I wonder why in most circuits the 6.3V heater supply is fed with AC. The additional component count for a DC heater isn't that much, and it would save a lot of problems with hum. The only problem might be that if you go all the way for a regulated power supply, you end up with an odd voltage like 8.88V DC. But you could also work with an unregulated, but heavily filtered heater power supply; you can't get more than sqrt(2)*6.3V from rectifier.
Plus, when you're at it, you could also make a 'slow start' heater, which doesn't kick in directly with full voltage, which would prolong the lifetime of your tubes.
Correct?
Plus, when you're at it, you could also make a 'slow start' heater, which doesn't kick in directly with full voltage, which would prolong the lifetime of your tubes.
Correct?
You do sometimes get dc heaters used for DHTs, but the problem is that the dc heater voltage adds or subtracts from the grid bias, so you will get more emission from one side of the heater than the other.
I've not tried dc heaters on indirectly heated valves because if the heater wiring is twisted, and the feed centre-tapped, I've never found hum to be too much of a problem.
I've not tried dc heaters on indirectly heated valves because if the heater wiring is twisted, and the feed centre-tapped, I've never found hum to be too much of a problem.
You are right, a DC heater will eliminate hum from the filament circuit. The filament circuit is not the only source of hum in an audio amplifier. In fact for most power amplifiers it is pretty far down the list of hum sources, so many designers just take the simple approach and use AC filaments.
DC heaters are an advantage in preamplifiers where the signals are much smaller compared to the hum voltages. In a quiet phono stage DC heaters are almost mandantory.
The other area of current controversy is the use of DC heater voltage on directly heated tubes. Here there is no doubt that DC filaments are quieter, but there are disadvantages. When the filament is powered by DC there is a voltage gradient along the filament wire. In other words, one end of the filament is at a higher DC potential than the other. This means that the grid to filament (cathode) bias is higher at one end of the filament than at the other end. Therefore one end of the filament will conduct a greater share of the plate current than the other end. Operating the filament on AC will average the emission out over the length of the filament wire. Some claim that DC operation causes shorter tube life, and degraded sound.
From a theoretical standpoint it would seem that this effect would be worse on tubes that run on relatively high filament voltages, and have a high Mu, requiring a low bias voltage. In this case the filament gradient is a large percentage of the bias so that one end of the filament will be responsible for almost none of the plate current. The 811A or the 211 operating at low plate voltages (bias voltage near zero) comes to mind. Tubes like the 2A3 have a 2.5 volt filament and need -50 volts of bias, so this effect really doesn't matter.
I find that DC filaments are an advantage, and use them in most of my designs. There are some tubes, like the 833A, that just require it. The filament current is 10 amps, the magnetic field created by the 10 amp wiring gets into everything. You can hear hum in the speakers as soon as the filament transformer is energized, without plate or bias voltage!
There are a few designers that claim that tubes will last longer, and sound better when fed by a constant current source, rather than a constant voltage. I haven't been able to verify this, however a constant current source will by its very nature, be a slow turn on circuit, so tube life should increase.
DC heaters are an advantage in preamplifiers where the signals are much smaller compared to the hum voltages. In a quiet phono stage DC heaters are almost mandantory.
The other area of current controversy is the use of DC heater voltage on directly heated tubes. Here there is no doubt that DC filaments are quieter, but there are disadvantages. When the filament is powered by DC there is a voltage gradient along the filament wire. In other words, one end of the filament is at a higher DC potential than the other. This means that the grid to filament (cathode) bias is higher at one end of the filament than at the other end. Therefore one end of the filament will conduct a greater share of the plate current than the other end. Operating the filament on AC will average the emission out over the length of the filament wire. Some claim that DC operation causes shorter tube life, and degraded sound.
From a theoretical standpoint it would seem that this effect would be worse on tubes that run on relatively high filament voltages, and have a high Mu, requiring a low bias voltage. In this case the filament gradient is a large percentage of the bias so that one end of the filament will be responsible for almost none of the plate current. The 811A or the 211 operating at low plate voltages (bias voltage near zero) comes to mind. Tubes like the 2A3 have a 2.5 volt filament and need -50 volts of bias, so this effect really doesn't matter.
