Using transistor as anode switches
ggtnuminix
discrete transistors. I understand the portion of using a MPSA42 to
control the cathode side of a nixie; however, I dont understand why
you couldn't use the same approach to control the tube anode. Most of
the designs seem to use a MPSA42 and MPSA92.
I came across a discussion board entry which explained that:
"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. "
I'm not grasping the concept of why the MPSA42 method works on the
cathode side but not the anode side.
Any further insight/explanation would be appreciated.
Adam Jacobs
http://www.kpsec.freeuk.com/trancirc.htm
-Adam
--
You received this message because you are subscribed to the Google Groups "neonixie-l" group.
To post to this group, send an email to neoni...@googlegroups.com.
To unsubscribe from this group, send email to neonixie-l+...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/neonixie-l?hl=en-GB.
David Forbes
> I have been reading Jason Harper's article on driving nixies with
> discrete transistors. I understand the portion of using a MPSA42 to
> control the cathode side of a nixie; however, I dont understand why
> you couldn't use the same approach to control the tube anode. Most of
> the designs seem to use a MPSA42 and MPSA92.
The Nixie tube has two voltage levels to which your logic needs to exert
control. The cathodes are connected by transistors to the low voltage
(ground), and the anodes are at the high voltage of 180V.
A transistor's base, its logic input, must be at the same voltage
(within one volt) as the emitter, which is connected to either ground
for the cathode or to 180V for the anodes. The reason is that the base
and emitter are connected internally by a diode which conducts current
and has that 0.7V voltage drop that you read about.
The logic circuits all run near ground, so they can be connected
directly to the transistors which control the cathodes. But if you
connected the logic signals directly to the transistors at the anodes,
then there would be 180V passing through the transistor's base into the
logic circuits. This would destroy the logic circuits immediately, as
they can only handle about 5V.
The MPSA42 transistor is needed to provide a path for each anode control
signal to be translated from the 0V level to the 180V level that the
MPSA92 anode transistor bases require.
--
David Forbes, Tucson AZ
http://www.cathodecorner.com/
will
transistors and you are multiplexing, remember to clamp the cathodes
to a higher-than-ground voltage. I've had grief with ghosting, and
after all that soldering I may just use integrated driver chips.
chuck richards
side driven by a 4N35 optocoupler just for the sake of
doing the experiment. The 4N35 then isolates the logic from
"the high side".
There's a little experiment of mine that uses that technique
in order to operate the bi-quinary tubes. The "1" bit from
the 74141 controls two 4N35 optocouplers which select the
tube's front screen or back plate as the active anode.
Then, the bits 2,4,8 operate normally on the cathodes.
It's just an experiment to do, to show one method to isolate
from the +175 volts. I know there are other better methods.
Chuck
>>--
>>You received this message because you are subscribed to the Google
>Groups "neonixie-l" group.
>>To post to this group, send an email to neoni...@googlegroups.com.
>>To unsubscribe from this group, send email to
>neonixie-l+...@googlegroups.com.
>>For more options, visit this group at http://groups.google.com/group
>/neonixie-l?hl=en-GB.
>>
>>
$4.95/mo. National Dialup, Anti-Spam, Anti-Virus, 5mb personal web space. 5x faster dialup for only $9.95/mo. No contracts, No fees, No Kidding! See http://www.All2Easy.net for more details!
ggtnuminix
Chuck - I noted in Jason Harpers article he also references the use of
a high-voltage optocoupler (TLP627) on the anode side.
threeneurons
| the cathode side but not the anode side.
|
| Any further insight/explanation would be appreciated.
Put up a drawing, so you can reference what you don't get.
The most commonly used nixie driving, uses 'common emitter' stages on
both sides. Here's a drawing of the classically used ckt for driving a
mux'd nixie:
http://cid-f9db37b8211ce831.office.live.com/self.aspx/Snippets/Nixie_Muxd.gif
The Jason Harper anode drive Muntz's the circuit by putting a resistor
in the emitter leg (of Q2 of my ckt). This does double duty of
limiting both collector and base current, but you need better
knowledge of how transistors work to really understand th 'magic'.
Don't get rid of the 'bleeder' resistor !!!! Its there for speed.
If you think you can shortcut this circuit, by omitting Q3 (MPSA92),
and using Q2 as an 'emitter follower', it won't work. Think about it.
If its an emitter follower, with the collector (of Q2) to +180, and
the anode resistor to the emitter, you get squat. THINK ! THINK !
THINK ! The anode needs to get to 180V. That means the emitter needs
to be at 180V. Where's the Base ? It ain't going above 5V. Last I
checked 5V is less than 180V. In reality, the emitter won't go higher
than the base, which is only 5V. For a mental exercise, if the emitter
did get up to 180V, the BE junction would be in reverse bias, since
the base is less than the emitter, hence it would clearly be OFF. No
current flow.
threeneurons
http://cid-f9db37b8211ce831.office.live.com/self.aspx/Snippets/Nixie_...
using Q2 as an 'emitter follower', it won't work.
happens.
Adam Field
> cathode side but not the anode side.
>
> Any further insight/explanation would be appreciated.
This is a method I used for multiplexing, works well at ~2kHz. It
provides isolation and doesn't violate the transistor's Vbe rating (5V
on the Fairchild MPSA92) when driven from 3.3 or 5V logic.
threeneurons
> provides isolation and doesn't violate the transistor's Vbe
> rating (5V on the Fairchild MPSA92) when driven from 3.3 or 5V
> logic.
>
> 13KViewDownload
Yeah, that'll work. Downside is that the pulse duration is short, so
you have to compensate by increasing the peak current. With a pulse
that short, it'll will still be on the dim side, though that may not
matter in a low lit room (typical living room, den ...). You can up
the capacitance and the series base resistor a bit to increase your
time constant, allowing for greater pulse width. A small electrolytic
(1uf 250V) would do nicely (per anode).