asdf53 wrote on 2025-10-23, 17:53:

The left one (85U03GH) should work fine: […]
Show full quote

The left one (85U03GH) should work fine:

https://www.alldatasheet.com/datasheet-pdf/do … /AP85U03GH.html

If you want to know if a transistor will work on your PCB, look up the datasheet by googling its model number. It must be an n-channel MOSFET. Then look for one of these graphs:

Gate-to-source voltage vs. Rds(on)
Gate-to-source voltage vs. Drain-to-source current.

At a gate voltage of 4V, you need at least 4A of current, if the resistance of your motor is 1 ohm in the off state. The 85U03GH can do that, and at 4A, it would generate around 0.2W of heat, which is fine. There is a small chance that the switching speed of your microcontroller is not enough for this transistor, but I don't think it will be a problem.

If your microcontroller switches the transistor at a high frequency, the heat generated can be more than 0.2W (can be double of that or more). But it's difficult to calculate when we don't know so many variables. You should test the transistor for possible overheating when the motor is running - put your finger on it or use a thermometer.

So I could try to solder this one?

Yes, it's a good replacement for the old one. Solder it on and then test if it gets hot while the motor is running. As mentioned, it's not possible to calculate the heat dissipation before soldering it, because we don't know the switching current and frequency of the microcontroller.

Okay, I forgot something important in the calculation: The device is battery-powered, so you will have a considerable voltage drop during motor operation, and the battery voltage will drop further as it discharges. This means that the gate doesn't see the full 4V anymore, and the power losses of the MOSFET quickly add up. Assuming the worst case - the controller uses PWM switching at high frequencies and it doesn't use a mechanism to boost the gate voltage - then if the new MOSFET performs slightly worse compared to the original one, it could generate a lot heat and the battery will empty faster.

The best way to know if the replacement is still okay would be if you can identify the model number of the original MOSFET. It starts with STD60N, could it be "STD60NF55L"? That one is almost identical in performance to the MOSFET from your motherboard. The main difference is that the latter has a wider range for the gate threshold (1-3V compared to 1-2V), so there's a slight chance of it performing worse than the old one, but it shouldn't be too significant.

I manged to replace the transistor, but still not working. Now with the new transistor in place, what should I measure again to make sure if it is, or not, the same as before?

I have voltage on both motor wires ~ 4.12V.
Meaning, battery black wire + motor black wire = 4.12V. Battery black wire + motor red wire = 4.12V

How is this possible?

asdf53 wrote on 2025-10-24, 10:12:

The best way to know if the replacement is still okay would be if you can identify the model number of the original MOSFET. It starts with STD60N, could it be "STD60NF55L"? That one is almost identical in performance to the MOSFET from your motherboard. The main difference is that the latter has a wider range for the gate threshold (1-3V compared to 1-2V), so there's a slight chance of it performing worse than the old one, but it shouldn't be too significant.

It is not possible to read it anymore at all, but I am 99% sure that it wasn't STD60NF55L, because it was something shorter

boby wrote on 2025-10-24, 14:46:

I have voltage on both motor wires ~ 4.12V.
Meaning, battery black wire + motor black wire = 4.12V. Battery black wire + motor red wire = 4.12V

How is this possible?

This is normal, it happens when the transistor doesn't turn on. No current is flowing, voltage at all points is the same - you're measuring the battery voltage.

What you need to measure now is two things:

With battery disconnected - resistance between pin 1 and 2
With battery connected and switch on - voltage from pin 1 to negative battery terminal

It is not possible to read it anymore at all, but I am 99% sure that it wasn't STD60NF55L, because it was something shorter

Okay, there are some others that start with STD60NF. But let's do that later.

One more thing, can you measure the voltage of the battery when it's not connected to the board? Still 4.1V? What is the nominal voltage written on it?

asdf53 wrote on 2025-10-24, 15:16:

With battery disconnected - resistance between pin 1 and 2

95.6 kOhm

asdf53 wrote on 2025-10-24, 15:16:

With battery connected and switch on - voltage from pin 1 to negative battery terminal

4,17V

asdf53 wrote on 2025-10-24, 15:16:

One more thing, can you measure the voltage of the battery when it's not connected to the board? Still 4.1V? What is the nominal voltage written on it?

