IBT_2 BTS7960 Motor Driver Review - Avoid

[whatsupskip]

The IBT_2 BTS7960 is a very affordable motor driver which claims to support up to an impressive 43A over a wide voltage. It can be controlled via 3.3V logic so it looked an ideal match with the ESP8266 or ESP32. However all is not as it seems.

Looking at the design of the board, the heat sink and the terminals I never thought it would be able to handle 43A, but I was only looking to handle 20A to 25A. However even this is well beyond what the IBT_2 board could handle. The issue isn't the BTS7960 controller chips as they seem to have all the features you would desire with over current, over temperature, under voltage and short circuit protection. However it was this high level of protection on the BTS7960 that lulled me into a false sense of safety. I thought that if the heat wasn't properly dissipated by the heat sink then the BTS7960 would just shut down. This might be the case, but designed caused other problems prior to the BTS7960 shutting down.

Only after my unit failed did I find out that the IBT_2 board had many design issues. The killer problem in my situation was the lack of a direct physical and thermal connection between the BTS7960 chips and the heat sink. Yes that is right the heat from the BTS7960 has to transfer through the PCB to the heat sink. Not knowing this I had setup a temperature probe on the heat sink to see if the IBT_2 appeared to getting too hot. Well the design of the IBT_2 attempting to transfer the heat through the PCB was my down fall. I had planned to place an additional heat sink directly on top of the BTS7960, but the heat sink I planned to use wasn't being cooperative, so I put that a side for the moment. I after running the IBT_2 for several times with more than 5 minutes testing in each cycle, I noticed that the electrolytic capacitor on the motor power input look to have swollen on top slightly. I wasn't sure if this was really the case, so I left it. The heat sink temperature was running below 60C, so I thought it was all running well within parameters. I think let the set up run for a couple of hours....

On my return to the set up I found there was no current flowing to the load. I also found the electrolytic capacity had an out of body experience. At first I thought there might have been a problem with the motor power supply I was using, but I doubt this was the problem. It certainly wouldn't have been too high a voltage as it only had a 9V DC output.

After doing some testing of the IBT_2 board I found that the half of the H bridge that was active was dead, but the other half still worked. It was at this point that I found the review on Amazon for the IBT_2 BTS7960 motor driver board with the explanation of the heat sink design fault. This review also points out several other major design or manufacturing problems with the IBT_2 board that are leading to boards being short circuited when leaving the factor. It might be just me, but I think the suggested solutions are a bit too much of an effort to rectify the IBT_2 if you are intending to purchase the board. If you have already purchased the board, then that is up to you.

I still think the concept of this motor controller is a great idea and the BTS7960 chips seem to be good. The problem is the implementation. I have seen another BTS7960 motor driver board on Aliexpress, but it also lacks sufficiently large terminals. It doesn't come with a heat sink and the layout would make it difficult to place a sufficiently large heat sink in the correct location.

[TD-er]

It looks rather small to be honest, so I'm not really surprised it would fail on more than a few amps.

Just curious, what were you running at 9V/20A?

[whatsupskip]

A reversible pool chlorinator cell. I managed to have an ESP8266 running ESPEasy drive the IBT_2 and was able to change directions (done to reduce the build up on the plates). I was able to vary the output using PWM, which is needed to allow maximum output with variable NaCl concentration. The original pool chlorinator has had problems (failed multiple times), but for the moment I will have to use the original SCR, diode and variable resistor in uni directional mode.

I still can't decide if I want to attempt using the BTS7960 again.

[Ath]

I still can't decide if I want to attempt using the BTS7960 again.

The complaint is about the crappy design of the board, the controller/FET is fine, according to specs. You may want to find (or design) a better board, that allows for properly mounting a heat sink to it.

[TD-er]

I just searched for the part and I find it fascinating to see only that board and not the datasheet of the H-bridges used on it.
Here is the datasheet of just the board: https://www.handsontec.com/dataspecs/mo ... Driver.pdf

What's also rather curious to see is that the H-bridge parts can only be cooled like this through vias in the PCB.
So there must be lots and lots of vias under the H-bridge to the other side to move away all this heat.
To me it would make a lot more sense to add thermal paste on the top and mount a heat sink on there using a clamp.
For the design like this, it is very likely the H-bridge parts are not exactly level compared to each other, so maybe thermal paste isn't the best idea, maybe a thermal pad which can be a bit thicker.
Or mount one heat sink per H-bridge, but that's more chalenging to mount.

I wonder why the capacitor has burst. Maybe it got too hot?


Edit: Found the H-bridge datasheet, when adding a "B" to the part number:
https://www.infineon.com/dgdl/bts7960b- ... 782ebd6b60
https://www.infineon.com/dgdl/BTS7960_D ... 3945006d5d

See page nr 9 (PDF page 10) of the datasheet.
Worst case scenario is the low side of the H-bridge has a resistance of 18 mOhm.
At 20A, the power dissipated in there is I^2 * R = 400 * 0.018 = 7.2 Watt.
That does require quite a beefy cooling in such a small setup.

[whatsupskip]

To me it would make a lot more sense to add thermal paste on the top and mount a heat sink on there using a clamp.

The existing heat sink looks like it could be mounted on top of the packages using the existing screw holes. I am just not sure that enough heat would be transferred from the top of the package as clearly it is meant to be transferred through the metal section on the back.

For the design like this, it is very likely the H-bridge parts are not exactly level compared to each other, so maybe thermal paste isn't the best idea, maybe a thermal pad which can be a bit thicker.

They look level, but I will have to look more carefully. It may also vary between individual boards.

I wonder why the capacitor has burst. Maybe it got too hot?

I am almost certain it was heat through conduction along the PCB.

[TD-er]

Well, then the PCB must have been very very hot as those caps also don't burst while soldering.
So I guess it must have been over the rated 80C (likely range) for a long time on the entire board.
This means the H-bridges themselves were at a lot higher temperature.

[ThomasB]

Bargain priced Chinese made boards often need modification before putting them to use. That's a hidden cost I've learned to accept.

BTW, popular boards like this one are often just copies produced by clueless offshore suppliers. So there's always the chance that some IBT_2 cloned boards might be better than the one you got. That could explain why the IBT_2 has been successfully used in a lot of Arduino projects.

Regarding the blown cap, it could have been produced by a sketchy manufacturer (counterfeit, fake specs, and/or poor quality). These are a long standing issues and the reason I avoid buying caps from China based suppliers. Since I'm in the USA, Mouser, Digi-Key and Newark are my trusted "local" sources for caps.

Long story short, consider reworking your motor driver board rather than tossing it. You'll probably end up with something that is more reliable than another random motor driver board from eBay / Aliexpress / Amazon.

- Thomas

[TD-er]

Well his "low side" FETs in the drivers seem to be burned, which makes sense as those have a lower current limit and higher resistance.
So if trying again, I would do it on a new board.
And you can remove the heat sink to see how well it makes contact with the back of the PCB and also how many vias were used to transfer the heat to the other side.