Head-Fi.org › Forums › Misc.-Category Forums › DIY (Do-It-Yourself) Discussions › An hybrid headphones amp... the other way around
New Posts  All Forums:Forum Nav:

An hybrid headphones amp... the other way around

post #1 of 4
Thread Starter 
I've been having some fun playing with tubes models in LTspice. The good old pcl805 used in the Audiovalve RKV particularly attracted me. They're cheap and plentiful but mostly they're powerful ! The pentode section can withstand 8W. Their triode section is medium ยต and said to be decently linear.

As long as one doesn't go crazy with the rail voltage, it's easy to get 70ma of SE class A out of a pair of PCL805. This gets us 0.9W into 300ohms, 100mW into 32ohms. At 600ohms... 50Vp-p.

In the typical tubes applications I quickly checked however (cathode follower, white cathode follower, etc), the output impedance achieved wasn't that great for low impedance cans. So I decided to go the Audiovalve way: adding sand to the mix to increase significantly open loop gain and then apply feedback. And so, here's the crazy beast. A jfet-bjt-tube amplifier with a total power consumption of 50W



The input jfets are cascoded by the two triodes. A ccs in the tail and a current mirror at the top improve gain and matching. The output tubes are triode strapped and sit on top of a bjt CCS. It could be servoed, but I choose not to, for safety reasons.

The HV power supply value isn't that critical: around 140-150V is fine. An isolation transformer is perfect and cheap for the job. It must provides at least 200ma.
The LV voltages supplies (heaters and negative rail) depends on the tubes you get, the value of the negative rail is actually non critical (but regulate it). 6.3V for ECL805, 18V for PCL805. A dual secondary xformer is the best choice.
LL
post #2 of 4
Alternatively:

1. Obliterate the first stage and just use a 1:2 step up transformer feeding into your follower, for A) a more simple circuit, B) galvanic input isolation, C) less heat generated / lost, D) quicker assembly, E) easier fault-finding, F) easier channel matching, G) more consistent performance.

Or...

2. With a gain of about 7.5 and an internal impedance of about 1400 Ohms a triode-strapped PCL805 could be used as a single stage, single-ended, parallel-feed Class A amplifier. Use cascaded / cascoded FETs (such as DN2540) as a Constant Current anode load, and capacitively couple the output to a decent transformer such as a Sowter 8665, and you're a winner! You just don't need to make it any more complicated.


: )
post #3 of 4
Thread Starter 
Actually, I had considered such designs. They're probably way more "orthodox" than mine. So don't take the following as a dismissal. Just the reasons why I finally rejected those two designs.

Quote:
Originally Posted by Leny View Post
Alternatively:

1. Obliterate the first stage and just use a 1:2 step up transformer feeding into your follower, for A) a more simple circuit, B) galvanic input isolation, C) less heat generated / lost, D) quicker assembly, E) easier fault-finding, F) easier channel matching, G) more consistent performance.
Who cares about A, C, D, E on this board ?

I'm not certain about F and G: the 30db feedback takes care of that (and makes tuberolling quite moot).

B isn't really useful to me.

This said, it isn't too bad an idea for sure, but the problem is a 50ohm output impedance.

Quote:
2. With a gain of about 7.5 and an internal impedance of about 1400 Ohms a triode-strapped PCL805 could be used as a single stage, single-ended, parallel-feed Class A amplifier. Use cascaded / cascoded FETs (such as DN2540) as a Constant Current anode load, and capacitively couple the output to a decent transformer such as a Sowter 8665, and you're a winner! You just don't need to make it any more complicated.
And cost skyrockets too !

We just save the negative supply and a pair of heatsinks (which I've got anyway on hand or can salvage for free), compared to the cost of two sowters.

That and the fact that I can enclose the output caps in the feedback loop to minimize their impact; something I can't do with the xformers (unless I add a serious gain stage).
post #4 of 4
Quote:
Originally Posted by 00940 View Post

Who cares about A, C, D, E on this board ?
You might well be right. I hope others chime in to help you take this project forward. Good luck with it...
New Posts  All Forums:Forum Nav:
  Return Home
Head-Fi.org › Forums › Misc.-Category Forums › DIY (Do-It-Yourself) Discussions › An hybrid headphones amp... the other way around