PeeCeeBee

Jump to Latest
P

PMI

Member
Joined 2013
ivanlukic said:
How high is the output stage current without 470R resistor (with only a pair of 1N4148 diodes)?
Click to expand...

Total quiescent Current: 130mA~150mA, including VAS current.
(from notes, depending on how the bias is set, and DC rail variation)

Sample measurement:

Total quiescent current 137mA
Front end: 15.5mA
DC supply: 35.1V

Test conditions: 35V-37.5V DC rail voltage.
Total current measured directly w. meter in series
Front end current measured across 10R resistor which is in series w. the 1N4004 diode, ~15mA.

I can adjust the AC voltage roughly, +/- 1~2V using a Variac (not exactly, I am using a monster of a Superior-Electric Variac bought for scrap prices on ebay). The DC volatage can be adjusted within +/-0.1 thanks to my power supply from MrEvil's thread, but it does drift w. temps.

Total bias (quiescent) current should probably be measured only when the main heatsinks reach stable temperature.

If someone can suggest other test conditions, I will be happy to take some measurements.
 
JOHN BALI

JOHN BALI

Member
Joined 2010
naf said:
Wow .. nice build Bli John,
Are those red elcos is elna cerafine?
Click to expand...
yes they are ;) got it from Mr Daniel

shaan said:
Hi John.

200mV on 10R will indicate 20mA instead of 200mA. Is the resistor really a 10R or 1R? Or is the meter showing a dot in the reading? Like '20.0'.
Click to expand...
JOHN BALI said:
I'm re-read again and found my total current is around 200mA
(200mV on 1R resistor you suggest)
I Vas 16mA
Fet Current almost 180mA with that 470 ohm resistor
Rail is +/- 33VDC before 1R safety resistor
Click to expand...
Oops sorry :scratch1:

shaan said:
SUPERKEEWOOT. :D

John, it'll be healthy for us if you finish it fast. Hard to wait. ;)
Click to expand...
Ok, by your request I'll do it faster as I can

PMI said:
Total quiescent Current: 130mA~150mA, including VAS current.
(from notes, depending on how the bias is set, and DC rail variation)
...
I can adjust the AC voltage roughly, +/- 1~2V using a Variac (not exactly, I am using a monster of a Superior-Electric Variac bought for scrap prices on ebay). The DC volatage can be adjusted within +/-0.1 thanks to my power supply from MrEvil's thread, but it does drift w. temps.
Click to expand...
I just remember that my father have one big variac too, I will use it hehe :D
 
JOHN BALI

JOHN BALI

Member
Joined 2010
So here we go... :)
I will try it tomorrow, no comp. caps & no zobel (not yet)
because I use BD139-140
I hope the test is OK, then I can compare it to his older brother :D
 

Attachments

  • IMG_0132.JPG
    IMG_0132.JPG
    130.6 KB · Views: 599
  • IMG_0134.JPG
    IMG_0134.JPG
    114.3 KB · Views: 579
shaan

shaan

Member
Joined 2009
Idefixes said:
Hi shaan, for my personnal cultur could you explain advantages to use 1k/47R with zobel over 2k2/100R alone?

Marc
Click to expand...

Hi Marc.

Personally, I don't see an advantage of the 1K/47R over 2K2/100R.

Both can be made stable, but it is likely that the 1K/47R will introduce oscillation unless some kind of compensation is used. And I am an old hater of capacitors in VAS stage, seems to be another non-linear function added in the equation. See the simulation attachments. The 1K/47R almost doubles the bandwidth. But also notice the peaking. Meaning we need some more small capacitors somewhere in the signal path.

In my experience the 2K2/100R config has always been stable without any extra compensation other than the purely resistive band-limiting induced by the higher value feedback resistor (something called 'autocompensation' if I am not wrong). I feel this is the better way to make it stable.

THD-wise, I doubt there will be any audible difference.

shaan
 

Attachments

  • 2k2.png
    2k2.png
    139.2 KB · Views: 639
  • 1k.png
    1k.png
    140.6 KB · Views: 620
Last edited:
Idefixes

Idefixes

Member
Joined 2003
audiyofield.wifeo.com/index.php
shaan said:
Hi Marc.

Personally, I don't see an advantage of the 1K/47R over 2K2/100R.

Both can be made stable, but it is likely that the 1K/47R will introduce oscillation unless some kind of compensation is used. And I am an old hater of capacitors in VAS stage, seems to be another non-linear function added in the equation. See the simulation attachments. The 1K/47R almost doubles the bandwidth. But also notice the peaking. Meaning we need some more small capacitors somewhere in the signal path.

In my experience the 2K2/100R config has always been stable without any extra compensation other than the purely resistive band-limiting induced by the higher value feedback resistor (something called 'autocompensation' if I am not wrong). I feel this is the better way to make it stable.

THD-wise, I doubt there will be any audible difference.

shaan
Click to expand...

Thanks Shaan.

Marc
 
JOHN BALI

JOHN BALI

Member
Joined 2010
shaan said:
:up:

Seems you're gonna be the first one with the double barreled gun...

I hope you are using 2K2/100R feedback network. If 1K/47R is used then the amp may oscillate without some bandlimiting.

:2c:
Click to expand...
Shaan,
thanks for your secret formula, it is now singing loud :)
I guess this is what I need, the older brother need to some modification

I've just test it with little speaker & lower voltage but I change the gain to 30dB :D :no: :spin:
not oscillated (lucky me, with 1K5//47R resistor)
DC offset is 11.5mV (will fix it latter)

I have a tips how to set zero offset, you see the two bended jumper right...
that is parallel with 18k resistor I use there, touch it with your finger & you now which one 18k's resistor need to be lowered value...

The bad news is the older brother have a short Latfet because I connect to wrong rails :mad:,
need to repair it too with your secret formula ;)

Now I need to finish another channel, I feel so excited...
 

Attachments

  • IMG_0144.JPG
    IMG_0144.JPG
    383.6 KB · Views: 593
  • IMG_0145.JPG
    IMG_0145.JPG
    256.1 KB · Views: 540
  • IMG_0148.JPG
    IMG_0148.JPG
    257.6 KB · Views: 517
M

masood

Member
Joined 2010
JOHN BALI said:
Shaan,
thanks for your secret formula, it is now singing loud :)
I guess this is what I need, the older brother need to some modification

I've just test it with little speaker & lower voltage but I change the gain to 30dB :D :no: :spin:
not oscillated (lucky me, with 1K5//47R resistor)
DC offset is 11.5mV (will fix it latter)

I have a tips how to set zero offset, you see the two bended jumper right...
that is parallel with 18k resistor I use there, touch it with your finger & you now which one 18k's resistor need to be lowered value...

The bad news is the older brother have a short Latfet because I connect to wrong rails :mad:,
need to repair it too with your secret formula ;)

Now I need to finish another channel, I feel so excited...
Click to expand...
dear what secret formula
 
shaan

shaan

Member
Joined 2009
Agh... what innovative speaker design John! :yikes:

Congrats. You are the first to build the dual FET PeeCeeBee made of edible sweet solids. :up:

Please upload the pdfs of layout, silkscreen and mask if you think we should eat... uuh.. build it too.

Again, what innovative speakers John, quite inventional haha...

:worship:
 
You must log in or register to reply here.
Top Bottom