[UntilThen]

Hmm priced really low isn't a thing for a SET amplifier due to all that iron, but maybe 1/3 the price of Nirvana with a better circuit.

I like minimalistic aesthetics, like this.



Anywho, I will build one and send on tour most likely, maybe mid 2021.

I was kidding about priced very low. I want those transformers and that layout is what I have in mind.
It will run with tube rectification and EL34 and equivalent based. That way, I can also use my EL12 spez tubes with adapters in addition to EL34.

It has to be powerful enough to drive Susvara

 

[therremans]

Nice job. Since the DV uses resistors to load the input tube, maybe try some carbon film resistors down the road (Kiwame, Riken Ohm), I've found these to have a positive effect in the past.

Okay, yes will do soon. @bcowen mentioned his Riken carbon films and was kind enough to send me a pair of 1ks. I was interested in hearing these Amtrans AMRG, so I have them on the way in 1k, 1M, and 33k.

Amtrans Resistor 2W AMRG Series - Link - Datasheet - More info

@L0rdGwyn, The Darkvoice uses a 30k resistor coming from the 220uf side power caps going to each plate triode on the 6sn7 socket (pins 2 and 5). What change could occur by using a 33k instead of the 30k? (lower amps?) How should I measure this change if needed? @bcowen mentioned adding a 300k in parallel to each 33k to equal a resistance of 30k. So that is an option if necessary.

 

[L0rdGwyn]

Okay, yes will do soon. @bcowen mentioned his Riken carbon films and was kind enough to send me a pair of 1ks. I was interested in hearing these Amtrans AMRG, so I have them on the way in 1k, 1M, and 33k.

Amtrans Resistor 2W AMRG Series - Link - Datasheet - More info

@L0rdGwyn, The Darkvoice uses a 30k resistor coming from the 220uf side power caps going to each plate triode on the 6sn7 socket (pins 2 and 5). What change could occur by using a 33k instead of the 30k? (lower amps?) How should I measure this change if needed? @bcowen mentioned adding a 300k in parallel to each 33k to equal a resistance of 30k. So that is an option if necessary.

What will happen is you will alter the bias point of the 6SN7 slightly, not enough to really matter. You will get a slightly flatter load line which means perhaps a slight improvement in second harmonic distortion.

With the measurements you sent, a 140V B+ with a 33K load on the 6SN7, the load line will look like this. 1K cathode resistor with 2mA plate current and 2V on the cathode (Ohm's law again, R = V/I = 2V / 0.002A = 1K resistor). Approximate bias point is the green dot.



So, the 6SN7 voltage will be able to swing up to the B+ voltage (140V) and down until the grid hits close to 0V where it will start to draw current (roughly 40V) at a quiescent bias point of 70V. What that means is the peak-to-peak voltage swing across the load line will be limited by the negative voltage swing - it can swing up from 70 to 140 (+70V) but can only swing down from 70 to 40 (-30V), so it will clip if driven more than 60V peak-to-peak (30+30V).

So what does 60V peak-to-peak get you in terms of power? Into a 300Ohm load, and assuming unity gain from the cathode follower output tube (which will be slightly less than 1 in reality). I'll skip the math, but it is around 1.5W.

At anything near 1.5W, your ears will be damaged, so the main point is changing the bias won't matter as it will not effect the degree of unclipped power available to your headphones in any noticeable way, unless you are listening at > 1.5W, which will make you deaf in short order

 

[JKDJedi]

What will happen is you will alter the bias point of the 6SN7 slightly, not enough to really matter. You will get a slightly flatter load line which means perhaps a slight improvement in second harmonic distortion.

With the measurements you sent, a 140V B+ with a 33K load on the 6SN7, the load line will look like this. 1K cathode resistor with 2mA plate current and 2V on the cathode (Ohm's law again, R = V/I = 2V / 0.002A = 1K resistor). Approximate bias point is the green dot.



