HV Power Supply Reverse Engineer

[zac]

Hey all-

  A while back I purchased this item on ebay, with the intent of using it's guts in another project. Upon receiving and shelling it, I discovered that it was not producing the claimed 120VAC, rather, it was producing about 170.3VDC! I am currently in the middle of a nixie project as well, and thought that i might be able to use the circuit for that, although i might rebuild it to a smaller footprint.

Does this look like a design any of you have encountered before? 

Some on-hand observations-

Under the heat-sink is a TO-220 package, the only marking is "D880" - I have serched it, and found a D880 transistor, 

The small cap is a 50V 10uF

The large cap is a 400V 2.2uF

What appears to be a mylar film capacitor is marked 2A4

you can't see it in the top down picture, and the markings are facing the wrong way for the side-view, but right in front of the heat sink/TO-220 component is a small ceramic disk capacitor, 4.7pF.

The leads bent off to the right are to an LED, to indicate when it is powered.

Slightly harder to see in the picture are the two holes where i removed the leads coming from the Vin attachment meant to be used in a vehicle cigarette lighter jack, so 12VDC in.

The leads on the left are the Vout. Originally, there were big clips at this spot used to hold the prongs of a standard us non grounded wall plug.

I have attached 4 pictures, one top-down, one bottom-up, and two from the side, one with a focal point on the near side, and one with a focal point on the far side.

Any insights?

[greg...@hotmail.com]

Low-cost 12VDC to 120VAC inverter; no idea what frequency it produces. You would need an oscilloscope to get a true idea what this inverter really does.

I prefer to purchase new parts that have datasheets; it might cost me a bit of money but having all the details saves you a lot of time. Remember, time is money, too.

All of my present nixie supplies are driven from the mains and they are NON-isolated. Just use an isolation transformer during debug & testing. I also recommend you use a voltage higher than 170V for nixies. Yes, they will fire at 170V, but as they age, their firing voltage will increase and you may not get reliable ignition. Another reason for using a higher voltage is that you will have less variation in the nixie current. Cathode current varies with supply-voltage, tube-condition, and even by digit. Using a higher voltage and a higher series resistor will reduce the variation in cathode current. If your supply is 170V, and the line-voltage drops by 6%, you will have a 33% reduction in tube current (brightness). If you run at 340VDC, a 6% drop in line voltage produces only a 7% drop in tube current.

My latest designs use current-feedback on the cathodes, so it's literally constant-current.

Since I'm in the US, our mains are 120V rms. Using a rectifier and capacitor gets you 170VDC. I use a doubler to get +340VDC, which is perfect. Tubes typically use 2mA, and have about 140V across them when lit, so a good starting value for 340VDC is 100K.   R = (Vsupply-Vtube)/Itube. Use the recommended current from the datasheet, and measure the tube's actual voltage-drop for various digits, and take the average. Better yet, do this for all of your tubes.

Nixies are damaged by excess current; higher voltage does not harm the nixie as long as you limit the current to a proper value. Using higher voltage does waste more energy, though.

Someday I will make a constant-current driver that wastes no energy, but I'm too busy with the nixie wristwatch right now (pcb goes out for fab as soon as I finish my simulations)

[zac]

Thanks for the reply!

I would go with a mains-driven supply, but my nixie project is portable. This forum is probably a good place to ask the question:

for a lightweight battery powered nixie display (6x IN-12, the battery will also have to power a few 595 shift registers and an atmega328), what type of design would you suggest? I only need 2-3 hours of run time. 

To get an idea of the project, you can visit my post on the Arduino forums. I had taken a sabbatical from the project for a bit to focus on a new job, but i am working on it again, and in the last week, have changed a number of items, so some of the info represented is incorrect. I will update that post when I have a moment to sit down and take some new pictures.

[greg...@hotmail.com]

The battery-powered design for my wristwatch uses a 3.7v Li-ion battery (1000mA-hr) to produce 180VDC at 7ma. I use a 50:1 transformer (Coilcraft LPR6235), and it will run somewhere around 150kHz. It's controlled by the onboard FPGA, and I have voltage-feedback. The power supply will actually go higher than just the transformer turns-ratio due to inductive effects. The transformer is rated for 100V, so it's going to be interesting to see if it breaks-down at 180V.....

If your area isn't as constrained (wristwatches are brutal....), there are a lot of other transformers available, such as those used in CFLs or laptop backlights. Coilcraft has a nice selection and they sell direct to you even if you only buy single units.

The circuit itself is simple: Transformer primary is connected to the 3.7V battery, and to a NMOS transistor. I used a low-leakage DMG3420U. The fpga turns the NMOS on for 2usec, then off. You have to be careful about how long the transformer is energized (otherwise you get really unsafe current), and keep the frequency low enough to minimize losses. I expect I will need to do a lot of tweaking once I get my PCB, but the tweaks are easy because it's an FPGA. The secondary winding goes to a bridge rectifier and small filter cap (0.1uf).

I use direct-drive, not multiplex.The 3-1/2 digit 7-segment display is rather large, so there is enough room on the PCB for each segment to have a small NPN transistor and emitter resistor, for current-control. The FPGA drives the NPN at 3.3V; with a Vbe drop of 0.7V, the emitter resistor is calculated from the voltage (3.3-0.7 =2.6v) and the segment current (335uA in my case). I use dual NPNs (PMBTA42DS) to save PCB area. Everything (except the neon display) is surface-mount.

[Jonathan Peakall]

Just a question: Did you put a load on it? Might be floating higher than spec with no load.

Jonathan

--
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 post to this group, send an email to neoni...@googlegroups.com.
To view this discussion on the web, visit https://groups.google.com/d/msgid/neonixie-l/d83515d8-30c1-48c6-944a-bf9d7bc0b95e%40googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.

[Adam Jacobs]

Holy cow... an automotive DC/AC inverter for 1 penny?
I'm not as brave as this man, not by a long shot. Every crazy cheap powersupply I've ever bought from China was a fire hazard. I have to assume that this thing is also a fire hazard. :)

-73 Adam W7QI

To view this discussion on the web, visit https://groups.google.com/d/msgid/neonixie-l/525718D8.8040107%40madlabs.info.

[zac]

Just for grins and giggles, I have powered all 6 of the intended nixies. My initial claim was incorrect, it fires without load at about 178V.

with all six IN-12's it maintains about 142V. a single nixie drops it to about 161.6V

nothing feels overly warm after running for 10mins or so, it was almost cool to the touch, no pyrotechnics today :(

[greg...@hotmail.com]

Nice that the tubes are running around 140V and the inverter can supply it. I think it's safe to say that if nothing is hot after 10 minutes, it should be fine.

I would expect that as the tubes age over months/years, you will find that some of them will no longer fire because the supply barely supplies enough voltage.

If you want to go ultra-geeky, get a dynamotor for your supply. There's a nice Redmond unit very similar to mine on Ebay right now. Very retro, and they sound really cool.

[Charles MacDonald]

On 13-10-10 10:18 AM, zac wrote:

> Under the heat-sink is a TO-220 package, the only marking is "D880" - I
> have serched it, and found a D880 transistor,

http://www.piclist.com/images/www/hobby_elec/gif/2sd880e.gif

Japanesse type transistors often omit the 2s in front of type numbers.
No doubt it is a 2SD880

http://www.futurlec.com/Transistors/2SD880.shtml

60Volts 30 Watts

--
Charles MacDonald Stittsville Ontario
cm...@zeusprune.ca Just Beyond the Fringe
http://Charles.MacDonald.org/tubes
No Microsoft Products were used in sending this e-mail.