That's why I have a plasma cutter 
I like hot metal work... tubes are in there somewhere...
I like hot metal work... tubes are in there somewhere...Yes...
I have "issues"
But then again i like the look that the neighbours give me when tho whole neighbourhood lights dim when i turn my system on
By all means, this is why this thread is in place...
I always tend to "overbuild" everything
Ok.
Do you really need abrupt B+ start when cathodes are already hot, and tubes start conducting? I don't think so. Charging coupling capacitors cause grid currents that cause abnormal anode currents. Most probably you need soft start, so inrush currents in power supply are limited, and limited are currents that slower charge interstage capacitors.
One of the way how I do that, I use a current limiting resistor (usually 20 Ohm 12W wirewound one, for 120V in amps up to 1 KW of power consumption). It is in series with all sources: filament, anode, bias, etc..., so all inrush currents summing together cause voltage drop on this resistor that is decreases as filter capacitors (all at once) are being charged.
Now, I don't need this resistor anymore. I can short it out by relay.
But how do I know when to short it out?
Simply. As soon as voltage after filament voltage stabilizer is enough for the relay. No need for additional current sensors, delay time calculations, and so on. When, and only when a voltage is enough to click the relay on, it is time to short out inrush current limiting resistor. All voltages go up proportionally, all inrush currents are summed and limited by a single resistor.
Do you really need abrupt B+ start when cathodes are already hot, and tubes start conducting? I don't think so. Charging coupling capacitors cause grid currents that cause abnormal anode currents. Most probably you need soft start, so inrush currents in power supply are limited, and limited are currents that slower charge interstage capacitors.
One of the way how I do that, I use a current limiting resistor (usually 20 Ohm 12W wirewound one, for 120V in amps up to 1 KW of power consumption). It is in series with all sources: filament, anode, bias, etc..., so all inrush currents summing together cause voltage drop on this resistor that is decreases as filter capacitors (all at once) are being charged.
Now, I don't need this resistor anymore. I can short it out by relay.
But how do I know when to short it out?
Simply. As soon as voltage after filament voltage stabilizer is enough for the relay. No need for additional current sensors, delay time calculations, and so on. When, and only when a voltage is enough to click the relay on, it is time to short out inrush current limiting resistor. All voltages go up proportionally, all inrush currents are summed and limited by a single resistor.
Yes. A single resistor in series with input switch and a fuse. And 12V relay powered from output of 12V filament stabilizer. When relay is on it's contacts shunt out that resistor.
Well, I think im onto what your saying...
easiest option will be a relay with the aforementioned resistor on the NC contact and wiring the NO straight to the LOAD. Latching of the relay can be achieved with a string of zeners
The timer is for the filaments, giving the tubes time to warm up prior to applying the HV across the tubes and avoiding tube poisoning.
Last edited:
In order to avoid tube poisoning don't feed it by pesticides.
Actually, HV delay was used in radars and similar high voltage high current devices where tubes with direct filaments were used, to avoid cathode stripping
So I'm on the right track trying to protect my matched quartet of Golden Lions
I'm afraid your Golden Lions are too weak for radar output.
Good luck!
Radars are another story, these guys are not on tough budget usaully and do not pay from they personal pockets for sure. Separated B+ delay might prolong tube life imho.
Moreover I am owning Rogue Audio M120 they are for sure very nice sounding however heat generated by KT88 quartes are enermous. imho push-pull parallel EL84 + hieff speakers is a way more practical solution (like Manley Stingrey).
Moreover I am owning Rogue Audio M120 they are for sure very nice sounding however heat generated by KT88 quartes are enermous. imho push-pull parallel EL84 + hieff speakers is a way more practical solution (like Manley Stingrey).
I live in Alberta...
Last winter we hit -57.8C !!!
So the heat generated from the amp is welcome
might even save on my gas bill LOL
Well -57С that's must be an experience. I've been in Glaciers Park few years ago - almost Alberta I believe - but summer time luckily Мarvelous nature!