So, the 6SN7 voltage will be able to swing up to the B+ voltage (140V) and down until the grid hits close to 0V where it will start to draw current (roughly 40V) at a quiescent bias point of 70V. What that means is the peak-to-peak voltage swing across the load line will be limited by the negative voltage swing - it can swing up from 70 to 140 (+70V) but can only swing down from 70 to 40 (-30V), so it will clip if driven more than 60V peak-to-peak (30+30V).

So what does 60V peak-to-peak get you in terms of power? Into a 300Ohm load, and assuming unity gain from the cathode follower output tube (which will be slightly less than 1 in reality). I'll skip the math, but it is around 1.5W.

At anything near 1.5W, your ears will be damaged, so the main point is changing the bias won't matter as it will not effect the degree of unclipped power available to your headphones in any noticeable way, unless you are listening at > 1.5W, which will make you deaf in short order

where's a RadioShack employee when you need one...

 

[therremans]

What will happen is you will alter the bias point of the 6SN7 slightly, not enough to really matter. You will get a slightly flatter load line which means perhaps a slight improvement in second harmonic distortion.

With the measurements you sent, a 140V B+ with a 33K load on the 6SN7, the load line will look like this. 1K cathode resistor with 2mA plate current and 2V on the cathode (Ohm's law again, R = V/I = 2V / 0.002A = 1K resistor). Approximate bias point is the green dot.



So, the 6SN7 voltage will be able to swing up to the B+ voltage (140V) and down until the grid hits close to 0V where it will start to draw current (roughly 40V) at a quiescent bias point of 70V. What that means is the peak-to-peak voltage swing across the load line will be limited by the negative voltage swing - it can swing up from 70 to 140 (+70V) but can only swing down from 70 to 40 (-30V), so it will clip if driven more than 60V peak-to-peak (30+30V).

So what does 60V peak-to-peak get you in terms of power? Into a 300Ohm load, and assuming unity gain from the cathode follower output tube (which will be slightly less than 1 in reality). I'll skip the math, but it is around 1.5W.

At anything near 1.5W, your ears will be damaged, so the main point is changing the bias won't matter as it will not effect the degree of unclipped power available to your headphones in any noticeable way, unless you are listening at > 1.5W, which will make you deaf in short order

Perfect. Thanks for another lesson. I really appreciate it, I love learning how the components work with one another.

 

[Mr Trev]

When I get my DV336se again and that will only happen when Drop drops it again. I have requested it !.... I will then try Mullard ECC33 in that amp.

ECC33 works in my previous GOTL and Studio Six which were design for 6SN7 driver input.... so it could possibly work for DV336se but that is like putting a Rolls Royce engine in Ford Cortina.

That would be freaking awesome!!! Esp. a Merlin engine

 

[UntilThen]

That would be freaking awesome!!! Esp. a Merlin engine

That's what we're talking about. @L0rdGwyn I want an amp that has the equivalent of this power.

 

[bcowen]

What will happen is you will alter the bias point of the 6SN7 slightly, not enough to really matter. You will get a slightly flatter load line which means perhaps a slight improvement in second harmonic distortion.

With the measurements you sent, a 140V B+ with a 33K load on the 6SN7, the load line will look like this. 1K cathode resistor with 2mA plate current and 2V on the cathode (Ohm's law again, R = V/I = 2V / 0.002A = 1K resistor). Approximate bias point is the green dot.



So, the 6SN7 voltage will be able to swing up to the B+ voltage (140V) and down until the grid hits close to 0V where it will start to draw current (roughly 40V) at a quiescent bias point of 70V. What that means is the peak-to-peak voltage swing across the load line will be limited by the negative voltage swing - it can swing up from 70 to 140 (+70V) but can only swing down from 70 to 40 (-30V), so it will clip if driven more than 60V peak-to-peak (30+30V).

So what does 60V peak-to-peak get you in terms of power? Into a 300Ohm load, and assuming unity gain from the cathode follower output tube (which will be slightly less than 1 in reality). I'll skip the math, but it is around 1.5W.

