Hi all, first post here!
I'm building an instrument tube buffer/DI, and the last stage is a cathode follower, so I need the heaters elevated. I'm using a 12 VDC wall wart SMPS power supply to power the whole unit, with a nixie style power supply inside the box, used for boosting the DC to get the B+ voltage.
When I try to elevate the heater voltage using a voltage divider from B+, the heaters get no voltage and no current. The 12 VDC input is referenced to ground at the input for the nixie power supply ground reference, so I'm guessing there lies the problem.
How could I get the heater elevation I need without losing the reference to 0 volts/ground that I need for the nixie power supply? Thanks in advance
I'm building an instrument tube buffer/DI, and the last stage is a cathode follower, so I need the heaters elevated. I'm using a 12 VDC wall wart SMPS power supply to power the whole unit, with a nixie style power supply inside the box, used for boosting the DC to get the B+ voltage.
When I try to elevate the heater voltage using a voltage divider from B+, the heaters get no voltage and no current. The 12 VDC input is referenced to ground at the input for the nixie power supply ground reference, so I'm guessing there lies the problem.
How could I get the heater elevation I need without losing the reference to 0 volts/ground that I need for the nixie power supply? Thanks in advance
You may want to look/move this to power supplies.
You will need to pay attention to a number of things:
* isolated - both in topology and pcb layout gapping
* rating of components - elevation isn’t a one sided effect, the components will need to be rated to cover the difference.
* current sharing - preventing damage by ensuring each smps knows it provide a % of the current required.
It unlikely that a basic commercial smps will have been designed with that in mind. If you look at the meanwell MPS series you’ll not the isolation but also it can be paralleled as it has current sharing connections. It can’t be connected to elevate without revising the components.
Have a look at the 1KW smps thread in power supplies.
You will need to pay attention to a number of things:
* isolated - both in topology and pcb layout gapping
* rating of components - elevation isn’t a one sided effect, the components will need to be rated to cover the difference.
* current sharing - preventing damage by ensuring each smps knows it provide a % of the current required.
It unlikely that a basic commercial smps will have been designed with that in mind. If you look at the meanwell MPS series you’ll not the isolation but also it can be paralleled as it has current sharing connections. It can’t be connected to elevate without revising the components.
Have a look at the 1KW smps thread in power supplies.
Can you confirm the nixie supply has its input 0V inherently connected to its 0V output (and not by some externally configured connection).
There are other small cheap isolated pcb-based converters on ebay that are 12V powered and can have their output configured for valve B+ applications. The 12V could then be connected to the heaters at an elevated voltage (relative to the B+ supply) that suits all the valves. But that then requires the 12V plugpack to have a floating 12V, and not be connected back to a mains supply earth pin.
As with mains hum loops, even smps allow high frequency ripple to couple around loops and can be noticeable.
There are other small cheap isolated pcb-based converters on ebay that are 12V powered and can have their output configured for valve B+ applications. The 12V could then be connected to the heaters at an elevated voltage (relative to the B+ supply) that suits all the valves. But that then requires the 12V plugpack to have a floating 12V, and not be connected back to a mains supply earth pin.
As with mains hum loops, even smps allow high frequency ripple to couple around loops and can be noticeable.
Last edited:
Thank you both for your replies.
The wall wart I use is isolated, the nixie kit I built isn't. It shares it's 0 V ground with the 0 V input. As for rating and current capabilities, the wall wart is rated for 2 amps and the nixie can supply 15 ma at 220V. I'm using a couple of 12a-7's, and the voltage divider I made only took around 1 ma out of the nixie, so both supplies should be well within limits.
I will have a look at the power supply forum when I have the time. A quick search revealed several threads pertaining to 1KW smps, so I have homework to do.
Yes, the 0 V input is connected to the HV 0 V output.
This is the kit I'm using: Tube-Town Store - TT VPump - Voltage Converter
They don't show the schematic anywhere, only the layout, but upon close inspection of the PCB, there is a visible trace that connects both input and output 0 V. I take from your reply that I need the nixie to be isolated from the input 0V to be able to do the heater elevation, right?
I will do a search on Ebay and Amazon to see what I find.
The wall wart I use is isolated, the nixie kit I built isn't. It shares it's 0 V ground with the 0 V input. As for rating and current capabilities, the wall wart is rated for 2 amps and the nixie can supply 15 ma at 220V. I'm using a couple of 12a-7's, and the voltage divider I made only took around 1 ma out of the nixie, so both supplies should be well within limits.
I will have a look at the power supply forum when I have the time. A quick search revealed several threads pertaining to 1KW smps, so I have homework to do.
trobbins said:
Yes, the 0 V input is connected to the HV 0 V output.
This is the kit I'm using: Tube-Town Store - TT VPump - Voltage Converter
They don't show the schematic anywhere, only the layout, but upon close inspection of the PCB, there is a visible trace that connects both input and output 0 V. I take from your reply that I need the nixie to be isolated from the input 0V to be able to do the heater elevation, right?
I will do a search on Ebay and Amazon to see what I find.
If you have a B+ at 220V, then the cathode follower cathode may not get to circa 180V, and only on a large signal swing, depending on the 12A?7. Given you are using 12Vdc for heater, and you only have one triode as a cathode follower, then make that triode the 6V-to-12V heater triode to also crib 6V about ground.
trobbins said:
The reason I didn't specify the 12a?7 type is that depending on whether I can elevate the heaters or not, I'd be more inclined to use a 12au7 or a 12at7 for the final circuit. I'd like to use the later, as the examples I have lying around are less noisy and have better tonal qualities than the 12au7's I have. The 12at7 is good for only 90 VHK though. Right now I'm running B+ at around 130 V with a 12at7, which is low but better safe than sorry I guess. amping the B+ to 200+ volts raises the noise floor too, whether I'm using a 12au7 or a 12at7. That was another reason for my desire to elevate, to maybe reduce the noise.
Thanks for the heaters wiring scheme idea, I didn't think of that! I just checked the unit and I actually wired it that way in the first place, without giving it any thought.
Although the 12AT7 has a nominal 90V Vhk limit, the 12AT7WA is service life tested at 100V (continuous) and also at 135V (cycling) - that may indicate that the heater cathode interface of a 12AT7 can normally sustain at least those levels for a reasonable service life, and possibly at higher levels for a reasonable period of time. If there was a brief breakdown of Vhk interface, it may cause no collateral damage.