Universal filament regulator - Page 4 - diyAudio
Go Back   Home > Forums > Amplifiers > Tubes / Valves

Tubes / Valves All about our sweet vacuum tubes :) Threads about Musical Instrument Amps of all kinds should be in the Instruments & Amps forum

Please consider donating to help us continue to serve you.

Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving
Reply
 
Thread Tools Search this Thread
Old 31st May 2011, 01:51 AM   #31
tomchr is offline tomchr  United States
diyAudio Member
 
tomchr's Avatar
 
Join Date: Feb 2009
Location: Greater Seattle Area
Quote:
Originally Posted by SpreadSpectrum View Post
Tom, it looks like you have to solder the pad on the bottom of that thing as well. Is that pretty easy to do?
You are correct. The DAP is exposed and is the dominant path for heat transfer. It will need to be soldered to the board. For the final boards, I will have a matrix of vias connecting the pad for the DAP with the ground plane on the bottom of the board (as well as the top plane) and likely use 75 um copper (2 oz) for better heat transfer.

For the smaller packages where you can see the pins, I generally find that they are pretty easy to hand solder -- especially on a board with solder mask. That goes for this bugger as well. I use a regular ol' heat gun. A good preheat of 15~20 seconds on the low setting, followed by whatever it takes on the high setting to get the solder to melt. It usually takes another 10~15 seconds. Then I make sure the part is centered and move the board away from the heat gun and allow it to cool.

I use regular 60/40 lead solder with rosin flux core. I pre-tin the pad before placing the part and heating the board with the heat gun. I also use plenty of flux. A Kester brand flux pen is $5 at Digikey.

The trick is to use flux and to be careful with the heat gun so the board doesn't get burnt. But it's not rocket surgery...

I suppose with the thermal vias connecting top and bottom layer, one could also use a wide chisel tip on a regular soldering iron. That would provide more control over the temperature. I'll definitely try that once I have boards made for this. I have left an opening in the bottom solder mask for this exact purpose.

With 5 V out @ 2.5 A, the LM3102 gets pretty hot. Last night I was measuring the temperature on the outside of the case to be 70~80 deg C. But I was also running a bit high on the switching frequency and my prototype board has no thermal vias.

~Tom
  Reply With Quote
Old 31st May 2011, 08:34 AM   #32
работник
diyAudio Member
 
Rod Coleman's Avatar
 
Join Date: Jul 2004
Location: Warwickshire UK
Tom, I am surprised to hear you found my comments in any way religious or
provocative of flames. That is not my intention at all, I assure you.

I regard this forum as a place for serious discussion and analysis of
technical matters, and would like to be able to offer criticism so as to provoke
discussion, not disputes. The account I gave above is mostly composed of
technical analysis, illustrated and quantified where possible.

Many contributors here know that I have been engaged in researching
Direct-Heated Triode filament supplies, and reporting here, for the best part of
a decade. Under such circumstances, I can hardly hold back from entering a
discussions on the subject, can I?

So, let's have an entirely non-religious debate on the subject! I promise not to
say anything even remotely smelling of flame. However, I don't promise to avoid
mentioning my listening impressions, and the listening impressions of others on
this forum, since the purpose of building our own audio is to give pleasure, by
way of music. Many people here (literally hundreds) have tried out my ideas, and
heating solutions that I offer, and expressed surprise at how much improvement
they make. I'll admit that my solution is not as convenient as a switcher, but
my goal is all-out sonic performance, at modest cost.

To begin: I think the big difference between us is this:

I say: That the filament heating current flows in a filament wire in a DHT. The
same wire is terminated for connexion to the heating supply. The same wire
is also coated with emissive [cathode] material, and emits electrons towards the
anode [plate], which form the anode current. One [or both] of the same
filament terminals is connected to the anode supply's return path.
Therefore, filament current and anode current flow in the same conductor.
Therefore noise and disturbance in the one current are added to the current of
the other. We can ignore the question of distribution across the length of the
filament [for now] - since even if the anode current flows across a tiny
portion of the filament wire, it will still MIX with the filament
current.... it has to because the filament current must flow through every
part of the filament wire, by definition. The diagram I posted [#23] illustrates
what I say here.

