MOSFET or something else?

[NaCl_H2O]

I am working on a mod design (Dual 18650 parallel) that will have a copper enclosure with a negative (-) battery path to the atomizer base, and I want the switch on the positive (+) side. All mod diagrams I can find use a Mosfet to switch the neg.

Is there an alternate to a Mosfet (or alternate wiring) to place the switch on the positive side?

 

[NaCl_H2O]

I am working on a mod design (Dual 18650 parallel) that will have a copper enclosure with a negative (-) battery path to the atomizer base, and I want the switch on the positive (+) side. All mod diagrams I can find use a Mosfet to switch the neg.

Is there an alternate to a Mosfet (or alternate wiring) to place the switch on the positive side?

Ok, so that question shows how little I know! and how much I hate searching through forums, but have done enough to realize I need a P-channel Mosfet, right?

Is wiring basically reversed from N-Channel diagrams?

Batt(-) to switch, and to atty body
Other side of Switch to P-Channel Mosfet Gate
Batt(+) to Source
Drain to Atty (+)

If so, any recommendations for a P-Channel Mosfet part#?

Thx!

 

[NaCl_H2O]

Sorry for the novice posts - I am much more of a craftsman than an electrical engineer (obviously). I am rapidly realizing that a good P-Channel logic level FET in a desirable package (TO-220) is likely unavailable?

Am I chasing a ghost here?

 

[dc99]

This should help. TO-220 would be hard to find. Notice the part number on the mosfet. Its not that hard to work with. I have used this one.

 

[NaCl_H2O]

View attachment 400433
This should help. TO-220 would be hard to find. Notice the part number on the mosfet. Its not that hard to work with. I have used this one.

Hey, thanks! I see it is also available in IPAK STD30PF03L-1, maybe a little easier to work with?

But the 70W max seems a bit light for this? I am doing a dual/parallel 18650 and trying to get to a reasonable low resistance coil

What have you built with this FET?

Thanks again!

 

[NaCl_H2O]

Found this one too?

MTP50P03HDLG ON Semiconductor | MTP50P03HDLGOS-ND | DigiKey

.. Actually having more luck tonight finding candidates - must be the glass of wine ?
http://www.mouser.com/ProductDetail...EpiMZZMshyDBzk1/Wi%2bDdTe3kt8oitXvQptl%2bWCc=
http://www.mouser.com/ProductDetail...GAEpiMZZMshyDBzk1/Wi5eGkr%2bPRUK3kW7q7tMcUpM=
http://www.mouser.com/ProductDetail...GAEpiMZZMshyDBzk1/Wi2Lis36WbXR%2bu8xuB3kMEfQ=

 

[twgbonehead]

Remember that 70 watts is the DISSIPATION in the MOSFET, NOT the amount of power it's providing. In other words, it's the amount you can burn up in the fet before you burn up the fet. The power consumed by the FET itself will be RdsOn*I*I, which should wind up being pretty small.

However, the first part has a limit of (I think) 28A continuous, the one you put out is rated for 50A; probably a better choice.

 

[WKS01]

But the 70W max seems a bit light for this? I am doing a dual/parallel 18650 and trying to get to a reasonable low resistance coil

The maximum power dissipate shown for a MOSFET is how much the unit itself can handle before it fails and has no impact on power seen at the coil. This would be found be using the drain current expected and the RDS_on resistance and plugging them into P = I² * R. General rule would be for the TO220 package 1.5 - 2 W is about all it can handle without an additional heat sink.

Looking at both the listed MOSFETs shown each have a fairly high RDS_on. Using 2W as the maximum heat for the MOSFET gives the following amp limits.

MTP50P03HDL = ~9A

STD30PF03LT4 = ~8.5A


Though it is fairly late for me so may work may be off, you may want to check it instead of taking my word.

 

[NaCl_H2O]

Remember that 70 watts is the DISSIPATION in the MOSFET, NOT the amount of power it's providing. In other words, it's the amount you can burn up in the fet before you burn up the fet. The power consumed by the FET itself will be RdsOn*I*I, which should wind up being pretty small.

However, the first part has a limit of (I think) 28A continuous, the one you put out is rated for 50A; probably a better choice.

Good point - thanks!

This is a bit like drinking from a fire hose for me, but I am learning with the help of you good folks ... but slowly! I want to build a nice looking, and different, mod, but also want the right safety built in. I know just enough electronics to be dangerous!

So tell me, with a dual/parallel 18650 unregulated mod, what happens if someone slaps a coil on it that exceeds the FET capacity (e.g. 50A). Or is that really feasible? I know in theory two perfectly matched & charged 18650s can dump 60A, but I assume that is not realistic. Plus, any safety fuses in the circuit would likely trip long before this much amperage was pulled off the batts - right?

Also - if two FETs are done in parallel, I understand that cuts the RdsOn in half - does it also impact the Amp rating?

 

[WKS01]

If someone exceeds the ability of the MOSFET there are two scenarios that can occur: 1. MOSFET fails closed 2. MOSFET fails open. It is difficult to say exactly which will occur. To start your build it would be best to first define what coils you will most likely build and then add a buffer. Once you know that finding a good MOSFET will be easier because you will know what you are looking for in the component.

As for dual 18650s my personal preference is to not go past 40-45A if I was pushing it. Too many variables that cannot be accounted for my piece of mind.