4.17V

boby wrote on 2025-10-24, 16:40:

95.6 kOhm

That's very strange. Could it be that for any of the resistance tests, you left the battery connected? Also for previous tests? Resistance must be measured with all power disconnected from the board. If there is still power, the measured resistance can be much lower than it really is.

If you're still measuring 100 kohm as before, it means that the new transistor's gate is also damaged - extremely unlikely. If you have 100 kohm at a node and remove a damaged part and it tests 100 kohm, the leakage from the node should be gone and you should get an almost infinite reading.

4,17V

Okay, this was just to confirm that you have no short circuit on the board.

asdf53 wrote on 2025-10-24, 15:16:

With battery connected and switch on - voltage from pin 1 to negative battery terminal

If you are now getting 4.17V at this point, that's a very good sign - it was 0V before. This means that now the microcontroller is sending a voltage to turn the transistor on.

With the switch on, while there is 4.17V at pin 1 of the transistor, measure the voltage from pin 3 to pin 2 of the transistor. If it's lower than 4.17V, it means that some current is flowing.

If the voltage is still high, it means that no current is flowing. Remove the battery and measure the resistance from the positive battery terminal to pin 3 of the transistor. If you get a high resistance, it means that either the diode or the motor is blocking the current.

boby wrote on 2025-10-24, 16:40:

95.6 kOhm

Back to this reading: if you are sure that the battery is disconnected, and you still get ~100 kohm from pin 1 to pin 2, do the following: Measure the voltage from pin 1 to the negative battery terminal. If it's > 0V, then short-circuit pin 1 to the negative battery terminal to discharge any capacitors, then measure resistance again.

If it is still 100 kohm, there is another possibility. Switch multimeter to beep mode and see if there is a connection from transistor pin 1 to any 100 kohm resistors on the board (they have 104 written on it). If yes, then your reading was correct, and there is a 100k pull-down resistor at the gate to turn the transistor off when the power switch is pressed.

asdf53 wrote on 2025-10-24, 16:55:

That's very strange. Could it be that for any of the resistance tests, you left the battery connected? Also for previous tests? Resistance must be measured with all power disconnected from the board. If there is still power, the measured resistance can be much lower than it really is.

Yes it is with disconnected battery.

asdf53 wrote on 2025-10-24, 16:55:

If you are now getting 4.17V at this point, that's a very good sign - it was 0V before. This means that now the microcontroller is sending a voltage to turn the transistor on.
With the switch on, while there is 4.17V at pin 1 of the transistor, measure the voltage from pin 3 to pin 2 of the transistor. If it's lower than 4.17V, it means that some current is flowing.

Sorry my bad. I get that voltage from pin1 to motor black wire. Sorry, got confused from so many things.
Still voltage between 2 & 3 is also 4.17V

asdf53 wrote on 2025-10-24, 16:55:

If the voltage is still high, it means that no current is flowing. Remove the battery and measure the resistance from the positive battery terminal to pin 3 of the transistor. If you get a high resistance, it means that either the diode or the motor is blocking the current.

It is 10 Ohm

asdf53 wrote on 2025-10-24, 16:55:

Back to this reading: if you are sure that the battery is disconnected, and you still get ~100 kohm from pin 1 to pin 2, do the following: Measure the voltage from pin 1 to the negative battery terminal. If it's > 0V, then short-circuit pin 1 to the negative battery terminal to discharge any capacitors, then measure resistance again.

It is 0V

asdf53 wrote on 2025-10-24, 16:55:

If it is still 100 kohm, there is another possibility. Switch multimeter to beep mode and see if there is a connection from transistor pin 1 to any 100 kohm resistors on the board (they have 104 written on it). If yes, then your reading was correct, and there is a 100k pull-down resistor at the gate to turn the transistor off when the power switch is pressed.

