Hoffman Amplifiers Tube Amplifier Forum
Amp Stuff => Tube Amp Building - Tweaks - Repairs => Topic started by: jjasilli on May 15, 2018, 11:42:22 am
-
Interested in limiting inrush current, both:
* HT, and
* Filament Supply
Any recommended solutions? Is there as preferred method? How to spec for voltage & current? Any published works for tube amps? Can't find good info.
For HT in a guitar amp I would not want the PS to become regulated.
For filaments KOC, TUT1, recommends a time delay relay circuit. Ch 2, pp 15 -16. Seems overly complicated and hogs space. Would an MOV or thermistor work?
-
This should get you started...
https://www.google.com/search?source=hp&ei=JBD7WpegJcPXzwKV1o2oDg&q=thermistor+current+limiter&oq=thermistor+current+limiter&gs_l=psy-ab.3..0j0i22i30k1l4j0i22i10i30k1.1956.30029.0.33470.26.25.0.0.0.0.1792.4546.7-2j1.3.0....0...1.1.64.psy-ab..23.3.4542...0i10k1.0.Pk1mQxvCVyo
Basically all you need is a NTC thermistor connected in series with the PT primary. Many of the old series filament tvs had a thermistor in series with the filament string.
-
Our friend Trobbins published something about
(http://el34world.com/Forum/index.php?action=dlattach;topic=23036.0;attach=69659;image)
http://el34world.com/Forum/index.php?topic=23036.msg247007#msg247007 (http://el34world.com/Forum/index.php?topic=23036.msg247007#msg247007)
https://www.dalmura.com.au/static/HV%20DC%20shunt%20load.pdf (https://www.dalmura.com.au/static/HV%20DC%20shunt%20load.pdf)
http://www.ti.com/lit/ds/symlink/tl431.pdf (http://www.ti.com/lit/ds/symlink/tl431.pdf)
Ciao
Franco
-
Thanks guys, but I'm still at "square one":
1. @sluckey: the general components I can find. But which one to select for the for the job @ hand: HT?; Heaters?
2. @kagliostro: two things:
a) Do the shunt regulators regulate B+ after the inrush period?
b) Re the Dalmura article: ANOTHER OPTION (bottom of web page). This flatly contradicts Tubenit's recent thread on Standby Switches. The Forum consensus there was that throwing a Standby SW (whether manual, or automated by a time delay relay) always causes a high current inrush, even after the heaters are warmed up. This article says the opposite. What's the truth?
-
I'm probably not understanding exactly what you want to do. But, I've used this "soft start" switch and it seems to work well. I don't know why you couldn't use the same thing for the filaments.
http://www.valvewizard.co.uk/standby3.jpg
-
Ciao JJasilli
a - I never experimented the circuit but reading the .pdf file, to me seems that there is negligible B+ control after the inrush period
it will be better to ask to Trobbins
b - also about this question it will be better to discuss with the author
Franco
-
@ John. Re filaments. They are a clever thing. Whether in an incandescent light bulb or a vacuum tube, filaments need to get so hot they glow. Also they are a type of resistor so they put a load on a circuit. Typical for resistors, the hotter they get the more resistance they have. But at turn-on they're cold and are darn near a short circuit with very low resistance. So they draw huge current. The trick is to get the resistance to match the flash of heat at turn on. Otherwise they burn out with no useful life. It took Edison years to get a working filament.
Anyway, the filament can't prevent the huge current inrush at turn on. It needs outside help. I'm trying to learn how to spec out such a device. The time lag here is as much as 30 seconds.
Re HT. There's a short time lag until the 1st filter cap charges and the power tubes start drawing current. This too causes a large inrush current at turn on.
@ kagliostro. I was responding to your reply, but only hoping you personally knew the answer. I think this might have inadvertently re-opened the Standby SW issue. I already have a 30 second time delay relay on my 100W tube amps, which the Dalmura article says is good.
If no one chimes in I'll post some PM's.
-
Is the concern only about power turn-on mains inrush current? Is the concern because mains AC circuit breakers are being tripped, or fuses blowing?
The circuit and article that K referenced are applicable to choke input filter B+ supplies, where B+ can initially be a lot higher than nominal - especially when using ss diodes or directly heated cathode diodes. They aren't directly related to inrush current concerns.
