Last March, I was sitting in our parts room with a brand-new Bafang H730 hub motor on the bench and a bafang rear hub motor wiring diagram on my phone. The diagram didn't match the motor's cable. That's when I knew the order I'd approved two weeks earlier was going to cost us more than the invoice.
I've managed procurement for a 14-person e-bike retrofit company for six years. Our annual component budget is around $180,000. I can quote you the difference between Bafang's older 9-pin and 13-pin connectors from memory. So when we signed a contract to convert 30 delivery bikes for a local courier network, I thought I had it handled.
The contract that started it
The courier wanted rear-hub motors, not mid-drive. Their logic: less chain wear and no risk of a rider shifting under load. They specified the Bafang H730 hub motor because it handles heavier loads and, in their words, 'is the boring reliable choice.' My lead tech said, 'You need the 48V version, and don't cheap out on the cable.' I nodded and went back to my spreadsheet.
In my head, the order was simple: 30 motors, 30 controllers, one spare harness, and a set of LM8LUU linear bearings for the battery slide mounts we were building. I also decided to add a timing belt replacement for our motor test stand. It had been slipping for a while, and I figured I'd kill two birds with one purchase order.
I knew the bafang-motor ecosystem well enough to source mid-drive kits in my sleep. But the H730 was a different animal, and I treated it like it wasn't.
The order I almost got right
Here's the thing about being the cost control person: my job is to find the best total cost, not the lowest price. But somewhere between comparing vendors and a launch deadline, I slipped back into old habits.
I found a supplier offering the H730 at a decent price. But their wiring harness was a separate line item, 30% cheaper than the official Bafang harness. The generic timing belt was also 40% less. And the LM8LUU bearings were on sale. (Note to self: a sale is not a saving if you don't need the parts.)
I typed 'what size is lm8luu linear bearing' into the supplier's search bar. The answer came back fast: 8mm bore, 15mm OD, 24mm length. I added them to the cart. That part, at least, was correct.
To be fair, the generic wiring harness looked fine when it arrived. The bag was sealed, the connector clicked, and the label said 'Bafang-compatible.' From the outside, a wiring harness is a wiring harness. The reality only showed up when our tech tried to bench-test the first motor.
Induction motors set the wrong expectation
Our shop also uses three-phase induction motors for the dust collector and the air compressor. Those are boring in the best way. Standardized terminal blocks, readable nameplates, and a wiring diagram that matches the actual product. If you know how to wire one induction motor, you can wire most of them.
That set my expectations for the H730. But a Bafang hub motor is not an induction motor. It's a brushless DC motor with a controller protocol, a speed sensor, a thermal sensor, and a brake signal all packed into one harness. The pinout matters. The wiring diagram isn't a nice-to-have; it's the difference between a 20-minute install and a three-day investigation.
The wiring diagram problem
Our lead tech connected the H730 to the controller, powered it up, and got nothing. No spin, no beep, no fault code. He checked the bafang rear hub motor wiring diagram I'd downloaded from the supplier's listing. It didn't match the motor's cable.
I called the supplier. 'That diagram is for the older H700,' they said. 'The H730 has a different connector layout.' And they didn't have an updated diagram in English. Honestly, I'm still not 100% sure why their diagram was wrong. My best guess is they copied it from an older model and never checked.
I called Bafang directly. Their support team sent the official H730 diagram within a day. (I should have done that first.) It turned out the speed sensor and thermal sensor shared a ground pin in a way that wasn't obvious from the supplier's photo. The generic harness had one wire flipped. One wire.
The timing belt replacement I also got wrong
You'd think I'd stop there. I didn't.
The timing belt replacement on our test stand was supposed to be a five-minute task. I ordered a belt from the generic catalog without measuring the old one. When the new belt arrived, the tooth pitch was 8mm. The old belt was 5mm. It fit on the pulleys but slipped under load. On a test stand that runs motors at peak torque, a slipping belt is a health hazard. We stopped that line and lost another afternoon.
That one was on me. I knew better. A timing belt is not a 'one size fits all' part; it needs pitch, depth, width, and tooth count. Measure first, order second.
The real cost of a bargain
Let's do the math, because that's my job.
The generic harness was $38 per unit. The official Bafang harness was $54. For 30 bikes, the difference was $480. That's real money, but it was nothing compared to the delay.
We lost three days. We paid $200 for expedited shipping. The courier contract had a late-delivery penalty. Our lead tech spent 18 hours troubleshooting instead of building. When I added it all up, the 'savings' on that harness cost us about $4,300 in labor, shipping, and penalties. The timing belt added another $30 in replacement cost and a half-day of lost shop time. The LM8LUU bearing was the right size all along. The bracket tolerance was the actual issue.
This is the part people forget: a 'bargain' only wins if it doesn't come back later. Most of the time, it comes back as a service call or a rework order.
What I'd do differently
First, for any new SKU—especially anything connected to a motor—I now require the official documentation before I approve the PO. If a Bafang hub motor doesn't come with a wiring diagram that matches the motor's serial number, we don't buy it from that source. That alone would have caught the H730 issue.
Second, we changed the parts intake process. Every mechanical component, from a timing belt to a bearing, gets measured against a spec sheet before it goes into inventory. If the seller can't provide a spec sheet with tolerance data, we pick a different seller. This sounds obvious, but when you're rushing, it's the first thing to slip.
Third, I updated our cost tracking spreadsheet to include a 'failure impact' column for critical parts. The formula is simple: unit price + expected failure cost + expected delay cost. For a $54 harness that could save $4,300 in risk, the choice is clear.
Granted, this process is more work than just clicking 'buy.' To be fair, it's also the work that keeps us from buying the same part twice. In my opinion, that's what efficiency actually means: not cutting corners, but eliminating the corners that cause rework.
The takeaway
If you're sourcing bafang-motor components, especially the Bafang H730 hub motor, don't rely on aftermarket listings for wiring diagrams. Go to Bafang's official documentation and verify the pinout before you order. As of May 2024, the official H730 page included the current wiring diagram as a PDF. Check it before you spend money on a generic harness.
If you're replacing a timing belt, measure the pitch, depth, width, and tooth count of the old belt. Don't guess.
And if you're wondering what size is lm8luu linear bearing: 8mm bore, 15mm OD, 24mm length. But the size isn't the whole story. Check your bracket tolerance before you blame the bearing.
This approach worked for us because we're a small shop with repeatable orders and a clear process. If you're a one-person garage doing a one-off conversion, you might not need this much paperwork. But the core lesson scales: the cheapest part only wins if it doesn't cost you time, labor, or trust.
I still believe in efficiency. I'm the guy with the cost calculator. But efficiency isn't buying the cheapest harness. It's doing the boring, up-front work so that when you plug in that H730 hub motor, it spins on the first try. That's the version of cost control I'm actually proud of.
Discuss this article with engineering