For there to be no current mixing, the filament heating current would need to
flow on a different conductor to the "cathode" of the tube..... but then we
would not have a DIRECTLY heated triode.

This is the great difficulty with DHTs! Because of current-mixing, they respond
to minute attention and care to the heating circuit, because the errors
in the heating circuit are imposed upon the the anode current.

We may have built a wonderful regulator for 5V 1.3A supply, and measure
artefacts in the voltage output to be 0.1%. Ok, so 5mV in 5V, great.
But, connect a 4 ohm load [=300B filament] and you get artefacts of 1.3mA,
imposed on a 60mA signal current - > 2%, of full output signal. That's the
reason dc-unregulated [and some regulated] filaments still hum!

With listening tests, this effect is obvious. To me and many others... so if
you don't measure it, well, come on - let's dig into the analysis and find out
where the energy is going. I am genuinely and openly interested to explore it.

Last edited by Rod Coleman; 31st May 2011 at 09:01 AM.
  Reply With Quote
Old 31st May 2011, 03:41 PM   #33
tomchr is offline tomchr  United States
diyAudio Member
 
tomchr's Avatar
 
Join Date: Feb 2009
Location: Greater Seattle Area
Rod,

I didn't find your comments provocative or inflammatory. But I'm not really interested in turning this into a discussion about constant current vs constant voltage.

Your entire hypothesis hinges on this mythical current mixing effect, that is unsupported by physics for reasons I explained in a previous post. That makes it rather uninteresting to me.

If you claim other regulators cause hum but yours doesn't; show me the data. Show before/after graphs of the noise/hum/emi from your amp. Back you claims up with data.

~Tom
  Reply With Quote
Old 31st May 2011, 04:08 PM   #34
работник
diyAudio Member
 
Rod Coleman's Avatar
 
Join Date: Jul 2004
Location: Warwickshire UK
Tom, the aim of my post was to offer a critical analysis of the switcher solution. I used the data you displayed, and calculated some likely effects.

The current mixing effect is supported perfectly by Physics - it's simply Kirchoff's Current Law: We have two currents entering/emerging from one conductor, therefore Kirchoff Rules, No exceptions.

But, if you won't entertain my criticism, fine. I leave it for others to make of it what they will.
  Reply With Quote
Old 31st May 2011, 05:19 PM   #35
M Gregg is offline M Gregg  United Kingdom
diyAudio Member
 
M Gregg's Avatar
 
Join Date: Jun 2010
Location: UK
Interesting,

So if we have a current from two different sources one smooth the other with hash on it added togeather at a common point the one will not modulate the other? (similar to mains supply with RFI riding on the top of it). If we split the mains supply we don't get a clean and dirty signal we have a current split equal to the two added togeather. (with hash on both)?


Regards
M. Gregg
__________________
What is the sound of one hand clapping?
  Reply With Quote
Old 31st May 2011, 05:46 PM   #36
diyAudio Member
 
astouffer's Avatar
 
Join Date: Jun 2004
Location: Pittsburgh, crumbling wasteland
Quote:
Originally Posted by Rod Coleman View Post
We may have built a wonderful regulator for 5V 1.3A supply, and measure
artefacts in the voltage output to be 0.1%. Ok, so 5mV in 5V, great.
But, connect a 4 ohm load [=300B filament] and you get artefacts of 1.3mA,
imposed on a 60mA signal current - > 2%, of full output signal. That's the
reason dc-unregulated [and some regulated] filaments still hum!

With listening tests, this effect is obvious. To me and many others... so if
you don't measure it, well, come on - let's dig into the analysis and find out
where the energy is going. I am genuinely and openly interested to explore it.
If your theory is correct then it should be measurable.

