Go Back   Home > Forums > >
Home Forums Rules Articles diyAudio Store Blogs Gallery Wiki Register Donations FAQ Calendar Search Today's Posts Mark Forums Read

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

Unexpectedly good EL84 amp
Unexpectedly good EL84 amp
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
Thread Tools Search this Thread
Old 16th August 2019, 11:39 PM   #151
john_tracy is offline john_tracy  United States
diyAudio Member
Join Date: Apr 2008
Location: Columbia, Missouri
I rebuilt this Eico HF-81 for a guy. He only wanted one input for his CD player. The tone controls were already bypassed. So out they went with the phono section. Added CCS fed shunt reg. for input and driver stage B+. There's something about shunt regulation that really picks up the sonics.
Attached Images
File Type: jpg IMG_3699.jpg (413.5 KB, 538 views)
File Type: jpg IMG_3706.jpg (318.7 KB, 536 views)
File Type: jpeg IMG_3704.jpeg (456.3 KB, 504 views)
"It is better to remain silent and thought a fool, then speak and remove all doubt." A. Lincoln
  Reply With Quote
Old 19th August 2019, 02:55 PM   #152
Lingwendil is online now Lingwendil  United States
diyAudio Member
Lingwendil's Avatar
Join Date: Apr 2008
Location: Oakley California
I think I may do shunt regulation on my next sweep tube amplifier- using the 12AV5GA and a DN2450 CCS fed VR tube for the screens. should be a nice way to go.
6SN7 Push-pull Flea Amplifier PCB files are now available!
  Reply With Quote
Old 20th October 2019, 04:45 PM   #153
chede is offline chede  Germany
... fascinated by amps
diyAudio Member
chede's Avatar
Join Date: Jan 2016
Location: In the Düsseldorf suburbs
Unexpectedly good EL84 amp
Default Tubecad EL84PP build documentation – optocoupler bias scheme

As already hinted at in my previous posts a few months back, I have been in the process of building this amp as well.

I wanted to build it in a modular fashion, so that later on I could try out different things. Maybe swap out the gain / concertina stage for a different front end, or try a different PP power stage, or a different biasing scheme …

Also, I wanted to try out a bias scheme a bit out of the ordinary. From my SS amp builds over at the Pass Labs forum, I am familiar with Nelson’s opto bias scheme, whereby an optocoupler is used in an automatic bias regulator for the output MOSFETs, for example as seen in the M2. I wanted to adapt this for the power tubes in this amp as well, and some SPICE simulation (my very first use of SPICE, actually) suggested that it might work. I adapted Nelson’s circuit to regulate grid-cathode voltage (fed from a negative bias supply) to keep cathode current constant, even under B+ variations (for example, due to line voltage variations), and always balanced, unaffected by tube aging.

Here is the little optocoupler automatic bias regulator module:

Unexpectedly good EL84 amp-optocoupler-bias-module-2-tubes-jpg

Unexpectedly good EL84 amp-testing-optocoupler-bias-servo-jpg

For the optocoupler to work, it needs a voltage of more than 1V, so for a typical bias of an EL84 a 33 Ohm current sense resistor would have to be placed into the circuit. I first tried it in the anode path, later in the cathode path. Simulation did not show any difference in regulation, testing worked well with both also. I chose to go with the cathode path because in that way I didn’t need to run the HV through the opto bias module. Maybe just a little bit safer.

I've also attached the circuit (in this iteration with current sensing in the anode connection) as a .pdf here.

The bias circuit has a time constant of maybe 15 seconds, so does only very slow bias adaptation. In static conditions, the bias scheme worked quite well both for a single tube and a PP pair.

