Suitability of ECL86/6GW8 in this design.
Hello All,
I recently gutted and old Viking console stereo for the 4 - 6GW8 in the output stage (among other things). I am wondering what your thoughts are in using these as substitutes in a slight variation of the design. The B+ supply I have on hand is +290VDC and I would prefer to replace these OT with something heavier than what came out of the original unit, something with UL taps instead of a separate supply for the screen.
I am also wondering what a good preamp stage with tone controls would look like for this amp...I have a PAS-3 that I can drive it with, but am more tempted to rebuild the original stage, adding a 12AX7 for the two triodes that I would move to the phase splitter.
Thanks in advance and happy breadboarding,
Andy
Hello All,
I recently gutted and old Viking console stereo for the 4 - 6GW8 in the output stage (among other things). I am wondering what your thoughts are in using these as substitutes in a slight variation of the design. The B+ supply I have on hand is +290VDC and I would prefer to replace these OT with something heavier than what came out of the original unit, something with UL taps instead of a separate supply for the screen.
I am also wondering what a good preamp stage with tone controls would look like for this amp...I have a PAS-3 that I can drive it with, but am more tempted to rebuild the original stage, adding a 12AX7 for the two triodes that I would move to the phase splitter.
Thanks in advance and happy breadboarding,
Andy
Andy,
If the output transformers you have don't have Ultralinear Taps then use Yvesm's original circuit. All the Baby Huey circuits grew out of this one. - See Post #48 for Yvesm's circuit.
http://www.diyaudio.com/forums/showthread.php?postid=552739#post552739
If you get replacement output transformers with Ultralinear Taps (Hammond 1608 or similar) then use this variant - See post #24 for schematic.
http://www.diyaudio.com/forums/showthread.php?s=&threadid=53969&highlight=
This amp went to a friend who is an opera fanatic. She uses it to drive a pair of Polk R5 and still claims she prefers it to all but the most expensive amps I've either bought or built since.
ECL86s are a seriously lovely tube - pity no one is remaking them.
Cheers,
Ian
If the output transformers you have don't have Ultralinear Taps then use Yvesm's original circuit. All the Baby Huey circuits grew out of this one. - See Post #48 for Yvesm's circuit.
http://www.diyaudio.com/forums/showthread.php?postid=552739#post552739
If you get replacement output transformers with Ultralinear Taps (Hammond 1608 or similar) then use this variant - See post #24 for schematic.
http://www.diyaudio.com/forums/showthread.php?s=&threadid=53969&highlight=
This amp went to a friend who is an opera fanatic. She uses it to drive a pair of Polk R5 and still claims she prefers it to all but the most expensive amps I've either bought or built since.
ECL86s are a seriously lovely tube - pity no one is remaking them.
Cheers,
Ian
Ian,
Thanks a tonne for the link to the variant with UL taps. This is exactly what I am looking for. I am considering a pair of Hammond 1608's for this tube as well, though in my experience, Hammond are good OT's, but certainly not the best. But they do have price and availability on their side (vs Sowter @ $180 US ea!)....I was also looking at a set of Dyna-clone Z565 for this amp, since I have heard good things about them, though they may be overkill given the power output here. I'll let you know what I come up with.
The tubes I salvages were from an old Dominion Electrohome Florentine console stereo...browsing the forums, I noticed that Planet10 (Dave in Victoria) posted a picture of exactly the same chassis that I have (though I think his was branded as a Viking Laurentian). In any case, the 4 6GW8's I got are from the "Selectron Tube Company" and are allegedly made in Japan. They all tested very strong at the old tube tester at one of the electronic supply stores that I go to, though I do wonder about the overall accuracy of the thing (its one of those self-serve testers with a red-and-green, good-and-bad scale which is pretty much meaningless). I wish I had something better since I would hate to build the amp only to have to re-tube.
Anyway, thanks again for your help and I'll try to keep you posted with whatever progress I make.
Andy
Thanks a tonne for the link to the variant with UL taps. This is exactly what I am looking for. I am considering a pair of Hammond 1608's for this tube as well, though in my experience, Hammond are good OT's, but certainly not the best. But they do have price and availability on their side (vs Sowter @ $180 US ea!)....I was also looking at a set of Dyna-clone Z565 for this amp, since I have heard good things about them, though they may be overkill given the power output here. I'll let you know what I come up with.
