So is this where EEs, or at least people who aren't dumb like me when it comes to circuits, hang out?
Because I have this problem: I have two loads, both about 50 Ohm, and I want the sum of power on both to be roughly constant (about 0.5 W -- sourced from a programmable PSU). I want to be able to change the relative distribution of the power across these two loads reasonably quickly -- within 1 ms.
How do I do this? Plus I need to control it via a PC (i.e., the distribution changes and is not know a priori), so something Arduino (or mcu in general) - based would be ideal.
I attempted what I attached in the pic -- switching an N-channel MOSFET with PWM. It seems to work reasonably well, the PWM is going at 250 kHz (both the same output but one is inverted) BUT, it seems, the PWM interferes with other instruments -- this is thing is part of a physics experiment. When I set the PWM duty cycle to something other than 0% or 100%, I see spurious signals all over the place on the rest of the instrumentation. Even when I disconnect the programmable PSU and short the wires (but not when I just leave them hanging -- something I don't understand at all). I suspect that the unshielded wires leading to the load (about a meter) act as an antenna, but the frequency still seems rather low.
(I tried lowering the PWM frequency to 125 kHz with not much effect. I cannot go much lower than that because there are actual processes that I measure happening at <50 kHz and I don't want the load to appear modulated.)
So, is there some reasonably simple circuit that can accomplish this? Or am I just doing something really stupid? My ideal solution would be: "there is an IC for that".
Thanks.
>>7653497
oh, ignore that 5 V on the PSU, that's just the drawing program being dumb
No? No one?
OK, I'll go and cry in the corner.
By myself. All alone.
;_;
The provided information is not enouth to help you more but telling you the standard emv stuff.
>>7653497
Try asking /diy/ but it's p. slow.
Just go to stack exchange.
>>7653497
>When I set the PWM duty cycle to something other than 0% or 100%, I see spurious signals all over the place on the rest of the instrumentation.
It sounds like you haven't added any decoupling capacitors. Add some decoupling capacitors.
Are your loads actually resistive (heating coils, incandescent bulbs, etc), or did you just measure the resistance of something else with a multimeter?
/diy/ would know, /sci/ doesn't know anything but abstraction
Wow, the thread is still alive. Thanks for the responses. I'll try /diy/. I didn't even know we had that board.
>>7654092
I'm not sure I understand. Decoupling capacitors, as in in series with the load? Or parallel to the load (i.e., a low pass filter)? I can see how a low-pass filter would help (I'm dumb for not already trying it) but where should I put the decoupling capacitor?
>>7654092
And yes, the loads are just few cm of resistive wire (it's a heater).
>>7655382
The LEDs there are just indicators. Could they be causing this?
And the Vcc of the MCU? Well, I haven't looked at it through an oscilloscope but it's a 5V voltage regulator (7805) powered from a 9V battery OR the 5V output from an Arduino. There's no difference between these two.
>>7655410
well, i usually try to prevent directly driving LEDs with an output. it's only like 10mA but still...
try removing them out for testing. why the hard(1k) pulldown? i'd use 100k or so.
>>7655451
also, did you try using another power source (for the power circuit aka heaters) for testing? do you have a big C(100u?) parallel to the psu close to the loads? tried a diode(4148) in series to prevent backflow to the power source?
>>7655451
>>7655461
I'm not really worried about the current (the PWM pin can source 50 mA) because the gate current on the MOSFET should be negligible if I understand this correctly, but I'll try to rip the LED out and see what it does.
The 1k pulldown -- no reason. I don't have a very good intuitive feel for what is and isn't a right value so I just put something there that seemed safe. Actually at first I forgot to put the pulldown there completely and the situation was no different.
I don't have a different PSU, but I doubt that that's an issue because I see the problem even when I disconnect the PSU and short the leads.
I don't have the capacitor there, which I see now is probably a mistake. And the diode will probably be useful too, because I've noticed some DC current through the load when the PSU is off and PWM is not fully on or off. I'll try tomorrow and see.
Thanks.
