Best way to step down 12v > 6v to power up the heater of valves?

[nguitar12]

Just wonder what is the best way to step down 12v > 6v to power up the heater of 6111 valves. LM317 seem generate a loy of heat. Currently I am looking at transistor zener regulator can someone please tell me if this will work for what I need?

[R.G.]

As I always do when presented with the question "what is the best...?", I will reply with "what do you mean by the word "best"? If you don't know what you mean by "best", you're really unlikely to get that.

First, vacuum tubes expect their heater voltage to be held to about ±10% variation for specified performance and longest expected life. Getting to better than 10% regulation is probably unnecessary. If you start with a *regulated* 12V supply that will vary less than ±10%, all you need is a resistor to drop the excess voltage.

The 6111 needs 300ma of heater current. So if you have 12V and want 300ma to flow into a 6V heater, the resistor is just the difference in voltages - 6V, by accident - divided by the current, or 6V/0,3 = 20 ohms. This resistor will dissipate 1.8W.

You could use your transistor regulator, OK, but now things get complicated. The only reason to use a regulator is to make up for an unregulated "12V" supply. So to do an OK (note - not "best") job, you have to know what the range of the "12V" supply is for all variations of line voltage and loading. It's worse than ±10%, or we would not be doing this exercise, right?    So what is it? You can't do a decent job until you know this. Well, you could just count on getting lucky, but I've gotten lucky many fewer times in my life than I wanted to.   

That being said, you now have to *design* the regulator. This involves knowing the transistor gain, the Vbe, the power dissipation (because something is going to eat that 1.8W!) and some other things.

If you can stand it, one way to avoid the morass of details is to use a 7806 three terminal regulator. This will still require that you make sure that your "12V" is at least 8V at the full 300ma loading, but it relieves you of many other issues, excepting heat sinking.

[nguitar12]

As I always do when presented with the question "what is the best...?"

Here is the factor I consider about in order:
circuit performance > heat problem > part count

I will accept the component in this circuit at can be touch by hand (not too hot) and I want to keep the part count low. Just wondering with approach will generate the least heat? LM317/transistor zener regulator/ 4W 20ohm resistor?

Thanks in advance.

[antonis]

Just wondering with approach will generate the least heat? LM317/transistor zener regulator/ 4W 20ohm resistor?

As R.G. clearly stated, the power dissipation will be 1.8 Watts (which will be dissipated in the form of heat) so there isn't matter of device type or anything else - ANY of them will have to dissipate 6.147322 BTU/h..

[nguitar12]

Just wondering with approach will generate the least heat? LM317/transistor zener regulator/ 4W 20ohm resistor?

As R.G. clearly stated, the power dissipation will be 1.8 Watts (which will be dissipated in the form of heat) so there isn't matter of device type or anything else - ANY of them will have to dissipate 6.147322 BTU/h..

Sorry for being nub. So you mean that they will generate same amount of heat?

[antonis]

So you mean that they will generate same amount of heat?

If you don't take advantage of the "excess" power, yes..!! 

[GibsonGM]

Dalton's law...I would be sure the 12V is well regulated, and use the dropping resistor.  Just how I roll; it's simple...

[anotherjim]

Is it not practical to wire the heaters in series pairs and if there's an odd one just have a resistor for it?
Can't we trust the transformer voltage well enough? Honestly, I don't know much about tube amps, but I've never seen anything that didn't rely on the transformers innate regulation quality and something like this...

Nowadays, you might only see at most 2 choices of supply selection for your part of the world, but they used to be very comprehensive and to non-technical folk, incomprehensible, so there must have been many amps rocking away with less than ideal voltages.

[amptramp]

You should look at the failure modes of whatever design you want to use.  It is theoretically possible for a resistor to fail short - I have just never seen it and I have been doing this for a long time.  If the ground pin on the regulator lifts, you get almost full voltage (minus a Vbe drop) applied to the filament.  Death within milliseconds (and I did fry a digital board once when a regulator ground pin went bad).  Go with the resistor.  It also has the advantage of limiting inrush current since the resistance of a cold filament is a lot less than a hot one.

[PRR]

> rely on the transformers innate regulation quality

Transformers don't regulate (except some oddballs).

