• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

Sennheiser HD 650 Question

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Here's a simple headphone amp I built to use with the Sennheiser HD 650. You can use any triode you want as long as it has reasonable gm. I wouldn't use a high mu tube; you're building a buffer not a voltage amp. You can use any of the small signal transformers with a primary around 15k and the secondary 600 or 300 and 600.


headphone_amp1.jpg
 
HD 650 specs say that Pmax = 500 mW. If I consider a power a bit less than Pmax, let's say 400 mW as a maximum power to be delivered to HD 650 by the amplifier then:

Umax = sqrt(P x Rh) = sqrt(0.4 x 300) = 11 Vrms

Thus, if I consider input signal of the amplifier as of 0.1 Vrms than the amplification of my amplifier should be 11/0.1 = 110.

Are my assumptions above correct?

Thank you,

Igor
 
nuhor68 said:
Also, if U = 11 Vmrs then SPL in the headphones is 103 dB + 20log(11/1) = 124 dB. Is this OK or too loud or not enough?

120dBSPL is generally considered the threshold of pain. So unless you are quite the masochist, I'd say it's enough :bigeyes:

For headphones like the HD6xx I usually shoot for good performance at 1V RMS (i.e., I measure THD and spectra there) and a maximum output at clipping of at least 5V RMS... 10V RMS is better and provides more headroom.

That said, if you look at RMS voltage at "average" listening levels it's usually well under 1V RMS...

BTW, most headphone amps (commercial) have gains around 10x, or 20dB.

Pete
 
sound levels

I know I am going to sound like an old party pooper , but I am always amazed when I hear people talk about sound levels they are considering listening to. I have measured what I listen to, and it is generally about 75dB. When I am feeling a bit wild and daring I will crank it up to 85dB. My background level seems to vary between 25 and 30 dB. May be it is just personal taste.

But hear, I mean, here, is something else to consider. The legal limit for workplace exposure to noise in Ontario where I live is 85DBA averaged over 8 hours. I believe this very simialr to most other jurisdictions. Here is a link http://www.e-laws.gov.on.ca/html/regs/english/elaws_regs_900851_e.htm#BK19 that defines what is meant by averaged over 8 hours.

Rearranging the equation given gives the following formula for the time of exposure, in hours, to a single sound level that will be equivalent to 8 hours of 85 dB :
t = 8 * 10 ^ (8.5 - .1*SPL)

so for 85 dB the allowable time is 8 hours

95 dB : 48 minutes, or time for 1 vinyl LP

103 dB quoted in the headphone spec above : 7.6 minutes, or time for 1 or 2 tracks

120 dB mentioned above as the threshold of pain : 9.1 seconds

and

124 dB calculated above as resulted from 11 volt input : 3.6 seconds.

I just think it is bizzarre that people will go willingly to some 105 dB rock concert, and maybe even complain it wasnt loud enough, or build 100W amps and hook them up to 95dB sensitivity speakers, but if their employer was to tell them they had to work in some noisy room at 90 or 95 dB they would be within their rights to refuse or demand hearing protection.

I think I will stick with my 75 or 85 dB levels.
 
nuhor68,

I'd have to agree with the other posters on this topic that I believe your assumption as to the amount of power needed to drive the headphones is way too high. Further, the drive voltage assumptions are off by a factor of 10.

In other words I assume the headphone amp will be driven by a CD player or i-pod type device. Both can easily output 2 VRMS; that is the spec. for the line level output. That said, I design for an input of 1V at full output to give me headroom. 1V will drive the headphones to full output for most people, so that is why I stated in my original post that you are not building a voltage amp, you are building a buffer. You are building a impedance matching device in the case of the CD player and headroom for the i-pod.
 
Re: sound levels

Robert McLean said:
I know I am going to sound like an old party pooper..

I've been spending a lot of time with the 24-bit/96 kHz live recordings available on the Internet Archive and constantly catch myself listening way too loud. A perfectly natural and clean level is often audible two rooms away when I put the headphones down to grab something from the fridge.
 
Re: Re: sound levels

rdf said:


I've been spending a lot of time with the 24-bit/96 kHz live recordings available on the Internet Archive and constantly catch myself listening way too loud. A perfectly natural and clean level is often audible two rooms away when I put the headphones down to grab something from the fridge.

There is a real tendency, especially with phones, to keep turning it up a little at a time. It is a good justification for a stepped attenuator in a headphone amp with a consistent setting. Mine only has three steps that I use, which keeps reasonable http://www.ecp.cc/less-pressivo-plus-plus.html


BTW, most headphone amps (commercial) have gains around 10x, or 20dB.

Most don't get turned past 9:00, either. They are almost always working as attenuators. There seems to be a psychological bias that we want to attenuate, so most amps are built so that they are never used anywhere near their may output.
 
The transformer I will use for my SE amplifier has two primary windings of 10 kOhm each and two secondary windings of 300 Ohms each. Thus, I could have nine combinations of primary/secondary impedances depending on how windings are connected: series, parallel or separate one is used. From these nine combinations five are different and assuming that HD 650 have impedance of 300 Ohm I could have five possible loads (RL) “viewed” by a tube:

40k (N = Nprim/Nsec = 11.55)
20k (N = Nprim/Nsec = 8.16)
10k (N = Nprim/Nsec = 5.77)
5k (N = Nprim/Nsec = 4.08)
2.5k (N = Nprim/Nsec = 2.89)

So, I am going to choose a tube. As far as I understand this choice also depends on RL/Rp ratio, where Rp – plate resistance. I saw different recommendations: 2, from 3 to 5, from 3 to 10, etc. What is a justification of such recommendations and where I could read about this in details on internet? Also, could somebody please recommend me good sounding tubes based on the above RLs and Ns?

