New DHT heater - Page 42 - 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 18th November 2011, 02:46 AM   #411
iko is offline iko  Canada
diyAudio Moderator
 
iko's Avatar
 
Join Date: May 2008
Location: Toronto
It is Andy's circuit. The only part that I use is the bias. My psu is different and I will change it again; I will wind myself some chokes for B+ filtering and plate load as well. Those should be simpler to make.
  Reply With Quote
Old 18th November 2011, 07:25 AM   #412
работник
diyAudio Member
 
Rod Coleman's Avatar
 
Join Date: Jul 2004
Location: Warwickshire UK
Quote:
Originally Posted by petavgeris View Post
2. In order to avoid this voltage drop, you can use e.g. an LT1083-4-5 in CCS mode right before the c.m. choke, not for AC filtration but for DC current stabilization only. Isolating the power supply is done through the cm choke, not the CCS! You have to bypass the CCS device with a very good sounding capacitor and this makes a rather shocking, unexpected sound improvement, quite close to the no-electronics-at-all solution described in #1. But you have no AC mains ripple rejection through CCS, so you still need to go with a differential choke. This way you can use more down-to-earth c.m. chokes. The extra power is consumed on CCS heatsink before the wire in C.M. choke gets heated, at thermal equilibrium power is mostly consumed on this cm choke's wire.
It is not surprising that you had to bypass the LT1083. This device generates too much noise for DHT heating. This may sound surprising, but just consider that (for example) 0.1% of current noise in a 1.2A 300B filament is 1.2mA. The current noise is DIRECTLY mixed with the anode current of the DHT (maybe 60mA). You can do the calculation for % of anode current noise! Maybe it is better than 0.1% - but it will never be small enough. The capacitor will damp some of this noise, but will not remove it.

My Filament Heater kit does not contain any reference (bandgap, zener etc) because these generate too much noise. The effective reference is simply a base-emitter junction of a low-noise transistor {a few nV/(Hz^0.5) in other words}.

As for the difference in sound, Paul Needs, and Gianluca and others, have directly compared the LT108x:

New DHT heater

The regulator also includes a passive Gyrator for mains AND high frequency filtering. Because it is passive (ie NO feedback loops at all) it performs sonically much the same as a choke. Of course, at low frequency it performs much BETTER than a choke. At VHF+ frequencies (which is very important for screening mains noise), one would need to build a choke with only few tens of pF across it, just to match the performance of the Gyrator.
  Reply With Quote
Old 18th November 2011, 09:24 AM   #413
работник
diyAudio Member
 
Rod Coleman's Avatar
 
Join Date: Jul 2004
Location: Warwickshire UK
We can put some honest numbers on the LT1083 noise, rather than just the hand-waving.

The data sheet specifies the noise as:

0.003% of Vout, Typical, 25 deg C, RMS value, 10Hz-10kHz.

When connected as a CCS, for say a 1,2A 300B filament, the regulator applies 1,2V across a 1-ohm resistor, so the output noise is:

Vn = 1,2V x 0.003% = 36uV;
and the noise current is translated by 1-ohm directly into amperes: 36uA.

So the CCS injects RMS noise of 36uA into your filament, where it is directly coupled to the cathode current.

For 300B at 60mA, this is RMS 0.06%
For 26 preamp at 6mA this is nearer to 0.5% (1.05A filament).

No wonder the difference is audible!

Remember, these are RMS values, so the peaks are bigger. For some samples of the chip, the noise will be worse, as it is not guaranteed.

And, if the test bandwidth were properly opened up to audio amplifier limits, the noise would be substantially worse.
  Reply With Quote
Old 18th November 2011, 09:55 AM   #414
diyAudio Member
 
Join Date: Feb 2010
Location: Athens Greece
Quote:
Originally Posted by Rod Coleman View Post
It is not surprising that you had to bypass the LT1083. This device generates too much noise for DHT heating. This may sound surprising, but just consider that (for example) 0.1% of current noise in a 1.2A 300B filament is 1.2mA. The current noise is DIRECTLY mixed with the anode current of the DHT (maybe 60mA). You can do the calculation for % of anode current noise! Maybe it is better than 0.1% - but it will never be small enough. The capacitor will damp some of this noise, but will not remove it.

My Filament Heater kit does not contain any reference (bandgap, zener etc) because these generate too much noise. The effective reference is simply a base-emitter junction of a low-noise transistor {a few nV/(Hz^0.5) in other words}.

As for the difference in sound, Paul Needs, and Gianluca and others, have directly compared the LT108x:

New DHT heater

The regulator also includes a passive Gyrator for mains AND high frequency filtering. Because it is passive (ie NO feedback loops at all) it performs sonically much the same as a choke. Of course, at low frequency it performs much BETTER than a choke. At VHF+ frequencies (which is very important for screening mains noise), one would need to build a choke with only few tens of pF across it, just to match the performance of the Gyrator.
Reality is exactly as you mention it!
LT devices produce noise. Cap bypassing was quite shocking to me, but you are right.
OK, the purity of the choke is probably unmatched but unfortunately you cannot go only with a choke as voltage drifts. Your solution at low frequency is definately better.
I would like to try it before this project ends. How would that be possible Rod?
Regards,
P.A.
  Reply With Quote
Old 18th November 2011, 10:08 AM   #415
diyAudio Member
 
Join Date: Feb 2010
Location: Athens Greece
Default LT1083

Quote:
Originally Posted by Rod Coleman View Post
So the CCS injects RMS noise of 36uA into your filament, where it is directly coupled to the cathode current.