The tubes I salvages were from an old Dominion Electrohome Florentine console stereo...browsing the forums, I noticed that Planet10 (Dave in Victoria) posted a picture of exactly the same chassis that I have (though I think his was branded as a Viking Laurentian). In any case, the 4 6GW8's I got are from the "Selectron Tube Company" and are allegedly made in Japan. They all tested very strong at the old tube tester at one of the electronic supply stores that I go to, though I do wonder about the overall accuracy of the thing (its one of those self-serve testers with a red-and-green, good-and-bad scale which is pretty much meaningless). I wish I had something better since I would hate to build the amp only to have to re-tube.
Anyway, thanks again for your help and I'll try to keep you posted with whatever progress I make.
Andy
Andy,
I rebuilt a Rogers Cadet III output stage into a "Baby Huey" and it sounds fantastic! It uses two ECL86/6GW8 tubes and no UL trafo just the original Rogers trafos which are rather slight affairs but I think the circuit excels in the Bass region. See my posts earlier in this thread for details - thanks to Yvesm & Ian for this circuit!
I rebuilt a Rogers Cadet III output stage into a "Baby Huey" and it sounds fantastic! It uses two ECL86/6GW8 tubes and no UL trafo just the original Rogers trafos which are rather slight affairs but I think the circuit excels in the Bass region. See my posts earlier in this thread for details - thanks to Yvesm & Ian for this circuit!
My build of the Baby Huey
Hi Gingertube,
I'd like to give you a feedback about my build of the Baby Huey amp.
I converted my Eico HF86 clone with Hammond 1650E iron to your topology.
In my build there's no global feedback, the cross-connect resistor value is 16k, the cap on the EL84-CCS is 470uF.
I can confirm several of your remarks you made in this thread:
- It just sounds terrific (read: balanced, subtle, precise, not tubey, very enjoyable to listen to during hours)
- It has no preference for particular music styles, they all sound good on the Baby Huey.
- It's a hard competitor to my other amps(2A3 push-pull with Lundahl OPT and Lundahl interstage costing 4x more)
- It's nearly indifferent to speakers (it mastered very well my 92db 3-way speakers and did an excellent job on my Fostex FE-103s
and Fe-168s speakers, the Fostexes usually are a bit critical on amplifier damping)
- It sounds better than the sum of its parts should.
The Baby Huey amp adds very little coloration on its own. I found out during tube rolling that that it shows very
clearly the sonic signatures of different ECC83 and EL84 brands often described by people on the net. In my case an industrial
Philips 12AX7S and a vintage Tungsram EL84 sounded best. The russian military 6P14P-EB types was a bit harsh in comparison but had
a very tight bass. The old NOS tubes all performed better here than the current production ones I had on hand.
(I didn't test the Mullard EL84 reissue although I'm tempted to do so)
I modified the periphery of the amp a bit (see the attached schematic):
- A dedicated reference voltage was added to the EL84's CCS to eliminate the minimal fluctuations (2..3 mA)
during operation of the amp.
- The ECC83's CCS was modified to a Morgan Jones CCS with improved caracteristics.
- Those reference voltages are generated by two LM317's.
- The B+ high voltage rail is MOSFet regulated to 300V with minimal ripple for each channel.
The B+ MOSfet regulation brought a subtle improvement in the stereo-imaging.
As for my CCS modifications i can't tell, as i never built the original CCS.
One of the biggest improvements in sound was the replacement of the 470uF cathode caps by Blackgates (I hate to say it...).
The Blackgates sounded awful for the first ten hours (so much that I first suspected the B+ Mosfet regulation and disconnected it).
After the burn-in of the Blackgates the sound was back to normal and then went up to excellent.
My distortion measurements with a Hameg HM8027/HM8037 distortion meter indicated:
Distortion: 0.05% at 0.5Vpp Output on 8 Ohms resistor (connected to 8 ohm tap on Hammond 1650E)
0.15% at 4Vpp Output on 8 Ohms
0.46% at 8Vpp Output on 8 Ohms
Clipping came in at around 10Vpp output
The distortion figures were very sensitive to the ECC83 sections being well matched (the balance pot. setting on the ECC83's
cathodes wouldn't change the distortion figures).
