I would like to know the best way to impliment a soft start b+ hv power supply. I do know about turning on my filament supply first then my high voltage. Is this good enough to prevent plate stripping. Does the filament supply need to ramp up slowly also? I read about using a thermister but I am not sure how to impliment them. Also the use of a 555 timer IC which I am also not sure how to impliment. I do not have room for the use of a rectifier tube. Could someone provide some information regarding this issue. I have been using solid state rectifiers on my amp for many years.
There is soft-start, then there is simple delay. I have not found any need for soft starting B+ when you apply filament voltage first.
This is one way to do it. My first version used a dual timer, where a large resistor was installed in series with the secondary center tap, for soft starting. Second timer shorted this resistor out for final operation. It provided no benefit, and I just got rid of it and went with straight delay.
Reason I like the delay method is it controls the magnitude of B+ when you have it loaded down. Most of my designs are light with the first cap, heavy towards choke input, so B+ can rise quite high without warm plates.
This is one way to do it. My first version used a dual timer, where a large resistor was installed in series with the secondary center tap, for soft starting. Second timer shorted this resistor out for final operation. It provided no benefit, and I just got rid of it and went with straight delay.
Reason I like the delay method is it controls the magnitude of B+ when you have it loaded down. Most of my designs are light with the first cap, heavy towards choke input, so B+ can rise quite high without warm plates.
Plate stripping is a myth. It just does not happen until you get into B+ voltages over 1KV. It is an issue on killowatt class radio transmitters but not home size audio amps.
That said the direct aanswer to your question is is place a "in rush" thermister in series with the B+. These can go in the AC side or DC side of the rectifier. I think DC is "cleaner". The device looks like a 100 ohm resistor when you first apply current to it at turn-on then afer a few seconds is lowers to about 2 ohms. This gives a soft start. They are cheap about $2.00 each but I got about a dozen free samples just for asking. You can select other numbers if 100R and 2R are not what you need, different time constants too. You find these in almost every medium to larger switch mode power supply so you can salvage a few from old TV and computers.
I think soft start is a good idea, not to prevent stripping but to safe the rectifiers and filter caps. They might last forever.
Last edited:
zigzagflux, Thanks for the info, I see the switch in the center tap secondary. So are you saying just delay the start of the B+ by applying filament voltage first and that is all you need to reduce plate stripping. Why is there two diodes in series for the fullwave center tap configuration. what was the current rating for the choke. Can the filament be withhout slow ramping of the voltage.
Valve Heater/Filament Regulator with Soft Start
Classic Valve Design - Original and Legacy Design PCB's
Classic Valve Design - Original and Legacy Design PCB's
As ChrisA says, the issue is inrush current not plate stripping. Plate stripping does not occur at normal audio amp voltages. Running with heaters on but no HT can cause cathode interface, so don't create a real problem by 'solving' a non-problem. For relatively low powers I would just include a suitable NTC thermistor in the mains wiring, but you really don't need even that with a conventional EI transformer. Putting the NTC on the primary side reduces by a small amount the initial current into cold heaters.
This assumes that your HT caps can take full unloaded voltage before the valves warm up.
This assumes that your HT caps can take full unloaded voltage before the valves warm up.
No, it prevents excessive B+ with choke input supplies before the tubes are warmed up. I don't think a thermistor in the primary, secondary, or wherever will adequately protect against this. The alternative is to install higher voltage caps, which aren't cheap.
Provides reverse voltage within diode ratings- the ballast resistors ensure equal voltage sharing.
Hammond 159S. Rated for 225mA.
For sure. Some of my builds are AC heated, voltage regulated, and current regulated. Never burned out a heater or filament. It's a choice you make. I will admit, I have one amp that uses (8) 6DE4's in parallel, and I use a CL-90 on the dedicated transformer primary. Not specifically to limit inrush on the heaters, but to control inrush on the 120VAC system. Personal preference, not a requirement.
Yes you are right. The delay is to allow you to save money by using under rated caps in the filter. The best solution is to spec the caps so they can handle the maximum voltage. I'm glad this did not turn into an argument over that stripping myth.
This is also the real reason Leo Fender put standby switches on his guitar amps, so he could use cheaper caps. (they were quite expensive) And then Jim Marshal made a near exact clone of the fender and
But today one can buy plug-in style caps that are rated for 450 volts for not much. and we can put two in series and handle 900 volts.
The thermister will address this problem if you design it correctly. You need to pick one with the correct time constant. Some have bigger slugs of ceramic in them and take longer to heat up. But really, just buy the correct caps. So it cost $5 more.
In rush limiters can also be used on the mains side of the transformer. But this will soft-start the heaters too. I can't think of a good reason to do that.
All this changes if we are talking about multi-killowatt amps with multi-KV B+ supplies for ham radio transmitters. That is a different thing.
If you ARE building a radio transmitter there are delay start relays made for this purpose. They look like normal tubes in octal bases and have glass ennvelope. The contacts close or open after a fixed delay from power up. Vacuum relays like these are reliable. They don't arc because there is no air to ionize
Depends; if you use electrolytics, then maybe a few dollars is all it takes to get extra voltage at high uF. I avoid electrolytics except in basic power supplies (like my timer circuit, which has no interface with my audio). Not a big fan of stacking electrolytics with ballast resistors, either; over a few years their leakage can vary significantly.
Also, just because you can install a cap suitable for the voltage doesn't mean your tube likes it. My 300B push-pull normally runs at 500V. Without significant load on the power supply (a bleeder resistor ain't gonna cut it) B+ would run upwards of 800-900V. I really don't want to expose $450 worth of 300B's to 900V. Nor do I feel comfortable running 2-450V caps at 900V; I don't trust 'lytics right at their rating, especially when long term they are running at less than that; the oxide layer tends to settle at a thickness suitable for the operating voltage.
I choose to install a $2 timer circuit and keep my films in the PS. The next guy is perfectly happy with a multi-section can 'lytic. Personally, I opine the former more reliable with a longer life expectancy.
Fundamentally each decision we make in the design process can affect the next; you need to make this timer /soft start decision carefully based on the entire amp.
Also, just because you can install a cap suitable for the voltage doesn't mean your tube likes it. My 300B push-pull normally runs at 500V. Without significant load on the power supply (a bleeder resistor ain't gonna cut it) B+ would run upwards of 800-900V. I really don't want to expose $450 worth of 300B's to 900V. Nor do I feel comfortable running 2-450V caps at 900V; I don't trust 'lytics right at their rating, especially when long term they are running at less than that; the oxide layer tends to settle at a thickness suitable for the operating voltage.
I choose to install a $2 timer circuit and keep my films in the PS. The next guy is perfectly happy with a multi-section can 'lytic. Personally, I opine the former more reliable with a longer life expectancy.
Fundamentally each decision we make in the design process can affect the next; you need to make this timer /soft start decision carefully based on the entire amp.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.