Circuit changes needed to switch higher voltage relays?

[Chris Wilson]

I wish to modify this circuit to switch relays with 24 volt coils with current draws of 190 mA. Is it practical and what would I need to change? The 2N7000's spec is at  https://www.fairchildsemi.com/ds/2N/2N7000.pdf? Thanks!

[minime72706]

What is this circuit supposed to do? It helps to know what the operational amplifiers are meant to do when determining what needs to be changed.

Is there a particular reason you want to make this change? i.e. maybe you only have 24V relays on hand?

EDIT: It looks like it might be for power sequencing or something - or for switching taps on a transformer as you turn some potentiometer. The lack of English comments in certain places makes it a bit difficult to make out. Depending on what it's supposed to do, there are likely very few changes that you need to make.

[mrkev]

Well basicaly, you would need just higher voltage on inputs for those rellays.
2N7002 is a basic mos, and different supplyers make it with a bit different parameters. So just make sure that part you have can withstand continuous current of 200mA (in case of that Fairchild it would be the NDS7002A version, not the standart one).

[SeanB]

The resistors for the LED's will have to be somewhere between 2k2 and 3k3, and the opamp supply will have to stay at under 15V, but otherwise nothing will change. The 15V supply will feed R5, R6, R7, R8, R9 and R14/15 along with the opamp, and can be anything from 12-15v, probably from a small 7812 regulator fed from the 24V rail. Otherwise a 100uF 25v capacitor on both input and output of the regulator will do fine.

[Andy Watson]

The dataheet says that 2N7000 are good for 200mA and 60V - so it should be ok. However, as others have noted, you will need to alter the supply to LEDs and the op-amp, and/or alter the arrangement of the gate drive. Since you've got more than 13V available you could consider using a beefier power mos fet - no need for "logic level" devices.

I'm curious about the capacitors on the outputs of the op-amps - looks like it's designed to cause instability!

[mrkev]

The resistors for the LED's will have to be somewhere between 2k2 and 3k3, and the opamp supply will have to stay at under 15V, but otherwise nothing will change. The 15V supply will feed R5, R6, R7, R8, R9 and R14/15 along with the opamp, and can be anything from 12-15v, probably from a small 7812 regulator fed from the 24V rail. Otherwise a 100uF 25v capacitor on both input and output of the regulator will do fine.

Doesn't LM324 work till 32VDC?
Oh, I see... The gate shouldn't go higher than 20V... Well personaly, I would use the low power LEDs to get that voltage down and so got rid of those diodes parallel with realays.

I'm curious about the capacitors on the outputs of the op-amps - looks like it's designed to cause instability!

I had a quick look and I missed that one. Those RCs at the output of op-amps are looking weird, plus using them without hysteresis could mean some oscilations. What is it for 

[minime72706]

The resistors for the LED's will have to be somewhere between 2k2 and 3k3, and the opamp supply will have to stay at under 15V, but otherwise nothing will change. The 15V supply will feed R5, R6, R7, R8, R9 and R14/15 along with the opamp, and can be anything from 12-15v, probably from a small 7812 regulator fed from the 24V rail. Otherwise a 100uF 25v capacitor on both input and output of the regulator will do fine.

Doesn't LM324 work till 32VDC?
Oh, I see... The gate shouldn't go higher than 20V... Well personaly, I would use the low power LEDs to get that voltage down and so got rid of those diodes parallel with realays.

I'm curious about the capacitors on the outputs of the op-amps - looks like it's designed to cause instability!

I had a quick look and I missed that one. Those RCs at the output of op-amps are looking weird, plus using them without hysteresis could mean some oscilations. What is it for 

Maybe things would be fine if one replaced that ridiculous 100nf "output capacitor" with a series resistor and a 18V zener diode? The gate should have a path to ground as well, but I suppose the zener can take care of that (I'd put a resistor to ground as well).

