Hi, I'm enbarking on my next project and would like some opinions from people on decisions of OTL tube designs. Specifically with low impedance headphones such as these Grados I have that are at a 32ohm level.
The two designs I am toying between are:
An OTL Tube Headphone Amplifier | HeadWize
No-Compromise OTL Tube Headphone Amplifier | HeadWize
I really want to just do the no-compromise design but it doesn't really seem to address the low impedance phones thing too much. The other design talks about modifying the the output resistor to stabilize the amp with low impedance phones, but I have some questions on gain an current draw with the amps ability.
Also i am not entirely adversely to transformer output designs but I am a little skepticle about accidental unplugs and if that would damage the output transformer. Also linearity and base response.
Does anyone have tips on successful all-tube headphone amp designs for low impedance phones?
The two designs I am toying between are:
An OTL Tube Headphone Amplifier | HeadWize
No-Compromise OTL Tube Headphone Amplifier | HeadWize
I really want to just do the no-compromise design but it doesn't really seem to address the low impedance phones thing too much. The other design talks about modifying the the output resistor to stabilize the amp with low impedance phones, but I have some questions on gain an current draw with the amps ability.
Also i am not entirely adversely to transformer output designs but I am a little skepticle about accidental unplugs and if that would damage the output transformer. Also linearity and base response.
Does anyone have tips on successful all-tube headphone amp designs for low impedance phones?
An OTL design would not be a "no-compromise" design for low impedance phones. Need transformer.
For reference, I have Denon D2000 (25 ohms) and have owned Grados (32 ohms).
You would use an 8 ohm speaker transformer. Everything would be the same as a speaker tube amp, except the power supply needs to be extremely quiet. You need DC heater supply for DHTs, though AC is fine for IDHTs.
No compromise and easy to build? Try the Tubelab SE 45 with the transformers in a separate enclosure. I can confirm this works really well. I've used the Salas High Voltage Shunt Regulator with it as well, but it is already quiet stock.
For reference, I have Denon D2000 (25 ohms) and have owned Grados (32 ohms).
You would use an 8 ohm speaker transformer. Everything would be the same as a speaker tube amp, except the power supply needs to be extremely quiet. You need DC heater supply for DHTs, though AC is fine for IDHTs.
No compromise and easy to build? Try the Tubelab SE 45 with the transformers in a separate enclosure. I can confirm this works really well. I've used the Salas High Voltage Shunt Regulator with it as well, but it is already quiet stock.
No way... Even with the fact that the output transformer is at 8ohms, it drives 32ohm phone drivers?
I also want to keep it strictly tube driven, I don't really want to use current regulators etc.
My example would be this:
Hi-end headphone amplifier
So with this design, I just need to source a plain old 8ohm output transformer?
I also want to keep it strictly tube driven, I don't really want to use current regulators etc.
My example would be this:
Hi-end headphone amplifier
So with this design, I just need to source a plain old 8ohm output transformer?
Yes, an 8 ohm output transformer. Use the lowest impedance you can get away with, that still provides plenty of voltage swing. Headphones are so efficient that if you use a 32 ohm output transformer, most tubes will have too much gain. I have even driven 600 ohm Sennheiser headphones and there was enough gain (out of a 300B amp).
I have made the amp that you referenced, with both the 5842 and D3a. It is a good amp. I used the 16 ohm tap out of my transformers to drive headphones. Note that the transformers referenced on his site is for 600 ohm headphones. You want 16 ohm secondaries for this design. My transformers costed $600 by themselves.
I have made the amp that you referenced, with both the 5842 and D3a. It is a good amp. I used the 16 ohm tap out of my transformers to drive headphones. Note that the transformers referenced on his site is for 600 ohm headphones. You want 16 ohm secondaries for this design. My transformers costed $600 by themselves.
Fascinating! Learn something new everyday. So when shopping for output transformers, 8 ohm secondary of course, but what about primary impedance? Lower the better? Primary inductance?
