I’m new to the T200 thruster and I am testing my Arduino control circuit before mounting the final setup on my surface vehicle. My power source is a 36V, 20Ah Lipo E-bike battery. My intent is to run 2 T200 thrusters at 18V by using a buck converter with each of the two Basic ESCs to step the original battery voltage down. The battery is capable of supplying 30A, so my plan was to limit the thrusters to 15A each. These are the max values that each buck converter is set to supply: 15A at 18V.
For my test, I am running a single thruster/Basic ESC setup and controlling the throttle with a 10K pot and Arduino Mega. After looking at the data plot that showed current draw relative to the PWM input values, I set the input value range from 1184 to 1816 microseconds. This range looked as if it would give me ~15A in the forward and reverse directions on the 18V curve.
When I attempt to test the thruster, I can reach roughly 50% of full throttle (~6 to 7A of current draw) but then the thruster appears to “bog down” before completely coming to a stop shortly after. The thruster emits a beeping noise similar to the initial arming sound and will not re-arm until the pot is returned back to its center “0” position.
My thoery as of now is that the thruster is starved of current. I’ve tried it with both of the thrusters that I have and both exhibit the same behavior. I also tried to adjust the buck converter in the test circuit to the maximum amperage setting (20A), but this didn’t solve the problem either. I figured that I wouldn’t achieve completely nominal results, but I also figured I would at least be able to exceed 50% throttle.
I apologize for the long-winded post. Any insights as to why the thruster is cutting out would be hugely appreciated.
Power conversion maintains power (minus some inefficiencies), so the current capacity of the input supply doesn’t necessarily match the possible current output from the converter. To determine the output current capacity, consider:
If you have a 90% efficient conversion from 36V to 18V then the 30A max supply output at 36V gets converted to \frac{36\times 30}{18} \times 0.9 = 54A max converter output (assuming the converter is capable of that, and there’s a load to draw it). In your case the two converters can supply 20A each in their max configuration, which requires a conversion efficiency of just 66.7%, so that should definitely be achievable (at least assuming steady state, and you may need to ensure the converters are sufficiently cool).
Seems reasonable, and good to see that you’ve considered the performance specifications in your evaluation
Are you testing in or out of water? The thrusters aren’t intended to run at high speeds or for extended periods outside water.
How fast are you ramping the throttle? If the converter doesn’t react fast enough to an increased current draw then the ESC can end up under-powered and could potentially restart (which sounds like what could be happening here, given the behaviour)
Do you happen to know the switching frequency of the converter, and how much smoothing it’s capable of? These can both contribute to the converter’s transient response/reaction time, which can effect the ESC (as in the previous point).
1.) I was testing one of the motors in the air, but upon doing some research, I saw that this could cause problems, so I set up a test tank. The air-tested thruster didn’t show this “cutting out” behavior while it was being operated in open air and since both thrusters showed the same behavior while being tested in the water, I am hopefull that the issue doesn’t lie with them.
2.) I was ramping the throttle at moderate speed, but I did notice that the cutout would happen sooner in the power curve if I ramped the throttle faster. It seems that you might’ve identified the problem. The ESC might be demanding more power before the converter can overcome the rise time. Maybe there’s a way to write a delay into the control code to compensate for the converter lag?
3.) The MOSFETs that the buck converters use are NCE8290s. Their data sheet is readily available online, and as far as I could tell, the switching frequency is listed as 1.0 MHz. The converter’s product page also lists a 50 mV voltage ripple at the output. I’m unsure if either of these would be problematic for the ESC, and I’m also not sure if there’s a way to smooth the output without an additional device.
When I tried to increase the throttle in a slow and controlled manner to avoid the suspected converter lag, I’m still running into the same issue. I did however find that the slow throttle increase resulted in more of a “dead-band” than an immediate cutout. The thruster would start accelerating normally until the current draw reached just over 4A. After that, the current draw would drop by about half to a little over 2A, and the thruster motor would sound as if it were laboring. This would continue as the throttle was increased, with the thruster making a higher-pitched noise, but not producing much more thrust before finally stopping.
Any other ideas as to what the problem could be? Much appreciated!
At least theoretically I don’t expect those characteristics would cause issues, but it’s hard to know without studying the performance and doing a more in-depth analysis.
Maybe something to do with resonances, or perhaps the converters or ESCs are overheating (both of which would be surprising given their ratings, the measured current levels, and the reasonably short time periods they’re active for in those measurements. Those ideas are shots in the dark though, and could very well not be the issue.