Inrush current is not usually a problem looking for a solution. A standby switch may or may not be related - that would depend on exactly the amp/circuit in question.
Inrush is a technical consideration when designing the AC fuse value and type, and if an NTC thermistor is being used. The linked doc discusses the technical design aspects of those two parts:
https://dalmura.com.au/static/Valve%20amp%20fusing.pdf (https://dalmura.com.au/static/Valve%20amp%20fusing.pdf)
Ciao, Tim
-
Inrush current is not usually a problem looking for a solution.
Ok, I'll consider it a non-issue. (I don't have any choke input amps.) Thanks!
-
Inrush current is not usually a problem looking for a solution.
I was looking for that solution awhile back, I was up against cap spec'd voltage +60-80vdc
I looked into a TD relay keeping a DL R in circuit for 30sec-ish. I solved the issue with proper caps. (900vdc :icon_biggrin:)
-
My well-pump start-up pulls 44 Amps, on one 120V leg of a 240V "100A" line.
Lights dim (and brighten!).
The wires are only rated 15A steady.
I see no reason to "inrush limit" that. It has worked that way since 1985.
Your amplifiers are not drawing 44 AMP inrush every time you flush.
-
I was wondering about improving tube life, especially in amps where tubes are stressed with hi voltage -- like 2X KT-88's @ 600V > 100W.
-
Filaments are rarely the part of the tube that fails, so I wouldn't worry about it.
Household incandescent lights use high voltage, requiring thin/long filaments for high resistance, which I believe makes them fragile. "Modern" (indirectly heated) tubes require lower temperatures than lightbulbs and older tubes, due to cathode material with good emission at lower temperature, which improves useful life and decreases chances of total breakage/failure.
As a result, I've dealt with maybe 10 bad tubes and only one of those had a failed filament. It was an old tube kicking around in the bottom of a combo amp for who knows how long before I got it, and I have every reason to think it was physically damaged, not failed due to heating/cooling cycles.
-
Watch this video from Blueglow Electronics. Explains why, how and where to use thermistors on valve amps.
A negative coefficient thermistor starts off cold at a high resistance (typical range between 10 ohms to 120 ohms ). When the mains electricity is switched on the thermistor begins to heat itself up, as it heats up its resistance drops away to almost nothing (takes about 2 seconds to gradually allow full current into the mains transformer). This action limits the inrush current and gradually increases the current flow into the HT and the fillament circuits.
Thermistors are normally installed just before the power transformer primary winding, inline with the positive lead after the mains switch and the fuse. They are cheap and simple to install, they take up very little room and the gradual start protects transformers, fillaments and the HT circuit filter caps. Especially a good idea in smaller amps diode rectified (so instant full power) and no standby switch. Before the valves get hot and start conducting current to the plate and away from the fillaments the voltage levels on the capacitors across the whole amp is scary to see on a multimeter. 400V filter caps can see plus 400V for a few seconds.
Be aware thermistors do get hot so fix them with some airflow space especially between the thermistor and other components and wiring.
-
Good video.
I installed a CL-90 in the stereo hifi EL34 amp I built last Spring (Tubelab SSE). It's in series with the black (line) primary of my PT.
I've been thinking about these components because I recently finished a guitar amp (AA864 Musings) that spikes up to 498VDC or so on the B+ nodes until the caps charge and the tubes heat up (several seconds of this higher voltage).
Also, my AC wall voltage is around 123VAC. I'm using 500V filter caps, but the inrush spike voltage is right there at their specced limit. As Mark Walker states in the Blueglow video above, it can't hurt to use one. So, if installing one could lower my mains voltage closer to 120VAC, and take some of the initial voltage slam off of my filter caps and tubes it sounds like a good idea to me.
-
In my view, this is a solution looking for a problem.
Design elegance --> simplicity.
Not sayin' that inrush current limiting is not ever an issue. It is. But, not for the majority of amps discussed on this forum.
http://www.valvewizard.co.uk/standby.html
-
...not for the majority of amps discussed on this forum. ...
Agree.
Few-second surges on e-caps are covered by specification. It takes time to kill an e-cap with overvoltage, and start-up surge is old news. Frequently a 450V cap was rated for 525V for 30 seconds a few times a day.