I find that DC filaments are an advantage, and use them in most of my designs. There are some tubes, like the 833A, that just require it. The filament current is 10 amps, the magnetic field created by the 10 amp wiring gets into everything. You can hear hum in the speakers as soon as the filament transformer is energized, without plate or bias voltage!
There are a few designers that claim that tubes will last longer, and sound better when fed by a constant current source, rather than a constant voltage. I haven't been able to verify this, however a constant current source will by its very nature, be a slow turn on circuit, so tube life should increase.
The definition of RMS is that it is the AC that gives the same heating power as the same voltage of DC. Thus, a 6.3V heater wants 6.3V AC RMS or 6.3V DC. The "designers" who slap a bridge rectifier followed by a 47,000uF capacitor onto a 6.3V winding and then feed the result to an EF86 would do well to note this fact.
I also find it's harder to get rid of the 120Hz noise plus transients than just 60 hum. I just bias the heater string about 30V about the cathodes. Proper lead dress really helps also!
If you want to run DC, then R-C-R-C can bleed voltage and tame the spikes. You can then bias this up in voltage. DHT's were really designed for AC heaters as noted above.
-Chris
If you want to run DC, then R-C-R-C can bleed voltage and tame the spikes. You can then bias this up in voltage. DHT's were really designed for AC heaters as noted above.
-Chris
I may be insane, but I don't like the way DC heaters sound. i put a little switching circuit in an amp to test the difference, and there was one... they didn't sound as "juicy" to use a technical term ;-)
I have had better luck with noise running an elevated reference voltage, like in the above post anyway, except in one really cramped layout where I did the ghetto, unregulated, 6.3v to bridge to large cap to .1r wirewound route, which gave me about 6.2v in that particular circuit.
I have had better luck with noise running an elevated reference voltage, like in the above post anyway, except in one really cramped layout where I did the ghetto, unregulated, 6.3v to bridge to large cap to .1r wirewound route, which gave me about 6.2v in that particular circuit.
Just an observation. In my line stage ECC88 preamp I am using AC heaters and I have absolutely no hum from it. I am also using all AC in my 807 parafeed amp which uses an ECC88 followed by a ECC81 into the 807 with a KT88 CCS load, again I am getting a tiny amount of hum. I have very sensitive tannoy speakers.
Then again I have a 5687 preamp which needed DC to tame the hum.
My default position is AC first and only use DC if absolutely essential. As said earlier the heater supply is low on the list of possable hum sources.
Shoog
Then again I have a 5687 preamp which needed DC to tame the hum.
My default position is AC first and only use DC if absolutely essential. As said earlier the heater supply is low on the list of possable hum sources.
Shoog
There you already have a totally interesting collection of thougts!
I would agree with Tubelab.com (post #3) that by and large filaments are not the greatest source of hum (transformer radiation is often worse!), at least not in my experience over dozens of tube amplifiers. You must not wire stupidly, of course. Often an arrangement where you earth through the slider of a small pot (few 100-ohm) will give you the best adjustment by trial-and-error - it is not always exactly in the centre.
I must stress that in my experience raw rectification-and-smoothing of the available 6V can give more problems than it solves. The resultant capacitor charging peaks are often sharper than sine wave sides, and in the end generate more noise than you would have had with AC. Unless you can smooth properly, don't go this route.
For pre-amps where one can serie the 12V 150 mA heaters it is a different matter - if you have the odd 50V available. If it is not blasphemy to you to use a semiconductor regulator (are you looking for results or appearance? - diodes are also semi-conductors) then some of the LM.....s available will yield excellent heater DC. I have used this for very low inputs, but the necessary voltage must of coarse be available.
I would agree with Tubelab.com (post #3) that by and large filaments are not the greatest source of hum (transformer radiation is often worse!), at least not in my experience over dozens of tube amplifiers. You must not wire stupidly, of course. Often an arrangement where you earth through the slider of a small pot (few 100-ohm) will give you the best adjustment by trial-and-error - it is not always exactly in the centre.
I must stress that in my experience raw rectification-and-smoothing of the available 6V can give more problems than it solves. The resultant capacitor charging peaks are often sharper than sine wave sides, and in the end generate more noise than you would have had with AC. Unless you can smooth properly, don't go this route.
