line reduced in size. It is a station whose duty cycle is dominated by setup, verification and waiting, and whose output is measured as much by what the batch taught you as by how many pieces survived it.
The published pages for these machines list a handful of figures. The page for the membrane electrode hot press bonding machine states a fitting error of 0.2 mm or less, a working temperature of 150 degrees C or less, a press pressure of 2t and a production cycle of 1.5pcs/min or more. The page for the membrane electrode vacuum hot pressing machine states a fitting error of 0.2 mm or less and a working vacuum of 60 Pa or less. None of those numbers describe a day in which you press thirty pieces across three formats with two material lots and one operator. So supplier support in this regime is not a service promise; it is a set of things handed over before your first batch runs. The most practical way to read a supplier is to walk one small batch through the station and listen for what they give you at each step.
Define the batch before you discuss the machine
A batch on a bonding station has two numbers that people routinely confuse: pieces per cycle, and pieces per shift. The published cycle figure describes a repeating cycle at steady state. It says nothing about the first cycle after a cold start, the cycle in which you stop to photograph a defect, or the cycle in which the layout is half empty because you only have six pieces of that format. Ask the supplier to break the cycle into its parts for your frame format: load, align, close, heat and dwell, cool, vent, open, unload. Then ask which of those the published figure includes, and whether it assumes automatic handling that you do not have.
The vacuum station needs one more line in that breakdown, because its page publishes no cycle at all. Ask how long the chamber takes to pull from ambient to 60 Pa, with what pump, and whether the chamber is pulled once per piece or once per batch. Ask what the leak rate looks like after the chamber closes with a load of membranes inside it, since a dry empty chamber and a chamber full of porous material are not the same volume of gas to move. In mass production, a few minutes of pump-down per cycle disappears into the schedule. In a small batch it is your day, and it is the first number that will decide whether you can run two recipes before the shift ends. A written cycle breakdown for your format and your part count is the cheapest document in this whole purchase: with it you can work out whether a batch fits in a shift, and without it every capacity claim you receive is a guess dressed as a specification.
The first piece after a stop is the piece the machine is judged on
Small batches start and stop, so the behaviour worth asking about is the behaviour at the beginning of a run. How long does the platen take to reach a setpoint from cold, and how is temperature across the platen verified, with what instrument, at how many points, with the platen empty or loaded? Can the machine hold a warm standby between batches, and what does that cost in energy and in the life of heating elements? What happens to the first pieces after an idle period, and are they scrap, or are they usable if you change one setting?
Then the number that decides whether a batch is usable at all: the 0.2 mm fitting figure. In mass production an alignment capability is proven once and then trusted. In small-batch work it is re-proven every time a new material lot, a new frame format or a new carrier film arrives, because the parts you are aligning have changed underneath the machine. Ask the supplier for a quick check your own operator can run in a few minutes with your own frames: a witness pattern, a calibration fixture, or a written procedure with a stated pass line. Ask in writing what the 0.2 mm is measured between, at which point on the platen, hot or cold, and with which instrument, so that your daily check and their acceptance test measure the same thing. A supplier who has worked with small batches will offer a first-piece procedure without being asked, because they have watched what happens when one is missing.
Ask what the platen will do when it is not full
Neither page publishes the size of the hot platen, so the first question is simply the usable area for your frame format. The question behind it is the one most suppliers have not prepared for: can this machine run a partial layout without changing the process? A half-empty platen is not a full platen with less to do. It has different thermal mass, different edge behaviour, and a different path for heat to move through the pieces that are there. In a vacuum chamber the load changes the gas path and the time to reach 60 Pa as well. So ask whether the recipe has to change when the layout changes, whether format inserts or tooling sets are offered for smaller formats, and what the supplier recommends when you genuinely press one piece at a time.
Ask specifically about a piece placed at the corner of the platen rather than the centre, because that is where a small batch actually lives: one odd piece, one trial strip, one part of a format you are phasing in. A supplier who can tell you what changes at the edge, and what they would set differently, has run their own machine this way. A supplier who answers that the platen is uniform everywhere is describing a simulation.
Separate the recipe from the hardware, in writing
The point of a small batch is usually to find a bonding recipe: a temperature below the stated ceiling, a force below the stated 2t, a dwell, a ramp, a pull-down and vent sequence. Those are settings, and a small batch exists to change them. Platen area, the alignment method, chamber volume and pump capacity are hardware, and no small batch will change them. Ask the supplier to draw that line for you in writing: the adjustable range of each recipe variable, who is permitted to change it, and which changes need a service visit. Ask whether recipes can be stored and recalled by name, so that the batch which worked on Tuesday can be run again on Friday without someone remembering.
