Biasing for an N-channel JFET

Thread Starter

jcbeck84

Joined Dec 20, 2008
46
I have a current source controller which is used to position a valve coil. Current runs 0-.95 amps and the voltage swings about 4-13 volts accordingly. I would like to use an N-channel JFET to just act as a signal to control a relay to tell me when there is power going to the valve and when there isn't. I have the gate of the JFET connected to negative valve control, the "source" connected to ground (for the relay circuit) and the positive, and the drain going to the relay and then the positive voltage supply. I also have diodes in place to prevent damage to the semiconductor when the inductive loads discharge. All seems pretty straightfoward, but my question is regarding where the two loops meet at the source of the JFET. To correctly bias the JFET gate the positive valve connection must be connected to ground on the relay circuit. Wouldn't having that ground connected to my positive valve control mess with the behavior of my valve operation?
 

Thread Starter

jcbeck84

Joined Dec 20, 2008
46
The JFET doesn't need to pass the current for the valve. I just need the voltage reference from the valve to close the JFET channel and cut off the current. The JFET I have chosen is this one: http://www.fairchildsemi.com/ds/PN/PN5434.pdf
It allows 30 mA at 0 gate voltage and the relay will draw about 22 mA. I have no idea what the specs for the valve are. I had to run flow tests on the pump while monitoring current and adjusting the gain just to find that my valve is fully open at .95 amps. The voltages at the valve coil are based on no significant loading of the pump but could theoretically rise to 24 volts as more load is put against movement of the valve.
 

Ron H

Joined Apr 14, 2005
7,014
What else is "ground" connected to (e.g., earth)?
I'm assuming you know that the FET will be off when the valve is energized.
Why did you choose a JFET as opposed to a BJT or a MOSFET?
 

Thread Starter

jcbeck84

Joined Dec 20, 2008
46
Ground is the negative terminal of the 24 volt power supply. There are misc other controls associated with the system as a whole connected to it but it's the lowest potential available. I do know the JFET will cut current when the valve is on which is fine, I can deal with it the other way around as well. I chose the FET over the BJT because it would require less components and I wouldn't have to skim any current off of my valve to generate my signal. As far as MOSFET's go I can't really say as I know a whole lot about them. Really the goal here is to turn the relay on and off as the valve receives power (on/off or off/on, doesn't matter) while maintaining as much of the original current through the valve as possible.
 

Ron H

Joined Apr 14, 2005
7,014
If the valve and its power source are truly floating, I don't see how the ground would cause a problem. If the valve and its power source are not floating, you would need to post a schematic.
 

Thread Starter

jcbeck84

Joined Dec 20, 2008
46
Ron, I'm not quite sure what you mean by floating. The valve controller acts as a current source. I can directly control the amount of current it sends to the valve via a seperate control potentiometer. I couldn't give you a schematic of the controller as it's a commercially made product and I can only analyze how it reacts to my inputs. It is also driven from the same 24V supply. I posted a schematic of what I'm trying to build just after my initial question.
 

Thread Starter

jcbeck84

Joined Dec 20, 2008
46
After some browsing of the term floating I can say that my power supply is in fact floating, but the - terminal has been jumpered to ground. This was done by someone else and I couldn't tell you why.
 

Ron H

Joined Apr 14, 2005
7,014
After some browsing of the term floating I can say that my power supply is in fact floating, but the - terminal has been jumpered to ground. This was done by someone else and I couldn't tell you why.
Isn't the - terminal the one you want to connect to the gate of the FET?
 

Thread Starter

jcbeck84

Joined Dec 20, 2008
46
I want to connect the gate of the FET to the negative side of the valve. There is essentially two loops: The current sourced which simply powers the valve, and the 24V sourced which is used to power the relay. Individually I know the loops function. It's their connection at the FET source that I'm having trouble with. Would it work like I want if I were to make the connections as I had posted in my original schematic?
 

Ron H

Joined Apr 14, 2005
7,014
I want to connect the gate of the FET to the negative side of the valve. There is essentially two loops: The current sourced which simply powers the valve, and the 24V sourced which is used to power the relay. Individually I know the loops function. It's their connection at the FET source that I'm having trouble with. Would it work like I want if I were to make the connections as I had posted in my original schematic?
The concern I have is that you said:

1. The negative terminal of the 12V supply connects to ground.
2. The negative terminal of the valve supply connects to ground.

This will short out the valve and its power supply.
 

Thread Starter

jcbeck84

Joined Dec 20, 2008
46
I went out and checked. There are several different input voltages (mine happen to be 480 and 0) to get a 24VDC output on the 0 and 24 terminals. There are tabs on either side of the terminal strip marked with the ground symbol that I'm assuming are simply grounded to the case of the supply. The case is then mounted inside another case that is then grounded to the grounding wire. The originating power comes from 480V triple phase and I don't mess with anything on that side of the power supply. The - terminal has a short jumper from it to one of these grounding tabs. There are already several relays and a host of other things being driven off of this circuit so I can only assume it is a functioning power supply. I have attached a schematic of what is existing and fully operational so maybe you can lead me in the right direction.
 

Attachments

Ron H

Joined Apr 14, 2005
7,014
Neither of the valve connections are at ground.
I guess I should have asked what the resistances are from each valve terminal to ground, measured with the probes attached in both polarities (4 measurements total). I'm concerned that there are active devices (e.g., diodes) that could be blown up if we connect one terminal to ground.
 

Thread Starter

jcbeck84

Joined Dec 20, 2008
46
One terminal does measure .75 Mohm with positive probe on the terminal and black on ground but OL with the probes reversed. The other terminal gives OL in both polarities. The terminal that did show a connection was the negative polarity side of the valve.
 
Last edited:

Ron H

Joined Apr 14, 2005
7,014
I'm posting what I believe you have suggested, but I'm not convinced the PN5434 will handle the relay coil requirements. Do you have a relay datasheet, or know how much current it requires?
I added 10k in series with the gate for protection from misconnection, transients, etc. This should not affect proper operation.
 

Attachments

Thread Starter

jcbeck84

Joined Dec 20, 2008
46
The relay I'm using is an omron relay. Rated for 24VDC at 1113 ohm or 22mA (approximately). I don't have time to look over the schematic you posted just yet but I should get back to you later tonight.
 

Thread Starter

jcbeck84

Joined Dec 20, 2008
46
That looks like what I was suggesting, but I also see what you mean about the valve now being shorted to ground so obviously this configuration won't work. Is there any configuration of this that will work? There's not a whole lot of parts to play around with.
 
Top