diyAudio

diyAudio (http://www.diyaudio.com/forums/)
-   Tubes / Valves (http://www.diyaudio.com/forums/tubes-valves/)
-   -   Rectified heater voltage problems? (http://www.diyaudio.com/forums/tubes-valves/25241-rectified-heater-voltage-problems.html)

eeka chu 3rd January 2004 10:36 AM

Rectified heater voltage problems?
 
Hello,

I've been messing around with some EL34 tubes the last few days. A lot of schematics I see just run the heaters directly from the ~6.3Vac take off from something like a Hammond transformer.

This is a bit too noisey for my liking so I am rectifying the Vac and passing it thru a Black Gate 1mF 50V capacitor for smoothing. The load resistor is roughly 5k but will be 10k in the final design.

What's making me wonder is that the rectified ac is appearing as roughly 9Vdc from the full wave 1N5404 bridge rectifier. No problem. But than when I attach the heaters, the voltage drops to 5Vdc as the load is added.

I had the transformer custom wound for me so I could keep everything seperate and simplified; it should be good for at least 3 amps. I am curious, do tube specficiations rate their tubes for the voltage that should be present with no load, or while the heater is in it's running condition? How close do people actually stay to the 6.3v recommendation and how many of you float the heaters around 50v or so to minimise hum?

Of coarse, with the lower voltage present, current rises, which is worrying me slightly as I had specified the transformer to run at 6.3v without a load. The supply must heat an EL34, which will require roughly 1.5 amps and two 12AX7's, which require a fraction of an amp each.

If a supply is going to be rectified and used to power heaters do designers usually accomodate for this in the input transformers voltages or do they just use a 6.3v heater transformer and live with the voltage differences?

I decided to go with just the Black Gate smoothing capacitor since a choke is going to start making things complicated and resistors are going to have to be very small to allow the correct current to flow.

In addition to these questions, I would also greatly appreciate any help with two others!

- Where in the UK sells well made, custom, mains transformers? I've used Newmark just recently and, while I'm not unhappy with them, I am not particularly impressed either.

- If my supply could possibly draw close to 3 amps should I switch the 1N5404's for something that can handle more continuous current to ensure there's no chance of them burning out over time?

I'm sorry if I've made some kind of stupid mistake somewhere along the way here... my mind seems to be in second gear mode at the moment and I'm making incredibly embarassing mistakes 24 / 7! :)

All the best,
John

fdegrove 3rd January 2004 11:49 AM

Hi,

Quote:

This is a bit too noisey for my liking so I am rectifying the Vac and passing it thru a Black Gate 1mF 50V capacitor for smoothing. The load resistor is roughly 5k but will be 10k in the final design.
Assuming 1 µF this way too low, try something in the order of 4.700µF or even 10.000µF.

6.3VAC FWB rectified would give a theoretical maximum of 8.82VDC so in order to bring both the voltage and ripple down you need a Pi cel filter.

When executed properly AC heating an indirectly heated powertube should not give you any audible hum.

Cheers,;)

tubetvr 3rd January 2004 11:52 AM

Hi,

Heater voltage is specified under load conditions. allowed tolerance is usually +-10%, even if it is probably better to be closer then that for best performance and longest life. 5V as you measure is way too low.

If you rectify 6.3V you probably end up with less then 6.3V DC unless you dont use a big cap due to losses in the diodes, according to PSUD I get roughly 6.3V using 10000 uF with 1A current draw and a transformer with normal spec. To use a black gate caps is to exaggurate a bit, a normal good cap will do, the larger the better.

Current should not be so much different with 5V or 6.3V.


The transformer has to be specified to give voltage under full load as it normally gives higher voltage under no load conditions. If you use DC for heater the transformer also has to be specified for higher AC current, normal is to calculate that AC current is 1.6 - 1.8 times DC current with a bridge rectifier and capacitors.