Your Power Company regulates the 117V (or whatever) so that your Incandescent Lamps do not run dim or burn-up in a few hours.

I've hollered at my electric company when lamps were awful bright and burning-out too frequently. They had tapped for a pretty steady 127V, in days when lamps targeted 120V.

If your wall-voltage is close-enough for incandescent lamps, it is more than close enough for tube heaters. Tube heaters are, for practical purpose, MUCH less fussy than incandescent lamps. Tube heaters do not work right AT the edge of disaster like incandescent lamps must to work well. (This is Receiving Tubes, the only kind you ever see. Mega-power transmitter tubes need special attention.)

The PT maker gets the transformer ratio (leverage) and sag so that nominal wall-power is just-right at the heater winding.

(We do sometimes end up a bit high with older transformers and rising wall-voltage. We in the US really got 110V in 1900 (well, 103-118), but it has been creeping up a volt a decade, some areas consistently 125V. Lamps of course are all dead before wall voltages wander-up enough to matter; it can be an issue with antique guitar amps lovingly restored long-long past their expected obsolescence. But from 115V in 1946 to 125V in 2016 isn't a real problem, usually.)

I would use a resistor. It is the SAME heat as any other plan unless you go to a LOT more trouble than it is worth.(*)

A significant resistance may mildly prolong tube heater life by limiting cold-start current. But small tubes usually fail of other problems (hiss, crackle, loose bits inside) before the heater burns-up.

I have not seen a simple failed-short resistor. It is possible with physical damage: crumple the chassis so it bites into the resistor. That was a fun (not) repair. Considering the MANY things to go wrong with regulators (especially hand-built jobs), resistor seems ultra-reliable to me.

(*)Actually 85% efficient switching regulators are now quite cheap from Asian eBay vendors. Some of them may be good quality and well documented.

[Vitrolin]

how many valves are you going to power up, if 2 or 4 i would connect 2 heaters in series, simple and no lost heat

where do you get your 12V from? if its a transformer maybe one with a center tap could be an idea

[Transmogrifox]

It has been pointed out that the same heat has to be dissipated somewhere whether using resistor or regulator and transistor.

How hot it gets is essentially thermal resistance times watts (same relationship as V=IR).  A good way to reduce heat is to mount something like 4 x 1Watt resistors on a copper clad board with a decent amount of surface area.  Evenly spread the power out over a larger surface area and the temperature goes down.

1 Watt is enough to start a fire if all concentrated to a point source (like an 1/8Watt resistor), but it's barely a measurable temperature change if applied to a water heater element.

[PRR]

> can be touch by hand (not too hot)

I bet you do not like to touch the 6111 when it is working. While the larger "mini" tubes of the same heat won't sear flesh, the 6111 is half the diameter and a bit shorter, must run hotter, and probably gives quick burns.

FWIW, the maximum temp for 6111 is 220 degree C, much hotter than boiling water. It won't run quite that hot on a stage, but still above "hot water", and will damage flesh.

So if you have one "danger" part already, what's wrong with another?

Also you generally do not go poking around inside audio gear while it is running. Touch a signal point, inject a big HUMMMM, blow your speaker or your ears. When debugging audio stuff you watch what you touch.

> Just wondering with approach will generate the least heat?

Same either way. The regulator is just a fancy resistor. It will adjust for input voltage and load current. However your input voltage is likely regulated, or at least stable-enough for incandescent lamps (+/-5% most days, +/-10% bad days). Your load is a dumb resistor (tungsten filament) and once warm its current/voltage doesn't change.

It's like: I need a 6 foot stick. The lumber yard only has 12 foot sticks. With sticks, we can just saw in half, but that's 6 feet of waste stick. Worse, with Voltage there is not an easy way to cross-cut a voltage. Linear regulators and resistors *both* act like wood-chippers. I can chip a 12' stick down to 6', but I get a pile of chips too.

Switchers "can" do another trick. I need a 6' 1x2. I buy a 12' 1x1. I cross-cut and then glue the two 1x1 into one 1x2. Aside from no waste, I also do not need as wide a rack on my truck to carry a 1x1 home. In electric, that's less current from the 12V source. However cutting 12V to two 6V at the same place requires complicated switching to "saw/join" and reactances to do it smooth.

