Just re igniting this threat as Felipe asked me for help on the loadline for the 307a. Using the valve kindly provided by Vegard Winge, I completed a series of traces with feedback (a la Schade).
In my perspective, the only way of using effectively the 307a in this circuit is by driving the screen low at about 125V and using a 6K5 OT (preferably) if you want to bias this stage about 50mA. You can go lower Zaa if current is increased, but that depends on what OT you have around. You can get 2-3W at 1% THD or less in this mode with about 100Vpp drive.
I personally would go for a simple 6C4C/2A3 instead.
Anyhow, here are the results in case you want to look at:
307a with schade feedback (Part II) | Bartola Valves
Cheers,
Ale
In my perspective, the only way of using effectively the 307a in this circuit is by driving the screen low at about 125V and using a 6K5 OT (preferably) if you want to bias this stage about 50mA. You can go lower Zaa if current is increased, but that depends on what OT you have around. You can get 2-3W at 1% THD or less in this mode with about 100Vpp drive.
I personally would go for a simple 6C4C/2A3 instead.
Anyhow, here are the results in case you want to look at:
307a with schade feedback (Part II) | Bartola Valves
Cheers,
Ale
Ale,
I totally agree with your approach. As per my latest experience with the push pull quad 4P1L, I would go for a PP 4P1L. You can easily get 3-4W class A1 from a PP 4P1L at very low THD. My amp sounds amazingly clear, detailed and warm. Can be checked out at simpl epleasuretubeamp.com.
Best,
Radu
I totally agree with your approach. As per my latest experience with the push pull quad 4P1L, I would go for a PP 4P1L. You can easily get 3-4W class A1 from a PP 4P1L at very low THD. My amp sounds amazingly clear, detailed and warm. Can be checked out at simpl epleasuretubeamp.com.
Best,
Radu
I agree that a p-p 4p1l is a great solution to the difficult to drive modern orthos, I have no issue expanding this thread to include sharing schematics, gain structure, and power output with such a design. Could prove a very nice option for HE-6 owners.
When it coms to p-p I personally prefer a hybrid with a p-p mosfet output since they are inexpensive and off the shelf (i.e. Lyr.) But never heard a real DHT p-p amp. Its just that going to all the trouble to DIY I want the SET sound.
I received a pair of finemt (sp.?) custom SE 5k:8+8+8+8 75mA Tribute transformers. I lucked into a pair of RCA 2A3's for cheap and got a decent deal on NOS guang 2A3's. First stage will be a simple E182CC with a gain of 24, just can't find a DHT with enough gain for the first stage of this 2-stage amp. Haven't decided between a Madia or a SSHV regulator or spec'd the power transformer. I have coleman regs ready, but haven't even started the complicated prefiltering.
I did hook up the tribute transformers to my 417A Spud amp, was able to turn up the bias and deliver considerably more power than the old smaller electraprints. The sound of the Tribute's are amazing, truly an opt designed for headphones with the special low loss core and low DCR's. Can't wait to hear them with the 2A3's, a commercial amp of this sort would cost a few thousand more, a rare case where DIY pays for itself if you have the time.
When it coms to p-p I personally prefer a hybrid with a p-p mosfet output since they are inexpensive and off the shelf (i.e. Lyr.) But never heard a real DHT p-p amp. Its just that going to all the trouble to DIY I want the SET sound.
I received a pair of finemt (sp.?) custom SE 5k:8+8+8+8 75mA Tribute transformers. I lucked into a pair of RCA 2A3's for cheap and got a decent deal on NOS guang 2A3's. First stage will be a simple E182CC with a gain of 24, just can't find a DHT with enough gain for the first stage of this 2-stage amp. Haven't decided between a Madia or a SSHV regulator or spec'd the power transformer. I have coleman regs ready, but haven't even started the complicated prefiltering.
I did hook up the tribute transformers to my 417A Spud amp, was able to turn up the bias and deliver considerably more power than the old smaller electraprints. The sound of the Tribute's are amazing, truly an opt designed for headphones with the special low loss core and low DCR's. Can't wait to hear them with the 2A3's, a commercial amp of this sort would cost a few thousand more, a rare case where DIY pays for itself if you have the time.
impedence ratio with headphones is completely different than speakers so keep that in mind.
8+8+8+8 give 5k:8, 5k:32, 5k:64 and 5k:128 which pretty much covers any headphone. The 5k:128 is basically the well received Woo5 AKG 1k1 output which also reportedly works well with a 30B and HE-6's since they have a flat impedance curve. The 5k:64 is perfect for sennheisers, beyers and other high efficient headphones. Then the 5k:32 for the Grados and Denons that need a large stepdown to not pickup slightest noises. I researched this for months and this is an ideal configuration for headphones.
I will take some DCR measurements when I get a chance.
8+8+8+8 give 5k:8, 5k:32, 5k:64 and 5k:128 which pretty much covers any headphone. The 5k:128 is basically the well received Woo5 AKG 1k1 output which also reportedly works well with a 30B and HE-6's since they have a flat impedance curve. The 5k:64 is perfect for sennheisers, beyers and other high efficient headphones. Then the 5k:32 for the Grados and Denons that need a large stepdown to not pickup slightest noises. I researched this for months and this is an ideal configuration for headphones.
I will take some DCR measurements when I get a chance.
Thanks for the info.
Can you share the cost of the Tribute tranny's.