The square wave response above 10kHz was not too nice on the Hammond 1650E (well we're without global feedback here), but i've seen
Hammond 1615's doing better on a very similar setup with 5902 subminiature pentodes. The 1650E seems to suffer for it's particular
winding scheme here.
As I think that this amp deserves a better iron, i ordered higher quality 8k-OPT's from my favorite german coil winder. So far they
didn't arrive, so I can't comment if the square wave response improves.
I'll attach some pictures of my amp with it's separate power supply. It's compact but a major PITA to work inside, i would
recommend using a larger chassis than I did. The heatsinks on the amp are much larger than they should be, but I had them on
hand and did like the optics.
To sum up: I was hooked on DHT-triodes, but a pentode amp done right made me rethink my convictions.
Thanks for sharing your work Gingertube! (and YvesM for sharing his ECL86 precursor)
Hi Gingertube,
I'd like to give you a feedback about my build of the Baby Huey amp.
I converted my Eico HF86 clone with Hammond 1650E iron to your topology.
In my build there's no global feedback, the cross-connect resistor value is 16k, the cap on the EL84-CCS is 470uF.
I can confirm several of your remarks you made in this thread:
- It just sounds terrific (read: balanced, subtle, precise, not tubey, very enjoyable to listen to during hours)
- It has no preference for particular music styles, they all sound good on the Baby Huey.
- It's a hard competitor to my other amps(2A3 push-pull with Lundahl OPT and Lundahl interstage costing 4x more)
- It's nearly indifferent to speakers (it mastered very well my 92db 3-way speakers and did an excellent job on my Fostex FE-103s
and Fe-168s speakers, the Fostexes usually are a bit critical on amplifier damping)
- It sounds better than the sum of its parts should.
The Baby Huey amp adds very little coloration on its own. I found out during tube rolling that that it shows very
clearly the sonic signatures of different ECC83 and EL84 brands often described by people on the net. In my case an industrial
Philips 12AX7S and a vintage Tungsram EL84 sounded best. The russian military 6P14P-EB types was a bit harsh in comparison but had
a very tight bass. The old NOS tubes all performed better here than the current production ones I had on hand.
(I didn't test the Mullard EL84 reissue although I'm tempted to do so)
I modified the periphery of the amp a bit (see the attached schematic):
- A dedicated reference voltage was added to the EL84's CCS to eliminate the minimal fluctuations (2..3 mA)
during operation of the amp.
- The ECC83's CCS was modified to a Morgan Jones CCS with improved caracteristics.
- Those reference voltages are generated by two LM317's.
- The B+ high voltage rail is MOSFet regulated to 300V with minimal ripple for each channel.
The B+ MOSfet regulation brought a subtle improvement in the stereo-imaging.
As for my CCS modifications i can't tell, as i never built the original CCS.
One of the biggest improvements in sound was the replacement of the 470uF cathode caps by Blackgates (I hate to say it...).
The Blackgates sounded awful for the first ten hours (so much that I first suspected the B+ Mosfet regulation and disconnected it).
After the burn-in of the Blackgates the sound was back to normal and then went up to excellent.
My distortion measurements with a Hameg HM8027/HM8037 distortion meter indicated:
Distortion: 0.05% at 0.5Vpp Output on 8 Ohms resistor (connected to 8 ohm tap on Hammond 1650E)
0.15% at 4Vpp Output on 8 Ohms
0.46% at 8Vpp Output on 8 Ohms
Clipping came in at around 10Vpp output
The distortion figures were very sensitive to the ECC83 sections being well matched (the balance pot. setting on the ECC83's
cathodes wouldn't change the distortion figures).
The square wave response above 10kHz was not too nice on the Hammond 1650E (well we're without global feedback here), but i've seen
Hammond 1615's doing better on a very similar setup with 5902 subminiature pentodes. The 1650E seems to suffer for it's particular
winding scheme here.
As I think that this amp deserves a better iron, i ordered higher quality 8k-OPT's from my favorite german coil winder. So far they
didn't arrive, so I can't comment if the square wave response improves.
