Go Down

Topic: Arduino controlled mosfet driverwont switch. (Read 3991 times) previous topic - next topic

esgeroth

I'm trying to switch a mosfet IRF3205 from an arduiono uno. I have a TC4227 mosfet driver to drive the mosfet. When connected, the mosfet will not turn off. So I wrote a blink sketch to the arduino and tested with an LED. If connected directly the LED will blink but if connected to the mosfet driver as in the attached schematic then it stays on. Am I connecting this thing correctly? Shouldn't the LED blink when connected to the driver chip?
Here is a TC4227 datasheet: http://ww1.microchip.com/downloads/en/DeviceDoc/20001422G.pdf

ReverseEMF

I'm trying to switch a mosfet IRF3205 from an arduiono uno. I have a TC4227 mosfet driver to drive the mosfet. When connected, the mosfet will not turn off. So I wrote a blink sketch to the arduino and tested with an LED. If connected directly the LED will blink but if connected to the mosfet driver as in the attached schematic then it stays on. Am I connecting this thing correctly? Shouldn't the LED blink when connected to the driver chip?
Here is a TC4227 datasheet: http://ww1.microchip.com/downloads/en/DeviceDoc/20001422G.pdf
For one, the IRF3205 is not really a Logic Level MOSFET -- it might work, but can't be relied on to function properly at 5V. BUT, your little LED experiment should have resulted in a blinking LED even when connected to pin7 on the TC4227.

Just for laughs, try connecting pin 4 to ground.

If that doesn't solve it, then check continuity on all your connections, and check if the voltage on pin 6 is steady at 5V.  And, if you have a scope, check for the waveform right at pin 7.
"It's a big galaxy, Mr. Scott"

Please DON'T PM me regarding what should be part of the Public Conversation -- Let it ALL hang out!!
Unless, of course, it's to notify me of a mistake.

outsider

Don't you need at least 8V on the TC4227 Vdd pin to drive a non logic level MOSFET?

TomGeorge

#3
Apr 22, 2018, 12:28 pm Last Edit: Apr 22, 2018, 12:40 pm by TomGeorge
Hi,

Can you please post a complete copy of your circuit, in CAD or a picture of a hand drawn circuit in jpg, png?
Reverse engineer your circuit, you may find your problem.
What voltage and load are you switching with the MOSFET?
The 4227 should be good to 4.5V Vdd supply by the specifications.

Can you please post a picture of your project, so we can see your component layout?

Do you have a DMM?
If so, measure pin 7 of the 4227 as you connect pin2  of the 4227 from gnd then 5V, with nothing connected to the output, or the inputs of the 4227.

Tom... :)
Everything runs on smoke, let the smoke out, it stops running....

Wawa

#4
Apr 22, 2018, 01:10 pm Last Edit: Apr 22, 2018, 01:14 pm by Wawa
That driver, when powered with 5volt, isn't much better than driving that mosfet directly from an Arduino pin.
That mosfet is not a logic level fet, and needs at least 10volt at the gate to fully turn on.

Power the driver with a 10-12volt supply, with enough ceramic capacitance on the VCC pin (see datasheet).

Mosfets are ESD sensitive devices. Handling the gate wrong will destroy them.
Test the mosfet by connecting gate to source, and measure drain/source with a DMM as you would measure a diode.
It should conduct in one direction only.

What are you switching, and at what (PWM) frequency.
Leo..

esgeroth

I should have mentioned that when I had the mosfet attached to the TC4227 I had 12v on the vdd pin.
Right now I've taken the mosfet out of the equation by setting up the little LED experiment in the schematic. If an LED won't switch on and off then the mosfet certainly will not.

With pin 4 of the TC4227 attached to ground the LED still will not blink.

If i disconnect the 5v from the arduino output to the vdd pin on the TC4227 then the LED starts blinking.

I'm using a sound card scope for now, if I examine TC4227 pin 7 there is nothing or just a bit of noise. Unless I disconnect vdd then I can see the switching signal on pin 7.

Uhh, maybe this is a bad chip?

ReverseEMF

#6
Apr 22, 2018, 07:37 pm Last Edit: Apr 23, 2018, 03:11 am by ReverseEMF
Don't you need at least 8V on the TC4227 Vdd pin to drive a non logic level MOSFET?
The simple answer is "yes".  But, there is a more complex answer involving how much Drain current you want to switch, and how much "on" Drain to Source voltage is tolerable -- as well as power dissipation, etc.  In a production environment, you probably wouldn't make these concessions, but in the "hobby" environment, it might make sense. 

For instance, if a non-logic level MOSFET is really cheap at, say, ALL Electronics, and say, it's designed for an ID of up to 10A, but you only want to switch 1A, and you can see, from the datasheet, that at 5V on the Gate, the Drain will easily handle 1A, then why not!  The only other consideration might be a higher gate capacitance than, say, a Logic Level MOSFET designed to switch 1A.  But that's only a possible factor, if speed is an issue.  A MOSFET with a VGTH that is higher than might normally, be considered too high to be "Logic Level", might be just fine at logic levels (at least with 5V logic) if it's not made to function at the extremes of it's capability.

