
Home  Forums  Rules  Articles  diyAudio Store  Gallery  Blogs  Register  Donations  FAQ  Calendar  Search  Today's Posts  Mark Forums Read  Search 
Chip Amps Amplifiers based on integrated circuits 

Please consider donating to help us continue to serve you.
Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving 

Thread Tools  Search this Thread 
17th August 2014, 01:14 PM  #1 
diyAudio Member
Join Date: Jun 2010
Location: Singapore

A Zobel Question on STK4241II
HI,
I am replacing the fried powerstage of a Sansui A60 with a STK4241 Hybrid Chip Amp. Rail volatges are +/52V. So I understand I can only use 80hm Load. The PCB is exactly like the typical application circuit below I already wired up a board except the zobel part. I have used 0.33Ohm 5Watts resistor. Is it okfor me to use a 10µH inductor there at the output instead of the recommended 3µH? I know that's quite a high value. But I have it ready in hand. I have the following from Panasonic. Specification from element14 as below CHOKE, 10UH, 3.9A, 0R026 Product Range: PANASONIC  ELC10D Series Inductance: 10µH Inductance Tolerance: ± 20% DC Resistance Max: 0.026ohm RMS Current (Irms): 3.9A Saturation Current (Isat):  Packaging: Each Case Dimensions: 10mm x 15mm Core Material: Ferrite DC Current Rating: 3.9A Inductor Case Style: Radial Leaded Inductor Type: Choke Coil Lead Spacing: 5mm No. of Pins: 2 SVHC: No SVHC (16Dec2013) Test Frequency: 10kHz Is this good enough as the inductor in series to the speaker load? Can I still use a 4.7Ohm resistor to by pass it? I assume I can also use another 4.7Ohm in series with.1uF to the ground across the speaker load just as in the circuit diagram? I read somewhere than when used with 4Ohm impedence loads(with out even considering the inductance/resistance of the speaker cables), 10uH will result in a 0.4dB attenuation at 20Khz? Will my high end be really affected? Or shall avoid the inductor whole together? But I assume it's necessary to avoid high frequency oscillations. Sorry even if I know a bit about the LC and RC filter equations I am a bit lost here on which all values to consider when calculating. My network theory fundamentals are rather weak Another question is what wattage should those 4.7Ohm resistors be? 
18th August 2014, 02:53 AM  #2 
diyAudio Member
Join Date: Jun 2010
Location: Singapore

No one wants to help me?

18th August 2014, 04:32 AM  #3 
diyAudio Member
Join Date: Oct 2008

I assume p/n is ELC10D100E.
Here's a link to the data sheet: http://www.mouser.com/catalog/specsheets/ELC_Series.pdf From a power dissipation point of view, it's rated at 3.9 Amps DC. Once again, strictly from IsquaredR dissipation, that says that you could dissipate 3.9 Amps RMS. That comes to be 3.9^2*8=121.7 Watts with an 8 Ohm load. With a plus/minus 50 volt supply and 8 ohm load, you can deliver max of about 150 Watts...That says, for power dissipation, in the worst case, it's not quite up to the task. Perhaps of greater concern is that the inductance will vary with signal swing...that will certainly increase distortion. Unfortunately, they give no curves for inductance versus current (or temperature, for that matter). so...you could use all the typical components and values around it, and it would be ok...just not totally blameless. That there's 10 uH...at 20 kHz, that looks like 1.25 Ohms at 20 kHz...there will be a little rolloff...about 0.1 dB (assumes 8 Ohm load). So, unless you're trying for ultimate blameless performance, they should work ok. 
18th August 2014, 07:22 AM  #4 
diyAudio Member
Join Date: Jun 2010
Location: Singapore

Djoffe, thanks a lot for reaffirming my decision. I do not see myself running this amp beyond 50Watts per channel. With the current railvoltage of +/ 43V I surely won't hit 100Watts even at max as there definitely will be a power supply sag also. And yeah I don't think I will use with a load with 8Ohm impedance at 20KHz either. It will be much more higher. So I guess I will be safe in this case.

18th August 2014, 07:34 AM  #5 
diyAudio Moderator

Inductors for stability such as this one are usually (always ?) air spaced. Just turns of wire with no ferrite core.
__________________
 Installing and using LTspice. From beginner to advanced. 
18th August 2014, 07:39 AM  #6 
diyAudio Member
Join Date: Jun 2010
Location: Singapore

So what happens if there is a ferrite core? Will it hurt stability?

18th August 2014, 07:48 AM  #7 
diyAudio Moderator

It can/will saturate on transient peaks... and nonlinearity... its just not the done thing
Your amp circuit is a little unusual. The 0.22 ohm is the one of the main components to aid stability by isolating the amplifier output from capacitive loading. The amp also runs at very high gain (56k/560 feedback factor) which also helps stability. The 0.1uf and 4.7 ohm "Zobel" network isn't technically a Zobel. Its on the wrong side of the inductor for that. It terminates the speaker cable at hf but not the amplifier (as a Zobel would) because of the inductor. The design seems to not need a Zobel (which is unusual) but the other factors I outline obviously render it not required.
__________________
 Installing and using LTspice. From beginner to advanced. 
18th August 2014, 07:57 AM  #8 
diyAudio Member
Join Date: Jun 2010
Location: Singapore

The circuit there is from Sanyo's Data sheet. And I am using it as is. The PCB I have, has provision for an inductor. But not for the bypassing resistor. So I am just soldering the resistor on to the inductor itself. Again the PCB I have doesn't have the RC filter(100nF, 4R7) provision so I am also doing a P2P for that from the output. I am doing all this after reading horror stories of STK chips oscillating. Anyway I am on my way to pickup an old Oscilloscope and function generator. Few hours later I will power up and see what do I see when fed with a 20KHz sine wave. Including the attenuation. I am happy to have 50W per channel. Because anything above STK produces massive THD like 10%.

18th August 2014, 08:00 AM  #9 
diyAudio Member
Join Date: Jun 2010
Location: Singapore

Oh yeah I wired it up now using 5W/4R7 carbon film resistor and the same 10uH coil.

18th August 2014, 11:53 AM  #10 
diyAudio Moderator

That is something that shows well in simulation.
Here is your set up and the response with a 10uh and a 3uh coil. And squarewave of 50kHz into 8 ohms. Also, here is the response at 5kHz with a squarewave using the 3uh value and when the amp is loaded capacitively. Look how the parallel resistor damps the ringing (last picture).
__________________
 Installing and using LTspice. From beginner to advanced. 
Thread Tools  Search this Thread 


Similar Threads  
Thread  Thread Starter  Forum  Replies  Last Post 
zobel question?  fazman  MultiWay  2  18th December 2005 05:11 AM 
Zobel Question  Soulwax  MultiWay  2  2nd May 2005 06:21 PM 
Zobel Question  poeee  Chip Amps  2  5th January 2005 06:29 PM 
zobel question  soundNERD  Chip Amps  17  10th November 2004 08:05 AM 
Zobel question  soundNERD  Chip Amps  12  29th March 2004 12:57 PM 
New To Site?  Need Help? 