On 10/10/2022 5:46 AM, sonnic...@gmail.com wrote:> On Friday, 30 September 2022 at 01:35:05 UTC+3, John S wrote: >> On 9/29/2022 2:43 PM, sonnic...@gmail.com wrote: >>> Hi all >>> >>> I am playing around with a solar lamp, pretty simple like this >>> https://www.pinterest.com/pin/540150549038731182/ >>> >>> However I do not (yet) have D2. >>> >>> D1 in my case is a Schottky 1N5818, and my battery 7.2V >>> I am surprised how much reverse current I can get through the diode. Without the solar cell, it does not turn on. There is enough current through D1 to the transistor to turn the whole thing off. >>> With the solar panel on it works as expected. >>> D2 will solve my issue, but I am also thinking that not completely as it will still get current through reverse D1 and forward D2. >>> >>> I made a test - Schottky diode 1N5818 -> 2K2 resistor -> white LED. >>> With the diode in reverse the current is 3uA, just enough to get a weak blue shade of the LED. >>> >>> Back to my solar lamp. It works with the solar cell so I do not have a big issue. There is 10 days to full moon then I will see the result of that. >>> But I could also just a classic 1N400x diode, which should not have that much reverse current. >>> In the same test I get no reverse current. >>> >>> The idea of D1 as a Schottky is to get most power from the solar panel. My loss seem to be minimal though the reverse current so it is acceptable. >>> >>> But what do people here think? >>> >> I think you could use a FET as a low-loss diode in place of D1. > > Hi > > I looked into and played around with it, and it works for power consumption units. However for solar units, or units with a power source on both sides it does not work.For that you need two in series, p-channel source-to-source.> I did found that during the night with colder temperature the leak current is small (~5uA for 1N5818/9) and having the resistor for the shut-down transistor at 470K I have no problems. > > > Read more here: https://www.rs20.net/w/2013/02/using-mosfets-as-blocking-diodes-reverse-polarity-protection/
Schottky diodes, reverse current and solar changers
Started by ●September 29, 2022
Reply by ●October 10, 20222022-10-10
Reply by ●October 10, 20222022-10-10
mandag den 10. oktober 2022 kl. 12.51.02 UTC+2 skrev John S:> On 10/10/2022 5:46 AM, sonnic...@gmail.com wrote: > > On Friday, 30 September 2022 at 01:35:05 UTC+3, John S wrote: > >> On 9/29/2022 2:43 PM, sonnic...@gmail.com wrote: > >>> Hi all > >>> > >>> I am playing around with a solar lamp, pretty simple like this > >>> https://www.pinterest.com/pin/540150549038731182/ > >>> > >>> However I do not (yet) have D2. > >>> > >>> D1 in my case is a Schottky 1N5818, and my battery 7.2V > >>> I am surprised how much reverse current I can get through the diode. Without the solar cell, it does not turn on. There is enough current through D1 to the transistor to turn the whole thing off. > >>> With the solar panel on it works as expected. > >>> D2 will solve my issue, but I am also thinking that not completely as it will still get current through reverse D1 and forward D2. > >>> > >>> I made a test - Schottky diode 1N5818 -> 2K2 resistor -> white LED. > >>> With the diode in reverse the current is 3uA, just enough to get a weak blue shade of the LED. > >>> > >>> Back to my solar lamp. It works with the solar cell so I do not have a big issue. There is 10 days to full moon then I will see the result of that. > >>> But I could also just a classic 1N400x diode, which should not have that much reverse current. > >>> In the same test I get no reverse current. > >>> > >>> The idea of D1 as a Schottky is to get most power from the solar panel. My loss seem to be minimal though the reverse current so it is acceptable. > >>> > >>> But what do people here think? > >>> > >> I think you could use a FET as a low-loss diode in place of D1. > > > > Hi > > > > I looked into and played around with it, and it works for power consumption units. However for solar units, or units with a power source on both sides it does not work. > For that you need two in series, p-channel source-to-source.https://electronics.stackexchange.com/questions/223935/understanding-an-ideal-diode-made-from-a-p-channel-mosfet-and-pnp-transistors
Reply by ●October 10, 20222022-10-10