The IRF3205 (that the OP says he/she is using), has a VGTH of 2V to 4V, and at 5V it has a range of Drain currents where it could easily be considered as "on".  From the datasheet,  at case Tj = 25°C, it won't be fully on for Drain Currents higher than around 6.5A - and even as high as 10A, if as much a 200mV can be tolerated in the channel  -- so, there's a good chance, it will work just fine as a "Logic Level" device, in the OP's application -- as long as the current demand is within these margins.

Transistors are Linear Devices.  The definition of "on" is a relative thing, and has everything to do with the end result.  "On" is defined by whether, or not, it's driving the load, within requirements, without frying in the process. Technically, a MOSFET, that can achieve a max of 50mΩ, can still be considered "on" if the channel is only at 1Ω, if that is enough to properly drive the load, and if the MOSFET doesn't get too hot, doing it.

Hobby electronics is all about out of the box thinking ;D
"It's a big galaxy, Mr. Scott"

Please DON'T PM me regarding what should be part of the Public Conversation -- Let it ALL hang out!!
Unless, of course, it's to notify me of a mistake.

ReverseEMF

#7
Apr 22, 2018, 07:40 pm Last Edit: Apr 22, 2018, 07:41 pm by ReverseEMF
That driver, when powered with 5volt, isn't much better than driving that mosfet directly from an Arduino pin.
Not necessarily.  What the driver provides is extra drive to overcome gate capacitance.  But, that's only an issue at high speed switching. 

That mosfet is not a logic level fet, and needs at least 10volt at the gate to fully turn on.
See my previous post.
"It's a big galaxy, Mr. Scott"

Please DON'T PM me regarding what should be part of the Public Conversation -- Let it ALL hang out!!
Unless, of course, it's to notify me of a mistake.

raschemmel

#8
Apr 22, 2018, 09:56 pm Last Edit: Apr 22, 2018, 09:59 pm by raschemmel
If someone has been able to find a datasheet for a TC4227 please post it because Google has nothing.

Are you sure you don't mean TC4427 ?

TomGeorge

Hi,
Please answer post #3.

Thanks.. Tom.... :)
Everything runs on smoke, let the smoke out, it stops running....

ReverseEMF

#10
Apr 23, 2018, 02:40 am Last Edit: Apr 23, 2018, 02:41 am by ReverseEMF
If someone has been able to find a datasheet for a TC4227 please post it because Google has nothing.

Are you sure you don't mean TC4427 ?
The OP included a link to a datasheet, and it is, indeed, for the TC4426/TC4427/TC4428 -- so presumably the TC4427.  In other words, typo in his/her "capture" and original post text.
"It's a big galaxy, Mr. Scott"

Please DON'T PM me regarding what should be part of the Public Conversation -- Let it ALL hang out!!
Unless, of course, it's to notify me of a mistake.

MarkT

You haven't connected the unused input of the TC4427, pin 4 - connect it to ground as indicated in the applications info section of the datasheet.

That MOSFET requires >=10V on the gate to work properly, it most definitely is _not_ logic-level.

Vgs(th) is _nothing_ to do with switching on, its about switching off. 

I suggect providing the industry standard 12V on pin 6 of the TC4427, and lots of ceramic decoupling for it
(1uF or more is good, within a few mm of the chip - note the need for decoupling is clearly stated in the datasheet, its absolutely vital for a MOSFET driver which handles large currents into a capacitive load).

What are you trying to do?  Do you actually need a MOSFET driver - a logic-level MOSFET might be easier
if you don't need high speed PWM.
[ I DO NOT respond to personal messages, I WILL delete them unread, use the forum please ]

raschemmel

I could be assuming too much but I believe the OP chose the TC-4427 because he thought "Logic Level Mosfet " meant Mosfet driver. Either that , or he thought that using a driver would guarantee that it is logic level, which may be true. I have used the TC4427 with 5V inputs and it worked, but I had +12V on Vdd.

ReverseEMF

That MOSFET requires >=10V on the gate to work properly, it most definitely is _not_ logic-level.
Like I said before:
Quote

For instance, if a non-logic level MOSFET is really cheap at, say, ALL Electronics, and say, it's designed for an I
D of up to 10A, but you only want to switch 1A, and you can see, from the datasheet, that at 5V on the Gate, the Drain will easily handle 1A, then why not!  The only other consideration might be a higher gate capacitance than, say, a Logic Level MOSFET designed to switch 1A.  But that's only a possible factor, if speed is an issue.  A MOSFET with a VGTH that is higher than might normally, be considered too high to be "Logic Level", might be just fine at logic levels (at least with 5V logic) if it's not made to function at the extremes of it's capability.
It's all about context.  In a production environment, the Iron Fist rules.  But in hobbyland, the lines are blurred.  They are blurred by lack of budget [i.e. no start-up capital, or research fund, or, usually, ability to purchase at price dropping quantities, etc.].  They are blurred by lack of customer service woes or potential for large-scale recalls, etc.  They are blurred by the "fun factor" and in most cases, the lack of anyone's life on the line.
But, I get your point about VGS(TH). In my mind I wasn't connecting the gate threshold with Logic Level capability.  Just that the threshold needs to, at least, be lower than the highest level defined as a logic "1" -- with enough margin for a credible switch action to occur [in the particular application].  But, I neglected to make that clear.
"It's a big galaxy, Mr. Scott"

Please DON'T PM me regarding what should be part of the Public Conversation -- Let it ALL hang out!!
Unless, of course, it's to notify me of a mistake.


Go Up