Dallas+SMUC=clock reset problem resolved.

Hello, All I think I have overcome the problem of resetting the clock. :v2_clapp; Loss; them at turning off the power, when resetting the machine, when freezing. Or resetting them, or whatever call it... In general, after studying the problem, I came to an almost unambiguous conclusion (respect then Kirill Frolov) that the clock freezes or is completely lost at the moment of reset. Mechanical, by pressing the reset button. With this reset, not a single impulse, but a pack. The solution is the following. (possible only, unfortunately, if you have ZXMC from CARO, or any other device that forms ONE, and not a pack, as the button does). I completely eliminated the mechanical reset. The reset was taken from 1 leg of ATMEGA (Reset), distributed using one logic body (LH1) separately on car, separately on SMAK, or rather separately on Dallas - because on Smuka the reset is in progress there's only Dallas. At the same time, in Dallas in the original, Reset goes to the 18th leg, and through 3 inverters 561LH2 (apparently for an easy delay) on the 13th leg - /CE. 13 the leg must be hung on the ground. In later models this cascade is already excluded - there is no this microcircuit (561), and the 13th leg is simply snot hung on the ground. Next, I completely removed the decoupling cascade of the main power supply and the backup one. I left only the diode through which backup power is supplied. I cut off the entire 24 leg (power supply). I left only the diode from the externalrecharge. I soldered the power from the computer through a diode (that is, it turns out that the power it just goes through 2 diodes - from the computer and from the battery). I soldered the battery - a coin cell battery 2032 - lithium, 3 volts. That's all. For the second day, the clock has been ticking with a bang. Resets do not affect it in any way, I already artificially drove the car, and was perverted in every way - pah pah, everything is ticking. In general, it would be interesting for someone to repeat it. And one more thing. In my opinion, 2032 type batteries are not enough. Under load, and connected in series through a diode there is a real voltage drop, and as a result, we have a Dallas backup voltage of something around 2.5 Volt, which in my opinion is on the verge of a foul - it might go wrong. I'm going to find a standard battery like on old motherboards 3.6 Volt. In theory, with all the losses, it will power Dallas with 3 volts, which should already it is guaranteed to support them, because, in my opinion, motherboard manufacturers do not They were fools and chose the optimal supply voltage. And it has more capacity. That’s it. In general, I’m waiting for some information from anyone. Confirming, or refuting my calculations..

Hello, Mikka_A Congratulations!!!

Hello John North Finally!! :v2_cheer:

Hello Orionsoft Schottky diodes have a much smaller voltage drop. you can still try Instead of a battery, use an ionistor.

Hello Sonic Son> So the half-pin button can be upgraded using a trigger or Son> monostable. yeah. I wanted to do this, but then I realized that resetting from the multicard was enough for me. especially since it is formed regardless of the state of Scorpio located.

Hello, The Exploited The> on Schottky diodes the voltage drop is much less. still possible The> try using an ionistor instead of a battery. yes, I already bought a battery, which were usually used in mothers - nickel metal hydride - 3.6 volts 60 milliampere hours - 60 rubles, all pleasure. Now on Dallas 3.2. I think that's enough.

Hello, Mikka_A Mik> The solution is as follows. (only possible, unfortunately, if you have ZXMC from Mik> CARO, or any other device that forms ONE, and not a pack like this Mik> does the button). So the half-pin button can be upgraded using a trigger or a one-shot device.