Under load (tested with a sine input and an 8R dummy resistor at the output transformer secondary), I actually saw a problem. With each tube biased independently, the regulator dialed back the bias current under signal. If I remember correctly, it was something like that at about 2W output, the bias went from 36mA to maybe 26mA. That was not the only thing; it didn’t go down on both tubes the same amount, but slightly differently, causing unmatched bias under load. So when I later did distortion measurements on the completed amp, with a constant 1kHz sine signal, the output power went slowly down with the time constant of the bias circuit to some significantly lower value despite of constant input value, and the distortion figures were quite high at anything over 1W …

As I will also show later when I post about the build itself, I did a variation on the bias scheme that now actually works quite well. I use a common sense resistor for both tubes, in this case 16R. One side of the 16R goes to ground, the other to the wiper of a 20R trimmer that is used to balance the cathode currents of the EL84 pair. The other ends of the trimmer connect to the cathodes via a 1R resistor for current measurement.

In this way, I’ve lost the automatic balancing that independent bias circuits would have provided, but the DC bias (cathode current) is nicely stable, doesn’t vary with line voltage / B+ variations, and stays almost constant under load up to clipping. In a bias re-check after a few days / weeks, the cathode current is still at the value I set it. So, I would call this a partial success for “opto bias” so far.
Attached Images
File Type: jpg optocoupler bias module for 2 tubes.jpg (658.4 KB, 354 views)
File Type: jpg testing the optocoupler bias servo.jpg (569.5 KB, 350 views)
Attached Files
File Type: pdf EL84_Rsense anode PC817A mod 190605.pdf (14.0 KB, 30 views)

Last edited by chede; 20th October 2019 at 05:11 PM.
  Reply With Quote
Old 20th October 2019, 04:54 PM   #154
chede is offline chede  Germany
... fascinated by amps
diyAudio Member
chede's Avatar
Join Date: Jan 2016
Location: In the Düsseldorf suburbs
Unexpectedly good EL84 amp
Default Tubecad EL84PP build documentation - modular design

As stated earlier, I wanted to build the amp out of modules. I separated the front end from the EL84 PP module, and did build a small B+ buffer module as well, and bias modules.

All modules have been built on small glass fiber experimental boards, and have been wired using the component leads as much as possible. I used a lot of PCB screw connectors, both for connecting the modules together, and for mounting components that I might want to change later via the connectors, like cathode resistors, B+ dropping resistors, triode / pentode / UL connection, etc.

For output transfomers, I chose the following Toroidy EL84 PP transfomers:
TTG-EL84PP - Tube output transformer [8kOhm] 2xEL84 Push-pull or similar - Shop Toroidy.pl
I have had very good experiences with Toroidy power transformers, and for PP, a toroidal output transformer should also work well, if done correctly. There have been positive reports here about the Toroidy output transformers also. And, in their chrome cans, they just look cool …

Here is the front-end module being tested:

Unexpectedly good EL84 amp-testing-gain-concertina-module-jpg

And here the EL84PP module:

Unexpectedly good EL84 amp-testing-pp-el84-module-jpg

Here, I used the independent opto bias current regulators for each tube, and you can see the output transformer connected at the left of the picture.

Because I use an external PSU connected with an umbilical, I built a B+ buffer board, using just MKP capacitors. For the main B+, it’s 120 uF, for the frontend, an additional 35 uF. The capacitors are WIMA DC-LINK (which is the same as their MKP4 series, as I understand it). The dropping resistor between main B+ and frontend B+ is connected “externally” via the screw terminals for easy voltage adaptation / dialing-in of the frontend.

Unexpectedly good EL84 amp-buffer-module-jpg

I also built a small distribution module for the power inputs from the PSU (B+, heaters, negative bias) to connect to the other boards. On this board, there are some decoupling capacitors added for the heater supply and the negative bias voltage. The heaters are connected to ground here. Audio ground is connected to chassis ground via a CL-60.