On Thursday, September 29, 2022 at 2:43:38 PM UTC-4, sonnic...@gmail.com wrote:> Hi all > > I am playing around with a solar lamp, pretty simple like this > https://www.pinterest.com/pin/540150549038731182/ > > However I do not (yet) have D2. > > D1 in my case is a Schottky 1N5818, and my battery 7.2V > I am surprised how much reverse current I can get through the diode. Without the solar cell, it does not turn on. There is enough current through D1 to the transistor to turn the whole thing off. > With the solar panel on it works as expected. > D2 will solve my issue, but I am also thinking that not completely as it will still get current through reverse D1 and forward D2. > > I made a test - Schottky diode 1N5818 -> 2K2 resistor -> white LED. > With the diode in reverse the current is 3uA, just enough to get a weak blue shade of the LED. > > Back to my solar lamp. It works with the solar cell so I do not have a big issue. There is 10 days to full moon then I will see the result of that. > But I could also just a classic 1N400x diode, which should not have that much reverse current. > In the same test I get no reverse current. > > The idea of D1 as a Schottky is to get most power from the solar panel. My loss seem to be minimal though the reverse current so it is acceptable. > > But what do people here think?That's not a bad circuit at all. Allow up to 100uA reverse current through D1- just as a hypothetical. This will circulate from BATT(+), reverse through D1, forward through D2, through R1 to Q1 base and turn it on, which turns Q2 and the LED off. Something you don't when the solar cell is off. To prevent this from happening, you shunt Q1 base with a resistor ( from base to GND) such that 100uA develops no more than 0.3V. That would be R=0.3V/100uA= 3kR, make that 3.3kR. Now, you should have no problems. When the solar cell is on, the loading of the shunt resistor is only about 0.6V/3.3kR less than 200uA which is not much at all, and leaves plenty of current to drive Q1 when the solar cell takes over..
Reply by ●October 10, 20222022-10-10
On Friday, September 30, 2022 at 11:50:33 AM UTC-4, Martin Brown wrote:> On 29/09/2022 23:54, John Larkin wrote: > > On Thu, 29 Sep 2022 11:43:33 -0700 (PDT), "sonnic...@gmail.com" > > <sonnic...@gmail.com> wrote: > > > >> Hi all > >> > >> I am playing around with a solar lamp, pretty simple like this > >> https://www.pinterest.com/pin/540150549038731182/ > >> > >> However I do not (yet) have D2. > >> > >> D1 in my case is a Schottky 1N5818, and my battery 7.2V > >> I am surprised how much reverse current I can get through the diode. Without the solar cell, it does not turn on. There is enough current through D1 to the transistor to turn the whole thing off. > >> With the solar panel on it works as expected. > >> D2 will solve my issue, but I am also thinking that not completely as it will still get current through reverse D1 and forward D2. > >> > >> I made a test - Schottky diode 1N5818 -> 2K2 resistor -> white LED. > >> With the diode in reverse the current is 3uA, just enough to get a weak blue shade of the LED. > >> > >> Back to my solar lamp. It works with the solar cell so I do not have a big issue. There is 10 days to full moon then I will see the result of that. > >> But I could also just a classic 1N400x diode, which should not have that much reverse current. > >> In the same test I get no reverse current. > >> > >> The idea of D1 as a Schottky is to get most power from the solar panel. My loss seem to be minimal though the reverse current so it is acceptable. > >> > >> But what do people here think? > > > > It's a horrible circuit! > > > That may be but it is just about the simplest one that a teenager today > might still be able to build from first principles and a scrap box. > > Having the ability to fine tune the light level where it comes on would > be nice to add. > > Commercial ones assume a low latitude and come on during the long > twilight of UK summer evenings fading out before it gets properly dark! > They die horribly every winter if left outdoors. > > The choice of battery voltage is too low for a white LED though. A 4.8v > battery and a nominal 6v solar panel would be a much better choice. > > R1, R2 are bit low though assuming transistors with a gain of at least > 100 and 50mA drive current you can get away with R1 = 220k, R2 = 2.2k. > (and 5x higher for a 10mA drive current) > > The better cheapest designs are along the lines of a Joule thief using a > single cell. They all turn on far too soon in the UK though.You think that cheap trash from China knows it's in the UK? It will malfunction regardless. They have the inverse problem with their phony controllers that are supposed to turn on at dawn. The sunrise somehow manages to sneak past their brilliance and the circuit stays off indefinitely. Wave your hand over the solar panel to create a dark/light abruptness and it comes on. Night lights, in addition to causing light pollution, have been found to be a significant disruption to insect populations contributing to their decline. If they're not fitted with a motion sensor, they should be prohibited.> > -- > Regards, > Martin Brown
