need help with my testing circuit (Solved)

Rasmus Godske

Member
Joined
Jun 7, 2018
Messages
39
I'm working on my testing circuit, for testing my cells but it have some weird behavior that I can't figure out...

When connecting my 5V(actually 5.8V) powersupply to the VCC and ground and no battery connected,the voltage at Shunt+ and Shunt- appears to be at1.7V with lots of noise(Picture 2). This is supposed to be 0V, when ChargeEn is a 0V and the mosfet is "closed", so I can detect when no batteries are connected. Also the red light indicating when the it's charging, are lighting up veery slightly, so it's clearly getting some current to run. I got the idea of enable/disable the TP4056 charger from BrettWatty, so I imagined it would work when I break the connection between IN- and ground with the mosfet.

Also when connecting a battery the TP4056 board heats up quite a lot, too hot to touch. It does seem to charge though, but it shouldn't get that hot. When using a USB to as power supply, it doesn't heat up anywhere close to that.


I have spent many frustratinghours trying to figure this out, but can't seem to find the cause of these problems... So I'm asking for help to find the cause of the strange behavior of my circuit.


Picture 1:
Circuit

image_nowvbe.jpg


Picture 2:
Oscilloscope with showing the voltages at shunt+(Yellow) and shunt-(Blue), while no batteries connected, and DrainEn, and ChargeEn is at 0V.
Frequency = ~100Hz
Amplitude =~300mA


image_nrcwib.jpg
 
I guess the problem is you switch the Ground of the TP, that doesnt stop it to suplly some voltage/Current into the other circuitry.
If the 1.7 V vanish when you cut the 5.8 V input to the TP it should be this.
Ifd it is, you have to switch plus or move the ground to the TP output.

Dont know why the TP heats up, that would need more info. Perhaps the 5.8 volt?
 
The 5.*V input should be not problem for the TP4056.

I looked up the pinouts for the TP4056 chip & the numbers on your diagram don't look right. See
https://dlnmh9ip6v2uc.cloudfront.net/datasheets/Prototyping/TP4056.pdf
Are the numbers in your diagram for a module vs the chip?

Is the ground points of the three sections called "drain" + "reverse polarity protection" + "18650 battery holder" connected to the same ground point as the charge enable FET Q8?
If so you probably have a path via the TP4056 even when draining.

I would not control the TP4056 like Q8 is shown, I would connect a 5V input via a 1k resistor to the TP4056 "CE" pin (charge enable), +5V = charging enabled, 0V = charge disabled.
I'd have pin 2 "IN-" connected to ground (0V)
(pin #'s from your diagram)

You might have damaged FETs - you can blow them easily with static electricity >20V on the gate.

LED1 should be dark at all times unless you put +5V on the "DrainEn" input.
With no power supply attached but battery inserted & with "DrainEn" input = 0V or disconnected, "Shunt-" should measure = 18650 cell battery +ve voltage.
 
Cherry67 said:
I guess the problem is you switch the Ground of the TP, that doesnt stop it to suplly some voltage/Current into the other circuitry.
If the 1.7 V vanish when you cut the 5.8 V input to the TP it should be this.
Ifd it is, you have to switch plus or move the ground to the TP output.

Dont know why the TP heats up, that would need more info. Perhaps the 5.8 volt?

Well if there isn't a place for the current to go, then the TP4056 board wouldn't get any power, but I guess the still current still have a path through the resistorR17 maybe, but if the voltage at B+ was 5V passing through R17 would be 5V/10k = 0.5mAh, which I imagine isn't the problem...

I could try with a mosfet on IN+, but I don't think I have any that I can use right now...

About the 5.8V input voltage, as Redpacket says I don't think that would cause it to heat up..


Redpacket said:
The 5.*V input should be not problem for the TP4056.

