I’m looking for two things: Immediately, I am looking for a good P-Channel MOSFET that works well with 5V systems. I got a few IRL520 N-channel FETs, which work nicely at those low voltages, but I haven’t found a P-channel FET that works at low voltages like that.
The second thing I am looking for is a good source for parts. Hobby Engineering has a good website and a fair selection, but far from comprehensive. Digi-key has…EVERYTHING, but their web site is horrible, and their catalogue is worse. For them, it seems that you need to know the part number to a very high accuracy, before you can order anything. Jameco has a pretty good variety, too, and not too bad a web site. I’ve ordered from all of these, plus Electronix Express. Are there other sites that people here like?
You might check BGMicro below. They often have some decent dreals on electronic parts. Below is their “last minute addition” pages. I see a P channel 60v 60a MOSFET there for $1.49.
Instead of searching the sellers, first search on the manufacturer’s sites to identify which part numbers will work for you.
Another major MOSFET maker is Vishay (Siliconix). I recently made a project using the Si49743 P-FET, which works well at TTL logic levels…
I’m not sue how the P FETs are supposed to work, but you probably can use one of the small generic PNP transistors to control one of your N FETs. I assume you have a control setup that sinks current instead of sourcing it.
A P-FET is generally what you use if you want to do hi-side switching. If you use an N-FET for hi-side switching, you need a ‘bias supply’ to provide gate voltage that goes several volts above the voltage that you are trying to switch.
Adding a transistor doesn’t help the situation…
Until then, Shaggy take a look at the International Rectifier IRF9Zxx parts ( xx is a number like 10, 14, 20, 24, 30, 24 ) as many of them have Vgs threshold specified -2 to -4 volts which would make them usable as logic level parts but with reduced specifications. An example irf.com/product-info/datashe … f9z34n.pdf
will probably work fine for a couple of amperes Id if you can give them -4.5 to -5 volts for Vgs.
In my satement the assumptions I stated are based on a control circuit that sinks current/voltage instead of sourcing voltage/current (which is where I usually see P types of transistors used). I also assume that the N-FET that reportedly worked was placed between the load and ground (I’ve read this is the prefered placement to avoid “voltage bias” issues). In this configuration the addition a simple PNP transistor might make the setup work. Saipan59 appears to have made an assumption that the FET could only be placed between the voltage source and the load. This may be true, but it is not stated in the origional post. So, whats your point?
Actually Petes assertion is likely based on the fact that 99% of the reason anybody uses a P-channel FET is FOR high side switching, and I would agree with that (mostly because I can’t think of anything I’ve ever designed or a design I’ve ever worked on where a P-channel FET comparable to an IRL520 N-channel FET was used for anything else.)
So if there was a point to be had or made for some reason it would be something like adding a transistor to a circuit trying to switch high side with an N-channel FET won’t get you far… unless you are using one of its PN junctions as a diode to charge the cap in your boost circuit genetating the bias supply to turn the N-channel FET on. Woo hoo!
Pretty much a restatement of the same potentially flawed assumptions based on the assumption that the power has to be high side switched. This requirement was not included in the origional post. I’d rather post a potential simple and cheap solution using an N-FET the person already has rather than doing a lot of data sheet “hanger flying” that may mislead the person into spending a lot of unnecessary $$$ ordering parts.
Note that the original question was about P-FETs, and I assumed that he already knew the difference between hi-side and lo-side switching.
Your alternative of doing everything lo-side is fine, but it could lead other readers to believe that an N-FET plus a transistor is somehow equivalent to a P-FET.
So, my subsequent post was intended to point out why the original poster might be wanting to use a P-FET. This clarification seemed necessary at the time because the Zoomcat post (above) included the phrases “*not sure *how P-FETs work” and “probably can use”, and “assume”, all in the same sentence…
Glad I generated some debate. I’m new to electronics, so some of the jargon went over my head, but let me describe what I am doing, perhaps somebody can suggest a better solution:
I have four whiskers on a robot, two in front, and two in back. Now, I have some pretty serious brainpower on this system, so I know that I can only run into something when I’m going forwards. Therefore, I don’t need four different I/O lines for these sensors, but instead only need two. For both the left and the right side, if either sensor trips, I don’t actually need to know which one it is. Since I am more I/O line limitted than anything else, saving two lines is valuable.
What I set up was that the sensors were pulled high, and would go low when the switch is closed. Thus I thought that I could use a p-channel FET such that current only flowed when the switch was closed. Each whisker is hooked to a different FET, so current will flow from either one, and light an LED when either switch is pressed. By using the N-channel FET, I have done the opposite: The LED is on unless the switch is closed, which would mean that BOTH switches must be closed.
On the other hand, as I was writing this, it occured to me that I might not need a FET of any sort. I guess I ought to test that idea, but once again, I’ll consider any suggestion thoroughly.
If I understand your description correctly:
An N-FET in the normal setup with the Source grounded will act as an inverter.
I think you’re right that you don’t need any FETs at all - if the switches are all wired to ground on one side, and all connected to the common LED on the other, then closing any switch will turn on the LED.
But I may not be interpreting what you’re doing right… If you could draw us a picture, we could get closer to a real answer.
Zoomkat you are reading more into my reply than I put there in your attempt to show how it can all (?) be done low-side. No one but you has said power HAS to be switched high side. No one else. What was said was that given the IRL520 as an N-channel FET to compare to, a P-channel device of the same characteristic ratings would generally be applied for the purpose of switching high-side. Realistically P-channel devices suck when compared against similar N-channel FETs and that is part of the reason people even bother to go about designing boost circuits to allow all N-channels devices to be used in a bridge. Anyway, what followed from that was a statement that an N-channel device plus a small transistor can not be made to perform the same function on the high side as a P-channel device can. Again no one said you have to switch anything high side, they said you can not switch high side with an N-channel FET and a small transistor without a bias voltage.
That’s cool and all but you have made an assumption that you can solve the undescribed problem with a simple and cheap solution rather than providing the answer to the question, which was looking for recommendations for a good P-channel MOSFET that works well with 5V systems.
Please show where I made this statement. It appears to be somewhat fantacy thinking on your part. saipan59 is the one that went off on a tangent about high side swiching with an N-FET based on an assumption he made on his own. Please reread the post to maintain accuracy.
This appears to be based on your assumption that high side switching was the only solution. I never mentioned high side switching in my setup, only low side with an N-FET which the person already had and had verified as adequate. The recomendations for a P-FET were pretty vague in my opinion, as no specific transistor was ever named.
Ugh, this is in my opinion pointless waste of time since we’re into one of these he said she said things where two people reading the same words can’t seem to get the same idea in their heads. Since the topic of the thread has been addressed and as Shaggy already has some good advice in answer to his question requesting a good logic level P-FET recommendation and how to go about finding some I am just going to let it drop.
Sounds like the front sensor switches need to be placed in parallel. If currently both need to be closed to light the LED, then they are probably in series.
