Go Back   Home > Forums > Amplifiers > Chip Amps
Home Forums Rules Articles Store Gallery Blogs Register Donations FAQ Calendar Search Today's Posts Mark Forums Read

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
Reply
 
Thread Tools Search this Thread
Old 27th May 2009, 07:55 PM   #1
diyAudio Member
 
Join Date: Mar 2009
Default Kit Confusion!

Hi,

I am trying to find a good quality kit for a stereo chip amp which can be run off of my existing 25V+25V 120VA toroidal transformer and cooled by a passive GPU heatsink (remnants of other projects).

I have looked at lm3886 based kits and found this one which seems to be reasonable. However, I have read some poor reviews of other HLLY products, but cannot find an opinion on their lm3886 kits.

Ideally, I would like to be able to drive 40W per channel through 4ohm speakers (already have them!) so to do this with the HLLY kit some high power resistors would be needed, but this is not impossible.

The HLLY kit will set me back £30 then an additional £15 for resistors and heatsinks, so something in the £45 range would be the desired target.

I'm new to the chip amp world so any advice from those experienced in this area would be greatly appreciated, thank you!
  Reply With Quote
Old 27th May 2009, 11:16 PM   #2
Minion is offline Minion  Canada
diyAudio Member
 
Join Date: May 2006
Well I don"t know that exact kit but You can not drive 4 ohm speakers with a LM3886 with a 2 x 25v Transformer .... After rectification you get about 35v DC which is good for 8 ohm speakers but not for 4 ohm , You need a 2 x 18v Transformer to drive 4 ohms .....

Also I doubt a GPU heatsink is going to be big enough for 2 LM3886 chips .....


  Reply With Quote
Old 28th May 2009, 12:03 PM   #3
diyAudio Member
 
Join Date: Mar 2009
I am aware that I would not be able to drive a 4ohm load with the lm3886, I would need to use a high power resistor in series and create an 8 ohm load for which the lm3886 outputs about 40W for the 30V DC I get out of my rectifier (actual measured value). Driving the 4ohm load directly is really wishful thinking...

The heatsink I have is this one pulled from a GeForce 7600gs, would it at least be suitable to cool a single lm3886 using thermal interface tape, or would something like arctic silver adhesive be needed?

The copper block measures 97mm*65mm*15mm and the aluminium fins are 52mm*32mm*13mm. If this heatsink cannot be used for both chips, I'm sure I can get my hands on another similar one.
  Reply With Quote
Old 28th May 2009, 02:48 PM   #4
diyAudio Member
 
Join Date: May 2008
Quote:
Originally posted by deadparrot
I would need to use a high power resistor in series and create an 8 ohm load for which the lm3886 outputs about 40W for the 30V DC I get out of my rectifier (actual measured value).
Yes, and you would waste 20 W in the resistor and make the LM3886 suffer unnecessarily.
With an 2x18 V transformer you should get ~40 W into 4 Ohm.

Quote:
Originally posted by deadparrot
The heatsink I have is this one pulled from a GeForce 7600gs, would it at least be suitable to cool a single lm3886 using thermal interface tape, or would something like arctic silver adhesive be needed?
Depends on how you use the amplifier. It loks too small for 40 W into 4 Ohm even for a single channel, but if you have it, try it. Thermal grease is preferable over thermal tape.
__________________
If you've always done it like that, then it's probably wrong. (Henry Ford)
  Reply With Quote
Old 28th May 2009, 04:27 PM   #5
kscharf is offline kscharf  United States
diyAudio Member
 
kscharf's Avatar
 
Join Date: Apr 2007
I purchased a pair of bare PC boards from an ebay vendor that are the same ones used in that kit. The boards are of good quality, the only issue I have with them is a long path for Rf. If you solder Rf directly to the pins of the '3886 or to the back of the the PC board that would solve the problem. The board spec'ed 22k for Rmute, the usual value is 10k; 22uf for Ci, the usual value is 47k and some people use something even higher (or leave the cap out). Also the resistor from the neg input to ground is 1k, and some circuits use 680 ohms, the input resistor is 1k and I've seen it as low as 220 ohms or left out.

Other than variations in part values the circuit is the same as the NS application example and the circuit board is not bad. I don't know how good the kit supplied parts are. You can buy the amp PC board for $2.50 each on ebay (I didn't look at the power supply PC board since I don't need them). You might be better off buying the raw PC boards and getting your parts elsewhere.
  Reply With Quote
Old 28th May 2009, 06:32 PM   #6
diyAudio Member
 
Join Date: May 2008
Quote:
Originally posted by kscharf
The board spec'ed 22k for Rmute, the usual value is 10k; 22uf for Ci, the usual value is 47k and some people use something even higher (or leave the cap out). Also the resistor from the neg input to ground is 1k, and some circuits use 680 ohms, the input resistor is 1k and I've seen it as low as 220 ohms or left out.
Rmute depends on the supply voltage. There is no usual value other than that some people are too lazy to do the math and use 10 k, which works from the smallest possible supply voltage upward. That happens assuming the increased current has no effect on the IC. Well, if it hadn't, why go to the length and work out such a formula for the datasheet?

Ci depends on the size of Ri and the high-pass frequency you want to achieve. The values of Ri and Rf have an influence on the performance. While some people go for low values to avoid Nyquist/Johnson noise, others prefer higher values to achieve low output DC offsets.

The same goes for Cin that depends on the size of Rin and the high-pass frequency you want to achieve.

Rb has a protective function. The lower its value, the less protection you have.

Rin should have a size that your source can easily drive, which means the bigger the better. On the other hand, the bigger a resistor is, the more noise does it introduce, so from that point of view, the smaller the better. And the bigger that resistor is, the higher is the DC offset voltage at the input. Practical values range from 10-100 k.

For good common-mode rejection you should strife for values that let both inputs "see" the same impedance. This means of course that any change to Ci, Ri and Rf implies a change to Cin, Rin and Rb and vice versa.
__________________
If you've always done it like that, then it's probably wrong. (Henry Ford)
  Reply With Quote
Old 28th May 2009, 08:52 PM   #7
diyAudio Member
 
Join Date: Mar 2009
Quote:
Originally posted by pacificblue
Yes, and you would waste 20 W in the resistor and make the LM3886 suffer unnecessarily.
A good point, I overlooked that completely! A new transformer is in order then.

Quote:
Originally posted by kscharf
Other than variations in part values the circuit is the same as the NS application example and the circuit board is not bad. I don't know how good the kit supplied parts are. You can buy the amp PC board for $2.50 each on ebay (I didn't look at the power supply PC board since I don't need them). You might be better off buying the raw PC boards and getting your parts elsewhere.
So the risk in buying the chinese kits is the component quality rather than the board quality? Is there a recommended seller/retailer for these types of items?
  Reply With Quote

Reply


Hide this!Advertise here!

Currently Active Users Viewing This Thread: 1 (0 members and 1 guests)
 
Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off


Similar Threads
Thread Thread Starter Forum Replies Last Post
mox confusion cowanrg Solid State 10 9th August 2005 11:11 AM
More(l) confusion mr_push_pull Multi-Way 2 29th May 2005 04:54 PM
SET Friendly Confusion Mike C Multi-Way 9 14th April 2005 04:29 AM
Noogie X-over confusion jmikes Multi-Way 4 6th December 2004 03:46 PM
I2S confusion MWP Digital Source 2 9th November 2002 10:36 AM


New To Site? Need Help?

All times are GMT. The time now is 04:35 AM.

Page generated in 0.10345 seconds (81.71% PHP - 18.29% MySQL) with 10 queries

Copyright ©1999-2012 diyAudio