In late June 2026, the buyer at a dry-type transformer manufacturer in Ohio sent us a short request. They needed 910mm EC axial fans (roughly a 36-inch fan) that could deliver at least 10,500 CFM at 0.61 inWG, at roughly 70% of the fan’s maximum rated speed, and they wanted a quote and a lead time.
Over the next five-plus weeks the conversation moved from airflow to control, and the evaluation closed in early August, before the project reached the sample stage. We’re sharing it because the questions this customer asked are worth asking before a sample ships on a large EC fan replacement.
Project snapshot
| Customer | Dry-type transformer manufacturer, Ohio, USA |
| Existing fan | A 910mm (roughly 36-inch) EC axial fan from ebm-papst, run by the customer’s own proprietary control program |
| Requirement | At least 10,500 CFM at 0.61 in.w.g. (about 17,840 m³/h at 152 Pa) at roughly 70% of maximum rated speed |
| Electrical and site | 380–480 VAC, 3-phase; panel mount; noise below 85 dB(A); occasional starts as low as -25°C |
| Volume and timing | About 40 fans per month; lead time of 4–6 weeks at most, weighted above price |
| Control | Fan speed driven by a 4–20 mA signal from an RTD temperature transmitter |
| Timeline | Late June to early August 2026 |
| Status | Evaluation closed in August 2026, before the sample stage; the proposed technical call did not take place |
The Request Came In as a Duty Point, Not a Fan Size
The first email didn’t name a fan model. It named a result at a specific operating condition: “Greater than or equal to 10,500 CFM @ 0.61 inWG @ Approx: 70% of Max Fan Rated RPM.”
It’s a useful way to specify a cooling fan for a dry-type transformer. The customer didn’t tell us the cooling class of their units, but as general background, transformer suppliers such as Hammond Power Solutions and Maddox describe a unit in the AA/FA cooling class as having two ratings: a self-cooled rating with natural air circulation, and a higher kVA rating with fans, using forced circulation of air. That forced-air cooling comes on once the self-cooled rating is exceeded. When the fans run, they work against the resistance of the enclosure and the transformer cooling system, not in open air.
Our first reply restated the requirement and set the method for everything after it: “This is a performance-point-based requirement, so we will need to check the fan curve at around 70% speed, not only the maximum airflow rating.”
The fan we put forward was a 910mm EC axial fan built for this inquiry, with a maximum rated speed of 1,300 rpm, 3,200 W rated power and a 380–480 VAC three-phase supply. Its datasheet lists a maximum airflow of about 24,000 CFM. That figure is measured in free air at full speed, and it says very little about this application. What mattered was where the curve sat at 0.61 in.w.g. and around 70% speed.
| Check | Value |
|---|---|
| Required airflow | ≥10,500 CFM (≈17,840 m³/h) |
| Required static pressure | 0.61 in.w.g. (≈152 Pa) |
| Required speed position | ≈70% of maximum rated speed |
| Maximum rated speed of the proposed fan | 1,300 rpm |
| 70% of that speed | ≈910 rpm |
| Speed at which our curve reading showed the requirement as achievable | ≈910–945 rpm, or 70–73% of maximum speed |
The last line is a reading from the curve data we had at the time, not a test result. The project never reached the sample stage. If you want the background on why the pressure point matters more than the headline number, see how to read a fan curve.
Why the 70%-of-Max-Speed Figure Is Part of the Spec
The customer never explained the 70% figure in their emails, so we won’t put a reason in their mouth. What we can show is how it fit their control logic. The diagram their engineering manager sent later marked the target operating point at “60%-80% (Approx)” of maximum speed, reached at 130°C and held from there up.
Our reading is that the spec puts the design point in the middle of the fan’s speed range instead of at the top. That has a practical consequence for selection: the fan’s full-speed capability has to sit well above the requirement. There are two ways to get there. You can go to a larger fan, or you can stay at the same size with a motor that runs faster. This project was the second case. The existing fan and ours were both 910mm; in our reply we pointed out that ours “achieves higher speed and the working condition is closer to your requirement.”
Outside the Range on Our Website
The EC axial fans listed on our website stop at 630mm, and the largest of them tops out around 14,500 m³/h of maximum airflow. A 910mm fan for this duty point was outside that range, so the fan we proposed was made to order for this inquiry rather than taken from the standard catalog. You can see where the standard range ends on our EC axial fan page.
The electrical and ambient operating conditions were not the hard part. The proposed fan took 380–480 VAC three-phase, and its operating temperature range of -40°C to +60°C covered the customer’s occasional starts at -25°C.
