Multiplexing nixies in a tube preamplifier

[Tim]

I am in the process of building an audio preamp using valves and I am using nixie tubes to indicate the volume control position.  I am using two nixie tubes to indicate the volume and they are being multiplexed via a PIC and 74141. 

My question to you fine folk is how should I provide power to the nixie tubes.  I have two options available to me:

1) Use the +200V rail being provided via a shunt regulator off the linear regulated +250V rail used for the valve HT.  (The +200V is used for providing power to the screen within a magic eye tube).

2) Use a separate SMPS HT source just for the nixie tubes (it is one of the tayloredge designed ones).

I have both options available to me, number one being more simple as the HT is already there, however I am concerned that the switching noise from the multiplexing will be heard on the audio output due to the multiplex switching noise propagating onto the HT supply (am I being overly paranoid, or will this not be seen due to the more than adequate capacitance being used on the HT rail).

Of course there is option 2, but I then have the concern of the SMPS radiating noise into the audio path.

Both options have there pros and cons and I think I may well resort to trying both and seeing what happens but I would love to hear your ideas and suggestions to which you think should be implemented.

Regards,
Tim

[David Forbes]

On 1/2/13 3:01 PM, Tim wrote:
> I am in the process of building an audio preamp using valves and I am
> using nixie tubes to indicate the volume control position. I am using
> two nixie tubes to indicate the volume and they are being multiplexed
> via a PIC and 74141.
>
> My question to you fine folk is how should I provide power to the nixie
> tubes. I have two options available to me:
>
>

> Regards,
> Tim

Tim,

If you can afford the extra four pins on the PIC and another 74141, then
you can run the tubes direct (non-multiplexed) and not worry about it.

If you will be multiplexing them, then the power supply will not conduct
the noise to any noticeable extent. You will need a high anode resistor
value, so you can insert a simple two-stage RC low-pass filter (10K
series, 0.1uf polyester shunt) between the power supply and the anode
resistors to eliminate noise from the anode power.

You will want to be careful about mounting the display unit away from
the input stage of the amplifier, to prevent radiated noise from getting
into the input stage where it will be amplified.

--
David Forbes, Tucson AZ

[Tim]

Hi,

Thank you for your reply David.  My reason for multiplexing was due to not having enough spare pins on the PIC, however having givin this considerable thought I have the IO pins that were being used to switch the MSD and LSD anode transistors and a third IO pin which was available to control the enable pin for the tayloredge SMPS, so by using these three pins allows me to to have 0 - 7 on the MSD which is great as I wanted 64 steps to indicate  the volume control potentiometer position. 

Sadly there is a but to this (is there not always) I need to be able to blank the digits which normally involves using all 4 bits of the 74141.  Now as I do not need 7 - 9 I was thinking I could shift all the digits along one position, so 0 in the nixie is connected to 1 on the 74141 1 to 2 and so on.  This results in being able to blank the nixie by sending 000 to ABC and having D permanently tied to ground.  This digit shift being easy to work around in the firmware.

I can position the SMPS in such a way that it will not interfere with the audio signal path so I would like to try and stick with it for the nixie HT supply as the valve HT supply has a delayed start to give chance for the soft started heaters to warm up. I want the volume indication to come on at power up with a possible count down on it to show the remaining warm up time. Additionally to this I am trying to keep the analogue and digital power rails separate from one another.

This is where my questions begin:  Can I leave the cathodes 7 - 9 floating or do they need to be tied to something? Will floating digits ghost? I would also like to keep my enable signal to disable the SMPS when the amplifier is in standby so can I use the now unused 0 output on the 74141 as a logic signal to drive the enable pin on the SMPS?  The only trouble with this is when blanking the nixie the SMPS will be disabled unless I use extra logic to look at the LSD bits?  (I guess it is quite environmentally friendly to turn the SMPS off when the digits are blank but if I was worried about this then I guess I would not be building a power hungry inefficient valve amplifier!)

[Michel van der Meij]

Hi Tim,

Not that I want to influence your design but it sounds a bit odd to have the nixies go from 0 to 63 in a valve (pre)amplifier. It is really related to something digital that IMHO doesn't really match a design of a valve amplifier. I think it is nicer if it would go from 0 to 99. If you use a few latches before the 74141 that would solve all your limited I/O problems. Another thing that comes to mind is using an IN-13 bar graph tube to indicate the position of the volume button.

Michel

on Jan 05, 2013, Tim <halfpin...@hotmail.com> wrote:

Hi,

Thank you for your reply David.  My reason for multiplexing was due to not having enough spare pins on the PIC, however having givin this considerable thought I have the IO pins that were being used to switch the MSD and LSD anode transistors and a third IO pin which was available to control the enable pin for the tayloredge SMPS, so by using these three pins allows me to to have 0 - 7 on the MSD which is great as I wanted 64 steps to indicate  the volume control potentiometer position. 