I'll attach some pictures of my amp with it's separate power supply. It's compact but a major PITA to work inside, i would
recommend using a larger chassis than I did. The heatsinks on the amp are much larger than they should be, but I had them on
hand and did like the optics.
To sum up: I was hooked on DHT-triodes, but a pentode amp done right made me rethink my convictions.
Thanks for sharing your work Gingertube! (and YvesM for sharing his ECL86 precursor)
Attachments
dupyv,
Thanks for sharing your experience. When we design and build an amp we are obviously predjudiced in its favour and so a second independent opinion is always valuable.
Could you post your schematic?
I'm in the process of building another incarnation of the Baby Huey, fixed bias outputs into Hammond 1608 Output Trannies. Choke input power supply and mosfet followers between the diffamp and the output tube grids. One monoblock fired up for the first time last night, the other is yet to be completed (I'll finish tweaking the first before tackling the rest of the buld of the second monoblock to minimise rework). So far I'm only at the sine wave into a dummy load and staring at the oscilloscope point. I did note that with B+ of 350V I was getting 13 Watts out and the 10 Watt rated 1608 was showing some waveform distortion below about 35 Hz due to core saturation. That was why I originally tried a 1650E (15 Watt rated) in the "Jessica" variant.
Cheers,
Ian
Thanks for sharing your experience. When we design and build an amp we are obviously predjudiced in its favour and so a second independent opinion is always valuable.
Could you post your schematic?
I'm in the process of building another incarnation of the Baby Huey, fixed bias outputs into Hammond 1608 Output Trannies. Choke input power supply and mosfet followers between the diffamp and the output tube grids. One monoblock fired up for the first time last night, the other is yet to be completed (I'll finish tweaking the first before tackling the rest of the buld of the second monoblock to minimise rework). So far I'm only at the sine wave into a dummy load and staring at the oscilloscope point. I did note that with B+ of 350V I was getting 13 Watts out and the 10 Watt rated 1608 was showing some waveform distortion below about 35 Hz due to core saturation. That was why I originally tried a 1650E (15 Watt rated) in the "Jessica" variant.
Cheers,
Ian
Maybe this was refernced before, but I just came accross this 2001 issue of Tubecad which has a long article on Partial Feedback & it's uses http://www.tubecad.com/march2001/
The same configuration as is being used in the baby huey circuits, I think
The same configuration as is being used in the baby huey circuits, I think
jkeny,
Yes it is similar to what JB calls partial feedback - the feedback works in the same way - that is we trade output tube gm for reduced rp. That is why The Baby Huey works best with higher Gm output tubes like the EL84. That hasn't stopped me using the scheme with lower gm tubes like 6V6 but it doesn't work as well with them.
dupyv,
Last night I had a bit of a mess around with the 6V6 version of the Baby Huey which is running Hammond 1650E Output Transformers. I run 6 Ohm nominal speakers off the 4 Ohm tap. I tried adding 10 Ohms plus 1, 2 and 3 uF zobels across the 4 Ohm tap. Without doing any formal testing but just listening I thought the 10 Ohm plus 2.2uF cleared up the top end a bit and the whole sound was a bit better balanced. I then decided to put on the Queen's Rock albumn. It sounded punchier without the zobel.
On one of the earlier amps (6V6 again but with Hammond 1608 in that case) I tried lots of zobels from anode to screen connections of the output transformer and in the end settled for just a 10 Ohm plus 2.2uF zobel across the 4 Ohm secondary.
I only bring this to your attention as something to try to see if you like it.
For your info I'm using 1000uF 25V Blackgate for Cathode bypass. Like you (I think) I originally thought that spending all that extra cash on an eletrolytic must be a waste of money. Like you I was converted once I tried them - they do take time to break in. its really easy to tell when they have finished breaking in - all of a sudden you have a whole extra octave in the bottom end. I have a couple of solid state amps which use a capacitively coupled bootstrap on the Voltage Amplifier Stage (VAS). Blackgates on them turned the amps from something fairy good to something quite stunning. In those amps it was not just the improved frequency response which stood out but the incredibly low noise of the Blackgate was very apparent. It was like a whole extra veil had been removed. I think this is noticable on the tube amps as well.