Yes, some choose to run MOSFETs in parallel to decrease the resistance and spread the amp load across the two components. You must still remember to verify your parameters fall within the safe operating range with the appropriate derating for amps, resistance and thermal load.

 

[NaCl_H2O]

If someone exceeds the ability of the MOSFET there are two scenarios that can occur: 1. MOSFET fails closed 2. MOSFET fails open. It is difficult to say exactly which will occur. To start your build it would be best to first define what coils you will most likely build and then add a buffer. Once you know that finding a good MOSFET will be easier because you will know what you are looking for in the component.

As for dual 18650s my personal preference is to not go past 40-45A if I was pushing it. Too many variables that cannot be accounted for my piece of mind.

Yes, some choose to run MOSFETs in parallel to decrease the resistance and spread the amp load across the two components. You must still remember to verify your parameters fall within the safe operating range with the appropriate derating for amps, resistance and thermal load.

Thanks - you have nudged my education a little further down the path - I need to paw through some more posts, but for now I think the wine has won and it's time for some sleep!

 

[drmarble]

It is harder to find p-channel fets with low rds-on at reasonable on voltages (3.0 to 4.5 volts). That's one reason many folks opt for the n-channel fets. The only hard part is you can't ground the battery negative to the box. It has to be switched through the fet. This isn't a problem if you are using a battery sled. It's only a problem if the batteries are stuffed into the box with the negative ends pressed against the side for the negative terminal.
You can do it with p-fets. It is just harder to find good ones and I gave up trying. So I can't suggest anything there.

 

[NaCl_H2O]

The maximum power dissipate shown for a MOSFET is how much the unit itself can handle before it fails and has no impact on power seen at the coil. This would be found be using the drain current expected and the RDS_on resistance and plugging them into P = I² * R. General rule would be for the TO220 package 1.5 - 2 W is about all it can handle without an additional heat sink.

Looking at both the listed MOSFETs shown each have a fairly high RDS_on. Using 2W as the maximum heat for the MOSFET gives the following amp limits.

MTP50P03HDL = ~9A

STD30PF03LT4 = ~8.5A


Though it is fairly late for me so may work may be off, you may want to check it instead of taking my word.

Ok, without the Wine influence, this makes more sense now

While these FETs may be rated higher, the package will fail at ~2W of power, and I verified your numbers for each.

I can mount the FET on a VERY large hunk of copper for a heat sink (however, it is connected to the batt neg(-)). Is the metal tab on the back of the FET connected to the gate or anything else that would keep me from this approach? Should I try to solder the FET or just use a thermal adhesive of some type?

With a "good" heat transfer, what can I assume about the MOSFET heat limit? Obviously, still searching for a lower RDSon, or maybe doing a parallel MOSFET?

 

[NaCl_H2O]

It is harder to find p-channel fets with low rds-on at reasonable on voltages (3.0 to 4.5 volts). That's one reason many folks opt for the n-channel fets. The only hard part is you can't ground the battery negative to the box. It has to be switched through the fet. This isn't a problem if you are using a battery sled. It's only a problem if the batteries are stuffed into the box with the negative ends pressed against the side for the negative terminal.
You can do it with p-fets. It is just harder to find good ones and I gave up trying. So I can't suggest anything there.

Definitely harder to find a p-fet!!!! Maybe I should just flip my batteries around? ;-)

 

[NaCl_H2O]

Any thoughts on this p-FET?

IPP80P03P4L-04 Infineon Technologies | Mouser
... this one is long lead, but these are available

IPP80N06S4L-05 Infineon Technologies | Mouser
IPP020N06NAKSA1 Infineon Technologies | Mouser
IPP020N06NAKSA1 Infineon Technologies | Mouser

I have looked at so many of these I am going blind - am I missing something? Are these good candidates?
... crap, somehow my filter got into the N-CHannels I knew it was too good to be true!

 

[NaCl_H2O]

Can someone educate me why on this MOSFET datasheet
IPP80P03P4L-04 Infineon Technologies | Mouser

It states Vds as -30 and Id as -80? Why the "- signs"?

 

[dc99]

Because your Vgate is less than your Vsource .

 

[NaCl_H2O]

Again, thanks to everyone for their help & knowledge. The IR 3034 is so far superior to any P-Fet ... I just need to look at changing to a design where I can switch the Neg!

... but you all knew that already!

Thanks again!

 

[dc99]

. SI7157DP-T1-GE3 Vishay / Siliconix | Mouser
Im tinkering with a few of these right now.

 

[twgbonehead]

Ok, without the Wine influence, this makes more sense now

While these FETs may be rated higher, the package will fail at ~2W of power, and I verified your numbers for each.

I can mount the FET on a VERY large hunk of copper for a heat sink (however, it is connected to the batt neg(-)). Is the metal tab on the back of the FET connected to the gate or anything else that would keep me from this approach? Should I try to solder the FET or just use a thermal adhesive of some type?

With a "good" heat transfer, what can I assume about the MOSFET heat limit? Obviously, still searching for a lower RDSon, or maybe doing a parallel MOSFET?

The tab is electrically connected. Note for example, in the data sheet for the PP80P03P4L-04 it shows the drain as "pin2/Tab". "Tab" refers to the metal tab, so your heatsink either needs to be electrically isolated from the tab (there's kapton tape that gets used for this, used to be mica insulators) or the design needs to allow the heatsink to be connected to the drain.

You can solder directly to the leads of these parts; try to do it as quickly as possible (although parts like this aren't that sensitive to heat).