Tube DIY Asylum

Do It Yourself (DIY) paradise for tube and SET project builders.

Return to Tube DIY Asylum

Message Sort: Post Order or Asylum Reverse Threaded

All pentode el84se no GFB

50.32.32.105
Posted on March 22, 2011 at 20:34:35
Paul Scearce
Audiophile

Posts: 471
Joined: March 1, 2002



I've built a number of different amps, including some zero feedback SETs, and some push-pull amps. I thought I'd try my hand at single ended pentodes. I'm aiming for an output impedance similar to zero feedback SET designs, without using any global feedback around the transformer. My speakers are pretty efficient--I'm currently getting plenty of sound from a 2.5 watt EL34 screenmode amp. I have several el84 tubes in my junk box, so I decided to try designing around them.After mulling over the data sheets, I think I've come up with a design I like.

The input tube is an EF184. The two 50kOhm resistors connected to its plate form the load and a local feedback network around the output tube. If I calculated right, this should work out to about 20dB of negative feedback. This brings distortion from the output tube to below 1%, and output impedance to a bit below 5kOhms. The 100kOhm resistor from the plate of the el84 to the cathode of the ef184 provides a bit of additional feedback. I think it works out to around 6dB. Since this loop includes the ef184, I'm hoping it will bring distortion from that tube down to around 1% as well as raising the damping factor of the amp to around 4.

The amp is still in the design stage, so I'm open to suggestions or comments.

 

Hide full thread outline!
    ...
Thanks for all the comments, posted on March 24, 2011 at 06:25:14
Paul Scearce
Audiophile

Posts: 471
Joined: March 1, 2002
 Thanks for the suggestions, they gave me some new things to think about.

I got out my number two pencil my calculator and the data sheets, and compared plate-plate feedback with plate-cathode feedback. I realized that more than half of the feedback in the plate-plate topology actually occurs in the unbypassed cathode resistor. Also by applying the feedback loop to the cathode resistor, instead of the plate, the current through the input tube is nearly constant, which should result in very linear operation of that tube.

By eliminating the E-linear loop, and combining the two plate resistors to form a single 25kOhm resistor to B+, and leaving the rest of the circuit the same, I should get all the negative feedback I want with much less distortion from the input tube. Since plate-plate feedback already includes plate-cathode feedback anyway, I don't see a down side.

 

RE: All pentode el84se no GFB, posted on March 23, 2011 at 10:42:24
reVintage
Audiophile

Posts: 271
Location: Eskilstuna
Joined: June 19, 2008



Hey Paul,
EF184 and EL84 is not a bad choice. But you made this little thingy a little to complicated and by that less functional. The EL84-plate to EF184-cathode NFB actually counteracts the function of the plate-plate feedback. About E-linear this makes feedback involve the OPT, something that you maybe should avoid too.

This basic one will work with typical triode characteristics and pentode effiency. If one also adds a gyrator instead of the EF184 plateresistor it will behave even better.


Brgds
Lars

 

RE: All pentode el84se no GFB: also for the lower output impedance, posted on March 23, 2011 at 12:47:53
Minussss3db
Manufacturer

Posts: 586
Location: so cal
Joined: January 28, 2010
 I see in your orginal post Paul you are asking for the good miracle of an output impedance as low as a SET...so..

The amount of feedback in the excellent LARS circuit above may not be enough...

so, LARS...??? could you please give a number from SPICE for the output impedance of your circuit...., and if this is not low enough...

then use Cathode EL84 to Cathode EF184 feedback...easy just put two resistors to sum to same value underneath EL84 cathode, smaller first underneath cathode, and then only bypass lower resistor...then cap couple with 25k resistor to start from EL84 cathode to EF184 cathode...leave EF184 cathode unbypassed or split it the same way as the EL84 cathode...

because there is less bypass on the EL84 you may lose a little deep bass, however because output impedance will go down, you should have better overall subjective bass,....and I believe two or more feedback loops where each feedback loop is applying less feedback...Sounds better...
so think on this, and perhaps Good LARS might build it into his Spice Model....also LARS what if the Stabilizing capacitor on the Screen of the EF184 went from 1uf to 10uf???? Often the very stable screen supply really does sound better on both driver and output pentodes in my low, weak and pathetic experience...What do you think about this Good LARS???

