Greetings All,
I'm wondering if someone could take a look at a schematic i made and answer a few questions i have. the 250 volts is connected to solar panels. i only show 2 heater but it will end up being 5 when i'm done. each heater will have its own ir2110. i don't understand the ground part. i think the vss and com on the ir2110 should go to the 5 volts dc black wire (ground). or maybe the 12 volt black wire (ground) ?? also the solar panels don't have a ground. just a positive and negative wire. would this schematic even work with no solar ground? as you can see i really don't understand what "ground" means. go easy on me, i know i don't know what i'm doing.

 

Greetings All,
I'm wondering if someone could take a look at a schematic....

Did you forget something?

I think we might answer the questions without it but it's so much easier to have a picture to look at.

In the meanwhile, remember that all voltages are relative, and electronic devices respond to relative voltages. In most schematics, "ground" is just a handy reference point such as the negative pole of a battery. It isn't necessarily genuine earth ground. A solar installation might need an actual earth ground and if it does, that might make a good reference point. So we're back to needing a schematic to sort out what is needed.

 

forgot to attach.

 

"GND is the negative connection for all the positive voltages, and the positive connection for the negative voltage." i read that in another post. i can wrap around that idea.

 

forgot to attach.

OK, from what I see, the 5VDC and the 12VDC share a common ground but as shown, the solar panel is floating. That won't work because the MOSFETs need about 10V between their gate and source pins in order to conduct fully. The only way to ensure that is to tie the low side of the panel to power-supply-ground.

"GND is the negative connection for all the positive voltages, and the positive connection for the negative voltage." i read that in another post. i can wrap around that idea.

You really don't have any negative voltages in your circuit, so for now you can just ignore that.

But as I noted above all voltages are relative. You can choose to call "ground" whatever you like. Picture an audio amp. For a variety of reasons it's convenient to call the chassis ground, and to anchor one pole of the speaker outputs to that. The voltage at the speaker then oscillates to positive and negative voltages, say ±60V, with respect to that reference. The speaker cone goes in and out. To provide the power to reach negative voltages, the power supply inside the audio amplifier must be able to handle negative voltages. But the designer could just as easily declare the chassis is +60V and the amp can produce a signal ranging from 0-120V. The speakers don't know or care what arbitrary voltage you label the chassis.

Now, in reality it does matter if the chassis of an audio amp has a voltage other than earth ground, because the human user might be grounded and might touch the chassis.

 

Backing up a bit, why are you using gate drivers and an H-bridge to turn heaters on and off? You can control a heater with a single MOSFET and you don't need a gate driver unless you're PWMing it. That still doesn't explain the complicated configuration.

 

perhaps its best to explain the best i can what i'm up to. seems only fair since your taking the time to teach me.

back in the 90's and early 00's some work was done on solar panels directly heating water. the idea is to use 5 to 7 different heating element (each with a different amount of resistance). they used photodiodes to measure the sun brightness and change to the proper heating element base on that. they used relays to turn the elements on and off. thats what i have setup now. but i don't use photodiodes, my computer reads the voltage,it knows the resistance. the computer can then figure out the watts. the dc relay thing really is big and loud when it switches on and off so i'm wanting to use mosfets. to get as many different resistance combinations as possible i need a high side and low side mofet on both sides of the heater elements. i think.

 

to get as many different resistance combinations as possible i need a high side and low side mofet on both sides of the heater elements. i think.

I don't see how that helps but I'm picturing the elements in parallel. If your heating elements are stacked in series, you're going to need something else. Is there a reason they can't be in parallel? Just connect all the heaters to the high side of the panel and use MOSFETs to control the path to ground (panel negative) for each one separately.

 

max resistance i can get is 67 ohms. anything else becomes expensive and uncommon. max ohms they used on their 200 volt system was 180 ohms. sometimes i need them parallel, sometimes i need them series.

power-supply-ground. i believe i know what you mean by that. but can you do that with solar panels ??

 

max resistance i can get is 67 ohms. anything else becomes expensive and uncommon. max ohms they used on their 200 volt system was 180 ohms. sometimes i need them parallel, sometimes i need them series.

Got it. In that case you might look up circuits people use for controlling individual cells in a battery pack or capacitors in a bank. Same problem - they need to be parallel at times and serial in others.

 

power-supply-ground. i believe i know what you mean by that. but can you do that with solar panels ??

Yes, the ground (properly, the negative side) can simply be connected to power-supply-ground.

But another way to do this, which may be required for this application, uses complete isolation between the two sides, just as a relay accomplishes.

 

its easy to supply ground in my case. the panels cable goes right by my house (earth) ground. so the idea would be to connect a ground wire from the solar panel to the house ground ??

 

its easy to supply ground in my case. the panels cable goes right by my house (earth) ground. so the idea would be to connect a ground wire from the solar panel to the house ground ??

No, I don't think that's necessary here. I'd follow the installation instructions that came with the panel. If they recommend grounding, go ahead and do that. But I don't think it enters into the circuit design.

 

panel instructions talk about grounding but they mean it more in a safety way. panel frames need to be connected to ground for safety. in the end i only have 2 wires a big red and a big black. i tried something before the relays. i had a logic mofet irl640 it wasn't strong enough to handle the panels so it turned on the irf740 mosfet. i worked on my 20 volts dc power supply, but when i connected it to the panels the irf740's began smoking. i believe a ground is needed. i think.

 

Connection to earth ground will not have any impact on the MOSFET issue under normal operation. (It might help protect it from damage.). A MOSFET usually gets hot because it’s not turned fully on or off but somewhere in between where it has resistance. If the current is too high and the MOSFET is not heat sinked, of course that can blow it up also.

 

k, i'll leave out the ground on the panels. what about the 12 volt power supply. i have 2 caps on it. for all 5 elements i would need 5 caps. do you think thats to much ??

 

k, i'll leave out the ground on the panels. what about the 12 volt power supply. i have 2 caps on it. for all 5 elements i would need 5 caps. do you think thats to much ??

I missed that earlier. You need just one between +12 and ground. The capacity you need there depends in part how quickly you are switching. I suspect a single 1000 microfarad is fine for your very slow project.

But you need to find a solution first for the heaters in series before you worry about sizing the components. What you have so far will not work with 5 heaters in series. I’ve seen one that may work and will post it if I can find it.

 

wow, i didn't even realize that it won't work 5 in series. i'll take another look.

wait a minute, why woun't this work. turn on all the high side mosfets, turn off all the low side mosfets except the very bottom one in the circuit diagram. only one way for the electrons to flow. though all the heaters to ground.

oh wait, i see what your saying. pretend the circuit diagram had all 5 elements. their no way i can just turn on element 2. electrons must go thru element 1 to get to element 2. for the love of pete. you would think i could figure that out. glad you caught that. would have been a bummer to wire this up just to realize its not going to work like i planned. if you'll look for the other circuit design i'll try to figure out the changes i'm going to have to make on this one.

 

So to be clear, the goal is to be able to have 2-5 in series, or 2-5 in parallel, or of course just one by itself ? Are there more requirements?

 

wait i see the problem. if all 5 heater were on the circuit diagram. it's impossible to turn on element 2 alone. electrons must go thru element 1 to get to element 2. if you'll look for the circuit design you saw before i'll try and fix this one. i think i know what to do.