Sadly there is a but to this (is there not always) I need to be able to blank the digits which normally involves using all 4 bits of the 74141.  Now as I do not need 7 - 9 I was thinking I could shift all the digits along one position, so 0 in the nixie is connected to 1 on the 74141 1 to 2 and so on.  This results in being able to blank the nixie by sending 000 to ABC and having D permanently tied to ground.  This digit shift being easy to work around in the firmware.

I can position the SMPS in such a way that it will not interfere with the audio signal path so I would like to try and stick with it for the nixie HT supply as the valve HT supply has a delayed start to give chance for the soft started heaters to warm up. I want the volume indication to come on at power up with a possible count down on it to show the remaining warm up time. Additionally to this I am trying to keep the analogue and digital power rails separate from one another.

This is where my questions begin:  Can I leave the cathodes 7 - 9
floating or do they need to be tied to something? Will floating digits ghost? I would also like to
keep my enable signal to disable the SMPS when the amplifier is in standby so can I use the now unused 0 output on the 74141 as a logic
signal to drive the enable pin on the SMPS?  The only trouble with this is when blanking the nixie the SMPS will be disabled unless I use extra logic to look at the LSD bits?  (I guess it is quite environmentally friendly to turn the SMPS off when the digits are blank but if I was worried about this then I guess I would not be building a power hungry inefficient valve amplifier!)

Regards,
Tim

On Wednesday, 2 January 2013 22:25:13 UTC, nixiebunny wrote:

On 1/2/13 3:01 PM, Tim wrote:

> I am in the process of building an audio preamp using valves and I am

> using nixie tubes to indicate the volume control position.  I am using

> two nixie tubes to indicate the volume and they are being multiplexed

> via a PIC and 74141.

>

> My question to you fine folk is how should I provide power to the nixie

> tubes.  I have two options available to me:

>

>

> Regards,

> Tim



Tim,



If you can afford the extra four pins on the PIC and another 74141, then

you can run the tubes direct (non-multiplexed) and not worry about it.



If you will be multiplexing them, then the power supply will not conduct

the noise to any noticeable extent. You will need a high anode  resistor

value, so you can insert a simple two-stage RC low-pass filter (10K

series, 0.1uf polyester shunt) between the power supply and the anode

resistors to eliminate noise from the anode power.



You will want to be careful about mounting the display unit away from

the input stage of the amplifier, to prevent radiated noise from getting

into the input stage where it will be amplified.



--

David Forbes, Tucson AZ

--

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.

To view this discussion on the web, visit https://groups.google.com/d/msg/neonixie-l/-/wzW1euMoJw0J.

For more options, visit https://groups.google.com/groups/opt_out.

 

 

[Tim]

Hi Michel,

This is a fare point, 0 - 99 would be allot nicer, I wonder if I should just bite the bullet and use an IO port expander IC on the nixie display PCB I dont think I have enough room spare on the PCB to implement the latches unless I can do it with one IC (do you have any that come to mind?).  This would solve all my problems in one go if I can find/make the space for it.  I will have a look to see if I can squeeze a SM package onto the PCB in the remaining limited space. 

I do like the IN-13 but am sadly too far into the design to make use of one.  I do however have a second project on the go which is an audio spectrum analyser using IN-9 's :o)  I just wish I had more spare time on my hands to spend more time on my many projects.

Regards,
Tim

[Per Jensen]

On 05/01/2013, at 00.03, Tim wrote:

>
> This is a fare point, 0 - 99 would be allot nicer, I wonder if I should just bite the bullet and use an IO port expander IC on the nixie display PCB I dont think I have enough room spare on the PCB to implement the latches unless I can do it with one IC (do you have any that come to mind?). This would solve all my problems in one go if I can find/make the space for it. I will have a look to see if I can squeeze a SM package onto the PCB in the remaining limited space.
>
> I do like the IN-13 but am sadly too far into the design to make use of one. I do however have a second project on the go which is an audio spectrum analyser using IN-9 's :o) I just wish I had more spare time on my hands to spend more time on my many projects.

Use a PCB5874 - I2C port expander with 8 bits, use that to control the 2 74141's.
You CANNOT blank the nixies with the 74141's you need to use anode transistor or disable the PSU.

The PCF8574 sits on the same bus as your RTC - Easy!

// Per.

[JohnK]

I would have thought that with 10 being the maximum loudness then having one that goes to 11 for extra loudness would be marvellous !

 

For anyone who hasn't heard the joke....

http://en.wikipedia.org/wiki/Up_to_eleven

 

John K.

[Jeff Thomas]

I would have thought that with 10 being the maximum loudness then having one that goes to 11 for extra loudness would be marvellous !

 

For anyone who hasn't heard the joke....

http://en.wikipedia.org/wiki/Up_to_eleven

 

John K.

Excellent point John. Audiophools will eat that up. "Dude, it goes to ELEVEN"


Regards, Jeff

 

[David Forbes]

Given this extra information, your multiplexed design makes sense. As long as you keep the Nixies physically distant from the amp input signals and tubes, and give yourself the option of steel shielding if needed, you should be OK with the Tayloredge supply and multiplexing.

