Constant current tube heating

What do you think about feeding the tube heater with a constant current source instead of a constant voltage source?

Let's say an ECC88 will take 6,3V at 365mA. If the heater is cold at startup it will draw much more current then after warming up. A constant current source will limit this "cold current" to 365mA maximum.
You could also put a few heater in series.

Any downsides?

Any downsides?

Click to expand...

Increased powersupply complexity?

[quote author="gyraf"]

Any downsides?

Click to expand...

Increased powersupply complexity?[/quote]

Yes, one additional resistor. :wink:

There are drawbacks. If the filament begins to deteriorate and develops hot spots (in regions of higher resistivity), or even just uniformly higher resistance overall, then you will tend towards a thermal runaway condition and hasten burnout, especially if the current source has a fairly high complaince voltage.

The optimal regulator for filaments IMO is one that regulates power. This can be done exactly by using an analog multiplier along with current sensing and a reference to regulate voltage, or done as a very good approximation with a synthetic matched impedance model. There was a brief mention of this last approach in AudioXpress a while back. I will try to find the reference.

EDIT: and if you were to use a three-term reg for a constant current source, use a 317 not a 7806. With the latter you are dissipating a whole bunch in the resistor.

it's all the rage with DHT (where the heater PSU certainly DO play a major part in the sound). i favour AC here. i think it's a waste with other tubes though

i use entirely passive DC heater PSUs with a boatload of capacity behind. my studio is at 225VAC ALL the time, so i don't need tight voltage regulation, which is important, the dif between 5,8 and 6,4VDC can make or break some designs. i don't like putting heaters in series and the huge start up current can bring some Vregs to their knees true enough. 78XXs suck anyway

there is a lot talk about that subject at diyaudio.com

[quote author="sismofyt"]...snip...
there is a lot talk about that subject at diyaudio.com[/quote]

Trying to lure us to the Dark Side of the Force ehh?

no, bewteen all the audio phoolery there is some VERY competent tube designers there, more so than here (yes, PRR is everywhere) and they're not so conservative as the pro audio world

AKG does it on the tubes in their mics. It slows down the warmup phase of the tube so you'll be waiting a few minutes longer to get sound, and a few minutes longer waiting for the clicking and popping as the tube warms up. It does make filtering the DC (to get the hum out) a bit easier, though.

You pretty much need DC filaments in tube mics - and with a lot of filtering at that.

oh yeah, slow ramping will help those precious DHTs last longer. a very valid reason, if you're running a quartet of WE300Bs

in AKGs case it doesn't matter with those shite mics they're putting out these dayz

no, bewteen all the audio phoolery there is some VERY competent tube designers there

Click to expand...

I second that. Lurking there you may find very interesting threads!

yeah, i have wet dreams about this forum merging with the diyaudio. we could get into some really deep tube (AND transistor) stuff and they could stop using magic water, wooden knobs, pure diamond cables and what not

we're actually not that alien to each other. it's all about great sound

[quote author="sismofyt"]in AKGs case it doesn't matter with those shite mics they're putting out these dayz[/quote]

Does that include the C60 and C61? These days, maybe, but those plus the C12A used constant current heating.

i hate all AKG mics with a passion, thin metallic sounding shit from C1000 to 414 with a general unpleasant tone. the old C12 is the most hyped op mic i've ever heard. first time i thought it was broken or just really outta shape. not so. don't like the C60/61 either

so maybe i'm a little bit biased

the 421 is the exception that proves the rule

> limit this "cold current" to 365mA maximum.

So?

> If the filament begins to deteriorate... hasten burnout

When was the last time we saw an indirect-heat receiving tube heater burn out? (I agree that $300 DHTs are a different beast.)

There was my H-P 200AB, but I suspect abuse (toss into dumpster) even more than the decades of power-on hours it lived through.

Before that, I have to remember series-string TV sets, in which the small tubes got "constant current" from the large tubes. If you got tubes from opposite ends of the current-draw spec, the big tubes over-fried the little tubes. That don't happen in constant-voltage feed.

