hi,
I have a simple circuit I'm trying to build. It's a 12V DC Motor (Stall current 20A) powered by 3 Lithium-Ion cells and switched on and off using an Arduino and a MOSFET (IRFZ34N). Datasheet here: https://www.infineon.com/dgdl/irfz34nspbf.pdf?fileId=5546d462533600a40153563b1a522202

I've attached a schematic I'm working on, and I have some questions about the passive components:
1. I have my pulldown resistor on the gate (R7) set at 1k. I've seen some online schematics with this as high as 10k, would it need to go that high?
2. I have a diode going from the negative output to the positive, but I wasn't sure what type of diode is needed here. I was thinking 1N4001, but I also have some Schotke diodes on hand (to be totally honest this is the one passive component I never understand).
3. Do you see any other passive components I need to add?

 

The basic circuit is OK, except:

The MOSFET is rated for 29A which isn't too much margin over the 20A it needs to conduct. I would go for 50A+
Also that FET really needs 10V on the gate to get it turned fully on and so keep the drain to source voltage and hence dissipation as low as possible. This will minimise the heat sink requirements.

The diode will pass the full 20A when the MOSFET switches off. The 1N4001 is rated up to 30A surge so again there is not much margin.

 

Awesome, thanks for your reply.

The motor takes 20A at a full stall, which I don't think will ever really happen. It's a very light-duty application so stalls would be uncommon. In my inital tests it never goes above 3 to 5 amps for normal usage. Should I still bump it up?

The basic circuit is OK, except:
Also that FET really needs 10V on the gate to get it turned fully on and so keep the drain to source voltage and hence dissipation as low as possible. This will minimise the heat sink requirements.

This is where I need some help. I was watching a "Great Scott!" video on youtube, and that video said the voltage of the gate needs to be higher than the 'Gate Threshold Voltage' (which is Min. 2, Max 4.) and it needs to be lower than the Maximum Gate to Source voltage (which is +- 20). I thought I was okay. Can you tell me how you got the 10V so I can start hunting down another FET...or if you have any recommendations for a better one.

 

Can you tell me how you got the 10V so I can start hunting down another FET

The motor will draw its stall current when you first switch it on but only briefly. In this case the drain current spec would be OK, but I still think you are pushing your luck with the gate threshold voltage. Note the conditions for 2V to 4V spec - Id 250uA - and the Arduino will output a bit less than 5V.
Look at the specification for Rds(ON) (the on state drain to source resistance) and see what gate voltage they measured it at. In this case, 10V and 16A.

 

This is why you should not get your electronic knowledge from youtube. The MOSFET will only conduct 250 uA at the gate threshold voltage— a far cry fom 20A.

Use an IRLZ24. That will be fully on at 5V.

Bob

 

If you look in the data sheet at the Vgs voltage used to specify the MOSFET ON resistance, you will know what Vgs is required to fully turn it on.

 

I recommend the IRLZ44. The ON resistance is specified as 25 milliohms maximum at 5 V gate to source, It is reasonably cheap, In through-hole packages you can get "logic level" FETs down to less than 5 milliohms without spending a lot. In surface mount (newer parts with better specs tend to be offered only in surface mount packages) you can get logic level FETs with sub-milliohm ON resistance.

In general, I recommend selecting FETs by ON resistance rather than current rating, particularly in circuits that operate at relatively low voltage. Many power MOSFETs, especially surface mount types, have "bullet point" current ratings that are unrealistic for continuous current because it is virtually impossible to remove heat fast enough to keep the operating temperature below the maximum. The rating is still meaningful for certain applications where huge current at low duty cycle and short ON time is required. Using ON resistance allows very simple calculation of "conduction loss" (as opposed to "switching loss" which is rather messy). 25 milliohms at 5 amperes works out to
(5 amperes) squared times 25 milliohms = 25 x 0.025 = 0.625 watts - easily handled by a TO-220 device without a heatsink
20 A of stall current gets it up to 10 W - definitely requires a heatsink if that current persists for more than a second or so

In most datasheets for modern FETs you will find at least one graph with a family of curves showing the typical current versus drain-source voltage (the actual voltage across the FET, not the supply voltage) for a range of gate to source voltages. These curves are for typical performance, not "worst case" but they give you a good idea of performance with the gate drive voltage you have available. Another approach is to consult the tabular data where you will find a maximum specification for gate-source threshold voltage, normally specified at some very small current, and a minimum specification for transconductance (ratio of change in drain current in amperes to change in gate-source voltage in volts - specified in siemens (inverse of ohms; used to be called mhos) abbreviated S. For example the IR datasheet for the IRFL44 shows a maximum threshold of 2 V and a minimum transconductance of 21 S. This means with say 4.5 V gate to source you have 2.5 V above the threshold, so you can expect a current of 2.5 V x 21 S = 52.5 A - assuming nothing else is limiting. This doesn't tell you everything, but certainly is strongly suggestive that for the 20 A of stall currrent, 4.5 Vgs will be sufficient.

Fully specifying a power MOSFET, as for most discrete semiconductors, requires a moderately large effort for critical or difficult situations. There are applications notes available form most of the big manufacturers of FETs to help you understand all of the parameters.