For pre-amps where one can serie the 12V 150 mA heaters it is a different matter - if you have the odd 50V available. If it is not blasphemy to you to use a semiconductor regulator (are you looking for results or appearance? - diodes are also semi-conductors) then some of the LM.....s available will yield excellent heater DC. I have used this for very low inputs, but the necessary voltage must of coarse be available.
Hi All,
Not sure but I think it's all been covered, I guess 6.3V AC become the "standard" and when valves ruled no manufacture was going to try to rectrify the heater supply and feed the filaments with DC not practical.
I like the idea of constant current for the heavier filament current and 10amps is only a small load for some of the high power transmitter tubes where inrush current can be a 100+ amps and causes stress on the tube structures, this is usually protected by filament supply impedence but is a cause of failure of these tubes.
Just a point, filament voltage needs to be fairly accurate for optimum tube life and function, so if you want to run AC filaments from DC for low level stages regulate it! and as the other posts indicate magnetic induction would likely be your bigest problem noise source in the higher filament current tube circuits. for higher current filaments use balancing pots accross the filaments to ground.
One final point DC filaments were usually folded not twisted (the few examples I can remember) I guess in an attempt to equalize the cathode emission as indicated.
Not sure but I think it's all been covered, I guess 6.3V AC become the "standard" and when valves ruled no manufacture was going to try to rectrify the heater supply and feed the filaments with DC not practical.
I like the idea of constant current for the heavier filament current and 10amps is only a small load for some of the high power transmitter tubes where inrush current can be a 100+ amps and causes stress on the tube structures, this is usually protected by filament supply impedence but is a cause of failure of these tubes.
Just a point, filament voltage needs to be fairly accurate for optimum tube life and function, so if you want to run AC filaments from DC for low level stages regulate it! and as the other posts indicate magnetic induction would likely be your bigest problem noise source in the higher filament current tube circuits. for higher current filaments use balancing pots accross the filaments to ground.
One final point DC filaments were usually folded not twisted (the few examples I can remember) I guess in an attempt to equalize the cathode emission as indicated.
tubelab.com said:
Some DHT tubes were designed to expect DC of a specific polarity on the filament. That the grid was designed to accomidate the differing biases along the length of the filament. Tubes like 1U4, 1U5, 3S4 and such small portable battery radio tubes were made like this. Don't know if any power tubes you'd use in an audio amp feature this, though.
Don't see any problems with that if you don't have a 5 volt winding or a seperate transformer. other option might be to find a suitable tube with 6.3v filaments. other more knowledgeable members might have reasons otherwise but ohms law works in this application.
R series Vdif/I fil V
dif is 6.3-5= 1.3 volts
I fil is the filament current.
R series is the required resistance in ohms
now there will be some variation in filament current from tube to tube so final R would be set by measurement at the socket of the actual filament voltage after an appropreate warm up time 1 to 2 minutes.
A series regulator on the other hand with DC on the filaments would be constant volts. horses for courses but refer to previous posts
R series Vdif/I fil V
dif is 6.3-5= 1.3 volts
I fil is the filament current.
R series is the required resistance in ohms
now there will be some variation in filament current from tube to tube so final R would be set by measurement at the socket of the actual filament voltage after an appropreate warm up time 1 to 2 minutes.
A series regulator on the other hand with DC on the filaments would be constant volts. horses for courses but refer to previous posts
Hi Robski666,
Try not to generate heat under the chassis if you can avoid it. It's that simple. Any heat wasted will also generate heat in the transformer.
I have some preamps that were operated with DC heaters using 12AT7 and 12AX7. In series yet from a 25V source fed from a selenium rectifier (died). You will find that designers from the earliest times were very inventive. The resistance line cord attempted to run the chassis cooler. Ballast tubes provided a similar benefit.
-Chris
Try not to generate heat under the chassis if you can avoid it. It's that simple. Any heat wasted will also generate heat in the transformer.
I have some preamps that were operated with DC heaters using 12AT7 and 12AX7. In series yet from a 25V source fed from a selenium rectifier (died). You will find that designers from the earliest times were very inventive. The resistance line cord attempted to run the chassis cooler. Ballast tubes provided a similar benefit.
-Chris
The question should really be "why DC heaters".
DC heaters show up only infrequently in commercial designs, and usually only when they are vital. When you're competing, you don't spend money unnecessarily.
They are a luxury for DIYers that create a lot more work, and sometimes problems. With the exception of mic and phono input stages, I cannot find an advantage in using DC for indirectly heated valve heaters.
