i have a SSR rated at 3-32v control voltage and 25 amps of load(i know extremely overkill, im interested in using it because its solid state)
i know most of you people go through all sorts of trouble making delay timers etc. But my idea is as i have a 6.3v heater system using a linear current limited power supply, i could perhaps use a voltage divider with equal resistance on both resistors, and then feed the divided voltage into the relay. the relay will catch when the voltage on the heaters gets warm enough that my recifier tube startup delay will be long enough that it will be all ready for conducting
anything wrong with my logic/thought process?
i know most of you people go through all sorts of trouble making delay timers etc. But my idea is as i have a 6.3v heater system using a linear current limited power supply, i could perhaps use a voltage divider with equal resistance on both resistors, and then feed the divided voltage into the relay. the relay will catch when the voltage on the heaters gets warm enough that my recifier tube startup delay will be long enough that it will be all ready for conducting
anything wrong with my logic/thought process?
Why not just use an LM317 in SLOW START circuit as per the datasheet.
For heaters you can even use it as a CCS with SLOW START.
No chunky relays.
This is how I approached the problem with my Borbelly clone.
The LM317 slowly applies heater power to the valves, then after approximately 30 seconds the B+ is applied to the HT rail.
For heaters you can even use it as a CCS with SLOW START.
No chunky relays.
This is how I approached the problem with my Borbelly clone.
The LM317 slowly applies heater power to the valves, then after approximately 30 seconds the B+ is applied to the HT rail.
Attachments
Last edited:
i know most of you people go through all sorts of trouble making delay timers etc.
Most of us don't understand why it can be needed.
Soft start - yes. But abrupt B+ application on hot toobs when interstage coupling capacitors are discharged? Hah...
Search for soft start, inrush current limiting, we discussed such matters here many times.
i already have soft start actually, i use a tube rectifier. but it doesnt quite take long enough to warm up and the tubes dont conduct for 5 to 10 more seconds after rectifier gets hot enough to light up my voltage regulator tubes. linear supply bring up the filaments in about 60 seconds total, which is probably nice for them anyways
so what im really aiming for is a delay isnt it?
so what im really aiming for is a delay isnt it?
Last edited:
i already have soft start actually, i use a tube rectifier. but it doesnt quite take long enough to warm up and the tubes dont conduct for 5 to 10 more seconds after rectifier gets hot enough to light up my voltage regulator tubes. linear supply bring up the filaments in about 60 seconds total, which is probably nice for them anyways
Why?
It was safety measure for radar output tubes that started generating high power RF energy when parts of their filaments started emission. As the result such parts were getting even hotter, so each time the tube was powered on, tiniest parts of filaments evaporated faster than the rest of filament. But such effect do not exist in audio amps.
Why?
It was safety measure for radar output tubes that started generating high power RF energy when parts of their filaments started emission. As the result such parts were getting even hotter, so each time the tube was powered on, tiniest parts of filaments evaporated faster than the rest of filament. But such effect do not exist in audio amps.
Ah i think i understand what your getting at now. your saying its just fine for the tubes to have HT come up before the tubes are even close to conducting.
Im a little bit worried about my filter capacitors too tho, there rated for 500v but i imagine 10 or so seconds at up to 565v isnt going to internally heat them enough to kill them
Ah i think i understand what your getting at now. your saying its just fine for the tubes to have HT come up before the tubes are even close to conducting.
Yes, and manufacturers of tubes thought similarly when specified max voltages on cold tubes higher than on hot tubes: they expected them to heat up then sag power supply.
Imagine, when interstage coupling capacitors charge initially through grid leak resistors positive voltage on grids don't cause grid currents. But charging through conducting go cathodes grids they cause significant stress to tubes.
Ah so it has the side-benefit of allowing my coupling capacitors some time to charge up pretty much all the way, makes sense. so in fact the timing of all my stuff is rather ideal, my bias supply comes up near instantly, then the b+ about 30 secs or what feels like it later, and then the filaments some time later
Ah so it has the side-benefit of allowing my coupling capacitors some time to charge up pretty much all the way, makes sense. so in fact the timing of all my stuff is rather ideal, my bias supply comes up near instantly, then the b+ about 30 secs or what feels like it later, and then the filaments some time later
There is a saying, "Better is enemy of Good"
I agree with payitforwardeddie. KISS.
However there are always extremists in any field.
