Some other Source Follower Configurations

[crt]

Someone asked about modifications to the DAO follower as built by Steen.

This is the version with the Taylor Current Source.
The 2SJ103 should have about 8mA Idss.

Depending on JFET Vgs variations, the resistor values might need adjusting. I should start with a 1k trimpot for R1 to set DC offset to zero first.

Hope you'll build this. Total dissipation about 15W.


Patrick

can you more specific about heat dissipation for each JFET and MOSFET

Thank's

 

[HIPCHECK]

Just built my Dao

Built 1 ch of my Dao amp. So far only 1 puff of smoke.I lost a J511 because I had it hooked up wrong! Thats sad news. I only ordered 4. I will have to order 1 more to make the 2nd ch.

I can not comment to much about sound, etc. I am waiting to make sure it is stable. I built it P2P and just want to let it settle in before I do any evaluation or drag out a set of real headphones. I will take it all apart and countersink the mounting holes, lap the sinks better and swap from sil-pads to mica and goop.

I have it hooked up to my Mini Aleph supply pcb. +/- 20 V with a CRC. total OVERKILL !!!!! I will look into building a supply soon. Pics will come later.

Steve.

 

[EUVL]

Sorry to hear that.

Normally you should try to test with a lab supply with the current limit on at say 400mA.
But then that would still not save your J511 if connected wrongly.

Perhaps avoidable when we finally issue our PCBs ...


Patrick

 

[HIPCHECK]

Yes, that is the plan.

I will not mention that I etched 6 pcbs. I noticed that I did not mirror image them and could not be used the way I planned. That is just a learning lesson as they say.

Steve

 

[b0ri5]

Mark is preparing to build the first proto.




Patrick

Great news.

Thank you again, and keep the good work up

 

[jameshillj]

Hipcheck,

For a really overkill p/supply, use one of Salas's "beefed up" shunt regs!

No suprise, noticeable difference.

For a lower power consumption, there is a double Cmultiplier design available but I don't think it's been designed or tested for 250mA - I did try a simplified double version of the CMx part of the basic Teddy-Reg (pink fish media) with the H44/45 transistors - quite reasonable results.

Works really well with a "Lightspeed Vol Control" - simple systems ....

... my 2 cents

 

[EUVL]

There are enough Cap Multiplier circuits around for Class A power amps that you can use for this.


Patrick

 

[pidesd]

one reg i d like to try is this one, or at least the cap X part:

http://www.diyaudio.com/forums/digi...e-nos-dac-using-tda1541a-369.html#post2584184

right now i m using a cap mult like the one shown in this post and think it souds good. i have nt tried many configurations though.

http://www.diyaudio.com/forums/digi...e-nos-dac-using-tda1541a-369.html#post2584184

 

[pidesd]

oops this one is the second link

http://www.diyaudio.com/forums/digi...e-nos-dac-using-tda1541a-371.html#post2607090

 

[jameshillj]

For the required current, you'll need bigger transistors that John has shown on post #3978

There is quite a difference in the "sound" of C-Mx, between transistors and Fets, particularly noticeable in headamps - it makes these simple 'critters' so interesting!

 

[pidesd]

can't a bd679 support more than around 600ma without problem?

 

[EUVL]

All depends on how much dissipation.
But probably yes with 600mA and appropriate heat sinks.

Have not used them before.
Maybe you should ask John Brown.



Patrick

 

[pidesd]

hi all,

in an attempt to make a simple headamp, i wired 1 cahnnel of 2 lu1014d in a follower config, exactly like the DCB1. now i thought it would be plug and play. the sound is not bad. but i get input offset of 150mV or more if Vds get past around 8V. are these jfet not optimal in this config to start with and must they be used with 3V vds only to sound good?

 

[EUVL]

If you want to drive them in triode mode, you need to run it in a cascode as shown in the DAO schematics.

The offset you can adjust by trimming one of the source resistors.


Patrick

 

[pidesd]

sorry, i ll reformulate.

i wanted to experiment with simple headamp using 4 leftover lu1014d only.

can i use them exactly like in the DCB1 and get decent performance, playing with current and psu voltage?

or must they be used in triode mode only like in the DAO (which sound sublime )?

 

[juma]

Cascode is there to keep the Vds fairly constant. That's the way to keep the LU1014 working in the narrow triode-like region of the output characteristic.

 

[TiEx]

Very interesting project there Patrick, nice job! I have a question regarding load resistance. Will DAO drive a 85 Ohm or 100 Ohm Orthodynamic headphones? I've tried to read this threadm but I found that load must be 25 to 67 Ohm for perfect results.

And will TCS or XCS can help DAO to drive bigger loads then 67 Ohm. 100 Ohm is maximum that I need. I've seen square wave with 200 Ohm load that gives overshoot. Can you put mesurements for 100 Ohm with standart DAO, DAO + TCS and DAO + XCS.

And last question is about es9018 with SEN V18 I/V, will this configuration work together with DAO amp?

 

[EUVL]

100 ohm // 200 ohm = 67 ohm.

Therefore, you can still get perfect square wave by paralleling the 100R headphone with a 200R power resistor at the output of the DAO.

For ES9018, R_iv is 680R. This should be low enough to drive the DAO.
But if you have long cables in between, then perhaps you might need a JFET buffer.
It's a matter of taste.


Patrick

 

[Woodenhead]

I was wondering if there was any news on how the prototype pcb build went?

 

[EUVL]

The PCB has been there since at least 6 months.

Mark wanted to build, but has no time.
All resistors are with him.
If nothing happens till October, I shall take over the build.

Terribly sorry to keep you waiting so long.


Patrick