Excess MOSFET Heat and Voltage Bleed

[waldo1979]

Hello,
     I'm building clock with four IN-4 and INS-1 Neon Tube colons. I have everything wired but i'm having some issues with the MOSFET heating up and an issue where voltage is leaking to cathodes. For example, anode2 is shared by a nixie tube (cathode1) and the INS-1 colon (cathode1). When i fire the colon, some voltage is bleeding through to the nixie tube.

    I'm wondering if i damaged the SN74141's. I didnt' have any sockets on hand and impatiently solder'd them directly.

    FYI, here's a link to my Driver software. Still a work in progress. If the C++ pattern ends up being too slow i'll re-write it in C-style code:
https://github.com/waldo1979/WitchingHour

-Waldo

[nonentity]

Hello,
     I'm building clock with four IN-4 and INS-1 Neon Tube colons. I have everything wired but i'm having some issues with the MOSFET heating up and an issue where voltage is leaking to cathodes. For example, anode2 is shared by a nixie tube (cathode1) and the INS-1 colon (cathode1). When i fire the colon, some voltage is bleeding through to the nixie tube.

    I'm wondering if i damaged the SN74141's. I didnt' have any sockets on hand and impatiently solder'd them directly.

    FYI, here's a link to my Driver software. Still a work in progress. If the C++ pattern ends up being too slow i'll re-write it in C-style code:
https://github.com/waldo1979/WitchingHour

-Waldo

Can you tell me a bit about your setup?

Do the tube and colon share a common anode as well?
Are they both resistored on the anode?

Try different anode channel per unit. And use the multiplexing to gate them.

[waldo1979]

Hello,
     I'm building clock with four IN-4 and INS-1 Neon Tube colons. I have everything wired but i'm having some issues with the MOSFET heating up and an issue where voltage is leaking to cathodes. For example, anode2 is shared by a nixie tube (cathode1) and the INS-1 colon (cathode1). When i fire the colon, some voltage is bleeding through to the nixie tube.

    I'm wondering if i damaged the SN74141's. I didnt' have any sockets on hand and impatiently solder'd them directly.

    FYI, here's a link to my Driver software. Still a work in progress. If the C++ pattern ends up being too slow i'll re-write it in C-style code:
https://github.com/waldo1979/WitchingHour

-Waldo

Can you tell me a bit about your setup?

Do the tube and colon share a common anode as well?
Are they both resistored on the anode?

Try different anode channel per unit. And use the multiplexing to gate them.

Below is a table describing my tube wiring. The nixies have a 10k resistor on the anodes and the neons are connected in series with a 47k resistor. When I illuminate the neon tubes (colon) and leave it on the Nixie->hour2 has a ghost. Using a different anodes have the same effect. if i illuminate in a loop with a delay the ghosting isnt visible.

The wiring to the tubes is AWG 28 (http://amzn.com/B007R9SQQM), which might be a little small. The tubes are socketed using sockets from Germany i bought off ebay: http://www.ebay.com/itm/331114980365

When the neon tube is illuminated, if i take a voltage reading from the cathode pins which are supposed to be in the off/low state and ground there is 140~ volts. I checked for shorts but didn't see any. Seems like the problem lies within the shift registers.

Anode1	Nixie->Hour1
Anode2	Nixie->Hour2, Neon Tubes
Anode3	Nixie->minute1
Anode4	Nixie->minute2
Cathode0	Nixie->hour1, Nixie->hour2
Cathode1	Nixie->Minute1, Nixie->Minute2, Neon Tubes

[nonentity]

Hello,
     I'm building clock with four IN-4 and INS-1 Neon Tube colons. I have everything wired but i'm having some issues with the MOSFET heating up and an issue where voltage is leaking to cathodes. For example, anode2 is shared by a nixie tube (cathode1) and the INS-1 colon (cathode1). When i fire the colon, some voltage is bleeding through to the nixie tube.

    I'm wondering if i damaged the SN74141's. I didnt' have any sockets on hand and impatiently solder'd them directly.

    FYI, here's a link to my Driver software. Still a work in progress. If the C++ pattern ends up being too slow i'll re-write it in C-style code:
https://github.com/waldo1979/WitchingHour

-Waldo

Can you tell me a bit about your setup?

Do the tube and colon share a common anode as well?
Are they both resistored on the anode?

Try different anode channel per unit. And use the multiplexing to gate them.

Below is a table describing my tube wiring. The nixies have a 10k resistor on the anodes and the neons are connected in series with a 47k resistor. When I illuminate the neon tubes (colon) and leave it on the Nixie->hour2 has a ghost. Using a different anodes have the same effect. if i illuminate in a loop with a delay the ghosting isnt visible.

The wiring to the tubes is AWG 28 (http://amzn.com/B007R9SQQM), which might be a little small. The tubes are socketed using sockets from Germany i bought off ebay: http://www.ebay.com/itm/331114980365

When the neon tube is illuminated, if i take a voltage reading from the cathode pins which are supposed to be in the off/low state and ground there is 140~ volts. I checked for shorts but didn't see any. Seems like the problem lies within the shift registers.

Anode1	Nixie->Hour1
Anode2	Nixie->Hour2, Neon Tubes
Anode3	Nixie->minute1
Anode4	Nixie->minute2
Cathode0	Nixie->hour1, Nixie->hour2
Cathode1	Nixie->Minute1, Nixie->Minute2, Neon Tubes

First off, if you don't have your neon lamps equipped with a resistor for EACH anode per each lamp, you will have a bit of trouble there.  I've seen ghosting and other odd results when multiple neon lamps use unresistored anodes.

[waldo1979]

Here's a picture of the ghosting i was talking about. This happens when i'm only illuminating the neon tubes (the cathodes for the nixie tube are all OFF/LOW state... shift register is 1/1/1/1):


Here's a picture with all the nixie + neon tubes illuminated with a 1ms delay. There's no ghosting but the MOSFET is too hot (laser temp is 140F). This is bound to cause the neighboring ceramic caps to fail.

[waldo1979]

Apparently the ghosting has nothing to do with sharing an anode. I finally finished my software and when i display 5:00PM there is a ghost in the first hour digit (which is disabled):


Based on other post's i've seen the shift registers are probably the cause.

[nonentity]

Apparently the ghosting has nothing to do with sharing an anode. I finally finished my software and when i display 5:00PM there is a ghost in the first hour digit (which is disabled):


Based on other post's i've seen the shift registers are probably the cause.

Well, normally when they ghost like that, if it's a driver chip issue, that usually means they always fire multiple tube elements, AND they exhibit the same error across more than one tube.  Meaning, if say, your first two digits share the same driver chip, both those tubes would ghost.

It seems like there may be a short on the cathode side, can you tell me which numerals ghost? if it's all of them, or just one or two?


If a driver chip is bad, that would mean any tube associated with it would ghost, or display a jumble of numerals.

[waldo1979]

It seems to be only the first two tubes that are ghosting which are connected to the same cathode/shift register. I did some debugging where i disconnected various cathode pins to determine if there was a short but that was uneventful.  I'll do some more testing and get back to you.