It beeps with one resistor marked with 104 and also one marked with 100

Is it possible that when you measured the resistance from pin 1 to pin 2 after you cut off transistor leg, you put the probe on the pin that was still on the board, and not on the pin that's on the transistor? The purpose of isolating pin 1 was to see if the 100 kOhm leakage comes from the board or from the transistor. Take the old transistor that you removed and measure again pin 1 to pin 2, it's probably much higher than 100 kOhm and the transistor wasn't damaged at all.

In any case, now we know where the 100 kOhm comes from, it's from the resistor which is normal. But the problem is not solved, there is still no voltage at the gate, meaning the problem is at the microcontroller or a part near it.

I remember this measurement from earlier:

Switch OFF Switch ON Pin1: 4.11V | 0.06V Pin3: 2.03V | same Pin9: 4.14V | 3.94 Pin10: 4.09V | 3.94 […]
Show full quote

Switch OFF Switch ON
Pin1: 4.11V | 0.06V
Pin3: 2.03V | same
Pin9: 4.14V | 3.94
Pin10: 4.09V | 3.94

This is interesting, because with the switch turned on, there is a quite substantial voltage drop at pin 9 and 10. For such a large voltage drop from 4.14V to 3.94V, it could mean that a large current is flowing, indicating a possible short circuit. The first thing you could try is to turn the switch ON and touch the microcontroller and any parts around it with your fingers. For example, the smaller three transistors above the microcontroller (Q2, Q3 and the other one), and the diodes. See if anything gets hot. Are pin 9 and pin 10 directly connected to the battery positive terminal?

Use beep mode of the meter to identify all points on the PCB that are connected to pin1 of the transistor. Where else is it connected to?

The next step would be to connect your multimeter in current mode to truly see how much current is flowing with the switch turned on. But that's a bit more complicated and possibly dangerous, so we'll do that another time.

asdf53 wrote on 2025-10-24, 20:17:

Is it possible that when you measured the resistance from pin 1 to pin 2 after you cut off transistor leg, you put the probe on the pin that was still on the board, and not on the pin that's on the transistor? The purpose of isolating pin 1 was to see if the 100 kOhm leakage comes from the board or from the transistor. Take the old transistor that you removed and measure again pin 1 to pin 2, it's probably much higher than 100 kOhm and the transistor wasn't damaged at all.

I am not sure anymore 🙁 The old one now shows around 2 mega Ohms

asdf53 wrote on 2025-10-24, 20:17:

See if anything gets hot. Are pin 9 and pin 10 directly connected to the battery positive terminal?

Didn't notice that anything gets hot, not sure how long should I wait.
Pins are not connected to the battery positive terminal

asdf53 wrote on 2025-10-24, 20:17:

Use beep mode of the meter to identify all points on the PCB that are connected to pin1 of the transistor. Where else is it connected to?

Without battery other connected points are those resitors around it, motor point and one pin on microcontroller.
With battery connected there are many points connected.

I dunno anymore. I think I will give up. If you want I can mail you everything for you to try, and you don't need to mail it back, as I will throw the device 😀

boby wrote on 2025-10-25, 10:47:

Without battery other connected points are those resitors around it, motor point and one pin on microcontroller.

Then locate the 100 ohm resistor that's connected to pin 1, and from the other side of it (the side not connected to pin 1), see if this resistor is connected to any leg of the microcontroller or any of the transistors and diodes around it. The 100 ohm resistor could be a soft-start resistor to make the motor turn on slowly, and it's connected to a gate drive component on the other side. Also test the resistance across the resistor itself to make sure it's not broken.

I dunno anymore. I think I will give up. If you want I can mail you everything for you to try, and you don't need to mail it back, as I will throw the device 😀

That's just how it is - when a voltage at one point is not there, you need to trace it backwards through all components that are connected to it, and that can easily involve testing 20 or more points. There's no way around that.

asdf53 wrote on 2025-10-25, 11:20:

Then locate the 100 ohm resistor that's connected to pin 1, and from the other side of it (the side not connected to pin 1), see […]
Show full quote

boby wrote on 2025-10-25, 10:47:

Without battery other connected points are those resitors around it, motor point and one pin on microcontroller.