BTW, as you are using EL 34's I assume that this is for a power amp? I have never used DC heaters in a a power amp with in-direct heated tubes. Normally I separate the heaters for driver tubes and power tubes. For the power tubes I just tie one of the terminals to ground. For the driver tubes I use a hum pot of 100ohm with the slider connected to ground, (or a DC potential if the input stage is a SRPP or similar), the hum po is adjusted for min hum on the amplifier output. This method give me hum levels that are below the high frequency thermal noise and so low that I cant hear it even very close to my ~100dB sensitivity Lowther horns.

Please try to make the amplifier hum free without using DC for the heaters its should not be a big problem.

Regards Hans

eeka chu 3rd January 2004 01:18 PM

Things just seem to have gotten easier... something has to go wrong...! :)

Thanks for the advice Hans and fdegroove!

It seems like I've been going the wrong way with all this. FDE, have you ever had any annoying levels of noise using AC for driver tubes or would you agree with Hans that the hum pot is easily enough?

Do you have any recommendations for smoothing capacitors? I know Elna's Cerafines have a great capacitance at low voltages, but they have been discontinued now so I would rather not use them if possible. Are there any other favourites for this part or should I just go with any moderately expensive brand?

Kind regards,
John

Kashmire 3rd January 2004 01:23 PM

I'll repeat John's comment again but with more straightforward language:

You lose voltage in the diodes. Open-load, the voltage should be 6.3 * 1.414 = 8.9. You won’t see the diode voltage drop open-load.

Once a load is applied, you are probably losing 0.7V per diode. Since the power has to travel through 2 diodes in a full wave bridge, you are losing 1.4V in the bridge. Now you are down to 7.5V usable voltage – even before applying full load.

You can’t convert 6.3VAC rms into 6.3VDC at full load. The power factor in the transformer will be awful, causing very high current spikes across the transformer and diodes. These spikes are much greater than the transformer's rating, so you’ll be losing additional voltage in the transformer’s resistive losses (otherwise known as regulation factor).

There’s where your voltage is going: diode forward voltage drop and inside your transformer due to a low power factor.

It is my suggestion that in order to use DC on heaters, the transformer secondary must output at least 8V rms. I use a 9VAC transformer for my driver stage tubes. I use a bridge rectifier and a voltage regulator, which accounts for at least 3.7V of losses.

To avoid losing 1.4V in the diode bridge, you can use synchronous rectification. I’d love to change every diode bridge in every piece of my audio gear to synchronous rectification. Over the next months, I’ll be developing a little circuit board to replace diode bridges with synchronous rectification. Synchronous rectification won't solve your power factor problem, though.

If you want more information on what the power factor is, and why it is affecting your voltages, let me know.

fdegrove 3rd January 2004 01:39 PM

Hi,

Quote:

FDE, have you ever had any annoying levels of noise using AC for driver tubes
Well, I'd use DC in phono stages and linelevel preamps.
In an amplifier using indirectly heated tubes I always use AC and never had hum I could measure or hear.

If you don't use a center tapped xformer you can create a virtual one by putting two 100R resistors in series across the filament winding and ground the midpoint.

Always use solid core wire no thicker than appropriate for the current it needs to pass, twist as tightly as you can and push it against the chassis away from other signal carrying wires and components.

Quote:

would you agree with Hans that the hum pot is easily enough?
That works very well too in addition of the above methods, just don't use the two resistors but connect the pot's wiper to ground.
This is often how it's done with directly heated tubes.

Regarding xformers in the U.K., have you tried Maplin?

Quote:

Do you have any recommendations for smoothing capacitors?
I won't waste money on them here, you can use ordinary Nichicon, Philips, whatever you can easily buy.
Indirectly heated valves don't have their heaters in direct signal path.
Target for the correct voltage and make sure they can draw the current they need.

When I use DC for the heaters I also regulate their supply so it stays constant and the rail is squeeky clean.