For 300B at 60mA, this is RMS 0.06%
For 26 preamp at 6mA this is nearer to 0.5% (1.05A filament).

No wonder the difference is audible!
.
0,5% noise in driver tubes of ~4-6mA anode current is way too much indeed! And in my case this is differential so the CMC cannot block it! Maybe it would be better to 'hide' LT between the two differential chokes in a CLCLC filament solution but again filament chokes cannot supress noise that much.

How would it be possible to evaluate your solution?
  Reply With Quote
Old 18th November 2011, 10:09 AM   #416
работник
diyAudio Member
 
Rod Coleman's Avatar
 
Join Date: Jul 2004
Location: Warwickshire UK
Quote:
Originally Posted by petavgeris View Post
Reality is exactly as you mention it!
LT devices produce noise. Cap bypassing was quite shocking to me, but you are right.
OK, the purity of the choke is probably unmatched but unfortunately you cannot go only with a choke as voltage drifts. Your solution at low frequency is definately better.
I would like to try it before this project ends. How would that be possible Rod?
Regards,
P.A.
Peter, you can click my name to send email/PM, if you like some information on the Regulator kits.
I have some PDF documents for the Regulator applications, with recommended transformer, rectifier etc (but you don't need that! even better with Tungar rectifier)!

Please include a return email address.
  Reply With Quote
Old 18th November 2011, 10:19 AM   #417
diyAudio Member
 
Join Date: Feb 2010
Location: Athens Greece
Default kit

Quote:
Originally Posted by Rod Coleman View Post
Peter, you can click my name to send email/PM, if you like some information on the Regulator kits.
I have some PDF documents for the Regulator applications, with recommended transformer, rectifier etc (but you don't need that! even better with Tungar rectifier)!

Please include a return email address.
I have observed that differences are quite more noticeable while using Tungar rectifiers. Maybe the absence of switching noise of the diodes leaves more clean room for such detection. So the slightest hint of noise is more perceivable.

I will drop you a PM right now.
Thanks,
Peter.
  Reply With Quote
Old 20th November 2011, 10:00 AM   #418
работник
diyAudio Member
 
Rod Coleman's Avatar
 
Join Date: Jul 2004
Location: Warwickshire UK
Quote:
Originally Posted by danFrank View Post
Hi Rod,
I know I have asked you before, but could you please PM me info on how I can order a set of these from you? Thank you. Again...
Hi, I sent some PM on Friday. If you haven't received it, please click my name to send some PM, or email. Plenty of information available about Filament Heating.
  Reply With Quote
Old 20th November 2011, 10:03 AM   #419
работник
diyAudio Member
 
Rod Coleman's Avatar
 
Join Date: Jul 2004
Location: Warwickshire UK
Quote:
Originally Posted by petavgeris View Post
<snip...>

you can use CCS solutions or complicated electronics circuitry. This is the more simplistic approach, no need to use cm chokes. It still sounds miles better that voltage regulators.

I
Good luck!
P.A. from Athens.
Hi Peter, I am happy that you also find a big problem with Voltage Regulation applied to Filament Heating.

Nothing is worse than voltage regulators in this position.

With ac-heating you have high levels of 50/60 and 100/120 Hz cross-products, but at least this corruption is simple and constant.

With voltage regulators, a reference is necessary, and it is difficult to remove the noise from these. With IC voltage regulators, like the 317, it is impossible to remove the noise, and these make a bad solution in all cases.

But the biggest problem with voltage regulation arises because of the interaction between the filament supply and the cathode current.

The voltage across the filament (5V for 300B) means that there is an anode-cathode voltage gradient along the filament, and a grid-cathode gradient, also. This imbalance in the triode's operating voltage causes a music-signal to be developed across the filament terminals.

Now, this music-signal across the filament terminals is applied to the feedback network of the voltage regulator, since it is in the same place as the filament heating voltage. In other words, (part of) your audio signal is mixed up in the feedback of a voltage regulator that is ony supposed to be heating the filament. This is very bad, because the music signal is maybe only microamperes to mA, but the filament current is 1,2A for 300B. All voltage regulators have transient response errors, which do not usually matter if they are 0,1% of the output current. But think about 0,1% of 1,2A (1,2mA) applied to the mA music signal!
Also, think that the Voltage Regulator will think that the music signal is an ERROR, and will try to correct it!!! This is a very bad outcome for the audio signal.

It gets worse. IC regulators like LT108x (and all other ICs) also have an open-loop gain which decreases rapidly across the audio band. In the closed-loop, this means that the error-signal increases with frequency. The effect of this on the filament circuit is that the audio signal will suffer frequency-dependent cancelling by the voltage regulator!!

This is the reason that voltage regulated filaments sound muddy and confused. Don't do it!
  Reply With Quote
Old 20th November 2011, 12:59 PM   #420
Gluca is offline Gluca  Italy
diyAudio Member
 
Gluca's Avatar
 
Join Date: Nov 2004
Location: Back to Italy
THIS ^
  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
how to wire 6.3VDC heater on a heater with center tap (12.6V series, 6.3V parallel)? jarthel Tubes / Valves 9 14th April 2012 06:36 PM


New To Site? Need Help?

All times are GMT. The time now is 01:56 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