hopefully jeff will chime in here. perhaps he's still thinking of trying this. I would be interested in the results and some details to the build, since i'm not fully able to translate a raw schemeatic into a working unit. I do like the description of 'simple and sweet'....
I would even try a tube design but they are not easily gotten down low in output impedence from what I understand. What I 'm looking for is a nice big open sound but with body. I don't need outrageous detail or extra ordinary 'soundstaging'. Just solid natural invovling sound. I will be pairing it with a shureV15II and a pass pearl phono section if that helps anyone get on board with the sound signature we are going for....
I would even try a tube design but they are not easily gotten down low in output impedence from what I understand. What I 'm looking for is a nice big open sound but with body. I don't need outrageous detail or extra ordinary 'soundstaging'. Just solid natural invovling sound. I will be pairing it with a shureV15II and a pass pearl phono section if that helps anyone get on board with the sound signature we are going for....
Last edited:
For headphones I've found nothing to beat a purpose-designed transformer, wound on a cheap easily available ferrite core (say an RM14, available from Mouser). That way you can use practically any amp and get below the output impedance you're looking for without any effort. A step down ratio of 10:1 gives an impedance transformation of 100:1 so any amp with a Zout lower than 200ohms will give you a low enough Zout, given a low enough copper loss in the trafo (which is easy enough to obtain).
I wish I knew what you meant by that....it sounds interesting......you wind your own transformer? How can you step down the output impedence this way? I don't follow. Is there an article or link that explains the technique?
I can explain it if you have enough interest as I've wound several trafos myself without encountering any difficulties, each time I've been more than satisfied by the dynamics and inky-blackness of the background.
To get started I'd need to know what signal levels you want to come out and what frequency (power) response you need. I generally aim for full power down to 40Hz but no lower. This gives 1/4 power capability to 20Hz - very few recordings (other than organ music) have such content anyway.
So what's the efficiency of your IEMs (dB per volt input) and what's the maximum SPL you're looking for?
To get started I'd need to know what signal levels you want to come out and what frequency (power) response you need. I generally aim for full power down to 40Hz but no lower. This gives 1/4 power capability to 20Hz - very few recordings (other than organ music) have such content anyway.
So what's the efficiency of your IEMs (dB per volt input) and what's the maximum SPL you're looking for?
definitly interested. 12 ohm Sensitivity: 94dB/mW. Do not need outrageous volume, i tend to listen at moderate levels. It's mostly basic rock n roll and jazz so don't need super low extension for classical recordings etc
the amp will be fed by the Pass pearl phono. it is my main source. I don't know the specs on it or if that is important regarding this.....
the amp will be fed by the Pass pearl phono. it is my main source. I don't know the specs on it or if that is important regarding this.....
OK so 1mW into 12ohms is 110mV, assuming you'd like to reach 114dB that's 1.1VRMS, 1.6V peak.
Trafo needs to handle a certain number of volt-seconds, assuming 40Hz is the lowest frequency that's 12.5mS * 1.6 * 2 / pi = 0.0127V-s.
RM14 has Ae = 2cm^2 and we'll assume Bmax = 350mT. Plugging these numbers into the trafo equation gives 91 turns for your secondary, 910 for the primary. I'll work out the wire gauges in the next post.
Trafo needs to handle a certain number of volt-seconds, assuming 40Hz is the lowest frequency that's 12.5mS * 1.6 * 2 / pi = 0.0127V-s.
RM14 has Ae = 2cm^2 and we'll assume Bmax = 350mT. Plugging these numbers into the trafo equation gives 91 turns for your secondary, 910 for the primary. I'll work out the wire gauges in the next post.
I looked at the RM14 bobbin dimensions from Ferroxcube, I'm hoping they are fairly standard but I'll check against Fair-rite sometime soon.
The winding area is 6mm deep by 18mm long, this has to contain both primary and secondary windings, best to divide the space equally between the two. I'd wind the secondary on first, this should use magnet wire with 0.7mm diameter, roughly 21AWG. You can use 22AWG if you can't get the metric size. There will be 4 layers, about 23turns for each layer though with 22AWG, being thinner you'll get more turns.
The primary is wound over the top of the secondary, I'd not worry too much about counting the turns just fill the trafo up with as much 0.2mm diameter magnet wire (32AWG) as it'll take. Of course if you prefer to count turns it'll make matching the left and right channels easier.
The winding area is 6mm deep by 18mm long, this has to contain both primary and secondary windings, best to divide the space equally between the two. I'd wind the secondary on first, this should use magnet wire with 0.7mm diameter, roughly 21AWG. You can use 22AWG if you can't get the metric size. There will be 4 layers, about 23turns for each layer though with 22AWG, being thinner you'll get more turns.
The primary is wound over the top of the secondary, I'd not worry too much about counting the turns just fill the trafo up with as much 0.2mm diameter magnet wire (32AWG) as it'll take. Of course if you prefer to count turns it'll make matching the left and right channels easier.
cold feet...am saving the info for later though, I do appreciate. I've got alot of electronics theory and learning to catch up on and have got a guitar amp that needs fixing, and a phono section. Thank you though.
This thread could have ended here. Counter Culture's reply is correct and complete.
Even on the assumption that you want open loop output impedance below 1 ohm (closed loop using feedback this specification is trivially obtained) all you need is an output stage of sufficient grunt. This too is trivial. Just a matter of scale.
The output impedance is usually just the value of the resistor in series with the output. If you need a lower impedance, all you have to do is bypass this resistor.
Any op amp output will be sub-ohm, and most open loop buffers will be around an ohm.
The main consideration then, if you want only unity gain, is whether you can countenance an output coupling capacitor ... which, for 12 ohms, will have to be pretty large.
The problem, if you insist on a direct coupled circuit, is over-current protection since you can't rely on the series resistance for this anymore (you removed/bypassed it...). Accidentally putting close to the rail voltage through your 12 ohm headphones will most likely fry them... and the amp. Even turn on/off transients are a serious concern.
Personally, I'd go with a MOSFET single-ended buffer, like the Szekeres. Simple, cheap, and safe.
Any op amp output will be sub-ohm, and most open loop buffers will be around an ohm.
The main consideration then, if you want only unity gain, is whether you can countenance an output coupling capacitor ... which, for 12 ohms, will have to be pretty large.
The problem, if you insist on a direct coupled circuit, is over-current protection since you can't rely on the series resistance for this anymore (you removed/bypassed it...). Accidentally putting close to the rail voltage through your 12 ohm headphones will most likely fry them... and the amp. Even turn on/off transients are a serious concern.
Personally, I'd go with a MOSFET single-ended buffer, like the Szekeres. Simple, cheap, and safe.
- Status
- Not open for further replies.