| arnoldc |
I made my first ever online order for a tube project some 4 years ago. That order is for a JE Labs 2A3 which was sent to Angela.com which also contained a pair of 8532 and 1626 - parts for the Darling amp.
I haven't put those tubes to use since the order arrived.
Yesterday, while I was waiting for my students to finish ther labs, I started scribbling something, which ended up like this-
(photo was taken after I completed the amp) |
|
|
| arnoldc |
My chassis has only one hole sized for for a 9-pin tube, so I cannot use my 8532 on this circuit, as they're 7-pin and single triode. My bias towards the 407A sonics on my preamp led me to use a close "cousin" - the 2C51 (while I also have the Western Electric 396A, it will be put to service later) which is a 9-pin dual triode with 6.3V filaments (unlike the 20/40V of the 407A).
I got good results using the diyAudio CCS in my other project, so why not use it again. While my "RS Components design tool" shows the CCS set at 8mA, the actual circuit had 10mA because I don't have a 125R resistor. Tonight, I will put in a Bourns multi-turn variable resistor so I can dial it in at 8mA.
SY was a big influence on my choice with regards to biasing, so the 2C51 was biased using Yellow LEDs which should give me 2V or so.
I decided not to re-invent the wheel and went ahead with the original bias scheme of the 1626- a shared 500R, bypassed by a capacitor.
I have some left-over power transformer and output transformers (5K:8/16R) from a previous 807 project which I re-used for this amp.
Just before midnight I was able to complete assembly, did some voltage measurements which came very close to my scribble. With that, I hooked it up to my speakers and played some music. |
|
|
| arnoldc |
Like most of my projects, there are no exotic parts, even though I'm tempted to use my Riken resistors and MultiCap RTX coupling caps.
You'll see on the photo that my 1626 cathode capacitor is a white-coffin bypassed with a generic 220uF electrolytic. The coupling caps are Auricaps, and the power supply is a simple CRCRC with generic electrolytics and white coffin as well.
By the way, there's 1K grid stopper to the grid of the 1626 (not in schematic) because my 220K grid leak resistors have too short leads :smash:
So how does it sound? It has a different character which I tend to like, sweet and mellow, for a change. I can't play really loud as my neighbors might complain but the vocals reproduction is glorious. It does not have the punch and kick of my 2A3 or 45, but it does bass deep enough to give me some satisfaction. |
|
|
| Sal Brisindi |
Great job on the amp, nice, simple yet elegant. I have parts to build a Darling so one of these days, month or years I will get to it... :-)
Regards,
Sal Brisindi |
|
|
| alexg |
That amp looks great!
Is the 2C51 similar to 5670? How much voltage swing you need to drive the 1626?
I would like to make one like your Darling!
Arnold, question: which in your opinion would sound better a 1626 or a 12b4 flea powered amp? Thanks. |
|
|
| arnoldc |
Hi Sal, yes, it took me four years to do this :D
Hi Alex, you're right about 5670. I just need some 60Vp-p to drive the 1626.
On the 1626 vs. 12B4, I can't comment because have not made a 12B4 amp yet, but you're tempting me. However, the 1626 with its ST-bottle is physically very attractive for me ;) |
|
|
| alexg |
| quote: | Originally posted by alexg
That amp looks great!
Is the 2C51 similar to 5670? How much voltage swing you need to drive the 1626?
I would like to make one like your Darling!
Arnold, question: which in your opinion would sound better a 1626 or a 12b4 flea powered amp? Thanks. |
I know you like the 12b4 and I thought about you yesterday while looking at a 12b4 amp circuit, producing almost same power as the 1626. :D |
|
|
| arnoldc |
| Which one is that? The one driven by a 6AU6? |
|
|
| alexg |
| quote: | Originally posted by arnoldc
Which one is that? The one driven by a 6AU6? |
I have not seen the one with the 6au6 driver, the one I was looking at is driven by half a 12ax7. It said to produce 0.9watt.
Am really tempted to do this 12b4 amp.
:D |
|
|
| arnoldc |
| Ah, you're talking about the Abraxas audio I believe. Did he designed it for full range use? |
|
|
| alexg |
| quote: | Originally posted by arnoldc
Ah, you're talking about the Abraxas audio I believe. Did he designed it for full range use? |
I believe he did. His impression of the quality of the sound, conforms with the opinions I read about the 12b4 on preamp duties, transparent and uncolored. |
|
|
| arnoldc |
I'm just curious because, the 12B4 has low Rp and he used 24K (not sure) OPT primary.