What about wattage, any correlation of size / safety margin to freq response? What do I need to look out for to obtain the sweet spot?
Also, like full on speakers, if something happens that cause the headphones to get unplugged and the secondaries see no load while under power, is it true it would damage the transformer?
What about wattage, any correlation of size / safety margin to freq response? What do I need to look out for to obtain the sweet spot?
Also, like full on speakers, if something happens that cause the headphones to get unplugged and the secondaries see no load while under power, is it true it would damage the transformer?
It is possible to damage the transformers with no load, because it can arc, but I tested this before and in the real world for audio use it is fine. To be safe, you would always have the headphones connected. Short circuit is fine, too. So it's not easy to damage the transformer.
Primary impedance would be 5k or 7k for 5842/D3a. You want 16 or 32 ohm secondaries if building this amp. The 8 ohm can be used for designs that can put out more power, but here we are talking about using driver tubes as power tubes. The transformers should be able to handle at least 30mA, which any SE transformer can do.
Primary impedance would be 5k or 7k for 5842/D3a. You want 16 or 32 ohm secondaries if building this amp. The 8 ohm can be used for designs that can put out more power, but here we are talking about using driver tubes as power tubes. The transformers should be able to handle at least 30mA, which any SE transformer can do.
ooh man, this is interesting info for sure! Is there a nice way to transition the transformer to a dummy load via switch as a mode for swapping of headphones? Or does everyone just power down the whole system every time something goofy like that needs to happen. I may have remembered wrong but I thought that no load on a transformers output saturates the core? Or is arcing a result of just that.
So how would I derive the 5k or 7k by the tube data sheet? That may be an elementary question but I didnt see any values on the data sheet that would relate to primary impedance ratings. (I saw plate resistance, but thats different)
I am perusing through Edcor's site for output transformers and im finding them all to have a screen grid tap. 40% or so. If a transformer has one, (although the schematic doesn't define using a screen tap) will utilizing it in this specific diagram improve linearity? Or will it just mess with me. and if yes, the #2 grid gets hooked up to that right?
Sorry if this sounds like amateur hour, I just don't want to mix up concepts with full sized speaker drivers with tiny headphone drivers. This is tremendous help!
So how would I derive the 5k or 7k by the tube data sheet? That may be an elementary question but I didnt see any values on the data sheet that would relate to primary impedance ratings. (I saw plate resistance, but thats different)
I am perusing through Edcor's site for output transformers and im finding them all to have a screen grid tap. 40% or so. If a transformer has one, (although the schematic doesn't define using a screen tap) will utilizing it in this specific diagram improve linearity? Or will it just mess with me. and if yes, the #2 grid gets hooked up to that right?
Sorry if this sounds like amateur hour, I just don't want to mix up concepts with full sized speaker drivers with tiny headphone drivers. This is tremendous help!
A general rule of thumb for tube amps is that you want a damping factor of 3 to 5. It affects things like frequency response, and lowers distortion as it is easier for tubes to drive higher impedance. Less damping factor will sound looser in the bass, while more will sound tighter (SS amps have very high damping factors). The 5842 has an output impedance of around 1.7k. So we multiply that by 3-5x to get 5.1k-8.5k. My choice would be 5k:16 ohm or 7k:32 ohm, whichever is easier to obtain. I would recommend the Hashimoto H-507S that I use myself at 5k:16 setting.
On transformers with a screen grid, you will just not connect it. It is for pentode tubes to behave like something in between triode mode and pentode mode.
On transformers with a screen grid, you will just not connect it. It is for pentode tubes to behave like something in between triode mode and pentode mode.
Last edited:
Ah! that brings a lot of sense to the table.
So I want to deal with Edcor for best quality bang for buck. I want to use their CXSE line to hit the 20 to 20k freq response. Also at the same time, I am interested in using the D3A tube upgrade specified on that site. On the data sheet, it is specified that the plate resistance is 1.9k. Edcor goes up to 16ohms secondary so if I multiply 1.9k x 3 I get 5.7k. However in the 16ohm catagory, the primary impedance is only 5k.
