I agree 100kHz seems high, but I don't see any downside if you can drive that OK. I also see no need for a filter : the only effect I can imagine is either too low (certainly not at 100k) or possibly a beat frequency with the multiplexing which i'd avoid carefully (or zero beat i.e. in sync, but with some thought about average brightness).
Interested to hear the result as the usual approach is a floating transformer inverter but I'd considered the h-bridge and intend to try it. I'm not convinced about the dc bias from your simulation though : i'd expect the dc to be simply half the drive waveforms (since they're positive only). I'd also expect to be able to change the DC level by adjusting the h-bridge drive proportions but it would have to be symmetrical, which requires a dead period. On Sat, Apr 15, 2023 at 9:24 AM ZY <[email protected]> wrote: > I'm playing around a bit more with some active filter designs so it won't > be load dependent, but 100khz seems a bit high for this application. Going > a lower frequency would require I select a different/larger transformer > since my current would be too high. > On Saturday, 15 April 2023 at 03:26:46 UTC-4 ZY wrote: > >> I wanted to try to avoid having a AC line in my design if possible. I did >> have to break out my old ECE notes, and I designed a different filter. >> Turns out my original RC filter requires a much larger resistor to limit >> the primary side current, but that raises input resistance too high for my >> 40 ohm loads. >> >> Using a different filter I had some more success: >> >> https://www.falstad.com/circuit/circuitjs.html?ctz=CQAgjCDsB0AcYCYCsBTAtLEAGaBOWuYALFrgGwDMFRSSkSZIuuISrWr6YYAUAG7hE4Mh24IQRTB3FgscudjiKpipP0HiKZTGJDJG08PIU4pSitGJgkuBGSQUkBSLgqqeAcw3DRQsLexsT28tHSESDkieAHdvSXAA+KweACdWBB0M9J0AjhIeMDIWBBJwLIiylRx5WWY6+oaWMGhSAixato6zfzIiESJuJCxIWoRIEDQcYiwKSHbIMgQqWDkRiamSWfnF5bkqRWYYiQV-cT6iBPFkgSIKcRK8u70sKuPwQMjoNViKwseZERBH5PP4SJ4IF5BLy3e6lGF6SBXIESXqXbJo5IAYxRF0QmHOGL0lghxlJ8gQ6HG1RGuBGiGYFFsGUIyKQWVOrAWGKOdEYePR-OSsV5lRAIkyUiOHPiDwkktiun0elKSqFyukDHV3OFXI54u1Yshes1HLV+v59j5WTVsplpUFPAANjjRWyJR8WsZMJMWgwiCMEORbrBqBQhMlnb9yqjBYpjCVFNMtmAFksKCssJQgmkRfzc7ljqlDe7zQX8mlLaLK6bCwB7QIQURIay0ViKC6GDgUHj1tiN56SNwIL5xvt6Nvd3vvcZgRwpthsKbjxstSDwZA8IA >> >> However the problem is the output voltage is somewhat determined by the >> load. So if a filament burns out it can cause a cascade failure on the >> other filaments as the voltage rises. I guess monitoring the AC line would >> be possible but this complicates my design a bit, and I'd need a monitor >> that could do negative currents/voltages if my bias is not high enough. If >> such an IC exists I'd be interested, but I don't really have access to a >> FPGA if that's what it takes. >> >> It also seems it's easy to hit resonance with particular filter LC >> values, boosting the voltage higher which feels a bit dangerous for the >> filaments. >> >> Maybe just not having the filter would be the safest approach here. >> However I'm not at all experienced in filter design especially when there >> is a low impedance load involved, so maybe there's a better approach that >> I'm not using. >> >> On Saturday, 15 April 2023 at 02:03:39 UTC-4 gregebert wrote: >> >>> I use a mains-powered 2.5VAC center-tapped transformer for my NIMO >>> clock, and apply a DC bias to the center-tap. Basically the same as your >>> approach, but not running at 100kHz. The DC bias comes from a DAC (with an >>> OP-amp buffer), so I can use software to control the offset, rather than a >>> pot. No need for LC filters, etc. >>> >>> The transformer is rated at 3amps, even though 6 NIMO tubes only require >>> 1.2amps max. Also, each tube has it's own fuse and series resistor to >>> reduce the inrush current. The software-controlled DAC is also necessary so >>> I can have the FPGA monitor analog signals to check the fuses, filaments, >>> and resistors, and that has to be timed correctly to the AC line and the >>> offset has to be different for making these measurements than for regular >>> operation. It was probably overkill, but I enjoyed the design challenge. >>> >>> On Friday, April 