Cheers,;)

eeka chu 3rd January 2004 03:51 PM

Quote:

You won’t see the diode voltage drop open-load.
This is one of those moments where I should be kicking myself... :eek:

Quote:

I use a bridge rectifier and a voltage regulator, which accounts for at least 3.7V of losses.
Do you hand make a regulator for the supply yourself or buy a prefabricated version? I read an article by one amp designer who mentioned that he used the LM723 for his supply regulation. Is it worth putting together voltage regulation circuits myself or are the prefabricated ones just as good?

Quote:

If you don't use a center tapped xformer you can create a virtual one by putting two 100R resistors in series across the filament winding and ground the midpoint.

That works very well too in addition of the above methods, just don't use the two resistors but connect the pot's wiper to ground.

This is often how it's done with directly heated tubes.
This isn't directly related to my initial question but... I was reading about this idea ealier and found someone asking about whether or not it could be used on a directly heated cathode. One of the replies suggested that unless the resistors where 0.1% precision values and the transformer was perfectly balanced then it would allow hum thru and would also damage the sound quality. I guess an adjustable hum pot solves this.

Quote:

I won't waste money on them here, you can use ordinary Nichicon, Philips, whatever you can easily buy.
That sounds like a good idea. Would you use similar capacitors in the high voltage supply for the plates or would you go with a more expensive brand? I know Black Gate make the Wkz series specifically for the LC filter in HV supplies. It's 220uf + 220uf but it's retail is something around 55 pounds without VAT if I remember rightly.

Compared to Nichicon's, would I be right in thinking Black Gate believe they can charge more for their Wkz series due to some improvement in their transient response or filtering abilities?

My thanks to everyone who's helped out so far,
John

fdegrove 4th January 2004 12:04 AM

Hi,

Quote:

Is it worth putting together voltage regulation circuits myself or are the prefabricated ones just as good?
It's a ******** filament supply for crying out loud...Relax, if it doesn't hum it's O.K.

Quote:

I read an article by one amp designer who mentioned that he used the LM723 for his supply regulation. Is it worth putting together voltage regulation circuits myself or are the prefabricated ones just as good?
The off the shelf regs are fine unless someone's going to tell me that heaters ar feeding back into their cathodes that is...get a grip, please...???

Inversely, you can tap signal of a heater circuit....ah, ah....

Sorry if I sound rude but I start to get fed up with elemantaries...

Buy a book, read Tubecad, in short educate yourself...

Make an effort and learn, maybe one day we'll be buying your stuff?

Ah well,:xeye:

eeka chu 4th January 2004 10:38 AM

Quote:

Sorry if I sound rude but I start to get fed up with elemantaries...

Buy a book, read Tubecad, in short educate yourself...

Make an effort and learn, maybe one day we'll be buying your stuff?
No problem, I understand what you're saying! I don't like asking questions like these, but I'd been having quite a lot of trouble finding accepted answers to them.

I'm currently making my way thru the Naval electronics manual. It's good, but then it gets to points, like heater supplies, and it does nothing other than tell you that tubes need one; well, perhaps slightly more.

Is Radiotron's 4th Edition worth the $40 - 50 it's currently selling for? I'm having a bit of trouble finding a copy in the UK. Postage from the US is substantial for a book that weighs this much.

I would greatly appreciate any links to sites online specifically related to tube theory if you can suggest any others.

Regards,
John

Original Burnedfingers 4th January 2004 12:27 PM

Quote:

It's a ******** filament supply for crying out loud...Relax, if it doesn't hum it's O.K.

John

There comes a time in everyones life when they need to put trust into what others say. I will be blunt and too the point. Frank has the credentials, experience,knowledge to know what the hell he is talking about. Electronics is like a game of chess with Frank except he is always 5-6 moves ahead of his opponent. Frank won't steer you wrong. If I understand him correctly he is telling you not to waste your money on caps/components that aren't going to help you. If you don't have hum or hear hum don't put yout money into extra parts when you don't need them. Take that money and put it into other areas in the amp that will show a sonic/performance difference.

J


All times are GMT. The time now is 06:56 PM.


vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright ©1999-2014 diyAudio


Content Relevant URLs by vBSEO 3.3.2