I think constant current drive is a bad idea. Constant-voltage drive is fairly self-limiting for a filament with resistance that increases with temperature, constant current leads to much wider deviations.

If you think the start-up surge hurts tubes, try it. Put a tube on a 10 second timer, off-on-off-on..., and wait for it to fail. I bet a buck you can't kill a tube in a decade this way; one failure is a fluke. I'll bet $20 that you can not kill two tubes in one year this way. Plain receiving tubes like 12AX7 or 6DJ8 or 6L6, not battery tubes, not WE300B. Your choice of timing: maybe 2 seconds and more cycles, maybe 20 seconds so the heater goes full-cold on each start, I don't care. Total population say a dozen (I don't want to hear that you abused 10,000 tubes and two failed). Open-heater or H-K-short failures only. 1935-1985 US/EU/Japan production or a good USSR factory, not the worst Chinese low-bid junk.

C12 is the most hyped op mic i've ever heard

Click to expand...

I also used to think like this... I´ve tried it in percurssive instruments, vocals, and bass and it sucks on all those sources... But you must hear it in a string ensemble. WOW! Best mic ever, for this.

For all the other sources, a ELA-M251, a U87 or even any SP mics with chinese 797 capsules wuld be MUCH better than a C12.

Also, those vintage mics varies a lot, but in the studio I use to work, they have six C12...

[quote author="PRR"]> limit this "cold current" to 365mA maximum.

So?

> If the filament begins to deteriorate... hasten burnout

When was the last time we saw an indirect-heat receiving tube heater burn out? (I agree that $300 DHTs are a different beast.)

There was my H-P 200AB, but I suspect abuse (toss into dumpster) even more than the decades of power-on hours it lived through.

Before that, I have to remember series-string TV sets, in which the small tubes got "constant current" from the large tubes. If you got tubes from opposite ends of the current-draw spec, the big tubes over-fried the little tubes. That don't happen in constant-voltage feed.

I think constant current drive is a bad idea. Constant-voltage drive is fairly self-limiting for a filament with resistance that increases with temperature, constant current leads to much wider deviations.

If you think the start-up surge hurts tubes, try it. Put a tube on a 10 second timer, off-on-off-on..., and wait for it to fail. I bet a buck you can't kill a tube in a decade this way; one failure is a fluke. I'll bet $20 that you can not kill two tubes in one year this way. Plain receiving tubes like 12AX7 or 6DJ8 or 6L6, not battery tubes, not WE300B. Your choice of timing: maybe 2 seconds and more cycles, maybe 20 seconds so the heater goes full-cold on each start, I don't care. Total population say a dozen (I don't want to hear that you abused 10,000 tubes and two failed). Open-heater or H-K-short failures only. 1935-1985 US/EU/Japan production or a good USSR factory, not the worst Chinese low-bid junk.[/quote]

Well, maybe with tubes it is not as severe a problem. Brad Plunkett did get burned badly years back, according to his glowing account (sorry somebody stop me!) by driving pilot lights with constant current.

[quote author="sismofyt"]yeah, i have wet dreams about this forum merging with the diyaudio. we could get into some really deep tube (AND transistor) stuff and they could stop using magic water, wooden knobs, pure diamond cables and what not

we're actually not that alien to each other. it's all about great sound[/quote]

I was really mostly kidding. I haven't spent much time there but I did find one thread (about adding a current source to a Jung super regulator or something) to sound positively like medieval scholastics arguing about angels on pinheads. No one, neither legends nor lesser folk, was willing to let the slightest detail go undiscussed. On the other hand Bruno Putzeys's comments about the nature of so-called "digital" amps were a breath of very fresh air indeed.

EDIT: speaking of the (off-topic) devil, check out that forum's still-active thread: Difference betweeen Class D and "Digital Amplifiers"

More on-topic, note that the latest thread in this forum, requesting recommendations of headphone amps, has a link to an article about synthesizing a given output impedance amp. This, seen as a power supply, is the way to get approximately constant power into a tube heater. The effective output resistance should match the warmed-up resistance of the filament.