Lm3886 Pcb

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Even those N/C pins help...

soundNERD said:
yes, they can, i think. I mean they just sit there serving no purpose, so why not chop em off?
Yes, they can be snipped off, and the constructors using the dead-bug style of construction routinely do this. But they do serve a purpose. They help to build a stronger mechanical coupling between the chip and the PCB, and this could be important for those who don't use any other means of coupling between PCB and heatsink. (Not using any other brackets/attachments is always a bad idea, but for Gainclones, many constructors seem to be doing it.)
 
knabsol said:
Thanks alot roadkill! :)
About the inductor around r7, is it about 10 turns with a 1mm wire around..hm..5-6mm?
Hope I´ll try it this weekend!
Yes, this will work. I used a much more approximate method, just wound the wire around the 10 ohm resistor body. This will reduce the inductance, so I wound more turns (roughly 15, if I remember right)

soundNERD,
The traces are wide enough for me :) Well, actually I ended up soldering a few spare component leads parallel to the board along the power supply traces, just to be sure. The amp worked fine without this "beefing up" into a Caddock 4.7 ohm, 20W resistor kept in a mug of water, driven by a sine wave at 1 kHz (roughly 0.5Vpp... I don't have a scope!) for about 20 min. The resistor blew up later, but that's a diff story. Damn, I'll never get my hands on a power resistor like that one again! :bawling:
 
I have decided to opt for inexpensive 15-0-15V 500mA toroidals in my preamps just to keep the EMI around low-voltage high-impedance circuits as low as possible. These transfies are Rs.280 each, all inclusive, delivered from B'lore to Bombay. (An equivalent E+I in Bombay is Rs.50 each.).

I am not regularly reading this thread. Hence late reply.
Torodoils r not that expensive. I am regularly uwing them. In low watts it is costly. The 15W EI transformer is also not Rs.50/- as mentioned. It must be Rs. 80~90 with crngo lamination. The same in torodoil will cost Rs. 160/- max in Mumbai or Pune.
In higher watts there is hardly much difference in the cost. Iwas myself was making transformers for my own consumption but stopped from last many years. This is for info of my fellow indians who think that torodoils r very expensive.
Mahendra Palesha
 
palesha said:
Torodoils r not that expensive. I am regularly uwing them. In low watts it is costly. The 15W EI transformer is also not Rs.50/- as mentioned. It must be Rs. 80~90 with crngo lamination. The same in torodoil will cost Rs. 160/- max in Mumbai or Pune.
In higher watts there is hardly much difference in the cost....
Hello, Mahendra, good to hear from you. :)

The prices that I'd quoted are based on Visha and one or two others offering me 15-0-15V, 500mA E+I transfies at Rs.50 each. I've checked multiple times; this is what they quote. Of course, I haven't checked whether they transfy will actually handle 500mA without overheating or voltage droop. Maybe the figure will go up a bit if you want a bill with all Sales Taxes, etc paid. And the same thing in toroid is indeed Rs.170 or so in B'lore, but before duties and taxes. After adding excise duty, sales tax, and shipping costs, it comes to Rs.280 when shipped to Bombay.

Can you please point me to any sources who can sell me retail quantities of toroidals in Bombay or Pune?

Thanks and regards,
Tarun
 
There r two suppliers of torodoils with me . One is from Vardha & other is from Pune. Both r equally good. I will check with them & let u know about who is more interested in retail business. The Vardha party is even supplying to Mumbai.
I will be very happy help DIY like u.
Mahendra Palesha
 
palesha said:
There r two suppliers of torodoils with me . One is from Vardha & other is from Pune. Both r equally good. I will check with them & let u know about who is more interested in retail business. The Vardha party is even supplying to Mumbai.
As always, thanks. :) And even if the Vardha party can give me a pointer to any of their resellers, that'll be good enough. And actually, even Pune is okay with me, since I keep visiting friends and business associates there. Next time, I intend to try to meet you. :D
 
Do you know Shri Mahendra Dave of Cadence Audio? If you know him then perhaps you can request him or the Swamijee (Cadence designer) to participate in this forum.
Yes, I know Mahendra Dave. We know each other from our childhood. We were in the same school. I will check whether he is interested or not? I hv not met Swamiji an austrian. I think technically swamiji looks after all the things. Mahendra Dave & Ajay Shirke looks after production & marketing. Sometime back i hv some discussion with Mahendra Dave about speaker testing. But he told me that swamiji has setup all the things.
Mahendra Palesha
 
roadkill's output and ground traces are way too small, and it is not star earthed so there is the possibility of hum and/or oscillation - having made quite a good job of the v+ and v- traces with .1 or .15, he has neglected the fact that the ground trace and the output traces needs to be at least as wide, it is very likely that his amp is oscillation and so drawing it's rated power but not delivering it to the load hence not blowing up the output traces, rember those parts are capable of pushing about 4A rms each, so the supply and ground traces could well carry 8A continuous with 10A peaks, this means .1 at the very minimum.
 
roadkill's output and ground traces are way too small, and it is not star earthed so there is the possibility of hum and/or oscillation - having made quite a good job of the v+ and v- traces with .1 or .15, he has neglected the fact that the ground trace and the output traces needs to be at least as wide, it is very likely that his amp is oscillation and so drawing it's rated power but not delivering it to the load hence not blowing up the output traces, rember those parts are capable of pushing about 4A rms each, so the supply and ground traces could well carry 8A continuous with 10A peaks, this means .1 at the very minimum.
About the ground traces: no, you're mistaken in believeing that they carry the load current. The speaker's ground return wire does carry this current, but that goes straight to the PSU star ground point, not to the amp PCB. The amp PCB's local ground serves two purposes.
  • One is a common point for the onboard capacitors. These are small, only 470uF or so. They're there to compensate for the series resistance/inductance of the wires used to bring the power from the PSU to the amp PCB.
  • The other is that the amplifier's ground reference must be returned back to the power supply. This reference is what the input voltage is compared against.
The point about my output trace not being wide enough is probably valid, I guess it should be thicker than what I've used so far. Anyway, like I had mentioned earlier, I had driven roughly 45W RMS into a resistor load for 20 minutes with no ill effects on the amp or PCB. At the end of this test, the output load resistor blew up :) but the zobel network resistor was not even warm. I feel this is proof enough that the amp was not oscillating, even though I've not looked at the output with a scope (coz I dont have one at home)
 
One question.
If the second pin of LM3886 is not connected internally, could it be for example connected to V- net? This would eliminate that nasty half-loop in roadkill's pcb.
Alternatively, I suppose it could just be chopped off :) and the same result would be achieved
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.