<https://lh3.googleusercontent.com/-PBBb8HJADCs/Wl-aUfx0dpI/AAAAAAAAC4M/JdhcFh373sAuFtBvPz4SEMPdPtFU1z2GgCLcBGAs/s1600/IMG_20180117_192346%255B1%255D.jpg>
<https://lh3.googleusercontent.com/-EcZ9laxEN2E/Wl-Z3-srEuI/AAAAAAAAC4I/ZdWDoRjixFosyqFQvT-jT6kEf1Qbrg9wgCLcBGAs/s1600/IMG_20180117_194523%255B1%255D.jpg> Hi, I've been testing the popular anode switching circuit with one NPN and one PNP transistor. The circuit with my values is shown below: <https://lh3.googleusercontent.com/-KY9FZBzfztc/Wl-WzE8pZHI/AAAAAAAAC30/1jnJ6o2PVCQ0RVU6gYXqy1PjjmqiJTHPgCLcBGAs/s1600/anodeswitch.png> It is a fairly standard circuit. I've recalculated values to achieve lower current of the R25 resistor to limit dissipation on control elements. The circuit is repeated 4 times for 4 tubes. These are multiplexed with a blanking period implemented. During blanking, cathode switches are still on (so only anodes are blanked). Multiplexing frequency is 250Hz (so operations happen every 1ms). It generally works and switches the anodes, but... for some odd reason the PNP transistor (MMBTA92) decides to partially open whenever any other anode switch is open. I've checked the circuit with an oscilloscope and it seems that even with long blanking periods (over 0,5ms), the transistors will still falsely trigger. I've checked the bases and emitters and there is absolutely nothing, that could make the transistor open. I made few photos to describe the problem better: <https://lh3.googleusercontent.com/--vM3D_nYanY/Wl-YukXxTuI/AAAAAAAAC4A/vfpOgAxgGUodL-0DkbtF60sITIRxLydPQCLcBGAs/s1600/IMG_20180117_190738%255B1%255D.jpg> This is the voltage directly on an anode. As you can see, it is open properly during its intended period, and then opens up again when it's anothers tube time for opening. There are also two small deviations when next two tubes are firing, and then the cycle repeats. The problem gets far more visible when the blanking time is lower (here it's lower than 100us): <https://lh3.googleusercontent.com/-EcZ9laxEN2E/Wl-Z3-srEuI/AAAAAAAAC4I/ZdWDoRjixFosyqFQvT-jT6kEf1Qbrg9wgCLcBGAs/s1600/IMG_20180117_194523%255B1%255D.jpg> Now it all looks like it is driven incorrectly and transistors are being open by the program. But it is not the case, here is the oscillograph taken from Q14 base: <https://lh3.googleusercontent.com/-PBBb8HJADCs/Wl-aUfx0dpI/AAAAAAAAC4M/JdhcFh373sAuFtBvPz4SEMPdPtFU1z2GgCLcBGAs/s1600/IMG_20180117_192346%255B1%255D.jpg> As you can see, the signal at the base looks perfect! The signal appears only 1/4 of the time as supposed. This effect of course causes ghosting, which usually is solved by implementing the dead time and not closing the cathodes during dead time. I've implemented both those methods and still I'm getting fooled by MMBTA92s! The questions are: Did anyone encounter similar problem with this circuit and/or these transistors? Does anyone have any idea why do these transistors keep opening when they have no Vbe drop? And why on earth anode voltage is kept low during blanking time, but then rises again without a base signal? It would make alot more sense if the transistor didn't close at all! But it closes as supposed and then reopens. -- You received this message because you are subscribed to the Google Groups "neonixie-l" group. To unsubscribe from this group and stop receiving emails from it, send an email to neonixie-l+unsubscr...@googlegroups.com. To post to this group, send an email to neonixie-l@googlegroups.com. To view this discussion on the web, visit https://groups.google.com/d/msgid/neonixie-l/84396188-25bc-4baf-b20e-bbd20fb3c792%40googlegroups.com. For more options, visit https://groups.google.com/d/optout.