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 22nd April 2012, 05:41 PM   #31
diyAudio Member
 
Join Date: Jan 2008
Location: Southern California
I am away from home and posting with my ipad. I need to sketch it to understand it. I have done some similar traces but with local parallel feedback, it works because there is no grid current that comes into play and the voltage divider is only providing voltage. Got to go,need to check out from the hotel, but will pick it up when i get home tomorrow.

Thanks
  Reply With Quote
Old 22nd April 2012, 06:43 PM   #32
diyAudio Member
 
Join Date: Mar 2007
Location: Oxfordshire
Thanks Alfredo . I too am going to give it 12 hours . I am going to watch Inspector Montalbano I recorded weeks ago and have promised myself to watch . I went to Sicily 2 years ago and loved it . I am working on how to suppress silicon diodes switching noise . I have been told not to use 10 nF as it causes the energy to move down in frequency and ring . I have tried soft recovery diodes and suspect all they have is built in capacitance . I have tried a 4n7 470 R snubber ( smaller peaks , more of them ) . I have a suspicion of Schotkey diodes ( no minority carriers , oh yes , they still put out noise ) , I do have some . Might try silicon carbide types of the same . Thought I might mention this as someone reading this thread might have an answer . The devices soldered as close to the diodes as possible .

Whilst I remember . If needing a cheap and easy constant current source ( sink ) Semitec current regulator diodes are very useful and cheap .These are JFET's ( I assume ) . They are 100 V and the one I list is 4.5 mA . They are smaller than 1/2 watt resistors . They can be used in series for higher voltage . I suspect these devices are becoming a rarity so buy some now . Rapid Electronics in UK code 47-2608 . They can also be used in parallel for higher current . To replace an anode load is child's play if they suit an application . It gives the chance to listen almost instantly to a change . Often building a bespoke CCS will not do a better job . They especailly can be added to older transistor amps . I use one as a general purpose diode tester in series with a red LED from a 9 V supply ( it's in my test meter as an add on ) . This will measure any diode safely and quickly ( even white LED's ) . Can test base to collector and base to emitter on a transistor and reverse diode from emitter to collector . They have about 3 V of loss so not very useful on cathodes . On that subject I see LED's connected to cathodes . I have doubts about that .
  Reply With Quote
Old 23rd April 2012, 12:38 PM   #33
TheGimp is offline TheGimp  United States
diyAudio Member
 
TheGimp's Avatar
 
Join Date: Aug 2009
Location: Johnson City, TN
Nibel, look up "diode Snubber". The cap needs a series resistor to help suppress the HF noise. There should be application notes describing how to calculate C and R.
  Reply With Quote
Old 23rd April 2012, 02:10 PM   #34
diyAudio Member
 
Join Date: Mar 2007
Location: Oxfordshire
Thanks Gimp . The 470 R and 4n7 I tried was a generic value given for 1N4000 series ( 3900 pf + 470 R in the calculated example actually ) .The " drawn " graph given showed how a 10 nF ( which again seems to be a standard value advised for regulations by testing houses ) produces a series of diminishing ripples . The " drawn " example of a snubber working shows a sharp initial peak then nothing . The reality is a little different . I am sure it is better than just the 10 nF . I will have to work on this . The problem is that the the exact frequency is not easy to define ( the snubber I use about 72 khz if my maths is right ) . The hypothesis of the 100 kHz region seems about right . I might try a little neon tube across it in some arrangement , in this case there might be enough to fire it . I am not averse to using thermionic diodes , however not for this application . I have been a little lazy with this . I should try various combinations . Someone recommended the 100 R + 0.1 uf as used for switches . I dismissed it as being for a switch when seeing the values . Maybe I should try it . I have the power supply quiet to about - 110 dB . The switching spikes are noticeable on the scope . I might try some carbon composition resistors as they have low inductance . Some low inductance types say ideal for snubbers . If I get a spectacular improvement I will post it . The important thing to say is diode noise is measurable and not some piece of Audiophile folk-law . I do warn people don't take the solutions as gospel that are offered . I am convinced that soft recover diodes whilst accepting that they are good can be mimicked by a 100 pF cog/npo ceramic capacitor across the junction ( respecting voltage naturally ) . One thing said is if using soft recovery diodes no snubber needed . The oscilloscope says that's not true . I would imagine 470R and 470 pF would be a starting point . If there was ever something for the penniless Audiophile to play with this is it . The solutions will only be pennies . Most equipment has this type of rectifier .

I said earlier that Semitec current regulating ( CCS ) diodes can be used in series for higher voltages . If doing so check it out carefully as I can imagine sometimes there will be a kink in the curve as the devices crossover ( 100 V ) . They recommended a zener diode in series which may limit the usefulness a bit . However to get a feel for how it sounds it should work fine . Then build a bespoke CCS .
  Reply With Quote
Old 23rd April 2012, 04:42 PM   #35
diyAudio Member
 
Join Date: Mar 2007
Location: Oxfordshire
http://www.hagtech.com/pdf/snubber.pdf

Anyone wanting to follow the diodes snubber should read this . I did a brief test with neon tube and series resistor of 10 K . There was a slight effect . My own test seem to show results are less than expected . I will redo the tests with soft recovery diodes and non inductive resistors . These test with 1N4007 .
  Reply With Quote
Old 23rd April 2012, 04:54 PM   #36
TheGimp is offline TheGimp  United States
diyAudio Member
 
TheGimp's Avatar
 
Join Date: Aug 2009
Location: Johnson City, TN
Another thing I have found helps is a lossy ferrite bead on the lead of the diode as close to the body as you can get it.