> and I want to keep the part count low

Resistor.

> 4W 20ohm resistor?

Yes, double the 1.8W actual to buy the resistor, minimum.

But now you ALSO want "touchable", or at least "not instant blister". You want a bigger resistor.

4X the rating *may* give 4X the surface area and half the temperature.

However resistor makers tend to build small and use tough material to stand the heat. I am astonished how small some "1 Watt" parts are today. I don't want you to come home with a flea-size resistor.

The other train of thought: since in this case the tube Watts is same-as resistor Watts, get a resistor at least as big as the tube (surface area). A traditional "10 Watt" resistor is a little skinnier and a little longer than a 6111. I'd look at that and also a 20 Watt. Price/size does jump-up a lot 10W to 20W, your call.

[nguitar12]

Thank all I think I will go for the resistor solution and keep the circuit simple.

[anotherjim]

In my experience transformers do regulate. Not in the sense of holding to a set output voltage despite varying supply, but at least tending to mitigate change when the load is purely resistive. If input V rises, current must tend to increase, but internal losses (resistive + inductive) also increase, so secondary V can't rise as much. And internal losses reduce when supply V falls for the same reason and secondary V doesn't fall as much. I've never seen a transformer that doesn't behave like this.

[PRR]

12V to 6V with "96%" efficiency (low heat), two bucks:

http://www.banggood.com/Mini-DC-DC-Converter-Step-Down-Module-Adjustable-Power-Supply-p-920327.html
Mini DC-DC Converter Step Down Module Adjustable Power Supply
Price: US$1.55
Reviews     5 Star     68 (97.1%)
Input voltage: 4.5V ~ 28V
Output voltage: 0.8V ~ 20V
Output current: rated current is 2A, maximum 3A (Additional heatsink is required)
Conversion efficiency: 96% (Highest)
Output ripple: < 30mV
Switching frequency: 1.5MHz (Highest) 1MHz (typically)
4-star review: 'My tests show that consumption above 1A will "burn" the little guy. Test Conditions: Input 21.3V, output 11V / 1A provides a temperature above 80 degrees (measured 84.1 degrees Celsius (54 degrees above ambient, after 2 minutes of operation) which is far above what I would allow in my own designs)'

[PRR]

> tending to mitigate change when the load is purely resistive.
> If input V rises, current must tend to increase, but internal losses (resistive + inductive) also increase

R and L should be essentially linear over a wide range of voltage. So loaded output voltage should track input voltage.

It is also fairly routine to apply 12V to the 120V winding of an un-documented transformer to get the winding ratios and voltages without dangerous voltages on the bench. A "34V" winding will pretty reliably be a 340V winding at 120V input.

If this were not true, audio transformers would be fishy. They ARE fishy, non-ideal, but can show <0.1% linearity over a wide range of voltage.

If you get into "saturation", yes, sine-like voltage output does not rise much higher. Instead the input current soars and the transformer burns-up.

And of course this all excludes Sola and similar regulating transformers which are costly, heavy, hot, loud, and un-common.

[Hatredman]

In my experience transformers do regulate. Not in the sense of holding to a set output voltage despite varying supply, but at least tending to mitigate change when the load is purely resistive. 

I don't really know the correct term in english, please someone correct me.

But, as I understand, regulation is, indeed, only and exclusively "holding to a set output voltage despite varying supply".

"Tending to mitigate change when the load is purely resistive" is another feature called (i think) stabilization, which is a whole different thing.

[SatchmoeddieII]

They do make heat sinks for T0-92s T0-18s etc. The TO-92 heat sinks I have are a clip with a single pin on it. Solder that pin to a flag of copper foil. I'd use lead free solder or silver solder and DO NOT solder it with it clipped onto the transistor.

[vigilante397]

They do make heat sinks for T0-92s T0-18s etc. The TO-92 heat sinks I have are a clip with a single pin on it. Solder that pin to a flag of copper foil. I'd use lead free solder or silver solder and DO NOT solder it with it clipped onto the transistor.

Hi there Be sure to check dates when you reply to threads, this one has been dead over 5 years.