sheldon
The majority of headphone users have at one time or another, (perhaps quite a few times ?) tripped over their 10 feet long headphone cords and caused some kind of damage to the headphones or the wiring of the jacks and plug. A tip to sleeve, or ring to sleeve short circuit due to such an accident,would most likely be resented by the amplifier. Many headphone users even have special wooden stands etc. to help with the safety of their headphones.
SandyK
The majority of headphone users have at one time or another, (perhaps quite a few times ?) tripped over their 10 feet long headphone cords and caused some kind of damage to the headphones or the wiring of the jacks and plug. A tip to sleeve, or ring to sleeve short circuit due to such an accident,would most likely be resented by the amplifier. Many headphone users even have special wooden stands etc. to help with the safety of their headphones.
SandyK
Meier-Audio's way, good for sure but a simple switch to short the 120r output resistor takes less room.
Some amplifiers to be sure. I think this one is pretty safe as is.
Remember it already has 10R emitter resistors, and is cap coupled to the output. If more is needed, the 317 in the supply will shut down before the output devices overheat.
Sheldon
Sheldon,
I'd not realized that JWMiller is offering flatpack transformers. But digikey isn't actually keeping them in stock and it would require a bulk purchase to source them. Digikey does stock Signal Transformer semi-toroidal flatpack transformers, series LP. But if you do use a separate from the actual amplification unit ps, those semi-toroidal transformer's advantage of smaller
EM field is of less importance. Then an often more readily available split bobbin transformer works quite well.
For me, the major advantage of a separate power supply is that I can comply with 'best practices' wrt to safety, and not have to safety ground the conductive chassis of the actual amplifier, if the voltage is under 50vac or dc,. Dunno about you guys, but for me, a safety grounded signal carrying chassis and hum are often close to inseparable.
Regarding switching supplies, I've yet to meet audio component using one that hasn't benefited substantially (subjectively) from a reasonable power conditioner feeding it. A combination choke with 2 flanking X2 capacitors seems to make everything "mo' betta". While the lack of 'Y' caps does compromise potential common mode attenuation, keeping noise off the safety ground eliminates other potential problems, leastwise, for me. Indeed, many switching supplies do have similar filters incorporated into their inputs, but they are there for FCC compliance, and aren't particularly effective for audio use. One would think using the combo CMC on the output of an switching supply would handle both mains and switching supply noise [and the higher current draw minimizing potential 'Q' peaking from the filter as well as allowing more optimal filter characteristics], but my preference, subjectively, has been to filter at the input. This also works with switching amplifiers. Believe it or not, such a filter can actually make a NuForce or Spectron amplifier listenable!
FWIW,
Paul
Could we implement some sort of power limiting?
In the case of a shorted output, the 10R resistors would give some protection, but to be completely protected we would have to disconnect the feedback loop and then adjust offset, which is probably not very easy depending on the amount of feedback, and wouldn't be immune to temperature drift.
We could use an opamp-based limiter connected to the input of the OPS, which would go high or low a few volts in the case of overdrive. This is even improved by the 10R resistors.
- keantoken
In the case of a shorted output, the 10R resistors would give some protection, but to be completely protected we would have to disconnect the feedback loop and then adjust offset, which is probably not very easy depending on the amount of feedback, and wouldn't be immune to temperature drift.
We could use an opamp-based limiter connected to the input of the OPS, which would go high or low a few volts in the case of overdrive. This is even improved by the 10R resistors.
- keantoken
Last edited:
Andrew,
Quite right, I'd missed that....
KT,
With a 24V rail and half rail biasing, we'd have 12/10 or 1.2A through the upper device max, same through the lower, and this is well within the SOAR rating of the device (100mS, Vce 12V 2A). We don't need to worry about the output devices; they are fine. I'm not sure we would ever need to disconnect the fb loop - heaven forbid!
Hugh
Quite right, I'd missed that....
KT,
With a 24V rail and half rail biasing, we'd have 12/10 or 1.2A through the upper device max, same through the lower, and this is well within the SOAR rating of the device (100mS, Vce 12V 2A). We don't need to worry about the output devices; they are fine. I'm not sure we would ever need to disconnect the fb loop - heaven forbid!
Hugh
KT,
Besides the 5 ohm limiting resistor in the power line plus 10 ohm emitter resistor and half rail of 12V.
Worst case will occur when the transistor is biased to drop Vce of 6V across it, then the power dissipation would be a maximum at 400 mA dc and dissipation will be 2.4 watt. Besides there is a series cap in the output and this condition will only exist during square wave testing at half power.
This is a 12.5 watt device, it is well within the safe operating area. It would get pretty hot under these conditions if you have no heatsink at all and leave a white smelly mark on your finger.
Nico
Besides the 5 ohm limiting resistor in the power line plus 10 ohm emitter resistor and half rail of 12V.
Worst case will occur when the transistor is biased to drop Vce of 6V across it, then the power dissipation would be a maximum at 400 mA dc and dissipation will be 2.4 watt. Besides there is a series cap in the output and this condition will only exist during square wave testing at half power.
This is a 12.5 watt device, it is well within the safe operating area. It would get pretty hot under these conditions if you have no heatsink at all and leave a white smelly mark on your finger.
Nico
Well, Since no one else thinks it's a problem, that's pretty convincing for me.
I would calculate it nonetheless, but I don't know how when you add the emitter resistors to the equation. Not to mention, headphone power ratings cannot all be the same.
How does using those as emitter resistors affect the sound? Is it worth it from a design and subjective standpoint?
Thanks,
- keantoken
I would calculate it nonetheless, but I don't know how when you add the emitter resistors to the equation. Not to mention, headphone power ratings cannot all be the same.
How does using those as emitter resistors affect the sound? Is it worth it from a design and subjective standpoint?
Thanks,
- keantoken
Member
Joined 2009
Paid Member
That's a good question, I'll be interested in seeing answers to that one.
My thought is that because we're in Class A these resistors cause no problems (any resistor adds noise of course). I suspect they improve stability of the amp into different loads. They define to a large extent the output impedance (ignoring for a minute the additional output resistor we plan to add). To be honest, I've never really thought about them and it's a good thing to look at the different elements of the amp and ask 'why'.