IV-17 sockets WAS Re: New file uploaded to smartsockets
- Hi Chris,
A solution for the decimal point drivers could be using a UDN2981 or UDN 2982/2 instead of a bunch of transistors and resistors.
These are octal high-side drivers specified up to 50V and available in SMD.
UDN 2983/4 can even switch uo to 80V.
They can drive the segments (or in this case the decimal points) directly from the PIC since they are CMOS and TTL compatible.
With one of these ICs even all of the double decimal points of a quad IV-17 display could be driven.
I have used this IC for a VFD-Clock with 6 of the (7-Segment) IV-11, which are quite similar to the IV-17 in size and.
I too had no issues with uneven brightness using DC-voltage for the filaments.
In my IV-11 clock I have made a small SMPS with the popular MC34063 to get 4,5 V.
Then I set the filaments 3V above ground with a 3V power Z-diode in series to ensure full cutoff when the grid or the segments go to ground (which the filament then sees as -3V).
(IV-11s only have 1,5V filament voltage.)
Your first idea for the HV-Supply using a fet and a few passive components was perhaps similar to the solution by adafruit for their IV-18 clock.
With such an approach the voltage could be influenced with the duty-cyle of a PWM-output of the controller. So one could have another "feature" for dimming the VFDs.
This is of course not very important for a "showcase"-display, but if You have it in Your bedroom...
I have fitted nearly all of my nixie- and VFD-clocks with an LDR-regulated automatic dimming to extend the lifetime of the displays and for "bedroom-usage".
--- In email@example.com, "fixitsan2" <fixitsan@...> wrote:
> The current position is that the sockets are up and running. I have only made one socket so far and will need to assemble another on breadboard to make sure the communications and enumeration work as expected.
> The IV-17 socket hardware consists of a pic18f25k22, an HV5812 (or equivalent) and a couple of pnp/npn dirver pairs for the decimal points. There is an optional 32kHz crystal to produce a backup clock function which so far has seemed to be accurate to better than 1 second per day.
> The filaments are each connected to the 5V supply via a 56 Ohm resistor. Slightly wasteful, but it works well. The large differential between low filament voltage and high Vb+ (45V) does not produce any brightness gradient at all.
> The big hold up has been with me getting the Vb+ supply working off the pic. I've used an external FET, coil, 2 capacitors and a diode to get a reliable 40V output. But just recently i have found a company which can sell me supplies based on a popular switcher IC and with just a couple of component substitutions I have a reliable small low consumption 40V ready to plug into the main PCB. I have ordered 25 and these should be arriving soon.
> Then the last thing will be to draw the PCB. Now that I have included an output pin remapping function that should be pretty straightforward. I think I will have to look at using SMD components because I would really like to have all the electronics fit underneath the footprint of 4 tubes set out side by side.
This email message is a notification to let you know that
a file has been uploaded to the Files area of the smartsockets
File : /B-5971/README.TXT
Uploaded by : cyberb0 <radmeck@...>
Description : IF YOU DARE! - 04.21.2014
You can access this file at the URL:
To learn more about file sharing for your group, please visit: