Minimize MOSFET count

Go To Last Post
18 posts / 0 new
Author
Message
#1
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Hi All,

I am in the process of building a 3 valve controller. The valves are 24vdc, and have (1) negative lead, and (2) positives for open & close.

I am using a MOSFET with logic level gate drive.

I'm wondering if there is any way to avoid the need to have a seperate MOSFET for the OPEN and CLOSE outputs? This would reduce parts from 6 MOSFETS to 3, and save space, etc.

This is a one-off, so if there isn't a simple solution I can just use 6.

MOSFET I am using is:

The IRLZ14 provides the designer with the best combination of fast switching, ruggedized device design, low on-resistance and cost effectiveness.

Features:
Rated: 60V 10A
Dynamic dv/dt rating
Logic level gate drive
175°C operating temperature
Fast switching
Ease of paralleling
Simple drive requirements

TO-220AB package. Manufactured by International Rectifier.
G16764

Appreciate your time.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

How do you propose to use this mosfet when you require a high side drive (switching the 24V)? You need p-channel mosfets methinks.

I've done some work around Vallejo - nice area.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Can you provide a datasheet for your valve?

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Kartman, err, I'm not sure I follow.
This has logic level gate drive, and Vdss is 60v maximum.
I am looking at the following motor controller schematic at the bottom of
http://www.electronics-tutorials... as an example. What am I missing?
I'm not sure we're talking about the same Vallejo these days... :)

nleahcim, I just gave it back to my friend, will try to get that info.
It is an older version of this http://www.instrumart.com/Produc...
Basically, there are 3 leads for DC operation. One black, and 2 red. When red1 lead gets 24v, it opens and then stops. To close, red2 lead then needs 24v.

Cheers

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Your circuit description was a little ambiguous as to the polarity of the signal required to drive the valves, but I really don't see a much simpler design then 6 low side transistor switches.

It looks (if it's close to the data sheet posted) like the valve needs around 1 amp at 24vdc. If space is a concern, you could use smaller smd parts. Don't know how electrically noisy the valve motor is, but you may want some protection circuitry on those fets.

Another alternative might be some dual automotive high side switches. You'd just need 3 parts. Designed for these types of loads, you get protection and some feedback signals.

(one part for example)
http://www.nxp.com/documents/dat...

another, through hole, mouser stock
http://mouser.com/ProductDetail/...

-carl

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Since you need to ground the negative valve terminal and control the positive (high side) terminals with a switch, you cannot use IRLZ14 as it is a N-channel FET.

Or yes you can, but then you need to have 24V on gate to turn it off and say 30V on gate to turn it on.

For a high side switch, you need to use a P-channel FET, that are a bit but not much worse in specifications. But in that case, you need to somehow convert your 0V/5V logic level to 24V (off) to 18V (on) FET gate voltages. There may be FET driver chips that do the conversion for you, and they might even have a charge pump to enable the use of N-channel FETs on high side of the load.

It would be much simpler if the valve had common +24V terminal and you just needed to ground the open/close terminals.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

If I had to do this quickly, with parts that I have on hand, especially if it is a one-off, I would common the ON +'s & the OFF +'s and drive them with a relay, you could use 3 low side FETS to enable the valves -ve leads.
Alternatively you could replace the relay with two high side FETS for a total SS solution.

Charles Darwin, Lord Kelvin & Murphy are always lurking about!
Lee -.-
Riddle me this...How did the serpent move around before the fall?

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Carl, Jepael, and LDECRIES, thanks for the input.

Somehow I've really forgotten the fundamentals. Transistor Theory 101 or Forest Mimms where are you...

OK, I just reread the Wiki on MOSFETS,
and was only of limited help http://en.wikipedia.org/wiki/MOSFET,
however something clicked when I ran across the circuit symbols:

The circuit diagram I am using is a generic example, and shows a specific NPN,

This made me realize I don't know what type of fet IRZL14 is, so I found this:

They appear to match. But you all say the IRZL won't work for some reason. I am missing something, and now I'm wondering if its that my example circuit would not actually work with my valve in place of the motor?

The motor is shown with a constant +, and - is being switched.
My valves require constant -, and + is being switched.
This doesn't seem workable.

Sorry for the Transistors 101, clearly I need to go back to the reading room for a couple of days of study.
This seems to make some of your comments a bit more clear.

BTW, I was originally going to use relays, and then SSR's but wanted something with more lifetime cycles, so thought a MOSFET would be better.

I have 3 valves that are going to be controlled and activated independently, so am unable to common much I think.

It does look like the automotive TOPFETs are a good solution, but again how do I read if they are PNP as Jepael suggests? Or does that not appy with these devices? In the event they are difficult to find, do I just look for a TOPFET?

Appreciate everyone's help!

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

More transistor 101. with FETs, it's N channel and P channel not NPN / PNP. You're going to have to do some more research.