I have started working on my version of a DHT headphone amp.
Basing it on a Tubelab SE amp, but using Coleman regs for the DHT heaters.
Need to figure out where to get my output tranny's from.
And how much do they weigh?
I am starting to worry my amp will weigh a ton if I put it in one box.
For the Coleman regs, you need two of everything, starting with the power transformers. I want to use LC filters for the HV and coleman regs, but that's a lot of iron and weight.
Thanks
Randy
Can you share the cost of the Tribute tranny's.
I have started working on my version of a DHT headphone amp.
Basing it on a Tubelab SE amp, but using Coleman regs for the DHT heaters.
Need to figure out where to get my output tranny's from.
And how much do they weigh?
I am starting to worry my amp will weigh a ton if I put it in one box.
For the Coleman regs, you need two of everything, starting with the power transformers. I want to use LC filters for the HV and coleman regs, but that's a lot of iron and weight.
Thanks
Randy
SE is not better by default. It's a myth! It always depends on so many things......
In practice almost every time PP is better...
SE is not better by default. It's a myth! It always depends on so many things......
In practice almost every time PP is better...
I understand what you are saying but if we are letting numbers make our decisions a hybrid with a P-P mosfet output is just very effective for headphones if you want p-p. And one can say the some thing that Mosfet P-P is better than tube P-P "almost every time" (going off numbers/measurements.)
There is something special about SET sound, I don't claim to know what it is but I like it. Since I already have a 5W hybrid P-P headphone amp that I paid $300 bucks for it makes little sense for me to invest time and $'s trying to beat it with all tubes. Now a 300B or 2A3 SET headphone amp costs $3k to $5k. Plenty incentive to DIY if I want that sound.
BTW, The DCR of the Tribute secondary is only 1 ohm, vs the Electraprint which was 12 ohms. That's a significant gain in power transfer with 32 ohm headphones.
Did you also measured the primairy dcr? With dcr pr+sec and the windingratio (=25) its possible to calculate the copper losses. And with frequency response we can compare it with other transformers. (Ok, not all can be compared because the core is very special in your case)
Btw, 1 Ohm isn't high or low if you don't know what the total copper loss is. You have to calculate this to know if the power transfer is low or high.
Btw, 1 Ohm isn't high or low if you don't know what the total copper loss is. You have to calculate this to know if the power transfer is low or high.
Last edited:
SE is not better by default. It's a myth! It always depends on so many things......
In practice almost every time PP is better...
I've tried SE and PP on numerous occasions, like most of us here. I never got the vocal quality of SE on any PP amp I ever built, and since I listen to singers of all kinds, that seals it for me.
But second to SE, the most important thing for me is eliminating all coupling caps and cathode bypass caps. I can do this in SE with 4P1L and 26 in filament bias, and interstage transformers. But cathode bypass caps can be eliminated in PP amps with various kinds of CCSs in the tail, so that helps and can bring a PP amp pretty close to a SET. I suspect one way to go is SE first stage and driver and an interstage/phase splitter into the finals. Something good like a Tango.
I did a quick calculation and its a little tricky because i not know the primair dcr but i hope the current density primair-secondair is equal.
I also hope that each 8 Ohm winding has 1 Ohm dcr and not all together.
So i calculated 156 Ohm primair dcr and 0,25 for all sec. In this case your transformer has an efficiency of 94.1% (0,263dB).
If the all secondair windings are 1 Ohm together efficiency drops to 86.5% 0,63dB (not so good...)
Absolutely voices and strings, if I want P-P sound I have a 5W hybrid that anyone can buy off the shelf Lyr for 300 bucks, very little incentive to DIY.
I love the SET sound and it takes a big DHT project like this to get it with Hifiman orthos. Again I would not be pursuing this if power hungry Hifiman orthos weren't the best sounding headphones I've every owned.
I am away from the house and need to check the primary. It was less than one ohm per secondary. But this is important data as the transformer efficiency really drives the final quality of a high powered DHT headphone amp.
Anyway you look at these Tributes are much more efficient than the electraprints I had. Also this is a core that is on the cutting edge so to speak, I lost the link but there is a new 300B tube supplier in Japan and to showcase their new very expensive tubes they chose transformers wound with this core lam. I forget the metallurgy but basically it starts as an amorphous but then I believe it ends up crystalline after annealing. Basically its nano-technology which wasn't part of my material engineering course 25 years ago
With the brief testing I did with the 417A's the sound reminds me of nickel in a good way. So for not a whole lot of money its very likely to DIY a 300B/2A3/PX4 headphone amp that should very well outperform an Eddie Current/Moth/Woo that costs thousands. I basically chose the 2A3 because NOS tubes are available, I could switch to modern 300B tubes if I come up short on power which is very possible with 77dB/mW headphones. Of course the issue with DIY is time, and I'm not into chassis work during the summer
And you guys are going to be disappointed that I plan on driving the DHT outputs with a good IDHT like the E182CC (gain=24). I just can't justify a three stage headphone SET and it really reduces complexity using a single IDHT driver. I am a huge proponent of simplicity when it comes to headphone amplifiers.
Where I am stuck in the design phase is CCS + capacitive coupling verses trying DC coupling. I am leaning more toward a Maidia powered DC coupled amp, again reduced complexity. But the Salas Shunt does so well with my SPUD that I haven't ruled that out if I could get my head around stacked supplies.