Anyway...Peace, and Good Luck to all,
-3db

 

RE: All pentode el84se no GFB: also for the lower output impedance, posted on March 23, 2011 at 14:18:50
reVintage
Audiophile

Posts: 271
Location: Eskilstuna
Joined: June 19, 2008



Zout will be lower or at least on par with a triode-strapped EL84.

With a 10H primary OPT like 125ESE, -3dB will be just below 20Hz.

Forget about multiple feedback loops even it can seem nice in theory.

But if you want even better triodelike performance and a simpler topogology(initially shown by MJK), go for this one. MJKs prototype was DC-connected with 5881.
Brgds
Lars

 

RE: All pentode el84se no GFB: also for the lower output impedance, posted on March 24, 2011 at 13:09:23
Michael Koster
Manufacturer

Posts: 841
Location: Eureka, CA
Joined: October 24, 2007









I tried something much like this a couple of years ago and found it to have a nice harmonic structure and sounded great! The solid low end was refreshing to hear in a SET.

It sounds so good, I decided to build a pair of 6550 based amps using this circuit into my Klipsch La Scalas that I use for live sound.

The effective plate resistance is approximately 1/gm, or about 200 ohms for a 6L6 type.

Cheers,

Michael

 

This schematic has completely shaken my whole foundation of my audio beliefs., posted on May 18, 2012 at 05:37:17
TubeBuilder
 So many years I have been convinced that ultimate vintage dht system was the end of my journey and now I am totally unsure. Here is a single power pentode with lots of feedback directly coupled to some solid state devices that could out perform any 2 or 3 stage dht amp in every way. Considering the amount of money I have invested in tubes and time studying them. That really unnerves me. I guess I am going to have to build one of these amps.Or at least try. Problem is I am really not knowledgeable enough to build such an amp. I have some Telefunken R12P35 pentodes that would make a interesting tube for this sort of amp.

This is and old thread. I built a el84 SE amp around this time and it did sound good but not great. It is collecting dust currently. Whether that was my lack of know how or limitation of the technology I don't know. How good are these pentode amps really? 845 killer good? This is no time for modesty. Is there folks here that have retired there dht's for a amp like the one above?

 

RE: This schematic has completely shaken my whole foundation of my audio beliefs., posted on May 18, 2012 at 07:47:15
hum4god
Audiophile

Posts: 55
Location: wisconsin
Joined: October 22, 2004
 hi tubebuilder
sorry to be OT but this seems to be the only way to reach tubebuilder.
you emailed me re the radians and i tried contacting you but i am not sure if my email got spammed .
if you are still interested pls email me w/ your contact email address .
thanks
malcolm

 

RE: This schematic has completely shaken my whole foundation of my audio beliefs., posted on May 20, 2012 at 08:12:11
TubeBuilder
 Sorry Malcom I will check my email.

 

RE: All pentode el84se no GFB: also for the lower output impedance, posted on May 9, 2012 at 06:27:12
Nigel Peasrson


 
Schadeode Amplifier Michael Koster .Please contact me about 5881 amp . My friend and I have been looking at your amps and Alex Kitic's . If yours wins we would like to say you were it's designer ..

Nigel

 

RE: All pentode el84se no GFB: also for the lower output impedance, posted on March 24, 2011 at 08:28:39
PakProtector
Audiophile

Posts: 12363
Joined: May 14, 2002
 and that one will sound even better with a caocode as the input stage.
cheers,
Douglas

Friend, I would not hurt thee for the world...but thou art standing where I am about to shoot.