DC heaters show up only infrequently in commercial designs, and usually only when they are vital. When you're competing, you don't spend money unnecessarily.
They are a luxury for DIYers that create a lot more work, and sometimes problems. With the exception of mic and phono input stages, I cannot find an advantage in using DC for indirectly heated valve heaters.
WA2ISE:
Most tubes that were originally designed to operate on battery (obviously DC) power had the filament pins marked with a polarity. This includes some of the DHT's that find their way into modern audio equipment. You have to find a really old tube manual to see it mentioned though. When the AC world came to radio, this fact seems to be forgotten. I have seen mention of this effect in really old engineering text books (I have a few from the teens, and the 20's). There is mention that the filament to grid spacing was adjusted to account for the voltage gradient along the filament, and the filament pins were marked accordingly.
Dsavitsk:
Using a dropping resistor to arrive at a filaments correct voltage is actually a good idea, since it acts to limit the inrush current of a cold tube. Just remember that most 5 volt rectifiers have their filaments connected to the cathode. This means that your filament winding is connected directly to the B+ voltage of your amp. This is OK as long as you do not use the same winding to power other tubes in your amp. Most tubes can not take 300 to 450 volts filament to cathode. They will have a short unhappy life!
If you need to share the filament winding, use one of the many 6 volt rectifiers. If you need considerable power, two of the damper tubes work excellent. The other logical choice is a silicon rectifier.
Use of DC on a rectifier's heater is usually of no value, since the peak ripple voltage at the rectifier output is much larger than the filament voltage. Also the DC components (diodes, cap, regulator) will be operating at the B+ voltage, fried parts are easy to do!
Most tubes that were originally designed to operate on battery (obviously DC) power had the filament pins marked with a polarity. This includes some of the DHT's that find their way into modern audio equipment. You have to find a really old tube manual to see it mentioned though. When the AC world came to radio, this fact seems to be forgotten. I have seen mention of this effect in really old engineering text books (I have a few from the teens, and the 20's). There is mention that the filament to grid spacing was adjusted to account for the voltage gradient along the filament, and the filament pins were marked accordingly.
Dsavitsk:
Using a dropping resistor to arrive at a filaments correct voltage is actually a good idea, since it acts to limit the inrush current of a cold tube. Just remember that most 5 volt rectifiers have their filaments connected to the cathode. This means that your filament winding is connected directly to the B+ voltage of your amp. This is OK as long as you do not use the same winding to power other tubes in your amp. Most tubes can not take 300 to 450 volts filament to cathode. They will have a short unhappy life!
If you need to share the filament winding, use one of the many 6 volt rectifiers. If you need considerable power, two of the damper tubes work excellent. The other logical choice is a silicon rectifier.
Use of DC on a rectifier's heater is usually of no value, since the peak ripple voltage at the rectifier output is much larger than the filament voltage. Also the DC components (diodes, cap, regulator) will be operating at the B+ voltage, fried parts are easy to do!
Tubelab
damn good call, i hadn't had to consider tube DC supply rectifiers disposition re potential above ground for a looong time I should have
Chris,
good point on thermal management, that however is the black art that we all have to deal with.
I can remember the smell those selenium rectifier use to release with the smoke!
Robert
damn good call, i hadn't had to consider tube DC supply rectifiers disposition re potential above ground for a looong time I should have
Chris,
good point on thermal management, that however is the black art that we all have to deal with.
I can remember the smell those selenium rectifier use to release with the smoke!
Robert
> I wonder why in most circuits the 6.3V heater supply is fed with AC. The additional component count for a DC heater isn't that much
The last time people USED tubes (rather than just play with them), getting heater-quantity AC was easy, DC was tough.
Though the 6.3V convention IS the voltage of a car battery (back when they were 6V).
> when valves ruled no manufacturer was going to try to rectrify the heater supply and feed the filaments with DC not practical.
It was occasionally done after copper-oxide rectifiers and low-volt high-uFd electrolytics came along.
Though of course the BBC used batteries for everything from the start until long after WWII. Yes, high-voltage too! In broadcast centers the batteries were fed from motor-generators to keep them charged. A lot of outside broadcast gear was just battery.
> make a 'slow start' heater..., which would prolong the lifetime of your tubes.