When electric trolley-cars were proposed some pundits said you could not have two on the same track because start-up would fry everything. Frank Sprague put a half-dozen(?) trolleys on track and started them 1-2-3 on 30(?)-second intervals. The world did not end, nobody even noticed the added loads.
Do like Dr Frankenstein and Igor. Throw the switch! Your monster won't mind.
-
When I was a newbie in my career, I was sent out to cut out ALL the "Blue thermistors". directions in hand, hourly employee got me at least 80hrs OT :icon_biggrin:
I ask, was told; the FU :cussing: ing things keep blowing up and scaring the Sheet outta the techs and patients.
they sounded like an 'ol skool "black-cat" fire cracker
-
Points well taken. :icon_biggrin:
I did choose not to use a Standby switch after reading Merlin's thoughts, along with other discussions on the web.
But, I have to confess that being brand new to tube amp building I'm overly worrisome about my first creations. You want the very first ones you make to live and not die in infancy.
Though, the more I learn about all of those old Fender amps from the the past that ran the tubes beyond their ratings (like 5881's at 470V in the 6G6A, for example), and they didn't explode into fireballs I guess I'm just going to stop worrying.
-
tubes are pretty damn resilient.. if they can survive EMPs from nukes theyll handle inrush over 30 years.
Solid state rectifiers are more susceptible. They are all relatively cheap to replace. If youve properly fused the HT you should be good to go. Thats why we use slo blow fuses.
-
I'm probably not understanding exactly what you want to do. But, I've used this "soft start" switch and it seems to work well. I don't know why you couldn't use the same thing for the filaments.
http://www.valvewizard.co.uk/standby3.jpg (http://www.valvewizard.co.uk/standby3.jpg)
Yep - I use 180k 5W
-
A nice thing with inrush current surge suppressors is that they allow fast, rather than the typical slow blow fuses to be used.
Hopefully, in the case of a fault, that should result in the exposure to fault current being less prolonged before the fuse blows.
-
My limited knowledge and understanding leaves me with some questions and observations here.
It appears there are 2 separate and distinct albeit conjoined issues that are being addressed here.
1. Inrush current. Measured in amps.
2. Voltage regulation. Measured in volts.
As I understand it, inrush current is a phenomenon observed when a transformer or a capacitor is initially energized. When a transformer is energized, it can draw a large amount of current, perhaps several times its rated capacity until its core is magnetized and it operating resistance is stabilized. A capacitor does similar until it is fully charged.
This inrush current is typically allowed for in the design of the transformer and all downstream components, otherwise, there would be large scale component failures. The primary concern of current inrush is not downstream components but rather upstream components. If the supply for the device is not capable of carrying the large amount of current required instantaneously when the device is brought on line problems can occur. 1) a breaker or fuse can trip or blow, 2) a conductor can burn or overheat, 3) the generation source can be overloaded. A combination of these events may occur, not in the particular order listed.
A breaker tripping or fuse blowing is pretty typical and obvious. Many breaker manufacturers allow for short term transients and some even have adjustments for thresholds. A slow-blow fuse is a primitive time delay overcurrent protection.
Another symptom of inrush current is an initial voltage sag followed by a voltage surge. As a switch allows electrons to flow into the transformer or capacitor, the initial low resistance creates a voltage sag. As the transformer and capacitor reach operating condition, resistance is restored and the inrush current comes to a halt with a short voltage rise resulting before the circuit becomes stable as the conga line of electrons that has been rushing to fill the bar suddenly gets the door slammed shut and they all plow into each other.
Protection from these events must be addressed from both current and voltage perspectives if your goal is to extend the life of equipment. As Sluckey indicated, current inrush is most often addressed by use of Thermistors. Thermistors however are not voltage regulators and wil do nothing to prevent sag or surge of voltage in the circuit. To address Voltage sag, a capacitor or battery storage bank is typically used, and to protect from a surge a MOV is typically the most cost effective solution. A well designed and well regulated power supply typically incorporates a Breaker or fuse, a transformer or voltage regulator, a capacitor, and an MOV. In some cases to maximize protection, MOV's are installed both before and after the transformer.