Then locate the 100 ohm resistor that's connected to pin 1, and from the other side of it (the side not connected to pin 1), see if this resistor is connected to any leg of the microcontroller or any of the transistors and diodes around it. The 100 ohm resistor could be a soft-start resistor to make the motor turn on slowly, and it's connected to a gate drive component on the other side. Also test the resistance across the resistor itself to make sure it's not broken.

I dunno anymore. I think I will give up. If you want I can mail you everything for you to try, and you don't need to mail it back, as I will throw the device 😀

That's just how it is - when a voltage at one point is not there, you need to trace it backwards through all components that are connected to it, and that can easily involve testing 20 or more points. There's no way around that.

The problem is that there are no visible rails to easily back track it. Even on the other side, there are just few rails as I can see. Like the board is multi-layer one. Will try to go backwards from transistor, or forward from micro controller

boby wrote on 2025-10-25, 13:08:

The problem is that there are no visible rails to easily back track it. Even on the other side, there are just few rails as I can see. Like the board is multi-layer one. Will try to go backwards from transistor, or forward from micro controller

In this case, you have to go backwards - you know where the voltage ends, but you don't know where it starts. Start at the 100 ohm resistor, on the side that's not connected to the transistor (with "other side" i don't mean the backside of the board, I mean the other end of the resistor), and test where it goes. You don't need to see the rails, you can follow the path with your multimeter.

boby wrote on 2025-10-25, 13:08:

asdf53 wrote on 2025-10-25, 11:20:

Then locate the 100 ohm resistor that's connected to pin 1, and from the other side of it (the side not connected to pin 1), see […]
Show full quote

boby wrote on 2025-10-25, 10:47:

Without battery other connected points are those resitors around it, motor point and one pin on microcontroller.

Then locate the 100 ohm resistor that's connected to pin 1, and from the other side of it (the side not connected to pin 1), see if this resistor is connected to any leg of the microcontroller or any of the transistors and diodes around it. The 100 ohm resistor could be a soft-start resistor to make the motor turn on slowly, and it's connected to a gate drive component on the other side. Also test the resistance across the resistor itself to make sure it's not broken.

I dunno anymore. I think I will give up. If you want I can mail you everything for you to try, and you don't need to mail it back, as I will throw the device 😀

That's just how it is - when a voltage at one point is not there, you need to trace it backwards through all components that are connected to it, and that can easily involve testing 20 or more points. There's no way around that.

The problem is that there are no visible rails to easily back track it. Even on the other side, there are just few rails as I can see. Like the board is multi-layer one. Will try to go backwards from transistor, or forward from micro controller

Well, that is the problem too, as it beeps on both side of 100 resistor. However, it beeps only on one side, on the side of 104 resistor. Does this tell anything useful?

boby wrote on 2025-10-25, 17:23:

Well, that is the problem too, as it beeps on both side of 100 resistor. However, it beeps only on one side, on the side of 104 resistor. Does this tell anything useful?

Oh, I see. The resistor marked "100" is a 10 ohm resistor, and most multimeters trigger a beep with any resistance of < 30 ohm. In this case, it does not matter which side you choose. Put one probe on any side of the 100 resistor, and with the other probe, test for a connection to any microcontroller pin, transistors or diodes.

If it's only connected to the microcontroller, then it could be that the microcontroller itself is damaged. Look at the underside photo of the board, I can see corrosion at the pins of the battery connector and the pins of the large electrolytic capacitor, and their solder points look rough, pitted. Also the underside pad of the large transistor. If these points have been subject to voltage spikes that came from the motor, it means that all other components on the board could potentially be damaged as well.

You can do one more test - with the battery disconnected, measure the resistance from the positive to the negative battery terminal. If it's low, it means you have a lot of leakage current, possibly from damage to the board.

asdf53 wrote on 2025-10-25, 17:43:

Oh, I see. The resistor marked "100" is a 10 ohm resistor, and most multimeters trigger a beep with any resistance of < 30 ohm. In this case, it does not matter which side you choose. Put one probe on any side of the 100 resistor, and with the other probe, test for a connection to any microcontroller pin, transistors or diodes.

