[Glassman]

most likely wouldn't work.. 0.1uF is 1000x more than 100pF.. they're there to compensate the phase shift, rather delicate task..

 

[peranders]

Is it the one between opamp output <-> inverting input, if yes, the effect would be that the buffer would run without feedback, except for low frequencies, pretty pointless to make this cap big in other words.

A good NPO ceramic or polypropulene would be fine in this position.

 

[@sia@home]

Its just im again having problems finding stuff in Australia, without making a new order from digikey, which I may aswell do now, I think.

Hmmmm

I wasn't betting on such a large value working anyway.

 

[@sia@home]

Quote:

Originally Posted by peranders Is it the one between opamp output <-> inverting input, if yes, the effect would be that the buffer would run without feedback, except for low frequencies, pretty pointless to make this cap big in other words. A good NPO ceramic or polypropulene would be fine in this position.

I understand what you are saying, what I am saying, is i have 0.1uF caps here, that will fit, I don't know if I can get 100pFs anytime soon.

Will it work, and do its job? or going this large not let it do its job?

 

[amb]

Quote:

Originally Posted by @sia@home I understand what you are saying, what I am saying, is i have 0.1uF caps here, that will fit, I don't know if I can get 100pFs anytime soon. Will it work, and do its job? or going this large not let it do its job?

0.1uF is way too large a capacitance for that. There is no local source for 100pF silver mica or multilayer ceramic capacitors?

 

[skyskraper]

asia at home, what parts are you having a hard time finding? i find i can get 99% of what i need from altronics, rs, farnell, and futurlec (futurlec for transistors and other small bits that the others have, got all my gilmore stuff from them almost)

 

[skyskraper]

100pf polyester cap.
farnell number: 3834074
http://au.farnell.com/jsp/endecaSear...=3834074&N=401

edit:
RS number: 115-635
Wima 100pF polypropylene fkp2

 

[peranders]

Quote:

Originally Posted by @sia@home I understand what you are saying, what I am saying, is i have 0.1uF caps here, that will fit, I don't know if I can get 100pFs anytime soon. Will it work, and do its job? or going this large not let it do its job?

It will work but not as intended, see my previous post. I have a hard time to understand that you can't get a 100 pF in Australia fast enough.

 

[hypophysis]

I have purchased a 10 pF silver mica and put it in C6G position and also put the 100 pF cap on C14G - this configuration seems also stable.
So which one ist the (theoretically) better choice:
Leave out Glassman's C14G cap and put 100 pF on C6G, or populate C14G 100 pF and put 10 pF on C6G?

edit:
I have to correct myself. The config I've tried with 100pF C14G and 10pF C6G is a lot more stable than 100pf/100pF but there is still some faint noise when I turn on the amp and some buzzing when I touch the headphone jack, altough this disappears after some minutes of warming up.

 

[motu]

Which of the ad744 opamps listed on the analogue devices site should I buy for the PPA?

http://www.analog.com/en/prod/0,2877,AD744,00.html

 

[steinchen]

one of the dip-8 packaged versions
AD744JN / AD744JNZ / AD744KN / AD744KNZ
the k-types are higher grade

please note that you'll need a 5pF or 10pF compensation cap if you want to use ad744 in the ground channel (in non-inverting mode uncompensated only stable at gain greater than 2)

 

[motu]

How do I use the compensation cap?

 

[steinchen]

ad744.pdf datasheet

across pins 5 and 8, I mounted the cap directly on the opamp.

bypassing the output-stage of the opamp is highly recommended. easiest way to do this is to build an adapter-socket: bridge pins 5 and 6 on dip-8 sockets and bend pin 6 up before plugging the opamp into the socket and the adapter into the PPA. Be sure to remove the classA bias cascode if you use ad744 bypassed, I recommend to socket the R10 resistors

 

[Glassman]

no, there's no need for any additional compensation and the opamp should not be simply put in place of the previous, you should use the compensation pin instead of the regular output pin.. that means bridging pins 5 and 6 on the opamp socket and bending the opamp's output pin out of the way so that it doesn't touch the socket, you will also need to disable the class A biasing by the means of disconnecting the inline 1k resistor that's in front of the jfet cascode..

the guide is posted back here..

 

[__redruM]

I'm trying to "mod" my PPA v1.1 for the GDB, and am unsure how it differs from PPA v1.0. Here's my guess. My biggest point of confusion is C14. There isn't a C14 on v1.1. Is there really one on v1.0, or is this where the 'extra' cap from the kit goes? Lastly, do I still need to make the bridge on v1.1?

Here is my guess on the v1.1 procedure for L, R and G, can someone confirm it or fix it?

1. Remove R12
2. Install 'extra' cap in R12.
3. Bridge R12 buffer side to R14 buffer side.
4. Remove intersil buffers.
5. Edit Remove C3.
6. Install diamond buffers.

If I'm wrong, a schematic would help. Not the schematic to the buffer, but a modified ppa1.1 schematic.