 

RE: All pentode el84se no GFB: also for the lower output impedance, posted on March 24, 2011 at 08:54:24
reVintage
Audiophile

Posts: 271
Location: Eskilstuna
Joined: June 19, 2008
 Hey Douglas,
Matter of taste, my opinion is another ;-). Maybe depends on tube choice. But still, the concept whether pentode, cascode or FET is nice. Not sure which schematic you refer to in terms of sound, but the MOSFET for some reason gives the best overtone-spectrum with 2nd dominating and the rest gradually falling?!
Brgds
Lars

 

RE: All pentode el84se no GFB: also for the lower output impedance, posted on March 24, 2011 at 09:54:43
PakProtector
Audiophile

Posts: 12363
Joined: May 14, 2002
 Cascode MOSFET ov course. Same transfer function but with stabilized gate-drain capacitance.
cheers,
Douglas

Friend, I would not hurt thee for the world...but thou art standing where I am about to shoot.

 

RE: All pentode el84se Thanks LARS excellent circuit !!!, posted on March 23, 2011 at 12:06:39
Minussss3db
Manufacturer

Posts: 586
Location: so cal
Joined: January 28, 2010
 Dear Paul,

what LARS has put up here is excellent....this is a fully fleshed out amp with many good features...

the feedback from EL84 plate to the plate of the input tube is one of the best sounding, and LARS has detailed it better than my own simple suggestions...good bypass, good capacitor value choices, good screen stabilization value/choices...

so I think this is a worthwhile place to start for a circuit very similar to what you proposed, except with a lot of good things built in..


ok next

 

EF86 to EL84 w/GFB, posted on March 23, 2011 at 06:11:07
Triode_Kingdom
Audiophile

Posts: 10044
Location: Central Texas
Joined: September 24, 2006
 My main system at the moment consists of EF86s driving EL84s in SE pentode. NFB is the typical arrangement (from the OPT secondary back to the cathode of the EF86). I prefer to include the OPT in the loop; damping factor and products of core distortion both seem to benefit. The system measures very well, and if there's any loss of detail or low-level nuance as a result of GFB in these amps, it's not evident to me. I suggest you try it both ways before pronouncing judgement.

 

EF86 to EL84 Pentode ??? w/GFB ???... from TK ?, posted on March 23, 2011 at 09:27:17
Interstage Tranny
Audiophile

Posts: 3063
Location: Eastern
Joined: October 4, 2006
 What is happening to the triode kingdom you possess ? LOL !!!

It is always a good idea to include the opt, if it sounds best with that topology. Which SE opt tranny do you use with this amp ? If it were a super hifi opt, would you have tried the opt plate feedback ? The reason I ask is that lowering or removing feedback, when possible, always seems to open up the sound, more and more...ENJOY your pentodes and triodes; we all do !

 

RE: All pentode el84se no GFB, posted on March 23, 2011 at 05:00:06
PakProtector
Audiophile

Posts: 12363
Joined: May 14, 2002
 hey-Hey!!!,
have I got your schematic correct? It looks like the EL84 is only loaded by 40% of the primary. Second, consider using the E-Linear connection and eliminate the second B+ node and connect the EF184's plate resistor to the screen tap alone. I've not tried the plate-cathode connection yet; the E-Linear connection has provided enough FB by itself.
cheers,
Douglas

Friend, I would not hurt thee for the world...but thou art standing where I am about to shoot.

 

RE: All pentode el84se no GFB, posted on March 23, 2011 at 06:25:59
Paul Scearce
Audiophile

Posts: 471
Joined: March 1, 2002
 The el84 is loaded by the entire primary, and the plate feedback is taken from the 40% tap. I originally planned the amp to be true E-Linear, but it looked like distortion from the ef184 would be too high. If the input tube had been more linear at 20% feedback, I would have skipped the plate-cathode loop. Estimated distortion from the output tube is plenty low, and I'm sure I could live with a unity damping factor.