[Chris Wilson]

The purpose of this thing is to create a 4 stage delay (sequencer) for controlling an add on linear amp and its associated gubbins, to an amateur radio transceiver. Its relays change in sequence to do the various tasks so high power live RF is not being switched, or high power RF put into pre amps et cetera. Whilst I guess it was originally designed to be a fairly universal tool, I would rather not have relays driving other relays, and was wondering what it would take to make the device switch 24 / 27 Volt vacuum relays direct, instead of switching them via the DIL relays, which I would leave out. Thanks.

[SeanB]

Then change those 2n7000's to IRLML0100TRPBF, SMT but capable of driving the relay with ability to spare. SMT so you will have no problem mounting them. Just were the cheapest units from RS with 100V rating and over 1A rating in a search. 10c US each is pretty good.

[Chris Wilson]

Thanks SeanB and all who have contributed.

Sean, just to be 100% clear, just by changing the 2N7000's to the IRL devices, and upping the value of the dropper resistors for the LED's, I can apply 27 volts instead of 13.8 volts as the operating voltage for the board and the relay coils? The opamp will be happy with no other changes?

I wonder if a leaded device of sufficient power and voltage handling exists? I am comfortable with SMC work, but unsure if the "footprint" of the IRL devices means they will sit on the existing PCB pads? I am not clever enough to redesign the PCB layout, I am using an existing .eps file photo etch the boards. Many thanks!

[SeanB]

Board still needs 12V, and the diodes on the board need to be moved to the coils of the relays along with the LED's and the resistors for them so that this all operates off the 27V rail. The opamp and the board needs 12V to work, at 27V it will blow up all the mosfets by overdriving the gates. The mosfet can be almost any N mosfet you like with a 100V or more rating and a max current over 1A, in pretty much any package you are comfortable with. I just offered the above as a cheap drop in.

[minime72706]

I see a decent selection of TO-92 packaged MOSFETs that will tolerate the gate voltages and drain currents you need via mouser - going to check Farnell right now.

VN0300L (30V Vgs,max and 0.64A ID,max)
VN10KN3 (30V, 0.31A)
SSN1N45BTA (50V, 0.5A, bonus 450VDS,max rating!)

[minime72706]

I found Farnell's search interface obnoxious (same with Digikey actually; gate-source breakdown voltage isn't a searchable parameter), but I did find one device:

ST Microelectronics STQ3N45K3-AP
I_D,max = 600mA
V_DS,max = 450V
V_GS,max = 30V
     With respect to the above gate-source breakdown voltage, it's also self-protected via a back-to-back zener diode! Neat-o!

It's roughly $0.50 - $0.60 on Digikey, $0.91 (converted) on Farnell/Newark, and non-stock at Mouser.

WITH ANY REPLACEMENT TRANSISTOR, SINCE YOU'RE USING THE BOARDS THAT GO WITH THAT SCHEMATIC, WATCH FOR THE PIN-OUT. I CAN TELL YOU FOR CERTAIN THAT THIS PART DOES NOT MATCH THE PINOUT OF THE 2N7000
Personally, again, I think you can probably safely protect the 2N7000 with your own zener.

EDIT: also be aware that the threshold voltage is moderately high compared to the average. It's about 4.5V, which still shouldn't be an issue. It's just impressively high, in my opinion, for a tiny TO-92 device.

[Chris Wilson]

Thanks minime72706 and seanB. Great info. I have discovered a similar type of circuit, may I ask, do the 12V zeners outlined in red do the job of keeping the gate voltage on the devices down below their 20V maximum? I guess that's what they are for, but I would like to be sure I have their purpose understood correctly. Is it a standard way of using devices with a gate voltage below the signal voltage?
 Thanks

[mrkev]

Thanks minime. Great info. I have discovered a similar type of circuit, may I ask, do the 12V zeners outlined in red do the job of keeping the gate voltage on the devices down below their 20V maximum? I guess that's what they are for, but I would like to be sure I have their purpose understood correctly. Is it a standard way of using devices with a gate voltage below the signal voltage?
 Thanks

Well yes, they will do the job as you described. However usually (in this case, they are in fact the must as the vcc si quite high), it's mostly for protection. It's just always better to be on safe side in case that someone will connect the higher voltage...