Differential CCS'd headphone amp

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Trying to figure out a IRF610-based differential CCS'd headphone amplifier for 300 ohm sennheisers. Started from the SOZ v7 and looking at a few other IRF610 amps (pre and headphone).


I'll try to post a schematic but it's pretty much a SoZ 7 with stats like this:

Single 60v supply.
245mA CCS at the bottom. (122mA or so per IRF610)
82 ohm load resistors, putting the drains at ~50v.
610's sources are at ~35v.
Sources connected directly together and to top of CCS, which has the remaining 35v across it.


Bias/feedback network is 25k feedback resistors, 2.5k input resistors, 100k to ground for a bias voltage right at 40v. With the 82 ohm load resistors gain is around 6 at the moment. (Which is fine for me.)


The CCS has so much voltage across it because it is cascoded and has a few extra volts across the top FET for lower shunt capacitance.

Am I in the right ballpark? Having only 10 volts across the load resistors kind of sucks but I can easily get 28v p-p (+7/-7 to -7/+7) which is huge for headphones Cranking up the current a bit (180ma/side?) is possible I suppose. Given the regulating I want to do on the supply I'm already up around 90v @ 520ma or so though.

This is the first time I've messed with FETs & feedback (tube guy here) so especially advice on the 25k/2.5k/100k feedback/bias circuit would be greatly appreciated.
 
I'm not sure how much the Sennheisers can take, but it sounds as though you're well on your way to overkill. On the other hand, that's usually just about the time things start sounding good.
I doubt that you'll have a problem with voltage swing at normal listening levels.
There's not a single "right" answer to this. Get enough voltage and enough current and you'll have music. Add a little extra of each and it'll get more linear, i.e. the distortion will be lower. Something you might try is to put a test tone into the headphones at what you regard as a normal to loud level. Take readings of the signal going into the headphones. Write the numbers down and sit back to ponder them for a little while. If you only need half a volt or perhaps a volt to get a reasonable listening volume, it will put your mind at rest about whether you have enough voltage swing. At that point, you can begin meaningful experimentation as to what you can modify to make it sound better.
I used MTP7N20E MOSFETs (an On Semi part similar to the IRF610) biased by resistors to make a simple follower headphone amp. Sounds simply marvelous. Quoting from memory, the rails are +-25V and the resistors are 50 ohms, thus giving .5 A bias. The whole contraption weighs about 30 pounds and is gloriously impractical in the commercial sense. Does an excellent job on my Grados, though, and Grados are regarded as a brutal load in the headphone world because the impedance is comparatively low.
Headphones aren't that hard to drive. Just kinda wiggle a signal at them and they'll give you what you want. In my case, I just felt like overdoing it. You can make a simple circuit or a complicated one. Either way, you'll have music. Look at it this way--they charge several hundred dollars for headphone amps powered by a dinky little chip the size of your fingernail and a pair of 9V batteries. Just imagine how much it would cost to buy something like what you're building...

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