Rod I like your efforts posting schematics and encourgae you to continue, I'm sorry to say I don't have technical knowledge to discuss or help in yours or others designs.
It was not an elaborate attack, but an elaborate explanation of what I think about the power supply rejection of your circuit. If I didn't provide technical details anyone could say I pull things out of the hat.
And please stop implying this is personal. Nothing of the sort. It's healthier for you to accept the reality that not everyone on earth will agree with every single one of your circuits. I like your regulator idea very much and I also appreciate your contribution in other threads. Let's keep it technical. There we can deal in facts.
As far as more designs, I agree with you. A variety of ideas is better than too few. Let's explore and agree to sometimes disagree. I mean this in the nicest and friendliest way.
Hey guys instead to discuss why not agree in one type of headphones: for example I read or understand that high Z & high sensistive are easy to do with two 41PL, why not center first in one kind of headphones low or high impedance & low or high sensitive & do the things step by step.
N.B. it's only a suggestion.
N.B. it's only a suggestion.
Last edited:
Not a suggestion that it is personal. Perhaps your dislike of BJTs has you looking harder for faults than otherwise. Component-al rather than person-al.
But we have each had our say, and that is fine.
yes, let's do that. Bring on the design ideas and debate them in a friendly way.
Peace!
.
Thank you for the support, folks. It is appreciated!
I hope to find more ideas of interest.
Rod,
Your ideas are always inspiring, I'm tempted to open a new thread on the PP ideas discussed before. I guess there are many interested in applying some of your new concepts in 45 / 4p1l / 6c4c or other DHTs in PP.
Thanks for sharing as always!
Ale
Your ideas are always inspiring, I'm tempted to open a new thread on the PP ideas discussed before. I guess there are many interested in applying some of your new concepts in 45 / 4p1l / 6c4c or other DHTs in PP.
Thanks for sharing as always!
Ale
Thanks Ale!
Yes, there's probably a few ideas to move forward in the PP circuits, with DHT outputs. We'll need to get to work on these!
I simply don't understand all this fuss about using solid state technology where it really makes things better, especially if one can save a lot of money for better result!! What's the point in using tricky step-up input transformers where the result is far from guaranteed or poor devices like the 3A5 to drive a power triode? All this simply does not make sense to me.
If it has to be a true DHT amp then solid state should be banned tout-court. No SS CCS, no SS regulators, no SS rectifiers. Nothing at all! Would it be better???😀
If it has to be a true DHT amp then solid state should be banned tout-court. No SS CCS, no SS regulators, no SS rectifiers. Nothing at all! Would it be better???😀
Last edited:
Well, if you build a gyrator assisted driver or a DC filament regulator for the DHT and listened to them you may change your opinion....
Sand used properly can make great things 🙂
Sand used properly can make great things 🙂
I mind isn't about SS technology, the discussion was about how can affect the SS techonology proposed by Rod, Iko & Anatoly explained theirs reasons also Rod explained his reasons, I like debate with arguments I don't see the problem if people want use SS or all tubes each one can choose his way. That's the debate if sand is or not is used properly to improve DHT.
Last edited:
Sure, and my concern was about stability of regimes that can be achieved by common servo loop, and taming of added distortions by low load resistance of first tube and modulation of beta of the BJT that can be achieved by negative feedback loop. Also, pentode mode of the 1'st tube can be used in order to decrease distortions. One more concern is, low avalanche breakdown voltage of base - emitter junction of the transistor that either can damage it, or cause excess noises. It has to be protected against too high negative swing on input tube's anode (too high means about 5 V and more).
Low anode load does not produce added distortion.
Please see the characterisation of the triode PC86, measured for constant voltage (anode load = zero) - which makes it much easier to read.
This triode can be used in my shunt cascode 300B driver circuit, where an input signal of 0,5V pk gives 80V peak out. The whole swing follows only the blue line marked on the curve - ie well within the straight line.
Pentode is higher distortion than this - just look at the curves.
However, I would agree that pentode sounds better than 2 triode stages.
This is only too easy to solve.
1N4148 diode from base to emitter is all it takes to limit this voltage to 0,6V
Attachments
I've simmed the 4p1l-4p1l both dc coupled and ac coupled. The ac coupled starts giving up at 0 grid around .75 watts @ 1.5% THD into a 5k:38 load. DC coupled goes a little higher but the distortion gets so high as to question if dc coupling is worth it(as Iko said), with headphones distortion is not so easily forgiven as speakers. When you start adding real world losses from the opt and grid stopper/etc, the 4p1l-4p1l is probably a 0.5W amp at best into a 38 ohm load. I can do that with a single 6c4p1l spud amp.
So we I a don't think more gain from pentode first stage 4p1l or shunt cascode will be the silver bullet for that magic 1W clean output we need for HE-500, LCD2's, etc with the 4p1l in the output.
At this point I am think for these type headphones needing so much voltage that a p-p 4p1l design may make the most "sense" if we want all DHT. I think a concurrent thread would be good, it would be easy to just add a headphone tap to it. I know little of dht p-p, don't even know how to use coleman regs with a p-p stage. The other option is very high voltage 841 GM70 type all DHT for maximum linearity, swinging 200Vp-p with an all DHT SET amp is just difficult. A real eye opener to me was how these headphones take more voltage swing than most horn/fullrange driver setups as seen in the test thread.
Or we look closely at Rod's 4p1l-2A3 shunt cascode "SET".