The 6AU6->12B4 I mentioned used 5K OPT.
Why not make one, tonight!
:D |
|
|
| bloozestringer |
I've got the 2nd Abraxas 12B4. You can see it here:
http://www.pedalforroses.homestead.com/PowerAmps.html
I've heard the 1st one as well (the one on his website) on a set of PAWO's utilizing a DCA 4" speaker. It sounded very good! I've also heard my 12B4 side by side with Jef's D5 Darling on my Pi Two Speakers.
The 12B4 has a much tighter bass and clearer high end to my ears. The Darling did vocals a little better and the overall sound was a little more "wooly" for lack of a better term. Both sounded great though, just different.
My 12B4 uses a 14K primary on the OPT's.
I've got all the parts but a chassis and OPT's to do a 5Y3 rectified Darling using a 5965 driver. I really need to finish it. |
|
|
| kevinkr |
Hi Arnold,
Take a look at the dc darling sometime I think you'll find it an even better performer than the standard ac coupled darling. Another possibility is to run you current 1626 design with fixed bias and eliminate the wonky cathode bypass cap.. (My preferred choice in all amp designs.)
Very nice job by the way. :D
Another very nice tube to experiment with in the sub 1W class is the 171 or 71A, may be a little hard to find, but very nice sound. I drive a pair with led bias choke loaded 5842 in my flea power headphone amplifier. |
|
|
| alexg |
| quote: | Originally posted by arnoldc
I'm just curious because, the 12B4 has low Rp and he used 24K (not sure) OPT primary.
The 6AU6->12B4 I mentioned used 5K OPT.
Why not make one, tonight!
:D |
Am still debating if I want to do the Darling or the 12B4.
Can't make it tonight, don't have the tubes and the OPT! :D
All I have are some 6n1p, 12au7, 6dj8, 12j5 and 5670. I will have a pair of 12b4 next week though, I can "borrow" my SE KT88 OPT (5K primary) and borrow my 832A power transformers. :)
BTW, am trying to look for the 6AU6->12B4 amp on the net, can't find it. Can you point me to the site? Thanks. |
|
|
| arnoldc |
| quote: | Originally posted by bloozestringer
<snip>
The 12B4 has a much tighter bass and clearer high end to my ears. The Darling did vocals a little better and the overall sound was a little more "wooly" for lack of a better term. Both sounded great though, just different.
My 12B4 uses a 14K primary on the OPT's. |
Hi, thanks for the comparative overview between the Darling and 12B4. I'll ask my transformer winder to make me one with 14K primary. Nice amp you got there too.
| quote: | Originally posted by kevinkr
Hi Arnold,
Take a look at the dc darling sometime I think you'll find it an even better performer than the standard ac coupled darling. Another possibility is to run you current 1626 design with fixed bias and eliminate the wonky cathode bypass cap.. (My preferred choice in all amp designs.)
Very nice job by the way. :D
Another very nice tube to experiment with in the sub 1W class is the 171 or 71A, may be a little hard to find, but very nice sound. I drive a pair with led bias choke loaded 5842 in my flea power headphone amplifier. |
Thanks Kevin, and you know what? I've been searching for the 71A! No luck so far.
Last night I was listening to my Darling and on most of my music, it's running out of dynamics, it's just too mellow for me.
I've heard a friend's Darling and DC Darling and I preferred the DC version myself. But yes, why not? I haven't made a fixed bias amp in my life :xeye: so why not this one. Yes!
| quote: | Originally posted by alexg
Am still debating if I want to do the Darling or the 12B4.
Can't make it tonight, don't have the tubes and the OPT! :D
All I have are some 6n1p, 12au7, 6dj8, 12j5 and 5670. I will have a pair of 12b4 next week though, I can "borrow" my SE KT88 OPT (5K primary) and borrow my 832A power transformers. :)
BTW, am trying to look for the 6AU6->12B4 amp on the net, can't find it. Can you point me to the site? Thanks. |
Make both and compare! That schema is on my offline storage. I'll email it to you later.