If you need more power Shuguang EL156 from China has nice reviews, or Russian GU72 another option. It might be sufficient just have two PP output tubes instead of quarter KT88, simpler design, less tubes to replace. Output tubes wear out pretty quickly.
However as it was mentioned above transformers and decoupling capacitors are the most critical part. If you can afford supermalloy core OT and Jensen foil decoupling capacitors go for them. That is where sonic magic is coming from in tube amps actually.
Мarvelous nature!If you need more power Shuguang EL156 from China has nice reviews, or Russian GU72 another option. It might be sufficient just have two PP output tubes instead of quarter KT88, simpler design, less tubes to replace. Output tubes wear out pretty quickly.
However as it was mentioned above transformers and decoupling capacitors are the most critical part. If you can afford supermalloy core OT and Jensen foil decoupling capacitors go for them. That is where sonic magic is coming from in tube amps actually.
Yes, radars are another story, but in this case not about budgets. Cathode filament stripping happens with directly heated tubes working on such voltages producing such impulse powers like radar tubes do. They usually do not have negative grid bias supply; negative bias the tube gets by rectification of input signal on the 1'st grid. Since filament starts heating up unevenly hottest parts of it start emission earlier than the rest of filament. Cathode-anode current is enough to heat up the filament even more, so hottest spots heat up more and more: both by filament current, and by anode-cathode current which density is high because of uneven emission. With increase of temperature resistance of overheated parts become higher, that increases temperature and emission. It causes destruction of filament.
Indirectly heated tubes are completely different story. Such effects don't exist in them: filaments are protected from anode - cathode current by cathodes, so anode-cathode current can't affect them. Current densities are much less, negative bias applied, voltages are much lower. Application of anode voltage before heater voltage can't harm such tubes in their regimes.
But what really can harm them, abrupt switch on of B+ when they are already hot. Remember that coupling capacitors are discharged!
Charging through grids they create huge peak currents. What happens if positive voltages applied simultaneously to grids of output tubes in P-P amps?
Right, maximum possible anode currents, until coupling capacitors are charged.
Is it called "to protect tubes"?
I don't think so...
Well -57С that's must be an experience. I've been in Glaciers Park few years ago - almost Alberta I believe - but summer time luckily
If you need more power Shuguang EL156 from China has nice reviews, or Russian GU72 another option. It might be sufficient just have two PP output tubes instead of quarter KT88, simpler design, less tubes to replace. Output tubes wear out pretty quickly.
However as it was mentioned above transformers and decoupling capacitors are the most critical part. If you can afford supermalloy core OT and Jensen foil decoupling capacitors go for them. That is where sonic magic is coming from in tube amps actually.
This is gonna be a slow build as i don't want to take shortcuts and i want to do my "homework" prior to starting the build...
Im lucky enough to have someone in town that deals with tube amp repairs even though he's doing instrument amps mainly, but has a surplus of Russian PIO caps ( i believe the "brick" he showed me was 470uF @ 1.5KV ) and also high end teflons as well
After extremely busy Spring and Summer here at work, I'm back on this
Power would be fed by 1 or 2 toroid's ( depending on how much current is required ) on each mono block. This would be my choice as they are readily available and cheap while quieter than their EI counterparts
ANTEK AS-4T400
2 x 400V @ 0.5A and 2 x 6.3V @ 4A on each toroid.
I would say its safe to say that 2 of these would even have enough juice to even run quad KT120's as long as the output transformer don't saturate.
The only other transformer remaining would be the bias and output transformers.
I want to get the iron together before i start building the cases.
Power would be fed by 1 or 2 toroid's ( depending on how much current is required ) on each mono block. This would be my choice as they are readily available and cheap while quieter than their EI counterparts
ANTEK AS-4T400
2 x 400V @ 0.5A and 2 x 6.3V @ 4A on each toroid.
I would say its safe to say that 2 of these would even have enough juice to even run quad KT120's as long as the output transformer don't saturate.
The only other transformer remaining would be the bias and output transformers.
I want to get the iron together before i start building the cases.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.