At the moment it would take ages to explain why, so instead for now I will post the question.
Picture say 12 fans from a pc, in an array, the voltage 12V the I=max ~200mA, with another fan doing something else at 12V 1.2A max. 12 of the fans are driven by pwm @around 15 Hz and various duty cycles, one fan pwm 20KHz various duty cycles.
Playing around with this I obviously put snubbers in across the fans. The fans are not always on all together or at the same time, sometimes but not always. The single fan that sits on its own is on more than the others. While I havnt measured anything yet, but while doing something else i scoped a fan, and there is alot of pretty high spikes.
At the moment I am not using schotky diodes, but I will be later. Now the whole project once built will be mainly powered by a ATX psu for ease, and have a deep cycle large leisure battery for back up, we get alot of power cuts here (on average at least 2 a month lasting from 2 hours to 14 hours), the battery is there to give plenty of time to start up a small gen set and warm it up.
Battery back up will probably be for around 20-30 mins most times, on rare occasion upto 2 hours.
I was going to build a small mains powered trickle charger to keep the battery topped up and in shape, but I got thinking .
Would it be possible to direct some/all the back EMF into largish caps then into a trickle charge circuit to keep the battery topped up? There are several Pic's in my project so monitoring with a pic would be ok, one last thing, I dont want to use a special battery charging IC, I would prefer to build the whole thing myself.
Points to watch for and general comments wanted and welcome

ps. please refrain from stating the obvious! yes I am aware I am a foofing idiot

 

...
Would it be possible to direct some/all the back EMF into largish caps then into a trickle charge circuit to keep the battery topped up?...

Read my post #8 in this thread. What back EMF?

 

Hi Mike, I have read your post. I tried to make sure I was accurate and didnt leave anything out! but seems I did, When I was testing the fans especially the large one, there was no diode attached to the fan, I used a old FG that has a pretty good output current and toggled the pwm line direct with that, the back EMF I was talking about was seen without the diode. My intention originally was too use flybacks to protect the drive transistor's, then I had the idea of maybe using them to channel the back emf to a cap then into a trickle charger if that makes sense??
The scope showed some pretty high voltage spikes, I didnt bother measuring them as i was actually testing for something else.

 

Looking at it slightly differently, in order to trickle charge the battery, if i was doing it a more normal way, is it just a case of using a voltage reg set for 13.8V and maybe one used as constant current set to whatever figure a trickle charger uses (i dont know yet how much current it needs to trickle charge)?

 

There will be spikes on the collector of the transistor with no snubber diodes; not on the supply side of the load.

Instead of connecting the cathode of the snubber to the supply line, in-which case the current contues to flow through the motor winding, doing useful work. you could dump the current pulse somewhere else, such as into a big capacitor or battery. However, now that current is lost to the fan motor, and you will have to put more current into the fan motor from the main supply to make up for the loss. No something for nothing...

 

I wasnt after something for nothing as such, I thought because the voltage spike was pretty large i could use it to charge the battery, and be self contained. Otherwise I still have to build a circuit to raise the 12V to 13.8V anyway, so my thinking was more along the lines of using a regulator or probably 2?? to charge the battery and save myself having to build a boost circuit.

 

It would be like a poor man's boost smps...

 

It would be like a poor man's boost smps...

or a poor kids , actually you have made me think about all this. I have viewed the back emf as kind of wasted before, i know about the general operation of a inductor and fields collapsing blah blah blah. But i never realy considered that the current was actually used, I think I always viewed it as a pain in the neck you get rid of with a diode, the thoughts stopped there.
So I am on google and finding out as much as I can about inductive loads, what i would like, is to get to the point I know enough that i can work out which is better, use the back emf like normal and put back into the load, this meaning that i build a separate charging circuit for the battery, or will using the back EMF for charging and adding more current into the load would be better, maybe more economical. By economical I am not talking parts cost or that kind of thing, I am interested in which is most energy efficient, and more to the point I would like to learn how to work all that out!
So thanks for the input, it has made me stop and think

 

Here's a sim of an arrangement to get you started.
Use of steering diodes D1,D2 to control switch-over from PSU to battery as the fan power source, although simple, does have the side-effect that the battery never charges above the PSU voltage.

 

Here's a sim of an arrangement to get you started.
Use of steering diodes D1,D2 to control switch-over from PSU to battery as the fan power source, although simple, does have the side-effect that the battery never charges above the PSU voltage.

I'm not sure that D2 is needed? Depends on the PSU Voltage? Maybe use a Si diode so that the battery voltage can go higher?

 

psu is only 12V, the battery I have needs a charge of around 13.8V. its a large lead acid battery (marine leisure battery), I assumed I needed a higher voltage to charge it??
Having been read it might be easier to just build a charger, spent alot of time reading up on inductors, it all gets confusing and deep pretty quick , thanks for the sim, my go at a sim for it was a total fail

 

Here's a sim variation which allows the battery to charge above PSU voltage, to a level set by R1 and amplified diode Q1.

 

I'm not sure that D2 is needed?
...

No it's not. Neither is D1.

If the goal is to divert all the back EMF current into the battery you can replace D1 with a short circuit and totally remove D2.

It would also be a good idea to put a fast cap across the battery, like a 0.1uF or 0.22uF polyester.

 

If the goal is to divert all the back EMF current into the battery you can replace D1 with a short circuit and totally remove D2.

The goal is also to provide auto-changeover from the PSU to the battery.

 

hmmm I need to read a bit more, I will be back tho, and thank you for the suggestions!