Emission from filaments is not instantaneous, there is a delay from cold to operating temperature. What that delay is I don't know. I do know that it occurs a lot slower than the time between those 800Khz spikes, which should easily filter easily with a ferrite choke. If the filament behaved with the speed of an LED then I could believe you.

Analyze these ideas on paper all you want to. Eventually it comes down to real world measurements.
  Reply With Quote
Old 31st May 2011, 06:12 PM   #37
M Gregg is offline M Gregg  United Kingdom
diyAudio Member
 
M Gregg's Avatar
 
Join Date: Jun 2010
Location: UK
I think the only way to prove this,

would be to modulate the heater current / voltage and "Transmit" a signal into the O/P Tx. If this can be done then..


Regards
M. Gregg
__________________
What is the sound of one hand clapping?
  Reply With Quote
Old 31st May 2011, 06:20 PM   #38
работник
diyAudio Member
 
Rod Coleman's Avatar
 
Join Date: Jul 2004
Location: Warwickshire UK
Quote:
Originally Posted by astouffer View Post
Emission from filaments is not instantaneous, there is a delay from cold to operating temperature. What that delay is I don't know. I do know that it occurs a lot slower than the time between those 800Khz spikes, which should easily filter easily with a ferrite choke. If the filament behaved with the speed of an LED then I could believe you.
You don't need to consider the emission at all - I am not suggesting that the emission rate is modulated by filament heating-current noise.

Here's the noise mechanism:

The filament is a length of wire with a thin coating. There are emissions from all along the length of the coating - but how do the electrons get to each point along the coating? Along the substrate wire!

But this same piece of wire is the conduit for the heating current, therefore we simply face the situation where our signal current is sharing a conductor with a 1.3A heating current.

It's not fancy (or mythical) physics, or conjecture, it's just simple realisation that you have two currents - one sensitive and small, one big and burly - piped along the same dumb piece of wire.
  Reply With Quote
Old 31st May 2011, 06:27 PM   #39
M Gregg is offline M Gregg  United Kingdom
diyAudio Member
 
M Gregg's Avatar
 
Join Date: Jun 2010
Location: UK
Correct me if I have it wrong,

The idea is that "if the anode supply was dirty" we would hear it!
If the heater supply can modulate (not heat and cool) the current flow through the tube we will hear it.


Regards
M. Gregg
__________________
What is the sound of one hand clapping?
  Reply With Quote
Old 31st May 2011, 06:42 PM   #40
работник
diyAudio Member
 
Rod Coleman's Avatar
 
Join Date: Jul 2004
Location: Warwickshire UK
Quote:
Originally Posted by M Gregg View Post
I think the only way to prove this,

would be to modulate the heater current / voltage and "Transmit" a signal into the O/P Tx. If this can be done then..


Regards
M. Gregg
You are right - and Dmitry Nizhegorodov's work proves it. His study shows that even when you null out the fundamental signal (with the usual pot) you get 2nd harmonics of the filament current at the output!

Check out his measurements and see for yourself how clearly he proves that this is a modulation effect and not a thermal effect.

This correlates perfectly with widespread perception that ac heating sounds better than amateurishly implemented dc. With dc, you can't null-out the currents easily, so you get fundamental noise-modulation as well as harmonic modulation.

On Correlation Between Residual DHT Filament Hum and AC Frequency. Distortion-induced hum in directly-heated triodes.
  Reply With Quote

Reply


Hide this!Advertise here!
Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off


Similar Threads
Thread Thread Starter Forum Replies Last Post
Universal Tiger aged one Solid State 232 5th December 2012 11:09 PM
Noisy IC Universal Regulator dwalker0044 Power Supplies 4 14th April 2011 03:26 PM
Parallel vs. series filament supply - or another filament idea engels Tubes / Valves 11 13th May 2007 02:06 PM


New To Site? Need Help?

All times are GMT. The time now is 02:23 AM.


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