Unexpectedly good EL84 amp-distribution-module-construction-jpg

Unexpectedly good EL84 amp-distribution-module-underside-jpg
Attached Images
File Type: jpg testing the gain - concertina module.jpg (848.5 KB, 345 views)
File Type: jpg testing the pp EL84 module.jpg (636.2 KB, 348 views)
File Type: jpg B+ buffer module.jpg (513.3 KB, 348 views)
File Type: jpg distribution module under construction.jpg (488.5 KB, 345 views)
File Type: jpg distribution module underside.jpg (197.1 KB, 362 views)
  Reply With Quote
Old 20th October 2019, 05:10 PM   #155
chede is offline chede  Germany
... fascinated by amps
diyAudio Member
chede's Avatar
Join Date: Jan 2016
Location: In the Düsseldorf suburbs
Unexpectedly good EL84 amp
Default Tubecad EL84PP build documentation – amp construction

Construction was done into a 2U Modushop Galaxy case. All modules were mounted “to the ceiling”, using the perforated base that is available as an extra from Modushop mounted at the top of the amp, with holes for the tubes cut into it.

Case ground is carried separately on the umbilical from the PSU, and connected to the amp case near the connector. At the PSU, it is connected to PE / earth ground at the IEC connector.

Back panel drilled, with the PSU umbilical connector in the middle. I used 7 of the 9 pins the connector offers.

Unexpectedly good EL84 amp-backpanel-drilled-jpg

I drew up all holes in the back panel, the perforated steel base, and the top panels via a CAD program, printed the drawings out in original size, and used these as a drilling guide.

First fitting test:

Unexpectedly good EL84 amp-fitting-test-jpg

One side wired up (frontend not yet wired)

Unexpectedly good EL84 amp-partly-wired-jpg

Testing one completed side:

Unexpectedly good EL84 amp-testing-complete-jpg

And the internal view of the completely stuffed amp:

Unexpectedly good EL84 amp-internal-view-amp-jpg

I mentioned in a previous post that my optocoupler-controlled bias scheme didn’t work out as intended when each tube was independently biased. So I changed to biasing a pair of tubes together via a common cathode current sensing resistor of 16R, and dialing in the balance between the two tubes using a 20R trimmer. Here is that little circuit added to the amp:

Unexpectedly good EL84 amp-common-cathode-resistor-jpg

Now it’s time to talk about the PSU. I built this separately, originally for my Aikido build, from parts I had still laying around. For the Aikido, it is complete overkill; for this EL84PP amp, it actually works nicely

Unexpectedly good EL84 amp-psu-top-jpg

Unexpectedly good EL84 amp-psu-internals-jpg

Heater supply is via separate transformer, here a Toroidy 8.5V 80W unit. The heater voltage is regulated by an LT1084 with big heatsink – in the middle of the case, below the ventilation slots.

B+ is provided by tube rectification, followed by the big 10H choke in the lower left, followed by 2x 680 uF Aerovox capacitors. For the Aikido, I configured the HV supply as LCC, fronted by a 5C3S rectifier tube, and followed by a regulation circuit that provided approx. 240V.

For the EL84 amp, I use a GZ34 (less voltage drop / more current capability), and currently a 5 uF MKP4 as first capacitor for cLCC. Might need a bigger first cap here, as I am only getting 300-320V B+, depending on bias current and line voltage. The regulator is taken out of the circuit for the EL84 amp.
Then there is also the negative bias voltage regulated by another LT1084. In the Aikido, this powered the control circuitry and relays.

Here is the almost completed amp running in my system, still without the top plates:

Unexpectedly good EL84 amp-amp-running-top-plate-jpg

Amp at night (without top plate):

Unexpectedly good EL84 amp-amp-night-top-plate-jpg

The green glow that’s coming through the holes in the base plate is from the frontend cathode biasing LEDs. This is not intentional, and in the finished amp it’s totally unobtrusive

Attached Images
File Type: jpg backpanel drilled.jpg (931.7 KB, 346 views)
File Type: jpg fitting test.jpg (617.3 KB, 345 views)
File Type: jpg one side partly wired up.jpg (327.7 KB, 341 views)
File Type: jpg testing one complete side.jpg (198.3 KB, 353 views)
File Type: jpg internal view of amp.jpg (307.3 KB, 337 views)
File Type: jpg common cathode resistor.jpg (211.8 KB, 359 views)
File Type: jpg PSU from top.jpg (761.9 KB, 340 views)
File Type: jpg PSU internals.jpg (525.1 KB, 337 views)
File Type: jpg amp running - without top plate yet.jpg (507.6 KB, 340 views)
File Type: jpg amp at night without top plate.jpg (535.3 KB, 339 views)
  Reply With Quote
Old 20th October 2019, 05:17 PM   #156
chede is offline chede  Germany
... fascinated by amps
diyAudio Member
chede's Avatar
Join Date: Jan 2016
Location: In the Düsseldorf suburbs
Unexpectedly good EL84 amp
Default Tubecad EL84PP build documentation – first conclusion

