EEVblog Electronics Community Forum

Electronics => Beginners => Topic started by: eriknau on October 10, 2022, 05:30:50 pm

Title: LM4871 vfd filament driver circuit issue
Post by: eriknau on October 10, 2022, 05:30:50 pm
I have followed this design (http://magictale.com/2603/revisiting-vfd-psu-part-ii/) for a vfd filament driver circuit with the LM4871 chip specifically this schematic (http://magictale.com/wp-content/uploads/2015/05/LM4871_SquareWaveGenerator.png). It works great in the breadboard providing a square wave at ~3kHz frequency and vfd segments are crisp while non-lit segments are completely dark. Then I moved the circuit to a 2-sided PCB I made on a CNC and I'm getting weird noise I can't identify. Lit segments are still bright but those segments are dimly lit across other grids where they should be turned off. I had the breadboard and PCB circuits powered at the same source, from the 3.3V pin on a Teensy LC that is controlling the segments through darlington transistor arrays. The only thing I can figure is I have connected something that's causing the noise in the pcb design.
If anyone can see it I would be grateful for input!
This is my version which I tried to make the same:
(https://openblackboard.com/files/Schematic.png)
and the PCB design:
(https://openblackboard.com/files/pcb.png)
Title: Re: LM4871 vfd filament driver circuit issue
Post by: Zero999 on October 11, 2022, 04:19:37 pm
What sort of capacitors are on the PCB version? Certain types of ceramics are piezoelectric which means they vibrate, when an AC voltage is applied. Try changing C2 for an aluminium of tantalum capacitor, or swap C1 and C2, assuming C1 is polarised i.e. not ceramic.
Title: Re: LM4871 vfd filament driver circuit issue
Post by: eriknau on October 11, 2022, 06:57:53 pm
Thank you, that's a good tip. C1 is indeed electrolytic, C2 is ceramic and C3 I've swapped between ceramic disk and polyester. I will try some more swapping and see where it leads!
Title: Re: LM4871 vfd filament driver circuit issue
Post by: eriknau on August 09, 2025, 06:34:36 pm
Took 3 years of on-and-off troubleshooting but I solved this issue! It turns out the breadboard provided enough resistance that the voltage across the filament was in the optimal range of ~3.3V for the filament/segment differential. Putting the circuit on a PCB lowered the resistance, increasing the voltage across the filament to ~5V, causing filaments to glow that shouldn't be. I added a 10 ohm resistor to one of the filament signal wires and the vfd looks great. One thing that threw me off was the PCB would power the filament effectively if I connected it to the circuit with alligator clips. Now I realize the alligator wires were adding resistance as well.