Unless you are located in a third world country or California, where the grid is unreliable and dynamic, these concerns are probably not something you need to lose sleep over when applied to tube amplifiers. One of the beauties of Tube gear is their relatively large tolerance of dirty power and reasonable voltage swing.
I would liken this to an first time expectant mother. If a mother has never birthed a child before, she will think it requires a great amount of diligence and preparation, every last detail has to be planned and a lot of books are read, and doo dads purchased to create the theoretically perfect nest. Diet must be perfect, vitamins taken, doctors visited, birthing classes taken, and baby showers held. Despite all the hand wringing and consternation exhibited by first time mothers, birthing a child is something that happens all the time, most often with little fanfare, often in the most primitive and less than elegant manner and there is little empirical evidence to suggest that a well planned sanitary hospital birth is in some way superior to a birth in the backseat of a car when it comes to the long term effect on the child. General guidelines are sufficient, don't drink, smoke, or do drugs while pregnant, and don't engage in risky behavior before and chances are things will turn out just fine.
If you want that extra peace of mind that having the best OB/GYN care, a sanitary hospital bed and a staff on onlooking spectators can bring, then there is no compelling reason not to spend the extra money but it is still no guarantee that nature will not decide on an alternative course.
Most 3rd or fourth time mothers will likely tell you they went way overboard on their first pregnancy and to save your money for more important things.
-
Some of this is artifacts from the distant past. I think that way back, tubes and caps were relatively expensive. Interstage trannies were used instead of tube stages, until the price differential flipped. Also, in some places tubes and caps may have been hard to get. Protection against inrush current, even by way of a simple, current-limiting, series resistor was a more compelling need.
Another consideration is the filaments, which tend to fail first, even in incandescent light bulbs, due to inrush current at turn-on.
Even today, if you're spending several hundred dollars for a power tube, you may be motivated to take all possible steps to preserve its longevity. Hence, the inrush current concern is usually a hi-fi (not guitar amp) thing. Still, KOC devotes considerable space to, and recommends, soft-start circuits in guitar amps.
-
I'm probably not understanding exactly what you want to do. But, I've used this "soft start" switch and it seems to work well. I don't know why you couldn't use the same thing for the filaments.
http://www.valvewizard.co.uk/standby3.jpg (http://www.valvewizard.co.uk/standby3.jpg)
Yep - I use 180k 5W
Is there any reason why a person would not want to use a higher value resistor for the standby switch bypass resistor? Perhaps something along the lines of a 470K, assuming that you can find one of those with at least a 2 watt rating...
I don't know if the extra resistance would make much difference in slowing down the inrush current...
-
So, if I understand this correctly (a questionable assumption), the standby switch bypass resistor simply allows the downstream power supply to be energized at a lower voltage when the switch is "off." When the switch is "on," the restore is bypassed, and the downstream voltage is allowed to rise to operating conditions. This simplistic view assumes that the nodes have a sufficient path to ground that will not allow voltage to rise to nominal voltage when the standby switch is off.
The question becomes: What voltage do you want your nodes to be carrying while in standby?
-
...Another consideration is the filaments, which tend to fail first...
Really?
In my experience, a 'dead filament' type tube failure is a vanishingly rare occurrence.
Anyone else care to comment?
-
The only tube filament failures I've ever seen were in a series filament string such as an all American 5 tube radio, or an old BW tv with series filaments.
-
So, if I understand this correctly (a questionable assumption), the standby switch bypass resistor simply allows the downstream power supply to be energized at a lower voltage when the switch is "off." When the switch is "on," the restore is bypassed, and the downstream voltage is allowed to rise to operating conditions. This simplistic view assumes that the nodes have a sufficient path to ground that will not allow voltage to rise to nominal voltage when the standby switch is off.
The question becomes: What voltage do you want your nodes to be carrying while in standby?
I believe the standby switch bypass resistor does not limit the voltage going to the other nodes so much as it limits the inrush current so the nodes charge up slower than they would when you flip the switch. There's probably a bit of voltage drop across that resistor until you turn the standby on, but the node capacitors should be mostly charged up by then.
-
Interesting. So the standby switch really would not be a standby switch at all . . .
LOL. The graphic of that resistor came from Merlin's pages, and I am pretty sure he does not believe in standby switches.