Here is a picture of the finished amp with the top plates, painted in green hammertone lacquer:

Unexpectedly good EL84 amp-finished-amp-top-plate-hammertone-paint-jpg

So, how does it sound ? I like it very much. Very clear, strong, clean bass, very transparent, lots of detail, 3D, engaging.
The amp by itself is very very quiet on my 100 dB/W widebanders. Almost undetectable amounts of noise and hum with my ear in the loudspeaker cone.

As a preamp, I’m using Wayne’s BAF 2018 linestage at the moment. (My Aikido is missing a PSU ... ) When I switch that on, I hear some noise (tube rush ?) out of the speakers that is at some time more and at some time less. When the system is running for a few hours, it usually is rather quiet again. The whole system overall has a bit too much gain, and the volume control sits in front of the preamp, which has a bandwidth of more than 1 MHz …

I plan to experiment with two things: one, so far there is only a connection between audio ground and case ground via a CL-60 (10 Ohms). I want to try a 100 nF capacitor between audio ground and case directly at the gain stage input.
Second, maybe I should limit the bandwidth of the EL84 amp at the input, but have no good idea how to do that at the moment.
The amp has a bandwidth at 1W up to about 160 kHz for -3dB. Great iron At the lower end, it gets noticeably distorted on the scope below 10 Hz, and I guess that is to be expected .

In the frontend, I am using very nice NOS Voskhod 6N1P-EV. Nicer than other 6N1Ps I tried. Have tried other tubes, but didn’t optimize the circuit for them to meet John Broskie’s operating points (5687, E182CC, 6N6P, E80CC). So I guess these others didn’t really have a chance … anyway, I’m really happy with the Voskhods.

For the power stage, I first tried a matched quad of JJ EL84s at 34mA. Nice mid, good clean and strong bass, a bit subdued highs. Right now, I’m running a quad (not matched) of 1979 Reflektor 6P14P-EV at 37mA. A bit stronger bass and much more prominent highs, but sound in balance throughout the spectrum. A bit more punch, but still very clean and engaging.

At the moment, this is my favourite amp, even over my SissySIT that I recently built (ZenMod probably would not like to hear that ) My 11 year-old son says that this is the best amp I built so far, but then, he is using the system with the Xbox …
Attached Images
File Type: jpg finished amp with top plate in hammertone paint.jpg (507.5 KB, 344 views)
  Reply With Quote
Old 21st October 2019, 06:51 AM   #157
Nightpuma is offline Nightpuma  Norway
diyAudio Member
Join Date: Apr 2006
Location: Norway
I am very impressed @chede:-)
  Reply With Quote
Old 21st October 2019, 05:06 PM   #158
carlthess40 is offline carlthess40  United States
diyAudio Member
Join Date: Oct 2017
Location: Lake Mary Florida
What would the cost be for you to make one
For someone?
  Reply With Quote
Old 22nd October 2019, 08:06 PM   #159
chede is offline chede  Germany
... fascinated by amps
diyAudio Member
chede's Avatar
Join Date: Jan 2016
Location: In the Düsseldorf suburbs
Unexpectedly good EL84 amp
Default Cost of amp and PSU

Hmm ... that is really not easy to say, because for most parts of the PSU, I used stuff that I had laying around from upgrading my diy 2A3SE amp, or that I bought as "byload" inexpensively from eBay (like choke and Aerovox caps) when the seller had some other interesting items besides my primary auction win.

