Head-Fi.org › Forums › Misc.-Category Forums › DIY (Do-It-Yourself) Discussions › Couple questions about my pocket amp design
New Posts  All Forums:Forum Nav:

Couple questions about my pocket amp design - Page 2

post #16 of 50
2/13/12 at 8:09pm
Thread Starter 
Quote:
Originally Posted by Avro_Arrow View Post

This is what I was going to build to play around with.

I haven't designed a board for it yet but it doesn't really need one.

 

CMB Amp.png

 

I'm using it with a dual rail power supply or a split battery pack

so my bypassing is to ground (0 volts).


powerv3.bmp

Here is the modified powersupply design, I removed the big caps that connected the output of the buf634 to V+ and V-. Any suggestions? . My professor told me that the big caps at the output will mess up the feedback loop, is that true?

 

Reply
post #17 of 50
2/13/12 at 8:30pm

Your professor is correct.

Your power supply looks much better now.

 

When I am drawing in Eagle, I like to use the European non-polar capacitor symbol to

mean ceramic or film caps, the European Polarized capacitor symbol to mean Tantalum caps

and the American Polarized capacitor symbol to mean Electrolytic.

 

Like this:

 

caps.png

Reply
post #18 of 50
2/13/12 at 10:32pm
Thread Starter 

I build the power supply base on the sch. I am having trouble with the oscillation with the buf634.

oscll.bmp

It is VPP 1V @ 80MHZ (pretty bad :( )

 

Should I add a resistor between the opamp and buffer?

Reply
post #19 of 50
2/13/12 at 10:38pm
Thread Starter 

 

NewFile0.bmp

 

Problem solved by leaving the BW pin open :) now it is really nice


Edited by w62820616 - 2/14/12 at 12:23am
Reply
post #20 of 50
2/13/12 at 11:25pm
Thread Starter 

So I build everything. It looks amazing.

NewFile1.bmp

The one below is when under an inductive load

NewFile2.bmp

 

Idle current draw from the batteries is around 20MA and 40 ma for one channel so I guess 60ma for two channels. Is that too much?

Reply
post #21 of 50
2/14/12 at 7:28am

 

Quote:
Originally Posted by w62820616 View Post

Problem solved by leaving the BW pin open :) now it is really nice

 

You may like the results of using a resistor from the BW pin to V- this sets the idle current&bandwidth somewhere in the middle. 

The DIP-8 package of the BUF634 should NEVER have its BW pin shorted. It overheats the buf634 which sets off the self protection circuitry and then all hell breaks loose. 

 

Quote:
Originally Posted by w62820616 View Post

Idle current draw from the batteries is around 20MA and 40 ma for one channel so I guess 60ma for two channels. Is that too much?


If those current numbers are for the left & right channels I would be quite concerned that they are different. 

 

What do you expect them to be based on adding up your anticipated idle currents? 

Reply
post #22 of 50
2/14/12 at 10:57am

Assuming the 2k resistor as in my schematic:

 

With a +-15 volt supply, should be about 8mA through each side of the current mirror so 16mA total per channel.

For +-9 volts it should be around 4.5mA for each side of the mirror.

 

If the currents are different on each side of the mirror or between channels, try match the hfe of the transistor as close

as you can amongst what you have.

 

Reply
post #23 of 50
2/14/12 at 11:19am
Thread Starter 

Sorry I worded it wrong. I built one channel with the virtual ground module. The idle current is around 20ma and when I input a sine wave with a 32R load the current draw goes to 40ma. Therefore I conclude that the current draw per channel is 40ma -20ma =20ma and a complete system will be two channels with a virtual ground so that would be 60MA

Reply
post #24 of 50
2/14/12 at 11:21am
Thread Starter 
Quote:
Originally Posted by Avro_Arrow View Post

Assuming the 2k resistor as in my schematic:

 

With a +-15 volt supply, should be about 8mA through each side of the current mirror so 16mA total per channel.

For +-9 volts it should be around 4.5mA for each side of the mirror.

 

If the currents are different on each side of the mirror or between channels, try match the hfe of the transistor as close

as you can amongst what you have.

 

 


all the fluke and agilent meters does not have a hfe function :(((((((((((

 

Reply
post #25 of 50
2/14/12 at 3:09pm

Here is the long way...

http://www.diamondstar.de/transistor_matching_bjt.html

Reply
post #26 of 50
2/16/12 at 4:36pm
Edit: Sorry, I misunderstood the design intent 
w

Edited by wakibaki - 2/16/12 at 5:12pm
Reply
post #27 of 50
2/16/12 at 6:25pm
Thread Starter 
Quote:
Originally Posted by wakibaki View Post
 


noooooooooooooooooooooo it was helpful

Reply
post #28 of 50
2/16/12 at 9:14pm
Thread Starter 
Quote:
Originally Posted by Avro_Arrow View Post

Here is the long way...

http://www.diamondstar.de/transistor_matching_bjt.html



Thanks for the into. I built the test circuit, however, for my 3904 I have all of them around 9.75 V ( i notice that it will go up as the transistor heats up). For my 3906, I have all of them around 10.75V. Apparently, they have different hfe. Is that ok ?

Reply
post #29 of 50
2/17/12 at 9:57am

It probably won't make an audible difference. If you look closely at the waveform

on the oscilloscope it might show greater amplitude on one side. I just thought

that if you had a bag full you might want to sort through and find some fairly

close ones.

Reply
post #30 of 50
2/17/12 at 2:36pm
Thread Starter 
Quote:
Originally Posted by Avro_Arrow View Post

It probably won't make an audible difference. If you look closely at the waveform

on the oscilloscope it might show greater amplitude on one side. I just thought

that if you had a bag full you might want to sort through and find some fairly

close ones.



I bought 200 of them each from jameco. I randomly selected 20 of each kind and tested them. The result shows that all my 3904 are around 9.75 and 3906 are around 10.75.  Is the opamp feedback gonna fix the unbalance problem?

Reply
New Posts  All Forums:Forum Nav:
  Return Home
Head-Fi.org › Forums › Misc.-Category Forums › DIY (Do-It-Yourself) Discussions › Couple questions about my pocket amp design