During my Borbelly build #2 I made a few OVERKILL assumptions.
1. Valves are expensive and rare and should be treated with kid gloves.
If you think about the heater being just a low power light bulb. Most light bulbs fail when you turn them ON, they do fail in use but not as often.
In order to prolong their life it MIGHT be a good idea to turn them on SLOWLY and let them heat up SLOWLY.
2. Valves fail when they run out of electrons to donate from their cathodes.
OK - In HIGH POWER use it is beneficial to ensure that the valve is HOT before trying to draw current through it. This MIGHT prevent COLD STRIPPING of the cathode.
Both of these measures are largely immaterial in low power audio use and definitely OVERKILL in my Borbelly Pre-Amp.
But hey, I'm trying to preserve the few valves that are left in this world. I let the valves warm up for 30 secs and then apply B++
It's subjective. A good valve in a poor circuit will outlast a poor valve in an excellent circuit. I just wanted to give the amp a hint of being reliable for at least a few months without needing a tube change.
However there are always extremists in any field.
During my Borbelly build #2 I made a few OVERKILL assumptions.
1. Valves are expensive and rare and should be treated with kid gloves.
If you think about the heater being just a low power light bulb. Most light bulbs fail when you turn them ON, they do fail in use but not as often.
In order to prolong their life it MIGHT be a good idea to turn them on SLOWLY and let them heat up SLOWLY.
2. Valves fail when they run out of electrons to donate from their cathodes.
OK - In HIGH POWER use it is beneficial to ensure that the valve is HOT before trying to draw current through it. This MIGHT prevent COLD STRIPPING of the cathode.
Both of these measures are largely immaterial in low power audio use and definitely OVERKILL in my Borbelly Pre-Amp.
But hey, I'm trying to preserve the few valves that are left in this world. I let the valves warm up for 30 secs and then apply B++
It's subjective. A good valve in a poor circuit will outlast a poor valve in an excellent circuit. I just wanted to give the amp a hint of being reliable for at least a few months without needing a tube change.
I once did the same: stabilized filament voltage, with current limit slowly going up. And B+ stabilized, with time constant sensing current consumption. When tubes started consuming current B+ went up faster. As the result, it brought up B+ slowly, waiting for tubes to start consuming current.
Now I limit inrush currents of all rectifiers at once by resistor in primary of power transformer. It is enough to sense one of rectified voltages, because they all go up proportionally to each other. Then relay shunts the resistor in series with primary.
Now I limit inrush currents of all rectifiers at once by resistor in primary of power transformer. It is enough to sense one of rectified voltages, because they all go up proportionally to each other. Then relay shunts the resistor in series with primary.
Im a little bit worried about my filter capacitors too tho, there rated for 500v but i imagine 10 or so seconds at up to 565v isnt going to internally heat them enough to kill them
I don't think heat is the issue here. I wouldn't recommend going that high in over-voltage, as you may experience capacitor failure, and you may have times when you go even higher due to mains variation. You may also run in to reforming problems if you don't use the amp for an extended period.
so my initial worry and reason for wanting to implement the delayed b+ is at least somewhat valid, the b+ doesnt go past 450v when the tubes are conductingI don't think heat is the issue here. I wouldn't recommend going that high in over-voltage, as you may experience capacitor failure, and you may have times when you go even higher due to mains variation. You may also run in to reforming problems if you don't use the amp for an extended period.
I am ignorant with tubes but here is what I gathered from my reading.
a. filaments must be hot when high positive DC comes on. When using rectifiers this is mandatory to avoid damage
b. in pentodes (full pentode mode) the screen voltage must not come on before b+ as the delicate grid will try to aborb the entire electron stream and fail.
c. bias voltage must ideally come on istantly to avoid damage
My future configuration for a PP amp based on the GU81m tube is:
LV (filaments) : switch ---> slow start ----> full power
HV (anode ecc) : switch ----> timer delay (5 minutes for my large rectifier) ----> softsart -----> full power
Sounds complex but really only three modules are needed and all should go well.
a. filaments must be hot when high positive DC comes on. When using rectifiers this is mandatory to avoid damage
b. in pentodes (full pentode mode) the screen voltage must not come on before b+ as the delicate grid will try to aborb the entire electron stream and fail.
c. bias voltage must ideally come on istantly to avoid damage
My future configuration for a PP amp based on the GU81m tube is:
LV (filaments) : switch ---> slow start ----> full power
HV (anode ecc) : switch ----> timer delay (5 minutes for my large rectifier) ----> softsart -----> full power
Sounds complex but really only three modules are needed and all should go well.