Aside: For those of you not familiar with Blackgates - their noise level is 30 to 50 dB better than ordinary electrolytics, its really noticable.
Cheers,
Ian
Yes it is similar to what JB calls partial feedback - the feedback works in the same way - that is we trade output tube gm for reduced rp. That is why The Baby Huey works best with higher Gm output tubes like the EL84. That hasn't stopped me using the scheme with lower gm tubes like 6V6 but it doesn't work as well with them.
dupyv,
Last night I had a bit of a mess around with the 6V6 version of the Baby Huey which is running Hammond 1650E Output Transformers. I run 6 Ohm nominal speakers off the 4 Ohm tap. I tried adding 10 Ohms plus 1, 2 and 3 uF zobels across the 4 Ohm tap. Without doing any formal testing but just listening I thought the 10 Ohm plus 2.2uF cleared up the top end a bit and the whole sound was a bit better balanced. I then decided to put on the Queen's Rock albumn. It sounded punchier without the zobel.
On one of the earlier amps (6V6 again but with Hammond 1608 in that case) I tried lots of zobels from anode to screen connections of the output transformer and in the end settled for just a 10 Ohm plus 2.2uF zobel across the 4 Ohm secondary.
I only bring this to your attention as something to try to see if you like it.
For your info I'm using 1000uF 25V Blackgate for Cathode bypass. Like you (I think) I originally thought that spending all that extra cash on an eletrolytic must be a waste of money. Like you I was converted once I tried them - they do take time to break in. its really easy to tell when they have finished breaking in - all of a sudden you have a whole extra octave in the bottom end. I have a couple of solid state amps which use a capacitively coupled bootstrap on the Voltage Amplifier Stage (VAS). Blackgates on them turned the amps from something fairy good to something quite stunning. In those amps it was not just the improved frequency response which stood out but the incredibly low noise of the Blackgate was very apparent. It was like a whole extra veil had been removed. I think this is noticable on the tube amps as well.
Aside: For those of you not familiar with Blackgates - their noise level is 30 to 50 dB better than ordinary electrolytics, its really noticable.
Cheers,
Ian
Talking of the bypass caps. I have picked up and ran with an idea of Brian Becks which could be applied to the Baby Huey. If you take the cathodes of the bypass caps and tie them together, then attach a 1meg resistor from this node to earth, you have turned the output stage differential (ie current has no other choice than to travel from one tubes cathode to the other). This requires you double the value of the cap as they are now in series. Add a high quality film cap between the tube cathodes for good measure.
I noticed a big improvement to the overall sound after this mod. I would be interested in anyone out there lashing it up and reporting on their findings.
Shoog
I noticed a big improvement to the overall sound after this mod. I would be interested in anyone out there lashing it up and reporting on their findings.
Shoog
Gingertube,
My second thought on the blackgates was to eliminate them completely by switching over to fixed bias.
If there's no cap on the cathode, nothing should be added/removed sonic wise.
Or the blackgates add a sonic pattern to the sound that is pleasing to the ear.
Speaking of zobels, i'll wait for my new 8k-OPT to arrive from Germany,
and give them a listen before putting in the zobels. I expect this new OPT to behave a lot better on the measurement bench. If I can hear the difference is another story but due to the bigger core size, the lower octave should be extended.
Kind regards,
Yves
My second thought on the blackgates was to eliminate them completely by switching over to fixed bias.
If there's no cap on the cathode, nothing should be added/removed sonic wise.
Or the blackgates add a sonic pattern to the sound that is pleasing to the ear.
Speaking of zobels, i'll wait for my new 8k-OPT to arrive from Germany,
and give them a listen before putting in the zobels. I expect this new OPT to behave a lot better on the measurement bench. If I can hear the difference is another story but due to the bigger core size, the lower octave should be extended.
Kind regards,
Yves
Hi Yves,
That's what I've chosen in the original ECL86 implementation
40Hz square:
http://www.dissident-audio.com/PP_ECL86/Square_40Hz.pdf
10KHz square:
http://www.dissident-audio.com/PP_ECL86/Square_10KHz.pdf
No zobels nor UL taps but both NFB loops used.