This is the nice part of having a speaker that can be driven by half a watt. I can try all flea-powered amps with gusto. |
|
|
| mach1 |
The 12B4 sounds really excellent in parafeed mode using cheap line output transformers (10W or 15W) and a decent choke (handmade H3007 or Magnequest BCP-15 - the hammond 157G will also work). With 7-8k line transformers run it around 225 - 240B+ with a 1.2k cathode resistor - the 6au6 version operating point is a bit too low.
As previously mentioned, it has an entirely different sound to the darling. Very clear and punchy with a surprisingly potent bottom end. |
|
|
| arnoldc |
That's another thing I haven't done... Parafeed!
God, this DIY thing is getting to me again :D |
|
|
| alexg |
| quote: | Originally posted by arnoldc
That's another thing I haven't done... Parafeed!
God, this DIY thing is getting to me again :D |
Go Arnold GO, you know you want to do it! :D
I am already talking to Edrel for the irons of a parafeed 12b4. ;) |
|
|
| arnoldc |
Thanks for the encourgment, Alex :D :D
I'd like to do a Parafeed without the plate choke to keep things small. Again, I'd like to use the diyAudio CCS for ths duty. I hope it works :D
If this works, then I can parafeed my 417A headphone amp also :cool: |
|
|
| alexg |
| quote: | Originally posted by arnoldc
Thanks for the encourgment, Alex :D :D
I'd like to do a Parafeed without the plate choke to keep things small. Again, I'd like to use the diyAudio CCS for ths duty. I hope it works :D
If this works, then I can parafeed my 417A headphone amp also :cool: |
Ah ok, so I can get rid of the plate choke use plate resistor, and still have parafeed? Thanks for this info, my understanding was I need both the plate choke and the OPT. :xeye:
Thanks, I have been learning a lot! |
|
|
| alexg |
Thanks for the link.
I have been reading parafeed output topology and most of the materials I read states that I can "use smaller OPT since there will be no DC on it."
How smaller can it be? Let us say that the 12b4 has 35ma on the plate, what will be the specs of the SE opt?
Thanks. |
|
|
| arnoldc |
| I'm not in a position to answer that. You should ask your transformer winder. As an example, look at EXO-35 from MagneQuest. |
|
|
| alexg |
| quote: | Originally posted by arnoldc
I'm not in a position to answer that. You should ask your transformer winder. As an example, look at EXO-35 from MagneQuest. |
Thanks. My winder have never wound a parafeed OPT.
Thanks again, I will just try to compute/research the voltage and current after the capacitor on a parafeed circuit. |
|
|
| alexg |
| quote: | Originally posted by mach1
The 12B4 sounds really excellent in parafeed mode using cheap line output transformers (10W or 15W) and a decent choke (handmade H3007 or Magnequest BCP-15 - the hammond 157G will also work). With 7-8k line transformers run it around 225 - 240B+ with a 1.2k cathode resistor - the 6au6 version operating point is a bit too low.
As previously mentioned, it has an entirely different sound to the darling. Very clear and punchy with a surprisingly potent bottom end. |
Thanks for this info, I really want to try the parafeed 12b4, can you please share the schema?
Thanks. |
|
|
| mach1 |
I have built a few of these with different drivers. You can use virtually any triode with a mu > 35 to drive the 12B4. The 6EJ7 is available, cheap and sounds excellent triode strapped.
The circuit sounds good with non-boutique parts.
Start with a cheap 10 or 15W line transformer as output. I guarantee you will be surprised. They should cost you between $10 and $15 each.
I credit Tom Stoppleworth with the 12B4 parafeed idea. |
|
|
| alexg |
| quote: | Originally posted by mach1
I have built a few of these with different drivers. You can use virtually any triode with a mu > 35 to drive the 12B4. The 6EJ7 is available, cheap and sounds excellent triode strapped.
The circuit sounds good with non-boutique parts.
Start with a cheap 10 or 15W line transformer as output. I guarantee you will be surprised. They should cost you between $10 and $15 each.
I credit Tom Stoppleworth with the 12B4 parafeed idea. |
How much current do I need for the 12B4 on your schematic?
Thanks. |
|
|
| planet10 |
| quote: | Originally posted by alexg
I have been reading parafeed output topology and most of the materials I read states that I can "use smaller OPT since there will be no DC on it."