Now I limit inrush currents of all rectifiers at once by resistor in primary of power transformer. It is enough to sense one of rectified voltages, because they all go up proportionally to each other. Then relay shunts the resistor in series with primary.
This sounds the best idea to me, what type of values and times do you use for a particular power draw? I.e. what values (R and t) would you use for a typical 40W push-pull stereo amp>
I am ignorant with tubes but here is what I gathered from my reading.
a. filaments must be hot when high positive DC comes on. When using rectifiers this is mandatory to avoid damage
Not for lower anode voltage tubes, or lower current tubes, and not for directly heated tubes. If you have 2000 volt or higher tubes with indirectly heated oxide cathodes, and the tubes run a fair amount of current, it starts to be a worry. The higher the anode voltage and cathode current in oxide cathode tubes, the greater the worry.
b. in pentodes (full pentode mode) the screen voltage must not come on before b+ as the delicate grid will try to aborb the entire electron stream and fail.
Depends on the screen circuitry and the screen's allowable dissipation over time. Maybe it will hurt, maybe not. Depends on the screen system and tube.
c. bias voltage must ideally come on istantly to avoid damage
Right, for tubes that overdissipate without bias we want bias before or at the time anode and screen voltages appear.
My future configuration for a PP amp based on the GU81m tube is:
LV (filaments) : switch ---> slow start ----> full power
HV (anode ecc) : switch ----> timer delay (5 minutes for my large rectifier) ----> softsart -----> full power
Sounds complex but really only three modules are needed and all should go well.
People tend to lift little snips of truth from one application and apply it to other systems without the faintest idea why things are the way they are.
Directly heated tubes suffer no ill effects when HV applied before the filament power is applied. It is also doubtful you will ever see inrush damage to the tube filament on that tube.
Oxide cathode emitters, when very high voltages are used, tend to flash over if HV is applied at low cathode temperatures. Oxide cathodes also, if emission current density is high, tend to hot spot when very high currents are drawn before the cathode is fully warm. What this does is heat the already hotter parts of the cathode. There is also a problem of positive ion poisoning if the electron cloud is stripped away, with higher voltage tubes.
This is why lower voltage low current oxide cathode tubes, like 4CX150's, have a short "recommended" (but not required) holdoff time requirement, while high voltage hogh current tubes like 3CX10000B7's have a mandantory 5 minute heater time before HV is applied. Receiving tubes normally have no requirement at all to hold off voltage.
Thoriated tungsten tubes have none of the problems oxide cathode tubes do.
Delayed HV with warm up of small low voltage tubes just gives people something to feel special about, unless they have design errors they are patching (like unstable HV supplies that soar to unsafe levels before the tubes draw current).
Very informative reply. Thanks. However my ignorant approach is actually the most conservative. Aside from the fact i have power station level rectifiers which need at least 5 minutes to warm up any other design issue will find a solution in an overly cautious approach. Also there is a distinct setling time for the gu81m filaments. The entire tube makes the oddest sounds and vibrations before the metal parts expand and set in place. I'd rather give this time to the tubes before putting them in operation.
Inviato dal mio GT-I5800 usando Tapatalk
Inviato dal mio GT-I5800 usando Tapatalk
I had to cut the message short typing from my cell phone. I simply wanted to point out that in any given case 30 seconds warmup and softstart for the HV is not something that most people can't live with. It gives time for the filaments to properly warmup, it can help the user to identify problems with tubes (ie: bad socket tube contact, damaged filaments etc) I know it sounds unlikely but I really can't see what's wrong with allowing 15-30 seconds of warmup before HV comes on.
With gas rectifiers this is mandatory thus setting the main question of whether the output tubes benefit from it aside.
Also I don't see what's special in having a relay start....what is there to brag about? It is a small delay switch or even a small timer delay tube. Nothing exotic.
With gas rectifiers this is mandatory thus setting the main question of whether the output tubes benefit from it aside.
Also I don't see what's special in having a relay start....what is there to brag about? It is a small delay switch or even a small timer delay tube. Nothing exotic.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
- Home
- Amplifiers
- Tubes / Valves
- Delay B+ using relay?