What's also making it harder to calculate is the fact that when I order parts from Mouser, Reichelt, Arrow or somewhere else, I almost always order more than I need for that particular build - to have some spares, have more parts to match from, or just to build my inventory of parts for future builds or to whip up quickly some small circuit.

Let's try an estimate:

Toroidy TTG-EL84PP output transformers - 250 EUR incl. shipping
other transformers, choke and caps - maybe 200 EUR
two cases with the amp top plates cut to order - 100 EUR incl. shipping
all other parts and connectors - probably close to 200 EUR

So all in all, I guess it's around 750 EUR.

Connectors alone are about 60 EUR, especially the speaker connectors and the professional 9-Pin connector for the umbilical ain't cheap.

But then, I really built this amp as a platform to try more things later, which makes it much more complicated as necessary, and also more expensive - just for the amount of screw terminals alone ...
The PSU I already have adapted from usage with my Aikido, the adaptation has worked out quite well and easy. But it really is an overkill PSU.

And the Toroidy output transformers are just great ...

In order to just build the amp, I probably wouldn't go this modular route (again), way too much work; instead, probably try to find a PCB for this amp, or something like a Baby Huey, or a Red Light District amp, ...

That way it is probably much easier to build and less expensive.

Hope this helps,
regards, Claas
  Reply With Quote
Old 23rd October 2019, 07:10 AM   #160
Hilko Peter is offline Hilko Peter  Sweden
diyAudio Member
Join Date: Feb 2003
Location: Sjöbo, south of Sweden
Dear all,

This is my version of the EL84 amplifier. I tried to make it a small integrated amp and I'm pretty happy about it. As can be seen on the picture of the PCB, the 6N1P is LED biased while the EL84s use a LM317. I didn't expect The LM317 to get that hot by the 80 mA going through it so I had to mount a small heatsink on each afterwards. If I would have thought about it before i would have left more space for it on the PCB and the heatsink could have been fixed to the PCB (now its only connected to the LM317).
Output transformers are from Hammond (1650E) power transformer from Toroidy. Top and back and bottom plate are from schaeffer-ag.de (I do not have fun in drilling and cutting etc and they do it so much better for an ok price).
The rest is standard I suppose.
Buying all the parts new, and having PCBs and front panels made, doesn't make the project cheap, but it's a fun project, and I'm happy with the outcome.
It's nicely playing in our garden house/conservatory connected to a pair of super mini monitors ( W15CY001-OWI ) which makes a nice combo.

Attached Images
File Type: jpg IMG_2090.JPG (816.8 KB, 40 views)
File Type: jpg IMG_2091.JPG (782.9 KB, 54 views)
File Type: jpg IMG_2093.JPG (781.5 KB, 54 views)
File Type: jpg IMG_2095.JPG (462.9 KB, 47 views)
File Type: jpg IMG_2097.JPG (397.0 KB, 53 views)
  Reply With Quote


Unexpectedly good EL84 ampHide 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

Forum Jump

Similar Threads
Thread Thread Starter Forum Replies Last Post
FS: EL84 EI Yugoslavia & EL84 Tungsram NOS Bulk radiostar Swap Meet 3 29th January 2012 09:46 AM
what's a good layout for a quad EL84 amp on this box? freddi Tubes / Valves 3 18th July 2010 04:03 AM
1 pr. GE 6BQ5/EL84, barely good, $5 pjanda1 Swap Meet 2 23rd September 2008 10:32 PM
Speaker unexpectedly popping OUberLord Car Audio 10 1st September 2004 01:10 AM
triode => I.T. =>el84 good idea ? Jaap Tubes / Valves 9 7th January 2003 08:30 PM

New To Site? Need Help?

All times are GMT. The time now is 10:26 PM.

Search Engine Optimisation provided by DragonByte SEO (Pro) - vBulletin Mods & Addons Copyright © 2019 DragonByte Technologies Ltd.
Resources saved on this page: MySQL 14.29%
vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2019 DragonByte Technologies Ltd.
Copyright ©1999-2019 diyAudio