Tested the modified current source for differential pair as suggested by HH. Seems to be working fine. I incresed the current to about 16mA per device and I'm using 300 ohm drain resistors. The input mosfets are warm but ZVP current mosfet gets pretty hot (TO-92) case, so I had to use a small heat sink. The voltage is +/- 20V .I don't know why I'm delaying finishing this amp.
Attachments
Here's Harry's description of a circuit taken from another thread:
"Front End current sources
This shows the front end current sources for the Tex-X balanced Aleph-X type amp. The voltage reference is an LM329 with a voltage of 6.9 volts. This is biased by a couple of J501 current reference fets and series resistors from the negative 15v rail minimize the capacitance. Calculated rejection of audio band supply noise is about a factor of a few million. A typical zener and resistor combination gives a rejection of about a thousand fold for supply noise. The voltage noise for the LM 329 is also much lower than for a zener. I used Zetex ZVP3310s for the mosfets instead of the IRF 9910s for much lower capacitance and twice the simulated output impedance. A series 10 ohm resistor for each current source will also help to isolate any RF interactions with the front end IRF9610 diff pairs. Bias value is very stable after the mosfets warm up. A Rifa 100nF cap across the LM329 insures RF stability and filters any RF noise from supplies."
The pictured circuit is slightly different from the above, this is however how I did it.
"Front End current sources
This shows the front end current sources for the Tex-X balanced Aleph-X type amp. The voltage reference is an LM329 with a voltage of 6.9 volts. This is biased by a couple of J501 current reference fets and series resistors from the negative 15v rail minimize the capacitance. Calculated rejection of audio band supply noise is about a factor of a few million. A typical zener and resistor combination gives a rejection of about a thousand fold for supply noise. The voltage noise for the LM 329 is also much lower than for a zener. I used Zetex ZVP3310s for the mosfets instead of the IRF 9910s for much lower capacitance and twice the simulated output impedance. A series 10 ohm resistor for each current source will also help to isolate any RF interactions with the front end IRF9610 diff pairs. Bias value is very stable after the mosfets warm up. A Rifa 100nF cap across the LM329 insures RF stability and filters any RF noise from supplies."
The pictured circuit is slightly different from the above, this is however how I did it.
Attachments
Hi Peter,
I am very interested in the implementation of the LM329/129, and how it works in this circuit. I think the implementation of the LM129 is simpler using fewer components. Would you be so kind as to post a link to the thread in which Harry posted this circuit?
Are you going to make a daughter board for this circuit or will you go PTP?
Thanks,
Rodd Yamas***a
I am very interested in the implementation of the LM329/129, and how it works in this circuit. I think the implementation of the LM129 is simpler using fewer components. Would you be so kind as to post a link to the thread in which Harry posted this circuit?
Are you going to make a daughter board for this circuit or will you go PTP?
Thanks,
Rodd Yamas***a
Here is one link http://www.diyaudio.com/forums/showthread.php?s=&threadid=5392 and here's another http://www.diyaudio.com/forums/showthread.php?s=&threadid=3748&perpage=15&pagenumber=15
I didn't see the implementation of the LM129. Is there any link?
I'll do everything p2p.
I didn't see the implementation of the LM129. Is there any link?
I'll do everything p2p.
Yes, that’s it Peter, I remember it now. LOL… I miss HH. (Oops, sorry moderator.)
The data sheet from National has a simple circuit using an LM129 in place of a zener. Here’s the link.
http://www.national.com/ds/LM/LM129.pdf
I figured you’d go PTP.
Thanks again,
Rodd Yamas***a
The data sheet from National has a simple circuit using an LM129 in place of a zener. Here’s the link.
http://www.national.com/ds/LM/LM129.pdf
I figured you’d go PTP.
Thanks again,
Rodd Yamas***a
C
CryingDragon
Peter,
the implementation of an ultra stable, low noise current source like the one you are showing presents several technical challenges in terms temperature gradients and thermocouple effects in the components used. These may very well send your theoretical 100 ppm factor right out the window.
I have found the lm308a pdf on the national website to be very informative I will probably dig a little deeper than that later.
If you have to go to the trouble of setting this circuit up you may as well do it in the best possible way.
the implementation of an ultra stable, low noise current source like the one you are showing presents several technical challenges in terms temperature gradients and thermocouple effects in the components used. These may very well send your theoretical 100 ppm factor right out the window.
I have found the lm308a pdf on the national website to be very informative I will probably dig a little deeper than that later.
If you have to go to the trouble of setting this circuit up you may as well do it in the best possible way.
I used J505 (4.7mA). I just found it in my box, it was there for years, bought it somwhere locally. But I,m interested too where to get more. Maybe this link helps: http://www.findchips.com/
Front end board mounted in the amp, awaiting connection. Point to point is the way to go on all Alephs.
Note how output Caddock resistors are connected. The flat pieces of copper (Goertz speaker wire) are used for output buses and Caddocks solder directly to them. The short pieces of the wire coming from Dales connect directly to binding posts.
Point to point is the way to go on all Alephs.Note how output Caddock resistors are connected. The flat pieces of copper (Goertz speaker wire) are used for output buses and Caddocks solder directly to them. The short pieces of the wire coming from Dales connect directly to binding posts.
Attachments
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.