Yves.
I firmly second this idea !dupyv said:
That's what I've chosen in the original ECL86 implementation
Why not ? Tell us !
Could you reveal the name of the winder ?
Two samples from my own implementation:
40Hz square:
http://www.dissident-audio.com/PP_ECL86/Square_40Hz.pdf
10KHz square:
http://www.dissident-audio.com/PP_ECL86/Square_10KHz.pdf
No zobels nor UL taps but both NFB loops used.
Yves.
YvesM,
I knew you'd jump on the fixed bias issue...
I'll give the fixed bias a try if i manage to pull a cable from the PSU
to the amp and install some bias pots in that small chassis.
The OPT's I ordered are from Gerd Reinhöfer (www.roehrentechnik.de), type 53.15, he even pots them upon request.
I own some of his smaller units and they sound and measure excellent.
Craftsmanship is also outstanding.
As OPT winding is not his daily job, delivery is a bit slow though. You need lots of patience.
Kind regards,
Yves
I knew you'd jump on the fixed bias issue...
I'll give the fixed bias a try if i manage to pull a cable from the PSU
to the amp and install some bias pots in that small chassis.
The OPT's I ordered are from Gerd Reinhöfer (www.roehrentechnik.de), type 53.15, he even pots them upon request.
I own some of his smaller units and they sound and measure excellent.
Craftsmanship is also outstanding.
As OPT winding is not his daily job, delivery is a bit slow though. You need lots of patience.
Kind regards,
Yves
I built a ECL86 Baby Huey & love it's sound - now I want to have a go at an EL84 version but with some experimental twists (but I'm a novice here so advice appreciated)?.
- I would like to try some PS toroid trafos as OT - can anybody recommend specs: Pri,Sec voltages & VA rating?
- I would also like to try Shoogs idea of making the output stage differential - has anybody tried this? Ian, are Blackgates really that good for the money?
- Finally, but I don't think this is a runner, Has anybody tried to use a Beam Deflection Tube (like 6AR8) on the front end as phase splitter & vol control (see John Swenson BDT preamp) - I presume there is not enough MU in these tubes to work in this position
Datasheet Here: http://www.mif.pg.gda.pl/homepages/frank/sheets/093/6/6AR8.pdf
- I would like to try some PS toroid trafos as OT - can anybody recommend specs: Pri,Sec voltages & VA rating?
- I would also like to try Shoogs idea of making the output stage differential - has anybody tried this? Ian, are Blackgates really that good for the money?
- Finally, but I don't think this is a runner, Has anybody tried to use a Beam Deflection Tube (like 6AR8) on the front end as phase splitter & vol control (see John Swenson BDT preamp) - I presume there is not enough MU in these tubes to work in this position
Datasheet Here: http://www.mif.pg.gda.pl/homepages/frank/sheets/093/6/6AR8.pdf
Is there a reason that the triode/pentode sections in Yves' ECL 86 amp are crossed? Why does the triode section of valve 1 go to the pentode section of valve 2?
A propos of Baby Huey: will 5751s have enough gain? I've got some nice RCAs.
Ian, have you finished the fixed bias/source follower version?
I'm starting in on the Baby Huey. I've always liked the sound of EL84s in UL. Will compare it to the 6GK5 to triode EL34 PP monos I built based on a schematic kindly shared by Poinz of Audio Tropic...
A propos of Baby Huey: will 5751s have enough gain? I've got some nice RCAs.
Ian, have you finished the fixed bias/source follower version?
I'm starting in on the Baby Huey. I've always liked the sound of EL84s in UL. Will compare it to the 6GK5 to triode EL34 PP monos I built based on a schematic kindly shared by Poinz of Audio Tropic...
5751's should work fine.
One monoblock done and working, I need to finish off the second before doing final tweaks. Progress is slow.
At work I'm in the process of finalising testing, then install into the aircraft, of a new Laser Airborne Depth Sounder System which we have designed an built over the last 2 years. By the time I get home I've had electronics right up to the back teeth. This weekend comming up is a long weekend for us (Monday is a public holliday) so perhaps I can get more done then.
Cheers,
Ian