I’ve asked internally in case someone else has a better idea, but at this point I don’t have any particular suggestions beyond trying a battery/power supply that doesn’t require a converter, or potentially adding some decoupling capacitors to smooth out noise. If you’ve got access to one it may be worth probing the power wire with an oscilloscope to see if there are large fluctuations in the voltage level. It may even be possible to see something with a multimeter if the fluctuations are significant enough.
Can you share what buck converter you are using for this? It will be helpful to look at a datasheet.
I’ve run the thrusters off of a buck converter a few times. Most recently, I used a TDK I6A4W020A033V-001-R converter mounted on a custom PCB. I can share the PCB if you’re interested.
From what you’re saying, it sounds like the voltage output of your converter is quickly dropping off after 7-8A. I would recommend measuring the voltage output of the converter while running your test.
And, to be clear, you’re using one buck converter per thruster, right?
I thank you both for putting in so much effort to solve my problem.
Eliot, both the ESC and buck converter that I used for the test were essentially cool to the touch afterwards, so I don’t think that overheating is the issue.
Rusty, that’s correct, I’m using one buck converter for each thruster. The buck converters that I’m trying are cheap and are listed as “diymore” on Amazon. I doubt that diymore is the actual manufacturer, and their site doesn’t provide much in the way of datasheets, but here’s the link if you want to look:
My next course of action will be to monitor the voltage at the converter output as both of you suggested. I also do have a DC power supply that I can use without the converter, so I’ll give that a shot too.
I believe that I’ve isolated the buck converters as being the source of the problem. I tried measuring the converter output voltage as you suggested, and the voltage would fall to around 8 or 9V when it started cutting out.
After trying the same test with a DC power supply, I was able to use the full-throttle range.
At full throttle with 14V, the thruster launched water out of the test tank and got my socks wet. It was awesome.
I ordered a more robust buck converter with an output capacity of 30A at 13.8V. I’m going to try and get it working with the battery next.
Thanks again!
Nick
Final update: The new converter that I bought worked! For anyone reading this in the future, the takeaway here is don’t skimp on your buck converter if you’re using one.
Hi There! I found your post interesting, and I am asking you for a link now to the more robust 30A buck converter. I’m using T100 thrusters in my “FloatCam” rig, and unless I direct-wire individual 3S LiPos to the ESC’s I get the cut-out issue. I’m now using the 36V e-bike battery like you, and just back from my test … reaching near full throttle causes cut off/shutdown.
For the T100 thrusters I “should” have plenty of power using two of those buck converters … so I’m giving that a shot. Since my rig is basically designed to get a camera close to wildlife without scaring them … well, I would prefer not to need 100% throttle so much.
I’ll report back after I’ve rigged the replacements.
Sorry for the delayed response. I was using some cheaper adjustable 20A buck converters when I ran into those issues too. I still ran into trouble when I used one converter for each of my two thrusters.
If doubling up on the buck converters still doesn’t solve your issue, here’s the link to the more robust converter:
The voltage is fixed at 13.8V, but it is able to get both thrusters to full throttle without issue. The internals also potted, so it’s waterproof. Let me know if you have any other questions. Good luck!
Thanks Nick! Appreciate the link. I jumped the gun last night and purchased the two additional 20A units because this rig is for filming wildlife, and weather and opportunity might not wait. I’m crossing my fingers that they will work for me … in fact I plan to use all three … one for each thruster/ESC, and the current one just for the radios, rudder servo, and drone brain (using a DJI A3+Lightbridge to run a X5R).
Also thinking that perhaps since I’m using T100’s instead of T200’s, they draw less?
Did that one buck converter work for both your thrusters or did you patch two?
Seems like a cool project you have going! You may be right about the T100s, my understanding is that they will draw less current. A single buck converter that I sent you a link to powers both of my T200s. I have the throttle limits set so that each thruster will draw 15A at full throttle, to use the battery’s full output capacity. I’m powering the brains and other periphery electronics with separate batteries.
The T100’s max current at 12V should be ~11.5A, compared to ~17A for a T200 at the same voltage, so yes the T100 draws less. Note that 2 x 11.5A > 20A, so not too surprising that you’re running into issues at high throttles. Note also that the T200 is more efficient than the T100, so if the throttle is limited to draw 11.5A@12V its static forward thrust is ~2.9kgf compared to the T100’s ~2.36kgf (which is ~23% more).
In general, larger thrusters can draw more power but are also more efficient than smaller ones when operating with the same power draw, assuming they have a speed controller that can drive them well at a slow enough speed.
If efficiency and minimal noise/disruption are key for your use-case then you might want to look into our new/upcoming T500s, although general availability is still a month or few out at this point. Power draw at 12V is similar to that of the T200, but with ~60% more forward thrust and ~40% more reverse thrust, and larger improvements at higher operating voltages.
