I can't make PCB's: 1/2 p2p LM4781

[Beggar]

Hello, i thought i would post this pic of my three channel LM4781 amp i made today, I dont actually understand anything about 'gerber' files, so no one will make my pcbs, thus i have no choice but to do this...

It actually works, and works well too, no nasty noise etc so thats good, i think if i'd made 1 mistake it would have been easier to start again than to try and fix this 'birds nest'!

Anyway cant report on the sonics of the amp at all as ive not got it conencted to any speakers other than some test cones, that will all have to wait till its intergated with my "super amp box" (still to come!)

Feel free to laugh at my amp, i think its pretty funny

Nik

(Special thanks to John)

[digi01]

pretty crazy I like it.

do you set up the chip in standard stereo mode?

[Beggar]

Originally posted by digi01
pretty crazy I like it.

do you set up the chip in standard stereo mode?

Thanks

Its set up in triple mono mode, using it to for rear and centre channels on my amp (still in development )

Nik

[BC]

Hi Beggar!
I do alot of p2p myself. To me, it looks like a decent,clean job . PCB design is an art, however so is quality p2p. I hope you enjoy the sound soon. Never apologize for not having a PCB. However, a schematic is nice to see. Did you stay with Nat. Semi's?

[Beggar]

Originally posted by BC
Hi Beggar!
I do alot of p2p myself. To me, it looks like a decent,clean job . PCB design is an art, however so is quality p2p. I hope you enjoy the sound soon. Never apologize for not having a PCB. However, a schematic is nice to see. Did you stay with Nat. Semi's?

Thanks for the approval

Ive attached the schematic for y'all to look at, its the national reference PCB schematic with a few changes, I have got rid of the feedback caps and the input caps, as i wont have any low frequency, as its driven from a lowpass filter, i ditched the 10uf bypass caps for reasons based on my years of inexperience (looking at other chip amps ) I decided to keep the snubber networks in as this is my first venture into the world of the 4781 and wanted to try and avoid any HF oscillation going on.

But i guess thats more or less it

Cheers, Nik!

[BC]

Beggar,
A few suggestions. These are not criticisms!

1. Use monolithic X7R .1 uF power bypass caps. Far more temp. stable then standard ceramic.
2. parallel 2 1,000-1,200 low ESR power caps instead of 1 2,200 uF power cap. better transient response.
3. use 1/8 Watt resistors for input and possibly gain. Saves LOTS of space, especially with such a dense IC/circuit.
4. If you did not, tie input signal grounds to one point, then one wire to power ground.
5. Bring power rails to board perpendicular to board maybe 2 inches (50.8 mm ) from board. Reduce possible noise.

[Beggar]

Originally posted by BC
Beggar,
A few suggestions. These are not criticisms!

1. Use monolithic X7R .1 uF power bypass caps. Far more temp. stable then standard ceramic.
2. parallel 2 1,000-1,200 low ESR power caps instead of 1 2,200 uF power cap. better transient response.
3. use 1/8 Watt resistors for input and possibly gain. Saves LOTS of space, especially with such a dense IC/circuit.
4. If you did not, tie input signal grounds to one point, then one wire to power ground.
5. Bring power rails to board perpendicular to board maybe 2 inches (50.8 mm ) from board. Reduce possible noise.

Hi ya thanks for the suggestions
I wont actually be modding this cirucuit at all as its not really for 'hifi' however i agree with u on points 2 3, ive sortof done point 4 already
I'll do point 5 if i find it necessary

Point 1 concerns me slightly though, when you say temp stable what exactly do you mean, have i shot myself in the foot using standard ceramics? The reason i ask is because ive bypassed 10 opamps using these same caps!

Thanks, Nik

[BC]

Hi Beggar,
Here are 2 magazine articles discussing 100nF caps:

1. Electronics World November 2002, article by Cyril Bateman discussing distortion of 100nF caps. He talks about signal use not bypass. The only ceramic he recommends has a COG temp rating (I will say more later about that).
2.audioXpress December 2002, article by Tom Perazella discussing power supplies. At the end of his article he states using 100 nF monolithic ceramic caps as power bypass and says " It's also interesting to see how many audiophiles shun ceramic caps, when, in fact, they outperform all the so-called super foil caps in a bypass application".
Now COG. There are other ratings, but here is a standard format from stable temp change to poor %:
COG/NPO
X7R
X5R
Z5U
There are others! The point being, as temp increases, capacitance changes. Noise increases.
You can find COG 100nF ceramics, they are exspensive ( about $6.00 USD) and larger than other ceramics (about 13 mm wide).
Forgive me, but the ceramics you seem to be using are high "k" dielectric and have a poor temp response and high noise. I do not know this for a fact. You may want to research the types you are using. I suggested X7R because their change with temperature can be 15% ( I know, not great) but far better then others. Once again, if you consider size, cost and high freq. bypassing, they are very good. NOT for signal bypass! I hope others of the forum read this thread and contribute.

[microJean]

How much does the sink heats?

[Beggar]

Originally posted by microJean
How much does the sink heats?

Hi ya, I still havent got round to connecting the thing up properly and running it for a considerable time, Im considering putting a temp controlled fan on the sink (A CPU cooler globalwin FOP-32) if it gets to hot, but hopefully i should have something done tonight so i can report back on it for you )

Nik