Tube power transformer heater voltage

With a Antek power transformer
http://www.antekinc.com/content/AS-05T320.pdf

The heater winding is 6.3v with the input 115v, according to the datasheet.
I get 6.7v with 123v line voltage. 

I'd rather run the heaters low to extend life vs 8% high.
For DIY tube gear with AC heaters does it make sense to build in a voltage divider to drop the voltage 10%?

Heaters are designed to work with a supply up to 10% either side of the nominal figure so they will happily work up to 6.9V. 6.7V is comfortably within this range. However, if you want to limit the voltage to the nominal then you could also manage the inrush current as well by using a series NTC thermistor to do both jobs.

Cheers

Ian

Building a v76 and the EF804s datasheet specifies +/- 5%

Haven't used a thermistor ever - it is a temperature dependent resistor?

I have two 100R to ground on the heater already, I was thinking of adding 10R in series where the 6.7v supply comes in at the 10R and the tubes are supplied from the 10R - 100R junction. The tube H+ is then 100/(100+10) of the supply.

dmp said:

Building a v76 and the EF804s datasheet specifies +/- 5%

Haven't used a thermistor ever - it is a temperature dependent resistor?

I have two 100R to ground on the heater already, I was thinking of adding 10R in series where the 6.7v supply comes in at the 10R and the tubes are supplied from the 10R - 100R junction. The tube H+ is then 100/(100+10) of the supply.

Click to expand...

10 ohms will be far too large. The resistance of the two tubes in parallel is about 18.5 ohms (6.3V/340mA) so this will swamp the 100 ohms.

NTC thermistors are special silicon devices with a negative temperature coefficient so they are initially high resistance but as they heat up their resistance drops. I have to admit I have not used one myself but there are special types designed to limit inrush current and I have seen them mentioned in heater circuits for this purpose.

For the time being try a 1ohm  to start with. It will need to be rated for a minimum of 2W.

Cheers

Ian

You're right. The 4 tubes in the V76 pull 810mA total (at 6.3V), so the 10R would have far too much voltage drop.
1R looks better.

dmp said:

You're right. The 4 tubes in the V76 pull 810mA total (at 6.3V), so the 10R would have far too much voltage drop.
1R looks better.

Click to expand...

With four tubes even 1 ohm might be too big. It will drop 0.81 volts.

Cheers

Ian

Yes, that would be below the -5% lower limit. Maybe two 1R in parallel for 1/2 R
But what is the negative of too low H+?
Underheating in mics improves the noise floor and increases tube life.

dmp said:

Yes, that would be below the -5% lower limit. Maybe two 1R in parallel for 1/2 R
But what is the negative of too low H+?
Underheating in mics improves the noise floor and increases tube life.

Click to expand...

Under heating will reduce the noise but also reduces the gain (so the signal to noise rarely improves much) and increases distortion. Whether it makes any difference to tube life in a project studio set up is moot. Turning them on and off may shorten the life much more which is one reason I suggested an NTC thermistor.

Cheers

Ian

I have not experienced a negative from under-heating.  I question any significant reduction in gain, I haven't seen it.  You especially won't see it in a feedback circuit.  More detail would be required to make a solid decision along those lines.  Clearly though, you don't want to burn a lot of heat in a resistor either. 

> I'd rather run the heaters low to extend life

Heaters won't burn up in your lifetime. Tubes fail from crappy welds, dirty oxide, and leaks. The last heater failure I saw (not counting series-string TV) had been thrown 7 feet into a dumpster.

Also I figure 6.3% high, not 8%.

And this should be under load. If you are testing un-loaded you are ahead of yourself.

Isn't the current US standard 120 V? Why do companies keep making 115 V transformers?

I measured the h+ with two 100R to grnd and the four tubes hooked up (3xef804s and E80f). This is an almost complete build.
The H+ was 6.65v so yes, about 6% high.

CJ said:

if you use EF804s make sure you keep the heater voltage at 6.3 or below,

Ollie says that at 6.6 and above the carbonates burn off the cathode,

Click to expand...

john12ax7 said:

Isn't the current US standard 120 V? Why do companies keep making 115 V transformers?

Click to expand...

Yeah, the standard is 120v and my utility evidently is even high.
Seems Edcor and Hammond design for 115V
120v/115v nominal is +4.3% high

I remember reading somewhere about undervolting , might have been in one of Morgan Jones books .
It said in the case of low level preamp tubes ,where the plate runs below 100volts ,you can drop the heater voltage  by upto a volt or more on a 6.3volt tube without any problems . Undervolting the heaters is sometimes seen in tube mics ,an Oktava MKL2500  has has around 5.5 volts ,just over 10% low (6au6 equiv). I did also notice in special quality tubes there often limited to +/- 5%  on the spec sheets. I think way back in the early days of tubes when the fillaments were very fragile an prone to going open circuit a means of varying the filament volts was often used to vary the gain and help prolong tube life .

PRR is correct. Get your figure after the transformer is attached to its load, but tubes don't really care that much for perfection there. What's happening is you are heating up the Kathode so that it wants to share electrons to the plate. You want 777 degrees Celcius IIRC. Lower is written as giving lower emission but I've run 12.6VDC heaters at 9VDC and it worked great. I couldnt measure the temp inside but it certainly still worked just fine and looked just as amber inside. Granted, I wish it was between 11.6 or 13.6..

transformer heater voltage is usually speced for a certain current, go over and you have reduced voltage, go under and voltage rises. the DCR of the heater wind is a determining factor.

here is some blurbage from a book written by cavemen that still holds true today>

ruffrecords said:

Heaters are designed to work with a supply up to 10% either side of the nominal figure so they will happily work up to 6.9V. 6.7V is comfortably within this range. However, if you want to limit the voltage to the nominal then you could also manage the inrush current as well by using a series NTC thermistor to do both jobs.

Click to expand...

I've never seen NTC's used directly in the heater circuit. I think the reason is because they come with a minimum value of 5 ohms, which would result in too much voltage drops; could work if used individually (one NTC per heater), but would be costly. What I've seen often (and used regularly) is one NTC in the primary. Reducing the overall inrush current is also good for rectifiers and smoothing caps.

abbey road d enfer said:

I've never seen NTC's used directly in the heater circuit. I think the reason is because they come with a minimum value of 5 ohms, which would result in too much voltage drops; could work if used individually (one NTC per heater), but would be costly. What I've seen often (and used regularly) is one NTC in the primary. Reducing the overall inrush current is also good for rectifiers and smoothing caps.

Click to expand...

Types with a 25 degree C value an order of magnitude lower are now readily available:

http://uk.farnell.com/c/circuit-protection/thermistors/ntc-thermistors/in-rush-current-limiting-icl-ntc-thermistors?zero-power-resistance-at-25degc=0.5ohm

Cheers

Ian

ruffrecords said:

Types with a 25 degree C value an order of magnitude lower are now readily available:

Click to expand...

Interesting. I need to get deeper into them, what is the intended use. New products happen only when they're used in smartphones or computers these days. We can only try to grab the technological fallout of much more respected industries than audio.

abbey road d enfer said:

Interesting. I need to get deeper into them, what is the intended use. New products happen only when they're used in smartphones or computers these days. We can only try to grab the technological fallout of much more respected industries than audio.

Click to expand...

Most common use these days seems to be to limit in-rush current in SMPS.

Cheers

Ian