Well spotted. 47 ohms won't work here.

 

spikes maybe a good point, but 22V is not enough to kill your 7805, right?

I wouldn't normally think so, but then I would never expect a 7805 to be killed by reverse dump from the 5v 10uF on it's output. If it had 1000uF on the output and someone shorted the Vin pin to ground then ok, but that is not happening here.

It's pretty trivial to add a resistor and larger cap on Vin, and if there is a spike issue that may be all that is needed to save it.

RussPatterson, 200mA used by the 5v circuit sounds excessive (based on your circuit), I would try a few resistors and run everything ON (max Vout supply current needed) and see what resistor gives you 1v or 2v drop when everything is running.

If it actually does use 200mA then a resistor of about 10 ohms will drop 2v @ 200mA.

You should consider making the solar regulator a "short" type, you already have the FET and all it needs is a diode (between panel and battery) and rewiring the FET (so that it shorts panel + to ground).

When short circuited the solar panel will only produce a similar current to its full power current into the battery (current which we know your FET can handle), and the FET will short it to a very low voltage so there will be very little heat generated. The panel itself will run much cooler short circuited than open circuited.

 

So help me out with calculating the power dissipated in that 47 Ohm resistor. Assume max of 200mA for circuit (PIC, status LED, etc.)

So voltage drop would be: .2 * 47 = 9.4V. Something must be wrong because I need at least 7 to run the 7805.

Yes 200mA is too much,I thought you are using the mcu for only switching purpose.Anyway a 10 ohms 1/2 watt or 15 ohms 1 watt will do.

Good Luck

 

I wouldn't normally think so, but then I would never expect a 7805 to be killed by reverse dump from the 5v 10uF on it's output.

yeah, good point. I never managed to kill my 7805...

 

You guys are right. 40 or 50 mA is a better estimation of the max for the board. It's running at 20 right now (but not doing anything but blinking).

@THE_RB, Thanks for the info on the panel short. So take a look at the attached schematic. Is that what you intended? Thanks.

 

Question: on the recommendation for the resistor and 1000uF cap in front of Vin of the 7805. That should be a filter cap correct? Not a series cap? Thanks.

 

Hi Russ, the schematic change to short the panel looks ok, apart from the top of R2 needs to be at 12v when the FET is ON, ie when panel is shorted. I would connect the top of R2 to the battery side of the diode, so it is always at 12v.

Also your FET looks upside down. I assume it is an N FET? The D should be at the top, S to gnd.

If it is not an N FET then you need to change to N FET. You can probably change R2 to 10k, as you will only be switching this every few seconds or so and don't need super fast FET turn on. For that matter a small cap say 0.1uF across FET G to ground will give slower smoother switching and less chance of noise and induced switching currents.

Re the RC filter; yes a resistor before C1, and C1 of say 1000uF between Vin and Gnd.

 

I would like to use an N-FET, and tried many times. The issue I had was that sometimes it would not turn off because of the panel's negative-side voltage potential was often different than that of the battery. So, depending on panel, the Vgs on the N-FET would not be 0 when the PIC drove the gate low. The Source on the N-FET would still be a voltage well below that and it would stay on.

Ultimately I'd like to use a high-side driver IC that had an internal voltage boost so I could run an N-FET switching the high side on the panel, like I'm doing now with the P-FET. I'll put this in the simulator and see what happens with the Vgs.

 

Just for the record. I found this in the LM117 datasheet by coincidence:

"When external capacitors are used with any IC regulator it is sometimes necessary to add protection diodes to prevent the capacitors from discharging through low current points into the regulator. Most 10 μF capacitors have low enough internal series resistance to deliver 20A spikes when shorted. Although the surge is short, there is enough energy to damage
parts of the IC."

Interesting, I didn't know that it can be THAT problematic.

 

Wow 20A from a little 10uF. Who'd a thunk it?

 

Just for the record. I found this in the LM117 datasheet by coincidence:

"When external capacitors are used with any IC regulator it is sometimes necessary to add protection diodes to prevent the capacitors from discharging through low current points into the regulator. Most 10 μF capacitors have low enough internal series resistance to deliver 20A spikes when shorted. Although the surge is short, there is enough energy to damage
parts of the IC."
...

Maybe with a tantalum cap? I just grabbed my ESR meter and tested a handful of 10uF 16v electros, they were all about 1.5 ohms ESR, at low current (couple mA) and I expect MUCH more than 1.5 ohms at higher currents. But even if at 1.5 ohms worst case, to me that sounds more like 5v/1.5 = 3A (and probably MUCH less) and that is only if you can instantaneously short the input Vin.

In this case the OP had a 100uF on 12v Vin and 10uF 5v Vout, so the Vin cap will need to drain first before any current will flow back. It's very rare to blow a 7805 regulator when Vin cap is much larger (10x) then Vout cap, and there is nothing the OP has said to make us think he was continually shorting the Vin to ground anyway. I understand the need for the diode and why it's in the datasheet but I'm still not really buying it that it is the exact cause of the regulator failures in this case.

Russ, the NFET will work fine now you have the FET source pin at ground! Before you had the FET placed between 12v battery and solar panel +, so an NFET would have been very hard to drive there. Now it is easy.

With the circuit you just posted, just swap the FET D and S and replace the part with an NFET. And put that resistor to battery 12v.

 

Russ, the NFET will work fine now you have the FET source pin at ground! Before you had the FET placed between 12v battery and solar panel +, so an NFET would have been very hard to drive there. Now it is easy.

With the circuit you just posted, just swap the FET D and S and replace the part with an NFET. And put that resistor to battery 12v.

I think you're right. That's pretty cool. I've been working on driving that panel regulation; on and off for quite a while. This is fewer parts, less RDSon, more efficient, everything's better. I'll give it a try. Although it probably doesn't belong in this 7805 thread. I'll start a new one.

 

Alright, here's a new twist to the 7805 regulator puzzle. After reading ~12 V on the output pin for the controller I was working on (I've had this issue several times before) I got disgusted and left it on the bench. I never de-soldered and replaced it. Today, it works. 1.4V on the input and 4.8V on the output. I know that I measured the voltage correctly previously. I'm stumped, I would blame it on my cheap Harbor Freight meter but I don't know how it could malfunction like that.

So a week later it works???

 

Maybe it was a wiring short of some type? Although Vin and Vout are on opposite sides of the TO-220 pack 7805 so it's unusual to short Vin to Vout in the PCB.

Another guess would be that some current had leaked back from your 12v side (FET or motor driver) to the 5v rail and slowly charged the 5v rail up to 12v?

I'm just clutching at straws, it's probably just as likely your 7805 was/is faulty and is doing strange things.

 

hi,

putting the diode protection would improve the circuit, but here's the bumber, if your regulator does not last long as expected, there's a high probability that what you have bought is a FAKE. there's a lot of that in the market now.

moz

 

Thanks all for the input. I think there may be something to the idea that the 12V side is "leaking" back to the 5V through a FET, and also that the 7805 is flakey after the initial failure (if in fact it failed or there is something that lets the 12V back through the gate of one of the FET's).

I don't think it's a fake. I got those from Digikey.