How smaller can it be? |
Bottlehead uses a 70 V line matching transformer in the Paramour.
dave |
|
|
| mach1 |
| each 12B4 pulls around 25mA. |
|
|
| Tony |
| quote: | | My winder have never wound a parafeed OPT. |
Don'nt need to, you can use the 110/220 volt primary, 12volt multi-tap secondary 500mA traffo which can be bought for around 100 pesos and you are in business...
:D
using the 6volt tap will reflect around 10k at primary with an 8 ohm speaker load, wheras, using the 6volt tap will reflect around 2.7k primary...
you can even try the 10DE7 from Gerry, makes a 2 tube SET amp.:D |
|
|
| Shoog |
| quote: | | Don'nt need to, you can use the 110/220 volt primary, 12volt multi-tap secondary 500mA traffo which can be bought for around 100 pesos and you are in business... |
Toroidals work much better in this duty than EI's.
Shoog |
|
|
| dsavitsk |
| quote: | Originally posted by mach1
I have built a few of these with different drivers. |
You should connect the parafeed cap from the plate to the transformer primary, and the other side of the primary to ground (or the cathode for a WE connection) -- the way you have it, with the transformer connected to the plate and the cap to ground, charges the parafeed transformer like a cap, which is bad.
If you are looking for very good, very cheap parafeed transformers, check out the Hammond 119DA's. The downside is they are 600:8, the upside is that they work very well 2.5K:32, so are great for headphones. |
|
|
| planet10 |
| quote: | Originally posted by dsavitsk
connect the parafeed cap from the plate to the transformer primary, and the other side of the primary to ground (or the cathode for a WE connection) -- the way you have it, with the transformer connected to the plate and the cap to ground, charges the parafeed transformer like a cap, which is bad. |
I've seen that discussion before... as i understand it, it makes no difference which end of the trafo you put the cap, it still performs the same way.
dave |
|
|
| arnoldc |
| quote: | Originally posted by Tony
using the 6volt tap will reflect around 10k at primary with an 8 ohm speaker load, wheras, using the 6volt tap will reflect around 2.7k primary... |
Say that again? :xeye: |
|
|
| Dave Cigna |
| quote: | Originally posted by dsavitsk
You should connect the parafeed cap from the plate to the transformer primary, and the other side of the primary to ground (or the cathode for a WE connection) -- the way you have it, with the transformer connected to the plate and the cap to ground, charges the parafeed transformer like a cap, which is bad. |
With the parafeed cap in the ground leg, the transformer capacitance is biased up with some DC on it, but the AC should be the same either way. Some people feel that caps sound better with a significant DC voltage bias, which seems plausible to me.
In any case, it's easy enough to try it both ways and listen.
-- Dave |
|
|
| alexg |
| quote: | Originally posted by Tony
Don'nt need to, you can use the 110/220 volt primary, 12volt multi-tap secondary 500mA traffo which can be bought for around 100 pesos and you are in business...
:D
using the 6volt tap will reflect around 10k at primary with an 8 ohm speaker load, wheras, using the 6volt tap will reflect around 2.7k primary...
you can even try the 10DE7 from Gerry, makes a 2 tube SET amp.:D |
Tony,
I have seen some schematic using PP OPT (like the Hammond 125E on the End of Zen design), would this work better? I am planning to try the 12b4 circuit using my 832 PP OPT, it is around 8K primary.
My winder is willing to wind me a non-gapped OPT and am planning to use the plate chokes I used on my 7193 preamp on this 12b4 parafeed. Or I can use a CCS for plate load. :)
Ultimately, if this will turn out ok, I want to do a 12sl7 into a 6080 parafeed. :D
What do you think?
Thanks,
Alex |
|
|
| dsavitsk |
| quote: | Originally posted by planet10
I've seen that discussion before... as i understand it, it makes no difference which end of the trafo you put the cap, it still performs the same way. |
This is what I was told by Jack Elliano which is good enough for me -- as I understood it, connecting the transformer to the plate allows a huge DC across the windings, charging the transformer like a capacitor, which degrades its ability to act as a transformer. You'll have to call him for a better technical explanation than that. |
|
|
| Tony |
| quote: | Originally posted by arnoldc
Say that again? :xeye: |
ah, my mistake, so using the multitap traffos that are abundant and cheap in Manila, you can tap the 6volt winding and get around 10k primary impedance, whereas tapping at 12v you will get about 2.7k...
