开云体育

I just spent some time on DigiKey looking for a sufficient transistor. I plan on using a BJT, because its the only type I'm familiar with at the moment.

I found the MPSA43 to be a great candidate.
-Vce up to 200V
-Vcb up to 200V
-Veb up to 6v
-Ic up to 200mA

C - HV PSU -> collector
B - PIC output -> resistor -> base
E - emitter -> resistor -> anode of nixie tube

Now I think this would work. However, I then looked at some of the nixie tube designs in the Files section. What do you know, he used the MPSA42 also. So I've got some validity to that decision. However I noticed oz2cpu used two resistors for each set of on/off switches. I don't quite understand why. I would guess perhaps to protect the PIC I/O pins from any faults/leaking current. I know transistors somewhat, but not as much as I will after next semester. There is a lot I don't know, and this is a good example. Could someone please explain this to me. Thanks.

Also, whats with the 100n capacitor on the VCC/GNDs?

-Mike

On 7/2/10 7:03 PM, msilv3r wrote:

C - HV PSU -> collector
B - PIC output -> resistor -> base
E - emitter -> resistor -> anode of nixie tube

Now I think this would work. However, I then looked at some of the nixie
tube designs in the Files section. What do you know, he used the MPSA42 also.
So I've got some validity to that decision. However I noticed oz2cpu used
two resistors for each set of on/off switches. I don't quite understand
why. I would guess perhaps to protect the PIC I/O pins from any
faults/leaking current. I know transistors somewhat, but not as much as
I will after next semester. There is a lot I don't know, and this is a
good example. Could someone please explain this to me. Thanks.

Also, whats with the 100n capacitor on the VCC/GNDs?

-Mike

Mike,

The reason to use two transistors is that you need a level translator to isolate the PIC output at 0-5V from the anode of the nixie tube which sees up to 180V. Otherwise you'll find that you either can't turn on the anode or that you can't turn it off or that your PIC blows out.

The standard approach used in most nixie clocks is an NPN transistor driven by the PIC whose collector controls the base of a PNP transistor that gates the tube anode.

Here's my circuit:



You can use an MPSA42 for the NPN and an MPSA92 for the PNP; they cost less than the Zetex transistors I used.

I recommend the resistor values in the above schematic, as it worked flawlessly for me. Others have had good luck with different values.

--
David Forbes, Tucson AZ

On 7/2/10 7:03 PM, msilv3r wrote:

C - HV PSU -> collector
B - PIC output -> resistor -> base
E - emitter -> resistor -> anode of nixie tube


Mike,

Here's my circuit:



I recommend the resistor values in the above schematic,
as it worked flawlessly for me...

David Forbes, Tucson AZ

Use the circuit that David has shown. The resistor values can be varied a bit, but use them in all the places shown. They are all important.

The 'one transistor' circuit, you described is an 'emitter follower', which has NO voltage gain. The emitter will roughly follow the base, so the voltage will never exceed 5V of the base (actually ~4.3V). The voltage on the collector can be 6V or 200V, but the emitter will follow the base.

Remember for a transistor to turn ON there needs to be a current flow from base to emitter. This can only occur (for an NPN) if the emitter voltage is less than the base voltage, by the forward voltage drop of silicon. That voltage being ~0.7V at room temperature. An 'emitter-follower' circuit has current gain, but no voltage gain. In short, no worky.

Google it, if you don't believe me.