B1 Buffer Preamp

Jump to Latest
#183
But not so suitable when I read this:

Note that the DS1802 will not, working with shown supply (you can try to push it harder but you will do that on your own risk), accept the conventional output level of the CD players. It will certainly clip with usual 2V RMS and more, and with such sources DS1802 is unusable unless you decide to attenuate the signal before it enters the pot (in my view this approach does not have much integrity though). Clipping is instant and very unpleasant but generally all is well as long as it does not occur. So, a lower voltage at the pot input is a must. This, however, won’t be that bad for those using common base based or purely passive I/V converters in their CD players – almost always such kinds of stages will be happy to develop lower voltage.
Click to expand...
 
#185
GuidoR said:
Quote:
you can pretty easily use exact schmtc

Zen Mod: Why 16V and not 18 as in B1's schematic?
It's enough to add a LED to obtain 18V?

Many thanks
Click to expand...


naah
just think - it's not carved in stone ;

that schematic isn't originally intended for B1 , but for some Serbian NS10 project for newbies (for 24V); and l;ater someone asked me to rearrange things for 16-17V ......

more precisely - voltage on output will be approx . 8xUled + 1V35 , where 1V35 is Ube for darlington . it wasn't made for extra precise voltage , but rather for stability

if you use green leds , they can be anywhere between 1V9 and 2V1 , but - as with most project s, difference of 10% in PSU voltage is important just for newbies .......
 
#189
JFET grades

Why does the B1 (and Pass amps more generally) use GR or BL grade 2SK170s? I mean, the circuit obviously has to be designed to use some Idss grade, and Papa did design it to use GR or BL, but why? Is there something about lower Idss grade JFETs that makes them audibly superior to higher Idss parts, or is this just another example of Papa using the parts he has on hand, even though another part could have been used equally well in a slightly different design?
 
#190
B1 as a headphone amp?

A question somewhat outside of the discussions here.

Can the B1 be used as a headphone amp for high-impedance, high sensitivity headphones, such as Sennheiser HD600 (300 ohm nominal, 97db/mw). I do not think such headphones will need more than a couple of mA to reach very high sound levels. I have been using my DIY Aleph P with excellent results. And I almost never need to go past unity gain.
 
#191
Again I don't have a schematic at hand, but nevertheless:

Put in simple terms, GR, BL are used to reduce power dissipation as these parts only take reliably about 200mW of heat. No audio mystery here.

Preamps are designed to run high impedance loads (power amp) of at least a couple of kiloohms. I don't know the bias current of the follower in the B1, but it's probably too small for a headphone amp. Increasing it will very probably exceed the ratings of the used jfets*. If you want to build a nifty Jfet headamp, use this one - it's again a source follower:

EUVLs DAO headamp

http://www.diyaudio.com/forums/showthread.php?postid=1130805#post1130805

The drawback is that the used powerjfets are already obsolete, however www.audioprojekte.de still stocks them (as far as I know that is).

Have fun, Hannes

*alternatively you can parallel a large number of matched jfets - happy matching!
 
#192
h_a said:
Put in simple terms, GR, BL are used to reduce power dissipation as these parts only take reliably about 200mW of heat.
Click to expand...
Yes, I now that is a reason to use low Idss parts. Another is that you have low Idss parts in the drawer but would need to order higher Idss parts. My question is whether there is an audible advantage to using low Idss parts. Surely just saving a few hundred milliwatts of heat can't be the reason when we'll accept hundreds of watts of Class A heat in the pursuit of better sonics.
 
#193
Surely just saving a few hundred milliwatts of heat can't be the reason when we'll accept hundreds of watts of Class A heat in the pursuit of better sonics.
Click to expand...

Come on! Read more closely! The part does not survive reliably more than 200mW. Heat it up to 1W and it simply burns out. Even more and you get a puff of smoke.

Enjoy, Hannes
 
#194
You might wish to read Nelson's latest article "Leaving Class A" on the benefits of high bias.

http://www.diyaudio.com/forums/showthread.php?s=&threadid=125765

JFETs are not too different from MOSFETs in that respect.

As long as the dissipation of the device would allow, I personally prefer higher bias.
That means using V parts, e.g. 2SK170V.

If you read my thread on Pass discrete opamp, you might also notice that I was using BF862 with 18mA bias, in exactly the same follower configuration, for the output stage of the opamp.

You need to be able to handle the heat of course.

😉


Patrick


PS If you would believe Steen, then the DAO follower is certainly amongst the best headphone amp there is around. Ask him.

🙂
 
#199
output caps

Hi,

10uF output cap and 1K is about 16Hz for cutoff freq(?)

So we should not use anything below 10uF?

What if I used 2.2uF with a 4.7K resistor? is this close to same as 10uf and 1K without affecting performance?

trying to stick with lower value caps.

thx.
 
#200
filoter RC time constant

Bengali said:
10uF output cap and 1K is about 16Hz for cutoff freq(?)

What if I used 2.2uF with a 4.7K resistor? is this close to same as 10uf and 1K without affecting performance?
Click to expand...
Just multiply the two R & C values to get the RC time constant.
eg.
1k0 & 10uF = 0.01Seconds = 10mS
4k7 & 2u2F = 0.01034Seconds = 10mS

Frequency @-3dB = 1/2/Pi/RC
for a single pole passive filter fed from zero source impedance and supplying an infinite load impedance.

BTW,
10mS will cut some bass from the signal. Compare to 20mS and 40mS to hear the effect.
I recommend ~90mS for power amps and >=150mS for pre-amps.
 
You must log in or register to reply here.
Top Bottom