:D
@alexg,
6sl7 has too much gain imho, you will want something like 20ish mu. |
|
|
| mach1 |
| quote: | | connecting the transformer to the plate allows a huge DC across the windings, charging the transformer like a capacitor, |
Since the cap blocks DC I find that statement rather suspect. I don't understand where the 'huge DC' potential comes from. I have built quite a few parafeed amps and have mild preference for placing the cap in the lower leg. There are subtle differences, but nothing that could be attributed to the generation of 'a huge DC across the windings'. If this happened the core would saturate and the bass would suffer. This just doesn't happen.
Build both ways and see which you prefer.
This is a great budget transformer for parafeed use. It is roughly the same size as the ten watt unit but with a thicker stack. It only has one (8 ohm) secondary, which helps maintain sound quality.
http://www.altronics.com.au/index.a...a=item&id=M1115
Impedance over the whole primary measures 7.2 - 7.5k |
|
|
| alexg |
| quote: | Originally posted by Tony
ah, my mistake, so using the multitap traffos that are abundant and cheap in Manila, you can tap the 6volt winding and get around 10k primary impedance, whereas tapping at 12v you will get about 2.7k...
:D
@alexg,
6sl7 has too much gain imho, you will want something like 20ish mu. |
Edrel is winding me a 220V primary 9V secondary to be used for parafeed OPT.
If 6sl7 has too much gain, then I have a pair of 6j5 that I can use and a lot of 12au7 and 5670.
Thanks. |
|
|
| dsavitsk |
| quote: | Originally posted by mach1
Since the cap blocks DC I find that statement rather suspect. I don't understand where the 'huge DC' potential comes from. |
Potential, not current. The potential comes from the fact that one winding of the transformer is at plate voltage, the other is at ground. If you take two sheets of metal, put a gap between them, and apply a potential to one side while the other is grounded, you have a charged capacitor.
To answer the question you didn't ask, where you get saturation in a parafeed circuit due to DC current is when the cap is too big -- then the process of charging and discharging the cap pulls current across the primary. |
|
|
| mach1 |
I think if you look at the schematic you will see that one winding is at plate voltage, the other is floating.
Re point 2, I concur. |
|
|
| dsavitsk |
For safety reasons, I always learned to tie one side of the secondary to ground, so I assumed you did this too. If you don't, then I have a different, and distinct, objection to the schematic. :)
Anyhow, I don't care where you put your parafeed cap -- I was told by someone I respect, and who has forgotten more about transformers than most of us on this board will ever know, that one option was considerably better than the other, and that placing the cap to ground will hurt frequency response. I was simply sharing that bit of knowledge. |
|
|
| mach1 |
And you will find that Mike Lefevre of Magnequest has no such misgiving and concedes that many of his customers prefer results with the cap in the ground leg.
Big difference between 'knowledge' and opinion. |
|
|
| Dave Cigna |
| quote: | Originally posted by dsavitsk
Potential, not current. The potential comes from the fact that one winding of the transformer is at plate voltage, the other is at ground. If you take two sheets of metal, put a gap between them, and apply a potential to one side while the other is grounded, you have a charged capacitor.
To answer the question you didn't ask, where you get saturation in a parafeed circuit due to DC current is when the cap is too big -- then the process of charging and discharging the cap pulls current across the primary. |
The thing to be aware of here is that the capacitance between primary and secondary is always there no matter what DC potential exists between them. Likewise, that capacitance will always be charged/discharged by the signal in the same way, no matter what DC bias.
It's true that if you put the primary at 400VDC then the capacitance needs to be charged to that level, but it's done only once when you power up the amp, and then it's done directly by the power supply not by the output tube.
In the end, the capacitance seen by the output tube (as far as signals are concerned) will be the same either way. It works the same way as a coupling cap between stages inside an amp. If the cap is .1uF, then it's always .1uF. Putting one end at a higher voltage doesn't change that.
The above assumes that the capacitance is 'ideal.' In fact, it seems that some capacitors - purpose made or stray - behave more ideally when there is a large DC bias on them. Something to do with the dielectric being nonlinear as the field swings back and forth around (or near) zero. I don't pretend to understand the physics of the dielectrics, but it all seems entirely plausible to me.
In any case, more than one person has reported a slight but noticeable improvement in sound when the parafeed cap is placed in the ground leg. That's the most common report. The second most common is that there is no difference. I don't remember anyone saying that it sounds better in the input end, but I have not talked to everyone.
-- Dave |
|
|
|
