Switch to full style
A place to discuss member's DIY audio projects & post pictures/schematics. NOTE: There is a limit of 2MB per attachment, and a maximum of 3 attachments per message. If you need to post more than 3 attachments, just add another message.
Post a reply

300B filament question

September 22nd, 2018, 11:02 am

Curious if someone can answer this. It's quite basic but I have trouble with some of the basics, as we know. ;-)

So I'm playing around with single-ended 300B again. Decided to try DC on the filaments to reduce noise. I used a full-wave bridge block and a filter cap, very simple. I left the 25 ohm balance pot in place to create a center tap. Voltage dialed in just right and I listened for a while--still some residual DC, needs more filtering, blah blah. Then I wanted to compare to AC filaments again, so I removed the filter cap and simply connected the AC leads to the output tabs on the bridge, thinking I've just bypassed the bridge.

But have I? Are those diodes still doing something? Because now, when I use the hum balance pot I can null out all noise--it's almost dead quiet, even the DC component. I like it, but the question is, what am I doing by leaving a full-wave bridge across the AC filaments? And should my wife take my soldering iron away from me?

Re: 300B filament question

September 22nd, 2018, 11:15 am

Grover Gardner wrote:Curious if someone can answer this. It's quite basic but I have trouble with some of the basics, as we know. ;-)

So I'm playing around with single-ended 300B again. Decided to try DC on the filaments to reduce noise. I used a full-wave bridge block and a filter cap, very simple. I left the 25 ohm balance pot in place to create a center tap. Voltage dialed in just right and I listened for a while--still some residual DC, needs more filtering, blah blah. Then I wanted to compare to AC filaments again, so I removed the filter cap and simply connected the AC leads to the output tabs on the bridge, thinking I've just bypassed the bridge.

But have I? Are those diodes still doing something? Because now, when I use the hum balance pot I can null out all noise--it's almost dead quiet, even the DC component. I like it, but the question is, what am I doing by leaving a full-wave bridge across the AC filaments? And should my wife take my soldering iron away from me?


Well, I just asnwered my own question--yes, they are in play because the voltage is under spec. So out it goes. :-)

Re: 300B filament question

September 22nd, 2018, 11:22 am

Grover Gardner wrote:Curious if someone can answer this. It's quite basic but I have trouble with some of the basics, as we know. ;-)

So I'm playing around with single-ended 300B again. Decided to try DC on the filaments to reduce noise. I used a full-wave bridge block and a filter cap, very simple. I left the 25 ohm balance pot in place to create a center tap. Voltage dialed in just right and I listened for a while--still some residual DC, needs more filtering, blah blah. Then I wanted to compare to AC filaments again, so I removed the filter cap and simply connected the AC leads to the output tabs on the bridge, thinking I've just bypassed the bridge.

But have I? Are those diodes still doing something? Because now, when I use the hum balance pot I can null out all noise--it's almost dead quiet, even the DC component. I like it, but the question is, what am I doing by leaving a full-wave bridge across the AC filaments? And should my wife take my soldering iron away from me?


A full wave bridge provides 120-Hz. pulsating DC that needs to be filtered to smooth to straight DC.

Re: 300B filament question

September 22nd, 2018, 12:05 pm

Grover,
When you connect the transformer to the output tabs of the bridge rectifier, you have 2 diodes in series across the transformer. On one half of the AC cycle, these diodes are reverse biased and effectively out of the circuit. On the other half of the cycle, the two diodes clip the voltage at ~ 2 diode drops or ~ 1.2 volts. A 5V AC transformer has peak to peak voltage of 5v x 1.414 = 7.07v or referencing things to ground, the AC voltage will swing +- 3.535v. So on one half of cycle you'll be clipping the AC, which is why you're measuring the reduced voltage.
---Gary

Re: 300B filament question

September 22nd, 2018, 4:07 pm

GaryB wrote:Grover,
When you connect the transformer to the output tabs of the bridge rectifier, you have 2 diodes in series across the transformer. On one half of the AC cycle, these diodes are reverse biased and effectively out of the circuit. On the other half of the cycle, the two diodes clip the voltage at ~ 2 diode drops or ~ 1.2 volts. A 5V AC transformer has peak to peak voltage of 5v x 1.414 = 7.07v or referencing things to ground, the AC voltage will swing +- 3.535v. So on one half of cycle you'll be clipping the AC, which is why you're measuring the reduced voltage.
---Gary


...and that's exactly what I measured, Gary. Thank you for explaining it! :-)

Regarding the DC arrangement, I did add a filter cap to the output of the bridge. I'd need 6.3V to add a common-mode choke or extra RC filter. It was a quick-and-dirty experiment. ;-) It sounded pretty good but would need refinement.

When I built an SE 845 amp, I definitely needed DC on the filaments and used a CM choke to smooth it out. The sound was pretty good. I might pursue this with the 300Bs.

Re: 300B filament question

September 22nd, 2018, 4:09 pm

SoundMods wrote:
Grover Gardner wrote:Curious if someone can answer this. It's quite basic but I have trouble with some of the basics, as we know. ;-)

So I'm playing around with single-ended 300B again. Decided to try DC on the filaments to reduce noise. I used a full-wave bridge block and a filter cap, very simple. I left the 25 ohm balance pot in place to create a center tap. Voltage dialed in just right and I listened for a while--still some residual DC, needs more filtering, blah blah. Then I wanted to compare to AC filaments again, so I removed the filter cap and simply connected the AC leads to the output tabs on the bridge, thinking I've just bypassed the bridge.

But have I? Are those diodes still doing something? Because now, when I use the hum balance pot I can null out all noise--it's almost dead quiet, even the DC component. I like it, but the question is, what am I doing by leaving a full-wave bridge across the AC filaments? And should my wife take my soldering iron away from me?


A full wave bridge provides 120-Hz. pulsating DC that needs to be filtered to smooth to straight DC.


Ah, that was it. Thanks, Walt.

Re: 300B filament question

September 22nd, 2018, 5:12 pm

When I find it necessary to do DC filaments, I generally prefer to use a current source...

Roscoe

Re: 300B filament question

September 22nd, 2018, 7:07 pm

Please explain "current source" opposed to just xformers diode filter method

Re: 300B filament question

September 23rd, 2018, 12:28 pm

A current source (CCS) is an active circuit that maintains a constant current through the load at (ideally) all times. For a 300B that'd be 1.25A. For DHTs they have the advantage of eliminating the large current surge through the filament when you power on a cold tube. My stereo 845 amp has current regulated filaments. The easiest (but not necessarily best) way to build one is to use an adjustable 3-terminal regulator like an LM317 and wire it up so that it keeps a constant voltage across a small resistor in series with the load. Discreet solutions can offer better performance, and aren't all that difficult to build. Google is your friend.

Roscoe

Re: 300B filament question

September 23rd, 2018, 12:56 pm

Thanks
i will google and see if I can get a picture
Post a reply