14, 2023 at 2:25:57 PM UTC-7 ZY wrote: >>> >>>> Hello. I'm working on a driver for my VFD filaments. Say I need a >>>> Vfilament of 1.6V, and a current of 40mA, such as with a small VFD. >>>> >>>> So far I have a square wave driver set up, basically for my two >>>> filament pins I can alternate the two pins between 1.6 and 0 with some H >>>> bridge at maybe 100KHz. If I connect the filament load to this, it's the >>>> equivalent of using a 3.2V Vpp square wave biased at 0v. See Sim Link 1. I >>>> put two 40 ohm resistors as a stand in for 2 VFD tubes filaments in >>>> parallel. >>>> >>>> However, I'd like to add some sort of smoothing and DC bias. I was >>>> thinking of using a small transformer, such as this one: >>>> https://www.mouser.com/datasheet/2/445/750315240-3096542.pdf >>>> Applying a 100KHz RC low pass filter to the output, and approximating >>>> the turn ratio at 1:1 (it's 1:1.1 in the datasheet but ignoring that for >>>> now), I have a simulation below as Sim Link 2. >>>> >>>> The problem here is I seem to be wasting a huge amount of power just >>>> driving the transformer on the primary size. The plot on the very far left >>>> shows I'm delivering 2.2A to the transformer. I guess I could find a larger >>>> transformer with a higher inductance, or increase my switching frequency, >>>> but I'm not sure if I'm on the right design path here. >>>> >>>> Also, it seems like some of my issues stem from my LC filter. If I >>>> remove that, the simulation shows more reasonable primary currents of >>>> 100mA. See Sim Link 2. I don't really remember my ECE courses from a long >>>> time ago, so I'm not sure what is going on here. >>>> >>>> Anyone have any suggestions? >>>> >>>> >>>> Sim Link 1: >>>> >>>> https://www.falstad.com/circuit/circuitjs.html?ctz=CQAgjCDsB0AcYCYCsBTAtLEAGaBOWuYALFrgGwDMFRSSkSZIuuISrWr6YYAUAG7hE4Mh24IQRTB3FgscudjiKpipP0HiKZTGJDJG08PIU4pSitGJgkuBGSQUkBSLgqqeAcw3DRQsLexsT28tHSESDkieAHdvSXAA+KweACdWBB0M9J0AjhIYiQV-cSIyIgTxZNjS8sRMJCy6oOqyiuy25LS6RibujsKChpzxPuLm1kgerKHwLKq2+IQSCSkC3X09ZY35pekGTdFcgt3ZzAjT8fOxk7H5q6yai53lxeWm5IB7QIhRJGtaViKcqGDgUHhfNg-PRYSRuBDQNg4LCQvSAsEQ8BQcCOMD0QE4CAyRRYSDwZA8IA >>>> >>>> Sim Link 2: >>>> >>>> https://www.falstad.com/circuit/circuitjs.html?ctz=CQAgjCDsB0AcYCYCsBTAtLEAGaBOWuYALFrgGwDMFRSSkSZIuuISrWr6YYAUAG7hE4Mh24IQRTB3FgscudjiKpipP0HiKZTGJDJG08PIU4pSitGJgkuBGSQUkBSLgqqeAcw3DRQsLexsT28tHSESDkieAHdvSXAA+KweACdWBB0M9J0AjhIeMDIWBBJwLIiylRx5WWY6+oaWMGhSAixato7MNGbEWAYsCgRZMGosLOaseiIie2ItSARSGRbp2aR5skXhokVmGIkFf3FZ3eOggSIhvVKr8SWqw-BAyOg1WIrCvOuvoI+fkQSa4PIJeO43b73RYvA6nSrZeHJAAOIBEmUwaL0DBAbiiXkx+lRWEwCGhUQ+ZDOASQE1yqUOonKlMR4B4AGMJMy+pyqeJXsZ2oxmrAtHZ4IMiIRILpqu16XQhVkFQk+YcDjScuJledkrFtUqJlldaxIIqMcTEQdzvESnkpFahITbViDAdnU7Sjr2RCJCTPUbFMYIDgiIgMuQyKT8JJyIG5QB7QLB8AbWhsNgWVWqig8RNsZNLSRuBBvRQQCBanG51jPSDgRxgeg1nAV54thjIaujNuHfDNyw41lAA >>>> >>>> Sim Link 3, no LC Filter: >>>> >>>> https://www.falstad.com/circuit/circuitjs.html?ctz=CQAgjCDsB0AcYCYCsBTAtLEAGaBOWuYALFrgGwDMFRSSkSZIuuISrWr6YYAUAG7hE4Mh24IQRTB3FgscudjiKpipP0HiKZTGJDJG08PIU4pSitGJgkuBGSQUkBSLgqqeAcw3DRQsLexsT28tHSESDkieAHdvSXAA+KweACdWBB0M9J0AjhIeMDIWBBJwLIiylRx5WWY6+oaWMGhSAixato6zRGsiOyJILApEXEgyCDRq2AQKWG1ISAQ7OhE2SZbp2fnF5bGhxWYYiQV-cSIyIgTxZIEiCnESvPu9LCrj8EDI6DVYisKnmQiIK-Z7-CTPBCvIJeO4PUqwvSLT5Hc6XRCYJBZdFBAAOIBEmUwBL0DBAbiiXmJ+nxUIQSKivwuV2yzOSaRWlVYY1ZxyOmJy4g5p2BXMY2P5nOSsSFWWJ2KlzPijwkUiOumpyupCs1pOVwuSAGM9PDMHqsl9jBApmQRAhRhQsJJRrgsNcWu0USdykz5TwAPaBK3gJDWWisRSXQwcCj+8NByGSNxsHAyPThmMBtgQSDgRxgejhnAQVM4MjIWNgNxBkj4QuWMngHhAA >>>> >>> -- > 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 [email protected]. > To view this discussion on the web, visit > https://groups.google.com/d/msgid/neonixie-l/c9cc8a3e-a614-4c10-9017-7096b07f3cf7n%40googlegroups.com > <https://groups.google.com/d/msgid/neonixie-l/c9cc8a3e-a614-4c10-9017-7096b07f3cf7n%40googlegroups.com?utm_medium=email&utm_source=footer> > . > -- 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 [email protected]. To view this discussion on the web, visit https://groups.google.com/d/msgid/neonixie-l/CALiMYruPS51gezp7SrLi0yk4Q-cQOUQUHnMsJ7XkNLM_Gcx8%2Bg%40mail.gmail.com.