 

RE: All pentode el84se no GFB , posted on March 23, 2011 at 09:10:49
tubebuilder
 Siemens 2000 projector amps runs plate to grid negative feedback in their el84 SE amp. Cathode feedback around the output is much better then global and solves the damping factor and trans issues. I wish more trans where available with windings. You should search the 6bm8 club they have done just about every imaginable thing with SE pentodes 10 years ago. While the west where still laughing at little pentodes.

They used tubes and transistors to apply all sorts of feedback and direct coupling. I built a el84stc back then and its performance was stunning but with the feedback around the output tube being so severe and the complete lack of trans feedback(tiny trans) it was a little underpowered in the bass at the time I didn't have very efficient speakers. I should build another now that I am building horn speakers. I still use the stop diodes in the power supply and screen grids though.

If I recall correctly it had killer bandwidth only limited by the trans and perfect bass control again only limited by the trans and the most triode sounding amp I ever built unfortunately the most boring too. You can use 6al5 for stop diodes in the screens. Didn't know until yesterday according to the below site that stop diodes affect the op points significantly and the tube behaves more like a triode. I always just thought it gave better regulation. The power supply diode trick increases the damping factor subjectively incredibly and increases the separation in mono supplied stereo amps.The diode goes on the b+ wire of the output. You cant hear the diode I think because it isn't working hard.

My 4 year old daughter and I recently built a little 6aq5 SE amp with a lot of those tricks including the cathode feedback but only cap coupled. It was a sweet little dynamic amp. We named it Astro girl well she named astro girl despite the fact we only had astro boy decals for it. Shes a feminist you know. But a good first amp for her ;) Taking into account global feedback takes real skill to get right plate to grid and cathode feedback is easier to get right. I am no engineer neither is my daughter... Yet!

 

: All pentode el84se no GFB your feedback is nt the best, posted on March 23, 2011 at 03:15:17
Minussss3db
Manufacturer

Posts: 586
Location: so cal
Joined: January 28, 2010
 just take yur 50k plate resistor for the first stage right off the plate of the second stage and you will be happy, direct non global feedback...this is called an anode follower and it works very well...

also if you want more feedback, then you can direct couple, put a larger resistor underneath the output tube to 'fly' it relative to ground, and then make the cathode resistor in to a divider with the lower resistor bypassed, and then contect the cathode of the output tube to the screen of the input pentode...ok...

and it is so good to run the screen of the output tube from a nice stable voltage...

ok HAVE FUN,
-3db

 

Excellent!, posted on March 23, 2011 at 01:44:00
kurt s
Audiophile

Posts: 1137
Location: California
Joined: October 12, 2009
 I like this idea. I will have to quibble a tiny bit about the no GFB part. You're right in that it doesn't include the OPT, but the output of the 2nd tube goes back all the way to the input tube's input, and that is a form of GFB. Local only restricts it to one stage at a time.

However, I think this is a good idea. H-K Citation II used all pentodes and all NFB to sound great. You have a similar thing going, sans OPT, WRT NFB and pentodes. I think triodes are bad with NFB, not necessarily so with pentodes, a different sounding device.

And anybody that can design with a sketch pad probably knows a thing or two. :-) I used to do it that way to show schematics.

When someone asked "what should I use to draw schematics with?" I replied "a number 2 pencil." :-)

When this project finishes, I would like to know how it fairs against some triode zero NFB SETs.

 

Page processed in 0.034 seconds.

[ Top Top of Page ] [ Contact Us ] [ Support/Wish List ] [ Copyright © 2024, Audio Asylum, all rights reserved ]
[ General ] [ Speakers ] [ Tubes ] [ Vinyl ] [ Digital ] [ Hi-Rez ] [ Video Asylum ] [ Cables ] [ Tweaks/DIY ] [ Music ] [ Films ]