Then ask what the machine records, and for whom. In this regime the record matters more than it does in production, because each batch is an experiment and an experiment without conditions is not evidence. Ask whether the machine stores, per piece, the recipe name, temperature and force traces, the vacuum trace, the alignment result, the operator and the material lot, and in what format that leaves the machine. If it records nothing, ask the supplier to say so plainly so you can plan your own capture rather than discovering it at the first review. Twenty batches of a bonding recipe become a trend only if every piece carries its conditions; otherwise they are anecdotes, and the argument in the meeting will be about memory rather than data. A supplier who answers these questions in a paragraph has understood your situation. A supplier who answers that the machine is very easy to operate is quoting a manual at you.
Decide which station you own first
The two pages describe two stations in one sequence rather than two versions of one machine. The five-layer machine is described as heat pressing the catalytic layer together with the upper and lower frame films to form a five-layer membrane electrode. The seven-layer machine is described as heat pressing that five-layer electrode together with the upper and lower microporous layers to form a seven-layer one. The output of the first is the input of the second, and the two pages link to each other.
In small batches you rarely own both at matched capacity, so the useful question is which one you should own. If you buy finished five-layer electrodes, the second station is the one you need and the first is somebody else's problem. If you make them, the first station comes first, and the space and buffer between the two stations is planned on paper instead of discovered on the floor. Ask the supplier to state their recommendation for your actual part count in writing, even when the honest answer means selling you one machine instead of two. A supplier who will recommend the smaller purchase is more credible on every other line of the quotation, and a supplier who cannot explain why you would buy one before the other has not thought about your volumes at all.
The gap between batches is where small-batch machines go wrong
Between batches is a real operating state, and in this regime it deserves a written procedure. Should the platen be held at temperature or allowed to cool? Should the vacuum chamber stay under vacuum or be vented? How do membranes, catalyst coatings and porous layers behave after sitting overnight in a plant, and does a batch that waited need a different dwell or a different pull-down? What should be done before restarting after a week with no orders? Ask for these answers in writing and keep them with the machine, because the person who learns them by trial and error is usually the only person who knows them.
The same applies to consumables. Ask for a list with part numbers, replacement intervals stated in cycles rather than in calendar months, and prices. Chamber seals, protective and release sheets, heating elements and thermocouples all age differently depending on how often the machine is cycled and how often it is heated from cold. A part quoted as lasting six months means one thing on a station pressing thirty thousand pieces a month and something else entirely on a station pressing thirty a week, and the stock you should hold is different in both directions. Ask for a spares set sized to the batches you actually run, and ask how quickly each part arrives.
Who is allowed to run the station
A small team has one person who understands the press, and support in this situation should be measured by how quickly that person can proceed on their own, not by how quickly a reply arrives. Ask which changeovers and adjustments are operator-level and which require the supplier's engineer. Ask for the documentation set in the language your operators actually read, including the list of parameters the user is allowed to change and the list of those they are not. Ask for the escalation path in concrete terms: who answers, in what language and what time zone, what information they will ask you to send, and what you are expected to diagnose by yourself before anyone travels.
Ask, too, whether a remote session can look at the machine's own records. When a station runs a handful of batches a week, a fault that takes two weeks to diagnose costs a month of learning, and the difference between a remote look at the traces and a visit is often the difference between a lost week and a lost quarter.
What belongs on one page before you order
Ask for all of this as one document, before you commit: the cycle breakdown for your frame format including pump-down; cold-start and first-piece behaviour with the quick alignment check your operator can run; the answer about running a partial layout; the adjustable-versus-fixed statement and the format of the per-piece record; the between-batch procedure and the consumables list with intervals in cycles; a recommendation on which station to own first and why; and a statement of what your own staff may do without calling. The completeness of that one page predicts the rest of the project better than the layout of the quotation does. Blanks in it are not omissions to be chased later. Each blank is a place where the standard configuration does not meet your use, and every one of them will reappear as a change order at a higher price.
It is also worth being clear about what you bring, because a supplier cannot support a batch they cannot see. Bring drawings of your frame format with its tolerances, the materials you actually run, including the membrane, catalyst layer, frame films and porous layers with their lot information, the batch sizes and formats you expect over the next two quarters, your site's power, air, water, exhaust and space limits, and the acceptance wording you wrote before any trial. A supplier who asks for those things before quoting is doing diligence on your behalf. A supplier who quotes from the catalogue without asking has sold you a machine and left the project to you.
When the batches start repeating and the recipe stops moving, the conversation changes to takt, buffers and the automatic bonding line that appears in the same equipment category. Until then, the honest path is one station, your own materials, and a record you can read. Sending your batch requirements through the contact page and asking a supplier to answer that one-page list in writing is the fastest way to tell the two kinds of supplier apart.