The ST micro parts I linked to above (or equiv) will switch the positive leg and can be driven from an avr's output pin. Read the data sheets, add some protection diodes and possibly some noise suppression and you should be good to go. As long as the data sheet for the valve is not too much different from the one you posted.

It is interesting that the valve seems to specified for 24V dc OR ac operation (unsure if that is a completely different model no.) If this is true, it means that the polarity might not matter at all (Note that the data sheet does show a particular wiring for dc operation).

If it runs on ac, it means that you might be able to just use a 24v ac transformer for the valve and then use cheap triacs to switch the ac current. But that's another discussion.

-carl

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Carl, you are correct. I read that and just fell back on what I new earlier.

I did get the valve back, or actuator actually. Its an older Georg Fischer EA20, AC/DC.
Found a service manual http://safemanuals.com/user-guide-instructions-owner-manual/GEORG%20FISCHER/EA20
which shows the internals on Page 7 as below

I've got a decent 24vdc psu that will run the 3 of them, and would rather stick to dc at this point.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Looks good.
Current requirements are if anything a little lower on this rev. Still under 1 amp.

However you choose to drive the valve motor, it does not seem too big a deal. Unless you need ultra high reliability or fail safe operation.

-carl

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Depending on your application, You may want to look at the Taco EBV series of valves. EBV(electronic ball valve) are either on-off / open-closed. They have a brass body available in sweat or NPT. One major advantage may be that is 1/3 to 1/4 the price. Here is a PDF for the valves. We use them in an industrial enviroment and they have been very reliable. http://www.taco-hvac.com/uploads... . This would also simplify your control requirements. Note that they are rated for 24VAC, but they will work with 24VDC. The power input is feeding a FWB rectifier. The only down side, there is a PIC controller inside.

On the power requiremnt note: While power needs to be applied to keep the valve open, usage is minimal since the valve control motor is not running.

Roger

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Those Taco (love the name) valves look slick. I wish I'd known about them for a project about 6 months ago. I like the way they use a super cap to power the motor to close the valve on power loss. Too bad about the PIC.

I think the OP may be stuck with existing valves (which do look more rugged and environmentally sealed) for this project, but I'll keep the Tacos in mind for future applications.

-carl

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Thanks for the help, and suggestions.

Since you've all been so helpful, I'll divulge the not so secret project.

My friend is one of those all or nothing type of people. So we're both in marine/reef aquarium hobby, corals, etc.
One basic piece of equipment needed is called a skimmer, which is normally 16-24" talls and helps keep the aquarium water clean. You can build very efficient skimmers cheaply, and of course, longer is better.
So, he is getting a 350g aquarium, and whereas most people investing a couple of $K into such a system would get a tried and true pro skimmer for $500-800, my friend decides to build his own. And of course its now 20'+ tall, with pipe going from the sump under the tank stand, into the wall, up to the 15' ceiling, and into the attic another 5-8'. None of this requires valves though. The valves come in as he has decided to add whats called a surge tank to his tank. This is where you periodically dump x gallons of water back into your tank to mimic a reef enivronment of waves surging against the reef rocks.
So the valves will open randomly every couple of minutes, with either 1, 2, or all 3 opening at once to allow varying levels of surge.

Once I get the valves going, next will be figuring out how to make one of Sparkfun's strain gauges work as a digital scale to know when the surge tank is full and available to dump.

Thanks again for the help.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

If you wanted to save on one FET, you could have 2 pfets switching he top and 3 nfets switching the bottom... Think 2x3 or 3x2 matrix...
Not much of a savings though...

A float switch is probably easier and cheaper than a strain gauge...
If you want to punish yourself, you could try a capacitive level sensor...

Here's an float switch from ebay for about $1 ...
http://cgi.ebay.ca/Level-Sensor-Liquid-Water-Right-Angle-Float-Switch-/190458733922?pt=LH_DefaultDomain_0&hash=item2c5839a562

Michael

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Michael,

Thanks, however I think I'll stick with the less confusing automotive high sides per Carl.
The problem with float switches is that sooner or later the salt or algae growth will cause them to stick open/closed. One can always double up for redundancy, however having something outside the tank like a scale seems even better.

Thanks!

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

For non-contact liquid level sensing you might also consider ultrasonic or capacitive...

Capacitive is actually quite easy using an AVR with an analog comparator ...
I can post code and schematic if your interested...

If you choose the strain gauge method, Electric Goldmine has the unmounted sensors for $1...
They would require mounting to a bar of known deflection...

Michael

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Thanks again for everyone's help and patience. I finally went ahead and ordered the vnd10b from Mouser. I finally read something somewhere that clicked on the 'high-side' descriptor and realize my error.
And, ultimately with the MOSFET's I think I was just creating something like an SSR anyways, as I understand many SSR's have MOSFETs inside.

Thanks again.