> a few designers that claim that tubes will last longer, ..... when fed by a constant current source ... I haven't been able to verify this
Few of us have been around long enough long enough to have ANY valid data on tube life, at least for the good tubes from the 1950s and 1960s.
> the high power transmitter tubes where inrush current can be a 100+ amps and causes stress on the tube structures
Yes, the giant-size industrial tubes have problems we don't have to think about in receiving tube work.
> Some DHT tubes were designed to expect DC of a specific polarity on the filament. That the grid was designed to accomidate the differing biases along the length of the filament. Tubes like 1U4, 1U5, 3S4
The beach-radio tubes also have special needs (very different from transmitter tubes). You don't want 4 A batteries for a 4-tube radio, but if you feed them all from a common A battery you can't use self-bias on each tube. In home/farm radios it was common to use a C battery for grid bias, but that's an extra lump in a beach radio. The low-level tubes would all work with less than 1.5V grid bias, so in effect the A battery was also a C battery (if you got it the right polarity).
> Proper lead dress really helps also!
This is a lost art, and explains much of the modern fad to fool around with DC heat. As Shoog says, at line level with heater-cathode tubes, hum should not be any problem at all, if the wiring is done right.
> Try not to generate heat under the chassis if you can avoid it.
Very very true. My recollection is that classic tube gear did not quit from tube abuse, but from heat drifting the resistors and baking the capacitors. While we replaced a lot of tubes just because it was easy (and cheap TV sets did kill tubes), most failures were not tubes. And of those, NOT counting cheap series-string TV sets, almost none were heater failures.
The last time people USED tubes (rather than just play with them), getting heater-quantity AC was easy, DC was tough.
Though the 6.3V convention IS the voltage of a car battery (back when they were 6V).
> when valves ruled no manufacturer was going to try to rectrify the heater supply and feed the filaments with DC not practical.
It was occasionally done after copper-oxide rectifiers and low-volt high-uFd electrolytics came along.
Though of course the BBC used batteries for everything from the start until long after WWII. Yes, high-voltage too! In broadcast centers the batteries were fed from motor-generators to keep them charged. A lot of outside broadcast gear was just battery.
> make a 'slow start' heater..., which would prolong the lifetime of your tubes.
> a few designers that claim that tubes will last longer, ..... when fed by a constant current source ... I haven't been able to verify this
Few of us have been around long enough long enough to have ANY valid data on tube life, at least for the good tubes from the 1950s and 1960s.
> the high power transmitter tubes where inrush current can be a 100+ amps and causes stress on the tube structures
Yes, the giant-size industrial tubes have problems we don't have to think about in receiving tube work.
> Some DHT tubes were designed to expect DC of a specific polarity on the filament. That the grid was designed to accomidate the differing biases along the length of the filament. Tubes like 1U4, 1U5, 3S4
The beach-radio tubes also have special needs (very different from transmitter tubes). You don't want 4 A batteries for a 4-tube radio, but if you feed them all from a common A battery you can't use self-bias on each tube. In home/farm radios it was common to use a C battery for grid bias, but that's an extra lump in a beach radio. The low-level tubes would all work with less than 1.5V grid bias, so in effect the A battery was also a C battery (if you got it the right polarity).
> Proper lead dress really helps also!
This is a lost art, and explains much of the modern fad to fool around with DC heat. As Shoog says, at line level with heater-cathode tubes, hum should not be any problem at all, if the wiring is done right.
> Try not to generate heat under the chassis if you can avoid it.
Very very true. My recollection is that classic tube gear did not quit from tube abuse, but from heat drifting the resistors and baking the capacitors. While we replaced a lot of tubes just because it was easy (and cheap TV sets did kill tubes), most failures were not tubes. And of those, NOT counting cheap series-string TV sets, almost none were heater failures.
All true. The only audio tubes commonly in need of replacement were the output tubes. The rectifier occasionally and sometimes, a signal tube. Heck, contact cleaning before signal tubes.
Most times when I rebuild something it's resistors and caps. That and the cheaper tubes available today. New designs tend to abuse everything, like they didn't have a tube manual.
-Chris
Most times when I rebuild something it's resistors and caps. That and the cheaper tubes available today. New designs tend to abuse everything, like they didn't have a tube manual.
-Chris
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.