I found the LM5164 would blow up rather easily when connected directly to the DC Bus. Usually a diode is placed inline with the input of a buck when used on a motor controller so the input to the bucks are not used as part of the DC Link. Also the diode stops current flowing back through the buck in an uncontrolled fashion, which can potentially damage it and also powers up other devices on the DC Bus. This is especially a problem when voltages might remain on external inputs, USB, etc. The smallest package for the INA181 is SOT-23, so it's not those, not really that important for low side shunts what it is exactly as long as its fast enough.

I just realised, but those fets seem to be the same fets used as on the Flipsky 75100 & 75200 Alu PCB (HYG015N10NS1TA) edit and the EBMX

That 1AAD sot23-5 part you mentioned in the start is INA180A4 current sense amplifier by TI. Recently I used same part for my drone esc.

@de-bodgery

Жыл бұрын

Weird! Why would you need a current sense at the IO connector? It's not even close to the shunts.

@de-bodgery

Жыл бұрын

I think maybe you mean the 1DV?

@rahulkushwaha9500

Жыл бұрын

​@@de-bodgery with a fixed gain of 200 and those two 1ohm resistor in parallel near that IC, I would guess they must be monitoring some low voltage maybe 5V or 12V supply that they might have broken out into that connector. Maybe the output of one of the buck converter directly connects to the outside world and they are monitoring that to not overload the buck and shut it down if overload or short-circuit occurs. There is a thick track from the dc dc converter near that LDO to the connector.

@de-bodgery

Жыл бұрын

@@rahulkushwaha9500 So you think the 2 parallel resistors are acting like a current sensing shunt? It can't be much current.

@teamtriforceuk3985

Жыл бұрын

@@rahulkushwaha9500 Weird way to protect the output, there are purpose built e-fuse chips for this purpose which are very cheap.

Would you mind taking a look at these SFN60 Chinese LEDs? They seem pretty impressive

@de-bodgery

Жыл бұрын

This the only thing I found with that LED in it. www.aliexpress.us/item/3256805014801905.html?spm=a2g0o.productlist.main.1.89beQReLQReLJk&algo_pvid=4b6a75fd-51cf-4582-9b0b-086fb070077c&algo_exp_id=4b6a75fd-51cf-4582-9b0b-086fb070077c-0&pdp_npi=3%40dis%21ILS%21702.36%21702.36%21%21%21%21%21%402122457116831157124384927d0798%2112000032381896211%21sea%21US%21754968738&curPageLogUid=TVIoRX3hbwXD

The caps are on the top side of the shunt so they don't have the added inductance of the shunt. The capacitance will become less effective with every extra nH of inductance between the caps and fets.

@de-bodgery

Жыл бұрын

I think this is the only application of ceramics like this that I have ever seen. If what you say is true, what aren't others doing it too? Nothing from Trampa does this and that's about the only other controller I can think of with this level of ceramics.

@teamtriforceuk3985

Жыл бұрын

Yes it is better for reducing ringing, however it causes a huge spike on the current amplifier output which can cause wrong current measurements or false trip overcurrent protections. It's mainly a problem at high duty cycle.

@de-bodgery

Жыл бұрын

@@teamtriforceuk3985 I was posting this almost exactly to someone on Facebook earlier today.

This decision to install ceramic capacitors is primarily due not to the "fast" capacitance, and not to the mediocre RC filter, since the resistance of the shunts is a resistor

@Evgeniy_6KWT

Жыл бұрын

But at the same time, we have another problem - namely, the accurate measurement of the current, and possibly surges that can lead to the operation of the current protection

So what i see it is 375Aph controller, not 500 as claimed. Well, could be 500 either with small battery current or with exploded mosfets because of temperature runaway

@de-bodgery

Жыл бұрын

I'd say that's pretty close to my estimates too.

@Evgeniy_6KWT

Жыл бұрын

Ну почему же. Тут 500А будет, 0.3Сек но будет. Психи на самокатах 6ф убоксов снимают по 300А в пиках)

@vasilisk8944

Жыл бұрын

@@Evgeniy_6KWT but i cant make it like that if i want to have reliable device ;D

@de-bodgery

Жыл бұрын

@@Evgeniy_6KWT .3 seconds at 500 amps is not very useful! It's not even enough time to crank the throttle before you have to let off! 500 amps is just not realistic! It's hype marketing!

what's is your estimated max amp on this ?

@de-bodgery

3 ай бұрын

Something you can run at for minutes and miles at a time? 300 amps

In these videos of yours, it's clear that you're trying to find something negative about the torp controller, but you haven't really succeeded yet. Saying that the capacitor floats on a resistor of 0.1 micro ohms is total nonsense. 0.1 micro ohms is such a small resistance that it can't even be considered a resistor. If you understood how it works, you would know that for a high noise filter, inductance is much more important, and resistance doesn't have much impact, unless it's too high that it limits the current, which is not the case here at all. The resistor is a bit wider than the trace, and there's even a possibility that this reduced the inductance and therefore it performs better than if they had traced directly to GND.

@de-bodgery

Жыл бұрын

Repeating what I said in the video: The ceramics above the shunts was pointed out to me by a VESC engineer. Secondly: I'm always looking for negative aspects in any tear down I do. How is this new for the Torp controller? Thirdly: Voltage drop across the shunts is directly relational to current. At higher amps the caps "float" more.

@vasilisk8944

Жыл бұрын

@@de-bodgery i am okay with this weird capacitor-shunt scheme in regard of capacitor work. what is i dont like is that how these capacitors will affect measured current over shunt and how much extra ringing gonna be there

@de-bodgery

Жыл бұрын

@@vasilisk8944 AKA...don't do this.

@teamtriforceuk3985

Жыл бұрын

@@vasilisk8944 Yes it causes a huge spike on the current amplifier output which can cause wrong current measurements or false trip overcurrent protections. It's mainly a problem at high duty cycle.

@Evgeniy_6KWT

Жыл бұрын

@@vasilisk8944 ооо, не неожиданно тебя тут видеть:)

Torp is doing a way better job than KO Moto. Can it produce 600 pA at 50% FW.

@de-bodgery

Жыл бұрын

The power stage is pretty damned efficient! The logic board is VESC. OF COURSE it's way better than KO! With only 2 mosfets in parallel however, 375 amps is about all you can get from it continuous. I HATE that they hobbled VESC to make this thing! Many more people could be using it if they didn't!