Thank you,

Igor
 
Hi Igor,

With the primary inductance of your transformer wired for 10k ohms, the circuit shown above should work fine as drawn. I would suggest you build a power supply that has a filtered voltage of about 350V to 400V at least 20 mA.

Try the circuit, the transformer primary impedance, and the 6BL7 and see if you like it. Try the 5k or 20k primary impedance and see if you like it better. You can try other tubes such as the 7119 or 12B4. You might want to leave room on the chassis for two tube sockets a 9 pin and an 8 pin. Experiment and see what you like. You will need to adjust the plate and cathode resistors to work properly with your chosen tube. Consult the online tube manuals for this information. Note: the 6BL7 used as an example has a Rp of approximately 2200 ohms.

As for learning further through the resources of the internet, you are using a great one right here on DIYAUDIO. There are other forums also that can be a help. Read all you can. One of the most important things to learn in tube audio is how to work with load lines to determine the operating characteristics of a particular tube. I don't know of any web site that explains this, but any old tube manual should give instruction.
 
Re: sound levels

Robert McLean said:
I have measured what I listen to, and it is generally about 75dB. When I am feeling a bit wild and daring I will crank it up to 85dB.

No arguement on the loudness levels, but you do need some headroom. The dBSPL numbers are averaged over some period of time. I'm guessing if your system went into hard clipping on a sine wave at 86dBSPL, your music played at 85dBSPL average would not sound good at all.

I've heard various numbers thrown around for headroom, but to me 10dB would be a minimum.

As for headphones, the average voltage I've measured when listening at "normal" levels (probbaly 70-80dB for me) is amazingly small... 100-200mV for something like an HD6xx. But the peaks may be nearly 10x that. It does depend on the music... a lot of stuff these days is so compressed that it doesn't matter as much.

Pete
 
With the primary inductance of your transformer wired for 10k ohms, the circuit shown above should work fine as drawn. I would suggest you build a power supply that has a filtered voltage of about 350V to 400V at least 20 mA.

I have a power transformer with the following specs:

260 V - 0 - 260 V @ 35 mA
6.3 V @ 1A

I believe I could go with this in my project. I am going to use a rectifier tube and a filter with capacitor input.

As far as I understand I could evaluate PSU output voltage as 260 x SQRT (2) or ~360 V. My question is: if I would need a lower voltage than how I can change it? Simply using resistor voltage divider? Or there is a smarter way?

Thank you.

Igor
 
360V is perfect for the circuit above. Don't forget you will drop some voltage across the rectifier tube and any choke. If you end up with about 300V that will be fine. I would adjust the cathode resistor to drop half the voltage across the tube and half across the plate resistor.
 
Try the circuit, the transformer primary impedance, and the 6BL7 and see if you like it. Try the 5k or 20k primary impedance and see if you like it better.

I have just bought RCA 5R4GYA and 6BL7GTB and will start building shortly.

My output transformers have rated current of 15mA and maximum current of 30mA. As far as I understand I have to take this into account when choosing an operational point for the tube. So, do I have to choose OP to be at 15mA or I can do it at 30mA? And what could happen if I will choose OP current more than 15mA (or 30mA)? For, example transformer frequency response will be worse or not? Will burn down if more than 30mA? Please advice.

Thank you,

Igor
 
Notice that your output transformers are rated for DC current. The output transformer I used was not. I had to use parallel feed and a plate resistor. You have a very different transformer.

That said, there is no reason you can't build a great little headphone amp with those transformers. You can either build the amp as shown in my drawing and parallel feed the transformer or you can use the transformer with the plate tied to one end of the primary and the other end to B+.

If you decide not to parallel feed I would try for a current of 15mA to 20mA through the tube; measure across the cathode resistor, and vary the resistor to get the current in the correct range. You may have to bring the B+ down some to get the current in the correct range.
 
I would suggest you build a power supply that has a filtered voltage of about 350V to 400V at least 20 mA.

Does the current provided by PSU depends on the load? I have heard that one triode requires 10mA approx. So, if we have two triodes they will take 20mA from PSU? And this is limited by rated value of the transformer primary? If we have this as 200mA, does it mean we can supply current to approx 20 triodes?

Thank you,

Igor
 
Does the current provided by PSU depends on the load?

Ohm's law requires it.

I have heard that one triode requires 10mA approx. So, if we have two triodes they will take 20mA from PSU?

As the schematic shows each triode draws 8mA as biased in this application. So the total current draw from the transformer under these conditions is 16mA.

And this is limited by rated value of the transformer primary?

The power transformer secondary.

If we have this as 200mA, does it mean we can supply current to approx 20 triodes?

If the secondary is rated for 200mA at a particular voltage that's what it will deliver.
 
Dear Osprey,

Many thanks for your answers and explanations.

I am thinking about headroom and need an advice if I am doing it correctly. So, I made an operational point analysis using ECC99 plate characteristics, which are similar to 6BL7. I used a load line based on:

E = 300V
Zp = 20k
Zs = 600
RH = 300 (HD 650)
RL = RH x (Zp/Zs) = 10k

I chose Io = 15.5mA based on my transformer rated current and Ug = 4V. Then Uo = 146V (see PDF attached)

Let’s assume that input signal have average level 200mv RMS. From the PDF we can see that grid voltage can change from -0 to -8V (peak values) or from 0 to -5.6V RMS or with amplitude of 2.8V RMS. So,

20log(2.8/0.2) = 23dB

Does it mean that the amplifier has headroom of 23dB?

Thank you,

Igor
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.