Need Advice On The Best Type of Switch For This Application...

[audioguru]

All the AC switches in my home are labelled "ON", not open or closed.

 

[Reloadron]

Carl suggested earlier:

Make sure you add a protection diode (cathode to plus) across the motor to suppress turn-off spikes that could zap the transistor.

See the attached!

I have no clue how a microprocessor could have cost you $100?

All of your drawings still reflect a poor choice of MOSFET and yet you still persist, I fail to understand that? Using the correct MOSFET with the mentioned correct protection should be all you need. As to a DC SSR? Why? If you want isolation a few dollar opto coupler for the gate signal should work fine.

Just maybe if you would stop and listen to people you might stop cooking parts. If you don't understand when someone tells you something then ask for clarification.

Ron

 

[rs14smith]

Carl suggested earlier:



See the attached!

I have no clue how a microprocessor could have cost you $100?

All of your drawings still reflect a poor choice of MOSFET and yet you still persist, I fail to understand that? Using the correct MOSFET with the mentioned correct protection should be all you need. As to a DC SSR? Why? If you want isolation a few dollar opto coupler for the gate signal should work fine.

Just maybe if you would stop and listen to people you might stop cooking parts. If you don't understand when someone tells you something then ask for clarification.

Ron

Wow wow! Ron let me explain my friend, I hooked up that circuit before you all even posted suggestions, and considering the circuit worked just fine for a smaller version of the project, I assumed it would work alright with the large scale version. Of course it didn't, but hey, that happens...it's a learning experience. Since it did not, I posted my original circuit which one could maybe point out what's wrong with it, and then post suggestions of the cause of the issue.

I've only made this mistake once, so I'm not sure where you are getting the impression that I'm just completely ignoring your logic MOSFET post and continue to make the mistake over and over. It's just I'm "not quite there yet". I need to understand first what went wrong with the current circuit, "which I have thanks to Carl", THEN, I will look into a better MOSFET (logic FET).

Sorry I upset you, but I just prefer not to be attacked when trying to learn. I'm sure you mean well, however, your post has that type of tone. No worries though, it's all good

Anyway, I'm going to build a new circuit combining the one you just posted, plus the opto-isolator Carl suggested, and post it here to see if everything is wired correctly. It will not have the correct FET I will be using, but again, I'll get to that step next.

Baby steps

 

[Reloadron]

Sorry as my post though harsh was intended as constructive criticism. I wasn't aware of the sequence of events. If you have a difficult time finding a MOSFET let me know. I have several logic level flavors around here somewhere. You are in Tenn and I am in Ohio, I will mail you a few free gratis. Again, sorry if I came across harsh. Additionally for what it is worth I allowed smoke to escape from my share of components over a good number of years.

Ron

 

[rs14smith]

Alright, attached is the updated circuit.

I have a good feeling the optoisolator is hooked up wrong, and note I know it's not the same one Carl linked to, but multisim didn't have that one in it's database.

Anywho, I tried to apply this example that site provided as much as possible to this application but just started to get confused after a while: https://www.sparkfun.com/datasheets/Components/PS2532-Example.jpg

 

[Reloadron]

You want your circuit to resemble the attached. The way you have it drawn the MOSFET Gate is shorted to ground. Also your U2, when it is turned on effectively becomes a 36 volt short to ground. The LED side of the opto will work fine with a fwd current of 10 mA and a fwd voltage of about 1.2 volts. That would be about a 380 ohm series resistor. I used 330 as a common off the shelf part. This is your opto coupler data sheet. It's actually a quad so you might want to rethink that.

There may be a few other errors, I did not really look close at the MOSFET data sheet. Getting late for me as I am up at about 3:30 AM so you have a good evening.

<EDIT> OK, now I get the part numbers. I saw your comment about multi-sim. That makes sense. You are only working with the parts in the library that you have. </EDIT>

Damn, one more thing. With this setup the motor will run when the PIC output line is low. Look closely at what happens.

Ron

 

[rs14smith]

You want your circuit to resemble the attached. The way you have it drawn the MOSFET Gate is shorted to ground. Also your U2, when it is turned on effectively becomes a 36 volt short to ground. The LED side of the opto will work fine with a fwd current of 10 mA and a fwd voltage of about 1.2 volts. That would be about a 380 ohm series resistor. I used 330 as a common off the shelf part. This is your opto coupler data sheet. It's actually a quad so you might want to rethink that.

There may be a few other errors, I did not really look close at the MOSFET data sheet. Getting late for me as I am up at about 3:30 AM so you have a good evening.

<EDIT> OK, now I get the part numbers. I saw your comment about multi-sim. That makes sense. You are only working with the parts in the library that you have. </EDIT>

Damn, one more thing. With this setup the motor will run when the PIC output line is low. Look closely at what happens.

Ron

Alright, so I've modified my circuit to resemble yours, however, I'm unsure why you have bot R2 and R3. I understand we can't send 36v to the gate, so are those used to create a voltage divider of some sort to give an output of 5v to the gate?

 

[Jaguarjoe]

R2 limits the current thru the opto when it turns on. Its value will be determined by the opto. This resistor will also put 36v on the gate of the MOSFET when the opto is off. This will not be tollerated well by a 20v max MOSFET gate. To remedy this and use the circuit provided, you will need to supply R2 with a voltage lower than 36v (like 10 or 12) or make a voltage divider by placing a resistor across the opto to put 10-12v on the gate, or place a 10-12v zener diode across the opto. All of this is clumsy. The circuit will draw current when it's off and also while it's on. Not much maybe, but it still does.