I looked up the pinouts for the TP4056 chip & the numbers on your diagram don't look right. See
https://dlnmh9ip6v2uc.cloudfront.net/datasheets/Prototyping/TP4056.pdf
Are the numbers in your diagram for a module vs the chip?

Is the ground points of the three sections called "drain" + "reverse polarity protection" + "18650 battery holder" connected to the same ground point as the charge enable FET Q8?
If so you probably have a path via the TP4056 even when draining.

I would not control the TP4056 like Q8 is shown, I would connect a 5V input via a 1k resistor to the TP4056 "CE" pin (charge enable), +5V = charging enabled, 0V = charge disabled.
I'd have pin 2 "IN-" connected to ground (0V)
(pin #'s from your diagram)

You might have damaged FETs - you can blow them easily with static electricity >20V on the gate.

LED1 should be dark at all times unless you put +5V on the "DrainEn" input.
With no power supply attached but battery inserted & with "DrainEn" input = 0V or disconnected, "Shunt-" should measure = 18650 cell battery +ve voltage.

The numbers on the TP4056 symbol, is most likely wrong since, i made the symbol myself on EasyEda.

Yes all of the ground symbols are meant to be the same. Well when discharging I setChargeEn to 0V, to cut that ground path, if not the chip would attempt charging while the battery is draining.

I used the "CE" pin in my previous design, but I thought it was quite hard to solder a wire to it, so I went for this instead, also I want this to be as simple to make as possible. I want this to be something other people new in the hobby easily can create by themselves.

Hmm never thought of that actually, I thought they could take quite some abuse before getting damaged? I will try changing them, just to make sure.

I wasn't referring to the LED1, but the red led on the TP4056 board. The LED1 works as expected ::)

That's correct and that's also how it behave in that case, the problem occurs when the 5.8V powersupply gets connected...

Could the Q7 drain mosfet be left opened even though the DrainEn is at 0V? I'm wondering if that draws power from the TP4056, which makes it hot. Can I test if Q7 is opened or not?
 
Q7 will be open (ie off) when DrainEn is 0V and no current should be flowing via PR2.

After your response above, I'm thinking that if you do not have a 0V to the TP4056 and you have Q8 = off with ChargeEn = 0V, then the TP4056 doesn't work properly and is passing "charging current" to the battery.
The chip is confused & still "charging". Letting the ground float like this is a bad idea.
This happens because you have 5.8V on one side & the battery at eg 4V on the other but no proper connections & reference voltages.
You are not grounding the TP4056 properly when you are trying to disable charging. Use the CE pin or disconnect +ve to the TP4056.

Make sure the pin numbers do match properly - nothing works for long if you mix up the pins ;-)
 
Redpacket said:
Q7 will be open (ie off) when DrainEn is 0V and no current should be flowing via PR2.

After your response above, I'm thinking that if you do not have a 0V to the TP4056 and you have Q8 = off with ChargeEn = 0V, then the TP4056 doesn't work properly and is passing "charging current" to the battery.
The chip is confused & still "charging". Letting the ground float like this is a bad idea.
This happens because you have 5.8V on one side & the battery at eg 4V on the other but no proper connections & reference voltages.
You are not grounding the TP4056 properly when you are trying to disable charging. Use the CE pin or disconnect +ve to the TP4056.

Make sure the pin numbers do match properly - nothing works for long if you mix up the pins ;-)

This is exactly what i had said( or wanted to say), and I second that again.
 
After giving up on the idea of using Q8 to cut the connecting between "IN-" and ground and instead using the "CE" pin of the TP4056 chip as you Redpacket suggested it finally works. Both the noise and 2V on the postive side of the battery terminal disappeared, and the voltage is now 0V when no battery is connected. So you guys are right, this was the solution :D Although I hoped I could avoid solder a small wire to the TP4056 pin which is way to small to my soldering precision :p

I'll try to find a mosfet that can be placed between the VCC and the "In+" of the board, for now it works so I'm more than happy!