Noise was a stated requirement: below 85 dB(A). The sound pressure figure on the datasheet is a full-speed figure, and the fan was specified to run near 70% in service. Confirming the level at the actual operating point would have taken a sample measurement, and the project didn’t get that far. We’re not going to estimate it here.
Control was the open question.
The Control Questions, Answered One by One
The buyer said it plainly before agreeing to a sample: “The program we are using to control the EBM fans is proprietary, so we will not be able to plug it directly into your fans. We need to make sure we will be able to properly control the fans before we move forward.”
Over the following week, their engineering manager added the specifics. The table below lists what they asked, what we answered, and what we’d suggest you spell out in your own RFQ. Our answers describe the configuration of the fan proposed for this project, not every EC fan we build.
| What the customer asked | What we answered | What to put in your RFQ |
|---|---|---|
| Does the fan run off proprietary Modbus software, and can we enter our own operating points? | It communicates via standard Modbus RTU over RS-485, no proprietary software is required, and a PLC or controller can write the speed setpoint directly. We sent the Modbus register map for their engineers to review. | What will command the fan, and whether your current control program is tied to one fan brand |
| Can the fan achieve our operating logic from our 4–20 mA RTD signal? | Yes. The fan has a 4–20 mA current control input, and our engineering team checked the logic in their diagram against it. | The full control curve (temperature, signal and target speed), not just “4–20 mA” |
| Can we set a maximum RPM on the fan itself, or do we have to scale our signal? | The maximum speed can be configured on the fan, so the 4–20 mA signal doesn’t need to be rescaled. | Any speed limit you need, and where you want it enforced |
| They sent the datasheet for the temperature transmitter they use. | The fan can work with that transmitter’s 4–20 mA output and follow the control logic in their diagram. | The transmitter model and its output range |
| What can our control system read back? (we raised this) | Speed, current, power consumption, alarms and fault status, all over Modbus. (The datasheet we sent also shows a status relay that opens on a fault.) | The data points your controls or maintenance team need |
| What happens if communication drops? (covered in the register map) | The register map includes a fail-safe setting: stop the fan when the Modbus connection is lost, or leave it running, with an adjustable detection delay. | The behavior you need on loss of communication |
The control logic itself was temperature-based speed control. In the customer’s diagram, the fan stays off below 60°C, ramps linearly between 60°C and 130°C, and should hold the target point of 10,500 CFM at 0.61 in.w.g. from 130°C upward, with the transmitter’s 4–20 mA output spanning -50°C to 260°C. We also offered to take their transmitter, or whichever signal device they used, and run it with the fan in our facility before shipping samples. That test never happened. The actual input settings on the fan, such as the current at which it starts and the speed it holds from 130°C, were never configured, because the project ended before the sample stage.
For how 0-10V, PWM and Modbus control compare in general, see how to control an EC fan.
Where the Evaluation Stopped
After our answers in early July, the engineering manager asked whether they could set up a call with our engineering team. The topic was specific: “We mostly want to cover the fans programming capabilities.”
We agreed and asked for a date and an agenda. When nothing came back after a few days, we followed up in mid-July, including a note to the buyer asking whether the project was still under technical review or whether something else was holding it back. The engineering manager replied with their time zone (Eastern) and a preferred slot of 10:00–11:00 a.m.; then the buyer was out for a week, so we agreed to aim for the week after. A date was never set.
In early August the evaluation closed, before a sample was shipped and before the call took place.
Looking back, more of the thread was about control than about airflow. A candidate 910mm fan was put forward early; most of what followed was about whether it would take the customer’s control signal.
What to Put in Your RFQ for a Large EC Fan Replacement
This list comes from this one evaluation. It isn’t an industry standard, and it isn’t a full specification. Each item came up along the way, from the customer or from our side, and having all of it in the first email would have cut down the back-and-forth.
- The duty point: airflow at a stated static pressure, plus where on the speed range you want it (this customer said about 70%)
- Voltage, phase and frequency
- Mounting type and airflow direction
- Maximum noise level
- Lowest start-up temperature
- Monthly volume and the latest acceptable lead time
- The control signal source (4–20 mA, 0–10 V or Modbus) and the control logic or curve behind it
- Whether you need a speed limit set on the fan itself
- The data you need to read back
- What the fan should do if communication is lost
- Whether your existing control program is proprietary to your current fan brand
If you’re replacing a large EC fan, especially one beyond standard catalog sizes, send us the duty point (airflow at a stated pressure, and where on the speed range you want it) along with your control logic. Our engineering team will check the duty point against actual fan curves and tell you how the fan would take your control signal, before you get a quote. Our engineers reply within 1 business day.
References
- Hammond Power Solutions. What does “dry-type self-cooled force air-cooled transformer class AA (FA)” mean?
- Maddox. Guide to Transformer Cooling Classes