Resistor marked with 100 is connected (beside the transistor), only to pin12 at micro controller, nothing else.

asdf53 wrote on 2025-10-25, 17:43:

You can do one more test - with the battery disconnected, measure the resistance from the positive to the negative battery terminal. If it's low, it means you have a lot of leakage current, possibly from damage to the board.

01.3 when I set the dial to max 200 mega ohms, other values can't measure it

boby wrote on 2025-10-25, 19:00:

Resistor marked with 100 is connected (beside the transistor), only to pin12 at micro controller, nothing else. […]
Show full quote

asdf53 wrote on 2025-10-25, 17:43:

Oh, I see. The resistor marked "100" is a 10 ohm resistor, and most multimeters trigger a beep with any resistance of < 30 ohm. In this case, it does not matter which side you choose. Put one probe on any side of the 100 resistor, and with the other probe, test for a connection to any microcontroller pin, transistors or diodes.

Resistor marked with 100 is connected (beside the transistor), only to pin12 at micro controller, nothing else.

asdf53 wrote on 2025-10-25, 17:43:

You can do one more test - with the battery disconnected, measure the resistance from the positive to the negative battery terminal. If it's low, it means you have a lot of leakage current, possibly from damage to the board.

01.3 when I set the dial to max 200 mega ohms, other values can't measure it

Okay, then you have no leakage (at 200 megaohm setting, the meter generates a larger test voltage, turning the board components on which leads to a false low reading).

But there is still current flowing from pin 9 and 10 of the controller. I have another theory - if the controller uses a capacitor charge pump to boost the gate voltage, a shorted capacitor could lead to the current loss and no gate voltage. With the battery out, test the resistance across all SMD capacitors near the microcontroller (C2, C4, C6, C8 and any others). If any of them is unusually low (<100 kOhm), this could also be the problem. And also test the diodes near those capacitors in diode mode.

Edit: Thought about it and that theory is likely wrong. If the gate pin is connected to the controller, it has to be driven directly from it, there can't be an external boost circuit involved.

I'm out of ideas for now, sorry.

C2 around 130 kOhm
C6 not sure but more than 2 MOhm
C8 around 18 MOhm
C4 surprisingly around 1 kOhM

I hope I measured correctly.

I was also able to trace the connection from transistor pin 1 to micro controller pin 12. There is nothing in between except those resistors marked with 100 & 104.

boby wrote on 2025-10-26, 19:44:

C2 around 130 kOhm C6 not sure but more than 2 MOhm C8 around 18 MOhm C4 surprisingly around 1 kOhM […]
Show full quote

C2 around 130 kOhm
C6 not sure but more than 2 MOhm
C8 around 18 MOhm
C4 surprisingly around 1 kOhM

I hope I measured correctly.

I was also able to trace the connection from transistor pin 1 to micro controller pin 12. There is nothing in between except those resistors marked with 100 & 104.

The resistances look good (1 kOhm is to remove noise from power switch).

Back to pins 9 and 10. If the voltage drops when you turn on the switch, it means that current is flowing somewhere, and you need to investigate how much and where it's going to.

First test, not 100% sure if it works but try it: Put multimeter in diode mode. With battery connected and the switch off, test from controller pin 12 to battery negative. Repeat with the switch on. In theory, this will allow you to test if pin 12 is actively pulled low (low reading) or not (high reading).

Next, measure the current:

With the battery disconnected, write down which microcontroller pins are connected to the positive battery terminal. If pins 9 and 10 are not among them, find out where they go as well.

With battery connected, on one of the controller's supply voltage pins, measure the voltage. Turn the switch on, measure again. The measured voltage difference will allow you to calculate the approximate current, but if the current is too small and your meter is not sensitive enough, this will not work. Then you need to measure current directly with the multimeter in ammeter mode:

Board positive connector > battery positive
Board negative connector > multimeter positive lead > multimeter negative lead > battery negative

Another thing you should test is the large electrolytic capacitor. If it's dead, then the large initial current draw of the motor can cause the voltage at the microcontroller to sag, causing it to reset and getting stuck in an infinite loop. This would be consistent with the lower supply voltage you measure after turning the switch on. You could test this theory by holding a second capacitor on the battery connector pins while turning the switch on.