R3 helps snub oscillations that can occur at the gate. It is usually a few hundred ohms at most.

You can buy dedicated MOSFET gate driver optoisolators and skip all this junk.

What uC costs $100.00? It should not have died simply driving a MOSFET, logic level or not.

 

[rs14smith]

You can buy dedicated MOSFET gate driver optoisolators and skip all this junk.

I like the sound of that lol. Do you know of one that would be a good fit for my application? The main thing is of course it needs to be able to handle 36v but also 25A.

Thanks

 

[sujith34]

What voltage does it run at? Where is the 5 volts coming from?

 

[rs14smith]

What voltage does it run at? Where is the 5 volts coming from?

lol I posted this info already, but here it is again:

36v and 5v is coming from a micoprocessor which is an external power source separate from the 36v source.

 

[Reloadron]

You still have a few options. You can run with a Logic Level MOSFET like this one designed to run off a uC using a 5 volt gate for full on or you can run with a device like this. The problem with the latter being the maximum allowable voltage. Personally unless isolation is important I would just keep things simple and run with a MOSFET designed to run with a 5 volt gate voltage. However, your call on that note.

Ron

 

[audioguru]

Like most Mosfets, the one shown needs 10V on its gate (not just 5V) to completely turn on. You can buy a "logic-level" Mosfet if you want it to completely turn on with just 5V on its gate.

 

[rs14smith]

You still have a few options. You can run with a Logic Level MOSFET like this one designed to run off a uC using a 5 volt gate for full on or you can run with a device like this. The problem with the latter being the maximum allowable voltage. Personally unless isolation is important I would just keep things simple and run with a MOSFET designed to run with a 5 volt gate voltage. However, your call on that note.

Ron

I would like to use just the MOSFET, however, as I did before, when I ran the 36v ground back into the microprocessors ground, my microprocessor stopped working. I'm not sure if it was due to the large voltage or just a human error in the wiring. I did see a small spark in my breadboard when I did make the final ground connection from the 36v, which I believe may have caused it.

But without having both the 36v and the microprocessors grounds tied together, I'm not able to trigger the gate, so that's where I ran into issues, and Carl suggested the isolation chip.

 

[audioguru]

Most circuits do not work if their grounds are not connected togrether.
The spark you saw that caused damage was due to a wiring mistake.

Edit: Maybe the spark was the gate of the Mosfet being destroyed by the too high 36V then the 36V went into the output of the microcontroller.

 

[rs14smith]

Most circuits do not work if their grounds are not connected togrether.
The spark you saw that caused damage was due to a wiring mistake.

Edit: Maybe the spark was the gate of the Mosfet being destroyed by the too high 36V then the 36V went into the output of the microcontroller.

Well lets put the MOSFETs to the side one more time, and go back to the solid state relay idea, as I'm curious to how simple or complicated it would be to use one such as this for my application: D1D40: CRYDOM: Electromechanical

Would it be as simple as hooking my microprocessor to the inputs of the device, and hooking the motor connectors (36v) to the other side (output side) of the device? Or are additional components going to be needed? I see it states the input can support up to 32v, but what I do not see or know is, how much voltage is needed to close the switch or trigger the switch to close?

If this method is simpler, meaning fewer components are required, no soldering work needs to be done etc, I do not mind investing the money into it as I'm sure I can use it for other projects in the future. I'm just trying to view my options here again before I decide which route to go with. The simpler it is, the better in my specific case here.

 

[audioguru]

A D1D40 solid state relay will work fine if you can afford one.

 

[rs14smith]

A D1D40 solid state relay will work fine if you can afford one.

I actually just found this one below which is sooo much cheaper and meets my requirements...I think? lol

**broken link removed**

I've never done business with this company though, so hopefully they follow through if this solid state relay will work.

 

[crutschow]

Well lets put the MOSFETs to the side one more time, and go back to the solid state relay idea, as I'm curious to how simple or complicated it would be to use one such as this for my application: D1D40: CRYDOM: Electromechanical

Would it be as simple as hooking my microprocessor to the inputs of the device, and hooking the motor connectors (36v) to the other side (output side) of the device? Or are additional components going to be needed? I see it states the input can support up to 32v, but what I do not see or know is, how much voltage is needed to close the switch or trigger the switch to close?

You don't know "how much voltage needed to close the switch" because you apparently didn't look at the data sheet (the data sheet link is on your referenced web page). If you look at that data sheet where is says INPUT SPECIFICATIONS you will see that the control voltage range is 3.5V-32V and the input current is 1.5mA @ 5V. It also shows a diagram of how to connect it.

Yes, you hook the uP to the inputs and the motor to the outputs with no common ground required.

 

[rs14smith]

You don't know "how much voltage needed to close the switch" because you apparently didn't look at the data sheet (the data sheet link is on your referenced web page). If you look at that data sheet where is says INPUT SPECIFICATIONS you will see that the control voltage range is 3.5V-32V and the input current is 1.5mA @ 5V. It also shows a diagram of how to connect it.

Yes, you hook the uP to the inputs and the motor to the outputs with no common ground required.

Cool, thanks for the confirmation, and yes you are right, I did not look at the datasheet as they usually just confuse the mess out of me with all the graphs and etc. lol, but this one appears straight forward

I did not see a link for a datasheet for the cheaper one posted below that post, but I think I'm going to try that 25A version and see how it goes.