My question is: would having 700ohms difference below be an audible degradation of the bass? This would mean looser base, but is it within a tolerable/permissible range?
Or am I better off working with a 5842 tube where the dampening resistance is closer matched appropriately with Edcors products.
What have you found in sound between the 5842 vs D3A? With these factors, which rout is more pleasing?
So I want to deal with Edcor for best quality bang for buck. I want to use their CXSE line to hit the 20 to 20k freq response. Also at the same time, I am interested in using the D3A tube upgrade specified on that site. On the data sheet, it is specified that the plate resistance is 1.9k. Edcor goes up to 16ohms secondary so if I multiply 1.9k x 3 I get 5.7k. However in the 16ohm catagory, the primary impedance is only 5k.
My question is: would having 700ohms difference below be an audible degradation of the bass? This would mean looser base, but is it within a tolerable/permissible range?
Or am I better off working with a 5842 tube where the dampening resistance is closer matched appropriately with Edcors products.
What have you found in sound between the 5842 vs D3A? With these factors, which rout is more pleasing?
Oh excellent! That is good to know.
So if I worked with this transformer output of 16ohms, how capable is it to drive something with a much higher headphone impedance? Like a Sennheiser of 300ohms? What would happen?
And for thought sake, if I would want to switch the secondary of the output transformer to a 'standby' mode so I can swap phones, would I just have a switch to a high value resistor?
So if I worked with this transformer output of 16ohms, how capable is it to drive something with a much higher headphone impedance? Like a Sennheiser of 300ohms? What would happen?
And for thought sake, if I would want to switch the secondary of the output transformer to a 'standby' mode so I can swap phones, would I just have a switch to a high value resistor?
I don't know about driving high impedance cans, never tried. You certainly can if you build a traditional 2-stage tube amp. It's just an issue of getting enough loudness. High impedance headphones like more voltage swing. A 300 ohm transformer would solve this, or adding a power tube.
If it's an expensive transformer, I would build the switch. Otherwise, it's not a big deal. The conditions for arcing is not there.
If it's an expensive transformer, I would build the switch. Otherwise, it's not a big deal. The conditions for arcing is not there.
Last edited:
super awesome, thats fine by me.
So in terms of Power Supply, how forgiving is the B+ voltage? I want to do a simple unregulated design with resistors and electrolytics. How forgiving is the 200v B request? It looks like it comes close to the max watt line. Is this ok?
And does the 5842 or D3A require floating the filaments a 1/4 of the B+? I am planing to regulate the heater supply to DC and was wondering if there is anything I should be aware of.
So in terms of Power Supply, how forgiving is the B+ voltage? I want to do a simple unregulated design with resistors and electrolytics. How forgiving is the 200v B request? It looks like it comes close to the max watt line. Is this ok?
And does the 5842 or D3A require floating the filaments a 1/4 of the B+? I am planing to regulate the heater supply to DC and was wondering if there is anything I should be aware of.
Lol... I don't understand most of your questions. Unregulated is fine, it just needs to be clean power. You need a choke, maybe two of them. Keep in mind that your power line has over 10% distortion with a bunch of hash noise, and it is your power supply that makes the signal that you hear, so you will be hearing all that noise.
There's no point in using DC heater supply. These are not direct heated triodes. It will be quiet with twisted AC wires to power them. Don't float anything.
There's no point in using DC heater supply. These are not direct heated triodes. It will be quiet with twisted AC wires to power them. Don't float anything.
Last edited:
lol, at the time, the question made sense in my head. I think what I was trying to say was how forgiving is the B+, in terms of accuracy, but I guess that number will naturally fluctuate all the time when in use. Hence "unregulated"
What do you mean by 10%? Is this what will naturally happen if I go without chokes?
Are chokes something I should really shoot for?
What do you mean by 10%? Is this what will naturally happen if I go without chokes?
Are chokes something I should really shoot for?
