Discrete Custom Solid State Relay

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

Bo Deadly

Well-known member
Joined
Dec 22, 2015
Messages
3,266
Location
New Jersey, USA
UPDATE: Previous circuits posted here did not work in one way or another. I have since devised the following circuit that simulates well and seems to function properly so I removed the original question text and associated circuits.

The following untested circuit should allow switching in/out the load of a bipolar power supply (batteries in this case) using a single control pin (grounding the "enable" pin). The trick is to adjust the resistors connected to the mosfets to get properly above and below the particular mosfet Vgs(th).

BattShutdown2.png
 
Mosfets make nice low on resistance switches. I even made a crude pre-regulator using a mosfet as a pass element in a battery powered product.

There is also a mosfet configuration for battery powered circuits where it goes high Z for backwards battery polarity and very low forward resistance for proper battery polarity.

JR
 
JohnRoberts said:
There is also a mosfet configuration for battery powered circuits where it goes high Z for backwards battery polarity and very low forward resistance for proper battery polarity.
Don't tease me JR. What circuit? Would it save me a diode drop?
 
Ah, ok. That's not really applicable to my scenario. The diode that I would like to eliminate is the one between the smps and the battery and not between the battery and the load like that link describes. When the smps supply is removed, the battery will just drain backward through the body diode of the mosfet. So for now I guess I will live with the diode drop. The batteries are supposed to be 14.4V anyway.
 
Although it is somewhat applicable because it made me realize my original circuit was wrong. The mosfets have to be installed backwards so that the batteries don't just drain through the body diodes of the mosfets. That also means that the batteries will charge if supply is present even if the shutdown pin is grounded. But I suppose that's probably desired anyway - if there's supply, charge the batteries regardless.

Circuit in OP updated.

I suppose this circuit about as simple as one could hope for all-things-considered.
 
Can't say I understand what this circuit is supposed to do with the capacitors in series to the mosfets. Are the 1k resistors the load? Je ne comprends pas.
 
squarewave said:
Don't tease me JR. What circuit? Would it save me a diode drop?
Perhaps. I'm busy now but maybe I'll share the PS section of my tuner schematic later. I Have two mosfets in series with battery feed. One for polarity protection, and a second as a pass regulator.

I used two BSS84 MOSFETs in series the first backwards with gate grounded for battery polarity protection, the second normal as a pass device for a pre regulator. Note: my highest voltage input is around 6.5V DC (4xAA).

I also use 3 more MOSFETs as high side switches to multiplex my LED display 3x12

JR
 
volker said:
Can't say I understand what this circuit is supposed to do with the capacitors in series to the mosfets. Are the 1k resistors the load? Je ne comprends pas.
The capacitors are actually supposed to represent batteries. So this circuit is just trying to switch in / out two NiMH battery packs together with one control pin. And so far it seems to do that ok under simulation so I'm going to run with this circuit and see what happens.
 
JohnRoberts said:
I Have two mosfets in series with battery feed. One for polarity protection, and a second as a pass regulator.

I used two BSS84 MOSFETs in series the first backwards with gate grounded for battery polarity protection, the second normal as a pass device for a pre regulator.
Unfortunately the reversed mosfet trick won't work in my case. In my case the diode I would like to replace is between the SMPS supply and the batteries it's charging so that when the SMPS turns off, the batteries don't just drain backward into the charging circuit. With the mosfet trick, when the SMPS is turned off, what happens is that the mosfet doesn't shut off because the battery keeps the drain of the mosfet high letting the current just run free backwards. If the gate could be momentarily pulled up to the drain I think it would latch off. Tangentially that behavior might be used sort of like a triac. Ho hum. I guess I'm keeping the diode.
 
I'm not looking at the detail of this circuit. But just a note that it's often possible to use an IGBT where a MOSFET would work if it were not for the intrinsic body diode. Obvs don't get a 'co-pack' IGBT that has a diode incorporated.
 
I recently wasted some of my dwindling supply of brain cells on a solid state design to automatically steer AC power to an air moving blower in my house. I need to grab continuous power for summer air conditioning support, while for winter heat support, I need to grab power from my back bedroom wall thermostat. After some scribbling with sensitive gate triacs, gated by opto fets, etc, I have resolved to instead KISS... with a DPDT 120VAC powered relay (already ordered).

When energized from the wall switch for continuous cooling use, the relay disconnects the thermostat controlled feed, and grabs power from that wall switch. For winter heat, when wall switch turned off, the relay relaxes to grab power from back bedroom thermostat feed.

JR

[update- I installed the relay yesterday and screwed it down to the back box for the old ceiling heater. The relay hums louder than the blower. Perhaps because it is a cheap (Chinese) relay.  :'( If I can't get it quieter I may have to spring for a more expensive relay. /update]

[update2- Unscrewed the relay, wrapped it in bubble wrap, and rubber glove... now it is quieter than the blower... I don't expect a heat dissipation issue but we'll see.  [/update]
 
Circuit updated again. I think I got it this time. There were a number of problems with the previous circuits. This circuit can connect and disconnect a bipolar load and leakage is minimal (under 5uA when disconnected and under 100uA when connected) so it's nice for batteries which is what I'm doing.

Also note the BAT54 Schottky diodes reduce the overall voltage drop to under 400mV.

If anyone sees anything obviously wrong here please chime in. Otherwise, I'll move on to the PCB now ...
 
Back
Top