I am using a few 12V switching power supplies for heaters of my DIY tube amps. Because the heater resistance is too low, some of them will not start. Two of them blink for 10 - 30 seconds then run normally. I assume the blinking is due to current limiting of the supply. My question is whether start-up blinking will cause premature failure of the supplies or hurt the tube heaters?
I would look for some higher rated supplies. In my manufacturing experience, we would derate purchased power supplies by 50% for reliability. A power supply with constant current mode (not just voltage) would handle cold heaters just fine, with a nice slow warmup.. Just set it a little above the hot heater current.
If you have some margin for (hot) heater current, a series PTC thermistor might allow the power supply to start up without hitting the current overload. There's some voltage drop, but most supplies have a fine voltage adjustment, Maybe CL-30 - 2.5 Ohms, 8A, about 0.2 Ohms hot. https://f.hubspotusercontent40.net/...-325D-Thermometrics-NTC-Inrush-031814-web.pdf
If you have some margin for (hot) heater current, a series PTC thermistor might allow the power supply to start up without hitting the current overload. There's some voltage drop, but most supplies have a fine voltage adjustment, Maybe CL-30 - 2.5 Ohms, 8A, about 0.2 Ohms hot. https://f.hubspotusercontent40.net/...-325D-Thermometrics-NTC-Inrush-031814-web.pdf
Hi Tom,
Thanks for your reply. Your suggestion of PTC thermistor is what I did with a 6.3V supply that could not start at all. I asked this question just to see whether I can save some effort with those surplus 12V supplies (ranges from 1.5A to 4A). For 2 EL34 and 1 12AX7, the heater current is around 1.65A @ 12V. Interestingly, one of the 1.5A supplies actually started right up. However, it ran hot. I am a bit reluctant to use it.
The 6.3V supply was modified from a 5V 6.5A supply by adjusting the LT431 output resistors. I then added a power switch at the output. When the switch is off, a 5D-11 (5 ohm cold) thermistor is connected in series with the output. Once the heaters are warm, I turn on the switch to connect the output directly with the heaters. This is a two step operation. However, the thermistor heats up instantaneously. Only a few seconds is needed before switching over.
Thanks for your reply. Your suggestion of PTC thermistor is what I did with a 6.3V supply that could not start at all. I asked this question just to see whether I can save some effort with those surplus 12V supplies (ranges from 1.5A to 4A). For 2 EL34 and 1 12AX7, the heater current is around 1.65A @ 12V. Interestingly, one of the 1.5A supplies actually started right up. However, it ran hot. I am a bit reluctant to use it.
The 6.3V supply was modified from a 5V 6.5A supply by adjusting the LT431 output resistors. I then added a power switch at the output. When the switch is off, a 5D-11 (5 ohm cold) thermistor is connected in series with the output. Once the heaters are warm, I turn on the switch to connect the output directly with the heaters. This is a two step operation. However, the thermistor heats up instantaneously. Only a few seconds is needed before switching over.
That's why I suggested the 2.5 Ohm, 8A one. It's a larger diameter too, so will heat more slowly. And running at 20% of rating won't run that hot, so you can leave it in the circuit.
It depends on what you mean by "blinking". Many SMPSes will go into hiccup mode on overload. This means they'll try to start, the over-current protection will trip, and the supply shuts down. This repeats itself until the power is turned off or the over-current condition is cleared. After some number of cycles of this the filaments will be hot enough that they no longer trip the over-current limit, so the supply finally gets to start up.
While it isn't super elegant, the rough start shouldn't be an issue for the power supply. It could maybe be an issue for the filaments, especially if we're talking DHT filaments.
I added a soft start to my Universal Filament Regulator for this exact reason.
I like the NTC solution. You can always short it out with a relay once the tube filaments are hot. That way you get the benefit of the NTC even if the power glitches a bit.
Tom
While it isn't super elegant, the rough start shouldn't be an issue for the power supply. It could maybe be an issue for the filaments, especially if we're talking DHT filaments.
I added a soft start to my Universal Filament Regulator for this exact reason.
I like the NTC solution. You can always short it out with a relay once the tube filaments are hot. That way you get the benefit of the NTC even if the power glitches a bit.
Tom
I always over-spec the smps whenever I use them for heaters. I have a 12V 13A one that drives without hiccup a couple of 6P41S (~1A heater).