At anything near 1.5W, your ears will be damaged, so the main point is changing the bias won't matter as it will not effect the degree of unclipped power available to your headphones in any noticeable way, unless you are listening at > 1.5W, which will make you deaf in short order

Many thanks for the detailed explanation! With the (spec'ed) 103 dB sensitivity of the HD-6XX's at 300 ohms, I'd imagine that even 1/4 watt would be in ear damaging territory.

 

[bcowen]

where's a RadioShack employee when you need one...

You'd be lucky to find a RadioShack employee that knew the difference between a resistor and a flashlight.

 

[L0rdGwyn]

Many thanks for the detailed explanation! With the (spec'ed) 103 dB sensitivity of the HD-6XX's at 300 ohms, I'd imagine that even 1/4 watt would be in ear damaging territory.

Oh you bet! There is a very distorted (no pun intended) understanding of how much power a headphone actually needs. In reality, even the most "hard-to-drive" headphones don't even need 1W of power.

Let's look at an actually hard-to-drive headphone, the AKG K1000. 74dB/mW at 120ohm, as far as headphones go, that's about as hard as it gets (anyone else seen a less sensitive headphone?).

But 74dB is still a pretty reasonable listening volume at a thousandth of a watt. 0.5W with that headphone will get you around 102dB, which is damn loud! And this is close to the worst case scenario.

You just need enough power to get the headphone to a comfortable listening volume with headroom such that the headphone doesn't clip on the peaks and the amplifier maintains a audibly acceptable degree of distortion.

Amplifiers putting out 5W, 10W, 20W, etc. into headphones are absolutely insane power overkill. When it comes down to it, it is just marketing, power sells when it comes to amps, even if the headphone is using less than 1% of it.

Hope that makes sense, going to get off the soap box now.

 

[UntilThen]

You just need is enough power to get the headphone to a comfortable listening volume with headroom such that the headphone doesn't clip on the peaks and the amplifier maintains a audibly acceptable degree of distortion.

Heresy ! I want to cook a turkey on the amp.

Anyway I found another picture of my DV336se, Euforia (that is a Feliks Euforia by the feet) and La Figaro 339. Nostalgic ! It's like from another time dimension.

 

[L0rdGwyn]

Heresy ! I want to cook a turkey on the amp.

If you want to cook turkey, have someone build you a SET amp with these, 833C transmitting triodes just don't touch the plate leads or they'll kill you.

 

[bcowen]

If you want to cook turkey, have someone build you a SET amp with these, 833C transmitting triodes just don't touch the plate leads or they'll kill you.

I heard a Wavac amp using those tubes once. It was awesome. Not a lot of power, but with some reasonably sensitive speakers it was a killer sound.

Another turkey fryer: the Cary 211 amps. Not my pic, but I used to have a pair of these. No (external) danger of electrical shock, but the 211 tubes themselves reached close to 475 degrees F (as measured with an IR thermometer) at full operating temp. I found it best to turn them off and let them cool down for a minute or two before tube rolling.

 

[L0rdGwyn]

I heard a Wavac amp using those tubes once. It was awesome. Not a lot of power, but with some reasonably sensitive speakers it was a killer sound.

Another turkey fryer: the Cary 211 amps. Not my pic, but I used to have a pair of these. No (external) danger of electrical shock, but the 211 tubes themselves reached close to 475 degrees F (as measured with an IR thermometer) at full operating temp. I found it best to turn them off and let them cool down for a minute or two before tube rolling.

Wowzers, 211s in push-pull, that is a beastly set of monos. Could always grab the oven mitts for tube rolling not that I have ever done it.....

Speaking of transmitting triodes, in addition to those 3C24 tubes I mentioned in my thread, I picked up their bigger brother, the HK54. Not bad for $65! I was the only bidder, guess no one else wants these PITA tubes



By the way for the DV modders, I offered this to @bcowen some time ago, if anyone wanted to try a CCS load on the driver tube, for the cost of parts and shipping (and when I have some free time), I'd be happy to provide a set of boards ready to go. Would probably also suggest rebiasing the 6SN7 as well, which would just mean altering the value of the 1K cathode resistor. Question again is where the boards will fit, and some holes would need to be made in the chassis to mount the PCBs on standoffs.

 

[JKDJedi]