Opto Coupler for PWM control - Sanity Check

[Richard Scales]

I have some 7 segment Panaplex displays (SP-101) which I have a typical Cathode Current of 700uA for each of the large segments (7 per display).

If I wanted to deliver the 170V HV supply via PWM so that I could vary the brightness / duty cycle then would I be at liberty to drive all 6 displays via a single EL-817 (or equivalent) Opto Isolator?

My maths suggests that with all seven segments of all six displays illuminated that's 7 x 6 x 700uA = 42 x 0.7mA = 29.4mA.

From the specification of an EL-817 I note that the Collector current is rated at 50mA.

Does that suggest that I could be OK displaying 888888 and using a single EL-817 I could dim the display?

I should point out that my goal is not to actually dim the display at all really, I just wanted to reduce the duty cycle as far as I reasonably could in order to promote the life of the displays.

I noted that when I ran them continuously they actually got slightly warm - which I assumed was not ideal.

I am currently (!) using 68K Anode resistors with an HV supply of 170V. The displays have an Anode to Cathode voltage drop of 135V so I would expect (though have not measured) to cathode current to be something like:

=(170-135)/68K 

=35/68K

=514 microAmps

Does this make any sense?

... and if so, would driving the Anode HV via an EL-817 be a good plan?

Thank you for listening

Richard

[gregebert]

Fundamentally, it should work, but be aware you are now pulsing the displays. Some folks in this group have reported audible noise from multiplexing. Some panaplex displays require multiplexing by the nature of their pinout (common cathodes, separate anodes) so that would be a moot issue. Instead, you would adjust the duty-cycle of your multiplexing scheme.

Another option is to reduce the anode voltage under software control; how that is done depends upon the type of supply you have. If you are using a DCDC converter, you would change the resistance value in the feedback loop.

I wouldn't worry too much about the generated heat as long as your case allows the heat to escape via natural convection; you dont need a fan. Your display will consume 4W peak, and for a clock it's probably around 2W average. Given the large surface area, and black color, it's going to radiate heat efficiently which minimizes the temperature increase. I've checked nixie tubes with a handheld IR thermometer, and the measured temperature is on-par with what I would guess by touching it. I dont think I've ever measured above 40 C, and that's well-within the datasheet spec of 70C ambient temp.

[Richard Scales]

That's really helpful. Just to clarify. I don't actually want to dim the displays, in was really just trying to stop them from heating up. I'll check the specifications for temperature range. It is very easy for me to adjust the HV supply downwards which would then reduce the segment current. I shall experiment and report back Richard

--
You received this message because you are subscribed to a topic in the Google Groups "neonixie-l" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/neonixie-l/0f9uCa5EgsU/unsubscribe.
To unsubscribe from this group and all its topics, send an email to neonixie-l+...@googlegroups.com.
To view this discussion on the web, visit https://groups.google.com/d/msgid/neonixie-l/6150508f-c0fb-42ba-9c66-a7c864bcd384%40googlegroups.com.

[jf...@my-deja.com]

On Friday, February 7, 2020 at 8:59:19 AM UTC-8, gregebert wrote:

...Some folks in this group have reported audible noise from multiplexing....

I am one of the people who reported singing in my multiplexed B7971s.  More than 40 years ago, I also had a Panaplex/MM5314 clock that was sold as an evaluation kit.  It has been a long time, but I do not remember it producing acoustic noise.  Maybe the size of the digits makes a difference here.

[Instrument Resources of America]

The size of the digits would certainly have an effect on the particular resonant frequency.  I would think the larger, and therefore more massive digits, would have a lower resonant frequency.    Ira.

--
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+...@googlegroups.com.
To view this discussion on the web, visit https://groups.google.com/d/msgid/neonixie-l/9c95d902-f3e3-4a5a-915e-6cf6ac7997f8%40googlegroups.com.

[Mac Doktor]

On Feb 8, 2020, at 10:59 AM, Instrument Resources of America <IRACO...@HUGHES.NET> wrote:

I would think the larger, and therefore more massive digits, would have a lower resonant frequency.    Ira.

And then there's the age of one's ears.

Terry Bowman, KA4HJH
"The Mac Doctor"

"I've seen things you people wouldn't believe: attack ships on fire off the shoulder of Orion... beams...in the dark in the Tannhauser Gate. All those moments will be lost in time...like tears in the rain." — Roy Batty, Blade Runner

[Richard Scales]

I like the comment about the age of one's ears though when I sit at my desk I can hear B-7971's from about 8 ft away as I use a cross fade effect via smart-sockets - it is most clearly audible.

The same effect however on ZM1350's and SP-101's is not audible (to me!).

Regardless of all the above, I moved a stage closer to testing the code now that the display hardware is complete and in reality, as i am using PIR activation, The display is rarely actually on so my concerns regarding any rise in temperature affecting tube life would appear to be irrelevant!

Right now i am using a Teensy 3.2 to do all the display processing, I was using Arduino nano but I was running out of 'headroom' fairly quickly so I moved to Teensy 3.2 which is barely ticking over (I have yet to try Blue/Black Pill micros). It uses a WeMos D1 for NTP sync as well as random word generation, temp, pressure and local wind speed and direction reading. This all gets sent to the Teensy via a heavily modified 'GPS' style string.

Details will be forthcoming when i have finalised things a little more.

Richard