David Forbes

http://www.cathodecorner.com/

On Jan 4, 2013, at 3:20 PM, Tim <halfpin...@hotmail.com> wrote:

Hi,

Thank you for your reply David.  My reason for multiplexing was due to not having enough spare pins on the PIC, however having givin this considerable thought I have the IO pins that were being used to switch the MSD and LSD anode transistors and a third IO pin which was available to control the enable pin for the tayloredge SMPS, so by using these three pins allows me to to have 0 - 7 on the MSD which is great as I wanted 64 steps to indicate  the volume control potentiometer position. 

Sadly there is a but to this (is there not always) I need to be able to blank the digits which normally involves using all 4 bits of the 74141.  Now as I do not need 7 - 9 I was thinking I could shift all the digits along one position, so 0 in the nixie is connected to 1 on the 74141 1 to 2 and so on.  This results in being able to blank the nixie by sending 000 to ABC and having D permanently tied to ground.  This digit shift being easy to work around in the firmware.

I can position the SMPS in such a way that it will not interfere with the audio signal path so I would like to try and stick with it for the nixie HT supply as the valve HT supply has a delayed start to give chance for the soft started heaters to warm up. I want the volume indication to come on at power up with a possible count down on it to show the remaining warm up time. Additionally to this I am trying to keep the analogue and digital power rails separate from one another.

This is where my questions begin:  Can I leave the cathodes 7 - 9 floating or do they need to be tied to something? Will floating digits ghost? I would also like to keep my enable signal to disable the SMPS when the amplifier is in standby so can I use the now unused 0 output on the 74141 as a logic signal to drive the enable pin on the SMPS?  The only trouble with this is when blanking the nixie the SMPS will be disabled unless I use extra logic to look at the LSD bits?  (I guess it is quite environmentally friendly to turn the SMPS off when the digits are blank but if I was worried about this then I guess I would not be building a power hungry inefficient valve amplifier!)

Regards,
Tim



On Wednesday, 2 January 2013 22:25:13 UTC, nixiebunny wrote:

[Jon]

On Friday, January 4, 2013 11:06:35 PM UTC, zapro wrote:

>You CANNOT blank the nixies with the 74141's you need to use anode transistor or disable the PSU.

 

Actually it depends on the nixie and the '74141'. My experience is that IN-12 blank just fine using the K155ID1 Russian version of 74141. But I pick up from comments here and elsewhere that other nixies such as ZM1000 seem to be less forgiving. Compatibility depends on the leak current that the chip output transistors permit in the off-state (which will be influenced by the anode voltage you choose) and the sensitivity of the nixie.

 

Jon.

[John Rehwinkel]

> Given this extra information, your multiplexed design makes sense. As long as you keep the Nixies physically distant from the amp input signals and tubes, and give yourself the option of steel shielding if needed, you should be OK with the Tayloredge supply and multiplexing.

One of my favourite tricks for this is to use optocouplers for anode switches - they switch slowly (radiating fewer harmonics), and the level shifting is done for you by a light beam.

- John

[Tim]

That is exactly what I have done with the design I am working on.  It makes allot of difference.

Tim

[Alex]

Quick question,

Why would you need a soft / slow rise time on the anode cut off to reduce switching noise when the cathodes are toggled repeatedly by the 74141? Surely that introduces just the same switching noise that would be produced with just a high side switching FET or BJT? Or does the 74141 feature a slow transition on its output? I can appreciate the optical isolation for the high side switching as being a nice part of the opto-couple solution though!

- Alex

[John Rehwinkel]

> Why would you need a soft / slow rise time on the anode cut off to reduce switching noise when the cathodes are toggled repeatedly by the 74141? Surely that introduces just the same switching noise that would be produced with just a high side switching FET or BJT? Or does the 74141 feature a slow transition on its output? I can appreciate the optical isolation for the high side switching as being a nice part of the opto-couple solution though!

It's the sequence of events that makes it work. The cathodes are switched (hard) by the 74141, but no current flows (and noise is not radiated) as the anode isn't powered yet. Then the anode is turned on (soft) by the optocoupler, and the current rises gently, producing minimal radiated noise. When it's time to switch the digit off, the same thing happens in reverse, the anode is turned off (soft) by the optocoupler, smoothly reducing the current, then the cathodes are again switched while the current is off, thereby not producing switching noise.

- John

[Alex]

Ahh, of course. The whole topic being on multiplexing should really of been the clue there. silly me...

[Nick]

I've built more than a few amps, some with nixies (valve amps) and several with uP control.

Its not just the SMPS & potential multiplex noise - you also have the CPU clock and its contribution (just love all those square waves) - everything from the main clock to serial ports & I2C etc.

Of course its doable, but keep the analogue & digital parts of the system well apart - use a star ground and ensure no earth return or power supply is shared between the A & D parts.

Cheers