The amp is running in pure Class A so the B+ voltage will be perfectly constant. The point of regulation is to filter out the noise. With an unregulated supply, a choke filters out most of that noise. Without either of these, you will hear background hiss. The power supply makes a huge impact on the final sound quality.
Sweet!
Ok, so I have one last question;
I would like to have this tube rectified, also I am running into difficulty finding the right power transformer at Edcor that meets the proper ballpark of 180 - 200 V rms.
Is a possible to a bring down the Voltage out of the transformer by being selective with the tube rectifier? Since the amp is running pure Class A with a constant current, I can pick the tube rectifier of my choice to meet the required voltage drop at that current right?
Or was this how people typically used to design their amps.
So this leads me to this somewhat basic question: This SESS amp that I want to build, how much current is it going to draw total?
Ok, so I have one last question;
I would like to have this tube rectified, also I am running into difficulty finding the right power transformer at Edcor that meets the proper ballpark of 180 - 200 V rms.
Is a possible to a bring down the Voltage out of the transformer by being selective with the tube rectifier? Since the amp is running pure Class A with a constant current, I can pick the tube rectifier of my choice to meet the required voltage drop at that current right?
Or was this how people typically used to design their amps.
So this leads me to this somewhat basic question: This SESS amp that I want to build, how much current is it going to draw total?
It would draw about 40mA total. A standard CLCLC power supply would work well, with a pair of 10H chokes. Get a 150-0-150 VAC transformer, but with such a light load it will actually be 160VAC. Keep the power transformers a good distance away from the tubes and the output transformers.
Yes you can vary the voltage by using different rectifier tubes, but it will require a 5VAC tap to run rectifier tubes. I would just use solid state rectification and not need the 5VAC tap.
Edit: One more tip, and I learned this the hard way... if you are using a single chassis, get the biggest one that you have space for. Go way oversize on the chassis.
Yes you can vary the voltage by using different rectifier tubes, but it will require a 5VAC tap to run rectifier tubes. I would just use solid state rectification and not need the 5VAC tap.
Edit: One more tip, and I learned this the hard way... if you are using a single chassis, get the biggest one that you have space for. Go way oversize on the chassis.
Last edited:
Ah! thats very good to know. The 150-0-150 transformer from Edcor is perfect! It has all the secondaries to make the heater voltages. With the chokes and their DCR values, it should bring down the voltage to the appropriate amount.
Previously I made a pre-amp with a separate box for the power supply and I can totally hear the hum when I move the box in and out. If I combine the chasis for this headphone amp, what is a good rule of thumb for how far away I should place the power transformer and the choke from all the signal material?
And for the power supply, I kinda just want to get away with a 2 capacitors and a inductor in-between, Im thinking 10H. I was thinking of using a 47uf for the reservoir and 100uf for the smoothing. This should be plenty good enough ya?
Previously I made a pre-amp with a separate box for the power supply and I can totally hear the hum when I move the box in and out. If I combine the chasis for this headphone amp, what is a good rule of thumb for how far away I should place the power transformer and the choke from all the signal material?
And for the power supply, I kinda just want to get away with a 2 capacitors and a inductor in-between, Im thinking 10H. I was thinking of using a 47uf for the reservoir and 100uf for the smoothing. This should be plenty good enough ya?
A single 10H should be fine... if it's still not quiet enough, you can add another inductor. 47uF is too high for certain tube rectifiers. Use 4.7uF film capacitor instead.
If you draw less power from the power transformer and rotate it at a certain orientation (where magnetic fields cancel), then you can get away with closer placement of the transformer. 6" inches should be fine.
Don't forget bleeder resistors. It sucks to get shocked by 300V.
Other than that, sounds good!
If you draw less power from the power transformer and rotate it at a certain orientation (where magnetic fields cancel), then you can get away with closer placement of the transformer. 6" inches should be fine.
Don't forget bleeder resistors. It sucks to get shocked by 300V.
Other than that, sounds good!
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.