Here is a copy of the same article from Hagerman, that is a bit better image quality:


http://peufeu.free.fr/audio/articles...20Supplies.pdf

Start reading page 22 of this one for a nice simple explanation of the choice of values for C and R.

http://etd.uwaterloo.ca/etd/y6chen2005.pdf

Last edited by TheGimp; 23rd April 2012 at 05:09 PM.
  Reply With Quote
Old 23rd April 2012, 05:38 PM   #37
diyAudio Member
 
Join Date: Jan 2008
Location: Southern California
Quote:
Originally Posted by smoking-amp View Post
Timed out for edits again above.

You really need an auxilliary input on the curve tracer for a current probe to do this sanely.
Below is a general Idea of how my tracer works, it is kind of a solid state clone of the Tek570. I need the divider network not to draw any current and influence either the screen current or the plate current. Is your suggestion to put a mosfet source follower at the output of the divider (to gate), connect the Drain to the plate voltage and the source to the cap?

thanks
Attached Images
File Type: jpg UL setup.jpg (172.2 KB, 80 views)
  Reply With Quote
Old 23rd April 2012, 05:46 PM   #38
diyAudio Member
 
smoking-amp's Avatar
 
Join Date: Dec 2001
Location: Hickory, NC
Basically, but it needs all the extra CCS stuff mentioned earlier to DC bias the Mosfet, without loading down the screen or plate for the tracer accuracy. The Y output would still be plate (+screen) current as usual for plate curves (from the current sense R), unless you just want to see the screen V.

Looking at your tracer setup, it appears there could be a simpler way to measure the plate + screen current. What if the bottom of the divider gets moved over to the top of the current sense R? Then the R divider current is removed from the output display, but the screen current is still summed with the plate current thru the power supply.
__________________
I want a Huvr-Board!!

Last edited by smoking-amp; 23rd April 2012 at 05:53 PM.
  Reply With Quote
Old 23rd April 2012, 06:17 PM   #39
diyAudio Member
 
Join Date: Jan 2008
Location: Southern California
Thanks Smoking amp,

Wouldn't the CCS need to be ON all the time? My plate voltage swings from 0V to max Plate Voltage at twice line frequency.

I will have to wire some things up and test...
  Reply With Quote
Old 23rd April 2012, 06:48 PM   #40
diyAudio Member
 
smoking-amp's Avatar
 
Join Date: Dec 2001
Location: Hickory, NC
For the plate V to go below the screen V, you will need the CCS on the Mosfet follower drain, to a B+ (and the cap from drain over to the plate). Normally a DC B+, but here that causes problems with screen current occuring while the plate V is zero (although limited by the CCS at least). Could try connecting the CCS to the rectified plate AC with maybe a resistor down from there to the actual plate to give the CCS some operating delta V. (have to position your plate V pickoff for the tracer horiz. correctly then.)

If you do use a separate B+ for the CCS to the drain, it could be returned to the bottom of the plate current sense R to eliminate that offset. You may be able to then just use a simple R pull-up for the drain, since the variable current thru it doesn't register on the current sense R that way.
Yeah, looks like that is the easiest way, no CCS at all. Just a resistor pull-up on the drain to a separate B+ supply. Then the resistor pull-down, from the Mosfet source, goes to the bottom of the plate current sense R too, so it doesn't register either. Only the screen current actually going thru the plate supply return path registers thru the sense R.


Obsolete comments:
A simple resistive pull-down on the Mosfet source could similarly (to the R divider) just go to the top of the (plate) current sense R to eliminate the Mosfet current effect (beyond the screen current) on the tracer display. I think, still trying to figure where the extra CCS current goes then. No, not working, the extra CCS current will dump across the coupling cap to the plate? Or is it? Maybe that current loop (around the CCS loop essentially) doesn't disturb the operation. Will shift the plate V a little, but the tracer pickup is just monitoring the actual plate V anyway.
__________________
I want a Huvr-Board!!

Last edited by smoking-amp; 23rd April 2012 at 07:15 PM.
  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
Pentode to Triode Switch Sherman Tubes / Valves 40 6th February 2013 12:06 AM
Triode-wired small signal pentode question rongon Tubes / Valves 16 29th August 2010 09:10 AM
Pentode guys: important to decouple g2 supply between stages? leadbelly Tubes / Valves 4 27th September 2008 05:12 AM
Triode versus Pentode Ricky Tubes / Valves 98 20th March 2008 08:43 AM
Why no single triode OTL power stages? leadbelly Tubes / Valves 28 15th February 2006 06:17 PM


New To Site? Need Help?

All times are GMT. The time now is 11:22 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