Rod can you show a closer schematic with the darlington pair and other ss tricks so we can start simming a 4p1l to 2A3 all DHT amp , do you think that for stability it would have to be PCB based?
So we I a don't think more gain from pentode first stage 4p1l or shunt cascode will be the silver bullet for that magic 1W clean output we need for HE-500, LCD2's, etc with the 4p1l in the output.
At this point I am think for these type headphones needing so much voltage that a p-p 4p1l design may make the most "sense" if we want all DHT. I think a concurrent thread would be good, it would be easy to just add a headphone tap to it. I know little of dht p-p, don't even know how to use coleman regs with a p-p stage. The other option is very high voltage 841 GM70 type all DHT for maximum linearity, swinging 200Vp-p with an all DHT SET amp is just difficult. A real eye opener to me was how these headphones take more voltage swing than most horn/fullrange driver setups as seen in the test thread.
Or we look closely at Rod's 4p1l-2A3 shunt cascode "SET".
Rod can you show a closer schematic with the darlington pair and other ss tricks so we can start simming a 4p1l to 2A3 all DHT amp , do you think that for stability it would have to be PCB based?
Last edited:
Wait a minute! Firstly do you mean HiFiMan HE500? If so, this is one of the easiest loads I have ever seen,
Measured: HiFiMan HE-400 and HE-500 Headphones | Sound and Vision Magazine
In fact the frequency responses using a 5 ohm Zout amp and a 75 ohms Zout amp are virtually identical!
Measured: HiFiMan HE-400 and HE-500 Headphones | Sound and Vision Magazine
You can clearly see that an average SPL of 89 db needs just .18 Vrms which means less than 1 mW (from 300 to 10KHz)!
Having 100 mW one gets 109 db SPL! At 1W one is very close to the threshold of pain and permanent damage just after few minutes listening...such damage is cumulative, it can be slow but there is no way back!
The other model uses the same technology, I would not expect big difference. Am I wrong?
Last edited:
They lie on there sensitivty specs as all headphone manufactures do. The only headphone spec that is real is the max power or recommended power. The HE-500's recommend a 1W amplifier. The HE-6 a 5W amplifier.
Also if we build for 1W it needs to be a clean 1W.
I do like your idea of using a 150 ohm tap, since the headphone drivers have an extremely flat impedance curve, but to retain detail the secondary DCR starts getting in the way. A 75-120 ohm tap from a parafeed may be the answer (low DCR/induction secondary.)
I would be careful about 1W recommendation because usually people don't use a zero feedback DHT amp with very soft clipping! I would take that it as general thing. You might need that power with cheap amplifiers to stay as far away as possible from the clipping. People nowadays frequently use a smartphone and the 5-75 ohm drive is just to demonstrate is it possible if one doesn't need high spl.
One last thing is about perception of THD from the phisiological point of view. For triode like THD spectrum where you basically have dominant H2 and some H3 with all the others a lot lower (I would say this is the case when a typical SET THD is still within 3%) you have hearing thresholds. Human hearing is most sensitive to THD between 60 and 80 db SPL where you can pick up quite low harmonic distortion levels. As low as 0.1%. However around 80 db SPL human hearing starts to protect itself, closing, and sensibility to details and THD decreases almost exponentially. At 90 db the threshold is already at 2-3%, depending on the subject.
So having 0.75W at 1.5% is not bad at all. You can try by yourself.
Last edited:
I am reading the website you posted measuring the HE-500's. In their review of the HE-6 thay say the HE-500's are difficult to drive. Then they go on to say the HE-6 are superior to the HE-50', but require 5 watts (even more difficult to drive).
But reading the review of the HE-500 I think it may make sense to try a transformer with a 75 ohm tap, the concern of course as you say the low level details/distortion characteristics may be dominated by a 75ohm tap with a 75 ohm DCR and huge induction. This is where parafeed high nickel core can come in handy.
The sowter parafeed multitap headphone transformer may be perfect, it specs at 0.5W 220hz but a full 2W at 50hz, if only it was a little bigger.
A transformer with a 32 ohm center tap, could drive balanced 5k:64 which may very feasible with a single feed 4p1l. May be the answer I've bee looking for.
But reading the review of the HE-500 I think it may make sense to try a transformer with a 75 ohm tap, the concern of course as you say the low level details/distortion characteristics may be dominated by a 75ohm tap with a 75 ohm DCR and huge induction. This is where parafeed high nickel core can come in handy.
The sowter parafeed multitap headphone transformer may be perfect, it specs at 0.5W 220hz but a full 2W at 50hz, if only it was a little bigger.
A transformer with a 32 ohm center tap, could drive balanced 5k:64 which may very feasible with a single feed 4p1l. May be the answer I've bee looking for.
Ok, but why are they difficult? It seems to me because of their low efficiency. Typical HiFi gear headphone outputs, smartohpones and similar stuff are not designed for such low sensitivity. So you get at lot of distortion when you try to listen to music louder, as could do with more sensitive headphones! Instead you can taylor your amp to the HE-500. I think 0.75W with 1.5% will be quite good. Actually I would be more worried about THD at low levels where you can pick it up easily, as I told before. From this point of view 3A5, 2A3 and similar are not so good....
75 ohm is the Zout of the amplifier not the secondary impedance. If you want use such secondary you still need to provide the 4P1L with the right primary impedance, hence you have to have the right turn ratio if you want 5K primary load. Using a 5K:36ohm transformer your Zout is in the region of 12-15 ohm.
Last edited: