Nice work OS, glad that we can help you out. It was Harry's idea to move C1. Is it possible to post a new Sprint file so we can check the actual file again.
Did you add the positive rail led?
For me having both positive and negative rail leds is mandatory.
Last edited:
For LED I suggest:
https://hu.mouser.com/ProductDetail/Vishay-Semiconductors/TLHR4400?qs=W6eMvlWxUPxv3guYUdjVjA%3D%3D
low impedance type.
Sajti
Last edited:
With enough jumping through hoops I've figured out how to import the board file into Diptrace. In Diptrace I can add whatever images you like to the silkscreen, than I can export it back to Sprint with the images in the silkscreen. I figured out how to save them as a macro too, which is good because all of the importing and exporting is butchering the files!
Attachments
IPS standard -
(Below) spacings for the Wolverine C1 - 34mm is a big as she goes.
JW's solder bridge trick to jump the splits. UMS screw in the way.
To move it right would take up circuit space.
Attached is the standard IPS template. The one I'm using for the Spook !
C1 = 38mm with a 3'rd party IPS. To hit 43mm is possible with the input
euro vertical.
OS
(Below) spacings for the Wolverine C1 - 34mm is a big as she goes.
JW's solder bridge trick to jump the splits. UMS screw in the way.
To move it right would take up circuit space.
Attached is the standard IPS template. The one I'm using for the Spook !
C1 = 38mm with a 3'rd party IPS. To hit 43mm is possible with the input
euro vertical.
OS
Attachments
ZIP your chosen .LMK's , attach.
Edit , or just the whole .lay6 .... I'll make the macro's from the layout.
OS
Sorry I have been away building a router table.
OS - thanks for moving the cap and thanks for adding and moving the LED's!!!
If you have time, can you please reread Sandro's post and comment.
1. He pointed out that it is the common node (of the LTP) that the cascode uses as a reference. Using a 30V zener ( like Wolverine V1) is referencing
the emitter of Q7 (or Q8 C) directly ... -30V.
Using R5 is a distortion mechanism.
2. Q5 is a good idea if the amp is ideal (not abused).
Too many problems and extra components. The reliability and simplicity
of NO Q5 is worth the lost 1-2ppm. This is such a low power IPS ,
do we need 20 nA mirroring ? (below) is the 200nA "sloppy" basic mirror.
Reducing R20 to 2.2K gets us our 2ppm back ... so does using the proper
BAV20 model ! Pave it over with NFB !!
OS
Attachments
Last edited:
Looks Great OS, You need a outline on layer 0 that surrounds the pcb. This is what the pcb companies use to cut the board to not just sizing the properties of your pcb workspace within sprint layout.
Why only three output pairs? I think that more pairs would make this amp more universal. I need to drive 4 ohm 550W RMS load. 5-6 pairs would do and those who do not need that much power could use just 2 or 3 output pairs.
cheers,
I have the UMS source sheet in front of me. It would be easy to make a
UMS fully modular with the 2 row P/N/P/N. But that is for the future.
They did not want a modular this time .... I snuck one into this.
Nor SMD , but I also snuck that possibility in.
PS - 3 pair MT-200's can do over 300W with this design - I had one.
OS
UMS fully modular with the 2 row P/N/P/N. But that is for the future.
They did not want a modular this time .... I snuck one into this.
Nor SMD , but I also snuck that possibility in.
PS - 3 pair MT-200's can do over 300W with this design - I had one.
OS
Last edited: