Circuit Design Help

[golfmonke]

I'm new to electronics and I'm trying to create a circuit that will alternate between a negative signal and positive on the same line.

I've scoped out a circuit that i'm trying to reproduce....

It needs to alternate between -5v and +30V (although testing with 12v for now), it needs to be -5V for .8ms /then + for .2ms and it repeats every 80ms

At top i've got 2 power supplies, one at 12v and one at 5V with a common in between

I've posted a pic of my design below, i'd like to know if i'm on the right track, basically i'm thinking using some high side mosfets as switches, driven at the gate by an arduino (and some circuit), will figure that out later, but want to know if the rest of the circuit will work, assuming I figure out how to drive the mosfet gates.

Waiting on delivery of some N and P channel mosfets at the moment, or i'd just hook it up and give it a go myself..

 

[Pommie]

How much current is required and how accurate do the voltages need to be?

Mike.

 

[golfmonke]

The actual load is 3 ohm, when I scoped, it was pulling 8 amps max.

I don't think the voltages need to be all that accurate, but won't know for sure until I can create this driver circuit.

 

[ronsimpson]

Try #1
Too much information missing but ….. (8A ok need to change the mosfets to something that can handle the current with out a heat sink)
Assumptions: ESP runs on 3.3V and has a 5V to 3.3V on board regulator. It is OK to have the ESP run from the -5V regulator. 3.3V is too little to turn on a MOSFET.
M1 is a P mosfet. It type will depend on the current in the load. Use a 60V device.
M2 is a N mosfet. It type will depend on the current in the load. Use a 60V device.
When IO pins are low the mosfets are on. Use pull up resistors so during power up when the IO pins are in tristate the mosfets will be off.
Q1&Q2 will turn on/off M1. The pin IO1 moves 3.3V at most and is referenced to -5V. This signal is amplified to 15V and referenced to +30V
A1 is a level translator IC. Normally used to take 3.3V logic to 5V logic. It runs on the 0V and -5V supply.

 

[golfmonke]

I actually have something working at the moment using some SSR's , however my test circuit (at the moment with a 15 ohm resistor, and 3v/-3v) should only be pulling 200ma, but it pulls 1amp, but only when I connect the multimeter, not sure why.

 

[ronsimpson]

not sure why.

post your circuit.

 

[Pommie]

Note that is at any time both mosfets are on then smoke will appear due to the supplies being shorted. Testing your code could be exciting.

Mike.

 

[golfmonke]

Testing the code will be the easy part for me anyway as a software engineer. Let me draw up the circuit...

 

[alec_t]

it pulls 1amp, but only when I connect the multimeter

How do you know what the the current is when your meter is not connected?

 

[golfmonke]

This is just some simple test code and a longer delay so I can actually test and see what's going on........

Next step once this circuit is ironed out will be to write the code for the actual timing.

 

[golfmonke]

How do you know what the the current is when your meter is not connected?

Ah, was thinking somebody might ask that last night while falling asleep... I have 2 bench power supplies that display the current draw.

 

[golfmonke]

I do have the load resistor/multimeter connections part on a breadboard, let me post a pic of that, just in case what I have on the breadboard isn't actually the same as what i've drawn...

 

[golfmonke]

The white wire and the grey wire at bottom are where I connect the multimeter, the green wire with the alligator clip is connected to the common +/- connection between the power supplies, the purple and grey at the top right are from the outputs of the ssr's

 

[golfmonke]

I have 3 power supplies on the bench, the one on the left and right are being used. I was using the one in the middle to manually turn on/off the ssr's before adding the esp32 to the circuit... and yes in this pic the power supplies are off, but the ESP32 is running.

You can't see the connections on the power supply on right, because they are in the back.

You can see the grey and white wires on the breadboard here better where i'm hooking up the multimeter....

You can also see my $30 crapolliscope at the bottom of the pic , was easier said than done actually scoping what i'm trying to reproduce here , good thing I knew before hand roughly what I was looking for in terms of signal because a few settings off on volts/division or time/division and the signal was junk

 

[Nigel Goodwin]

Are the outputs of your power supplies earthed?, if so you're probably shorting one of them out by connecting them -ve to +ve.

 

[golfmonke]

Ya, I was thinking that too, what's the proper way to not earth the power supplies, I have some power cords with the ground connector snipped off that I was going to test....

 

[ChrisP58]

I actually have something working at the moment using some SSR's , however my test circuit (at the moment with a 15 ohm resistor, and 3v/-3v) should only be pulling 200ma, but it pulls 1amp, but only when I connect the multimeter, not sure why.

Sounds like you might be using the multimeter wrong when measuring current. Like you have the meter set to measure current, but then placing the probes across the load. Your schematic in post 10 shows it connected that way. That's OK to measure voltage, but not current.

To measure current, you need to open a wire and insert the multimeter in it's place so that the current flows through the multimeter.

 

[golfmonke]

Sounds like you might be using the multimeter wrong when measuring current. Like you have the meter set to measure current, but then placing the probes across the load. Your schematic in post 10 shows it connected that way. That's OK to measure voltage, but not current.

To measure current, you need to open a wire and insert the multimeter in it's place so that the current flows through the multimeter.

Is that irregardless of whether my power sources are grounded to earth or not?

How about for the oscilloscope? I have an inductive ammeter that sends a voltage signal to the oscilloscope, but what about for just measuring the voltage? I've heard that can be a problem if the circuit is connected to earth ground...

 

[alec_t]

When writing your code I suggest you include a delay between switching one SSR off and switching the other SSR on. This is known as 'dead time' and allows time for the SSRs to respond. Without some dead time you risk having both SSRs being on simultaneously (albeit briefly) and shorting the two power supplies.

I have an inductive ammeter that sends a voltage signal

Eh??

 

[golfmonke]

When writing your code I suggest you include a delay between switching one SSR off and switching the other SSR on. This is known as 'dead time' and allows time for the SSRs to respond. Without some dead time you risk having both SSRs being on simultaneously (albeit briefly) and shorting the two power supplies.
Eh??

Ok, i'll check out the specs on those SSR's for dead time.

EDIT: well it appears the datasheets on these are less than helpful.... what are some common dead time durations on SSR's?

It's a current clamp that sends voltage signals and hooks up to oscilloscope.

**broken link removed**