I've been working on this small print for over multiple weeks now, and I'm so appreciative for your help!
 
RasmusGodske said:
I'll try to find a mosfet that can be placed between the VCC and the "In+" of the board, for now it works so I'm more than happy!
You probably have them. :)

(Nearly) every Laptop BMS has N and P Fets in equal numbers. Conveniently as SO8.... and with 10+Amp rating.
 
I only have IRLZ44N, IRF540 and FQP06 I think and I haven't saved any of the laptop battery circuitry since we're very limited on space in our 50m2 apartment. My girlfriend are already complaining about all my electronics :p but when I get back to school I'll see, if I can find a suitable one in the lab :D
 
Good job there :)
Like Cherry67 says to switch the +5.8 line a Pch MOSFET will be what you need (the IRLZ44N, IRF540's are Nch). The FQP06 might be Pch but I couldn't find specs. You have a 10k resistor from the gate to the +5.8V & pull it to ground to switch on.
 
Redpacket said:
Good job there :)
Like Cherry67 says to switch the +5.8 line a Pch MOSFET will be what you need (the IRLZ44N, IRF540's are Nch). The FQP06 might be Pch nut I couldn't find specs. You have a 10k resistor from the gate to the +5.8V & pull it to ground to switch on.

FQP6 seems to be an incomplete number, see:

https://www.promelec.ru/pdf/FQP6N25.pdf

If your FQP06 is P or N depends on the following numbers and chars, like

FQP6Nxx is a N type.
FQP6Pxx is a P Type.
A well known decades old pattern of marking Fets, with the P6 identifying the current rating, the xx the voltage.

You might find this little gadget useful, you get it everywhere:
https://www.ebay.com/itm/LCR-T4-Meg...ec:m:mLakYynk0RqdZT62qrYQfdw&var=521083969550

The cheap and useful swiss-army knife for testing electronic components (and it would tell you the fet type :) ).

Be wise and buy one with casing, you know thr basic rule:

"The problem of electronic is always the mechanic" :D
 
Redpacket said:
Good job there :)
Like Cherry67 says to switch the +5.8 line a Pch MOSFET will be what you need (the IRLZ44N, IRF540's are Nch). The FQP06 might be Pch but I couldn't find specs. You have a 10k resistor from the gate to the +5.8V & pull it to ground to switch on.

That seems like a very simple change to make, I will for sure try that!

Cherry67 said:
Redpacket said:
Good job there :)
Like Cherry67 says to switch the +5.8 line a Pch MOSFET will be what you need (the IRLZ44N, IRF540's are Nch). The FQP06 might be Pch nut I couldn't find specs. You have a 10k resistor from the gate to the +5.8V & pull it to ground to switch on.

If your FQP06 is P or N depends on the following numbers and chars, like

FQP6Nxx is a N type.
FQP6Pxx is a P Type.
A well known decades old pattern of marking Fets, with the P6 identifying the current rating, the xx the voltage.

You might find this little gadget useful, you get it everywhere:
https://www.ebay.com/itm/LCR-T4-Meg...ec:m:mLakYynk0RqdZT62qrYQfdw&var=521083969550

The cheap and useful swiss-army knife for testing electronic components (and it would tell you the fet type :) ).

Be wise and buy one with casing, you know thr basic rule:

"The problem of electronic is always the mechanic" :D

That's very useful to know, most of the times I find the names of different components to very random, but for the N or P it makes good sense. Guess that's why IRLZ44'N' means it's a N-channel fet then :D

I'm gonna get one of those for sure. Never heard that quote before but I can relate haha
 
RasmusGodske said:
That's very useful to know, most of the times I find the names of different components to very random, but for the N or P it makes good sense. Guess that's why IRLZ44'N' means it's a N-channel fet then :D
Ah, not in this case. Its not THAT easy:D.


But there are IRLZ44N and IRLZ44 with only minor differences....
 
Back
Top