Amp3 #1 is done...

[mb1a]

** Updated to reflect completion (more or less...). Details on latest post below (23/5).


Jan: while checking my board, I tried to trace the connections for C1. One pin is on the +5, but the other seems to be NC? I expected it should be GND. Have I misunderstood something here?

Jan / anyone else: my 2nd AMP3 works beautifully, but the 1st board I attempted still has a problem.

- power on slight thump (normal); dc offset normal; completely quiet when no source connected
- with source connected, pretty good sound produced; no noticeable distortion or other problem, except:

There is a moderate level hiss (similar to white noise) on both channels. The hiss remains same level throughout. Sometimes there are clicks and bumps, but mostly its hiss only.

Trouble-shooting trail:

- Disconnect source from one channel; that channel goes quiet. Disconnect both cables -- both channels quiet.
- Connect shorting plug to one channel -- that channel starts to hiss(!).
- Tried other cables -- same result

I've traced and checked most of the incoming pins, and replaced input caps. No change.

Power supply is 12V SLA.

Any suggestions on where to focus my troubleshooting would be appreciated.

[lostboy]

Jan has sent out an e-mail concerning the C1 pad error on the AMP3 - the NC pad should go to ground. There is an earlier thread called 'Help with C1 pad' - in Jan's last post to this thread he explains all.

Chris

[mb1a]

Thanks Chris.

Anyone with suggestions on the tougher hiss and pop problem?

[bearman]

quote:Originally posted by mb1a
[br]Thanks Chris.

Anyone with suggestions on the tougher hiss and pop problem?

You may want to try cleaning your board with Alcohol and a soft fine toothbrush, to get the flux residue off between the legs of the chip.

Anyone correct me if I am wrong and it might do damage please.

[mb1a]

Thanks. That's a good suggestion. Did that quite carefully more than once. Don't think it's a case of flux or dirt.

I'm coming to the conclusion that it's -- believe it or not -- a couple of bad flying RCA leads! I've put this board aside, and I'm working on my 3rd AMP3. On this 3rd board there is intermittent hiss that can be corrected when the RCA phono connection is jiggled. Seems like the connectors may be bad.

[mb1a]

The first of my pcbs is in a simple alu case.

- Power transferred from a 7.2Ah SLA + 9,000uF to a 80VA toroid with 13,200uF; 9V transformer gave ~13.5V on load (unreg.). Would have liked 14V+, but can't be too fussy here.
- Electrolytics on pcb were lower ESR types (eg. Rubycon ZL)
- Very short lengths of pure silver solid core used for input connections (2-3cm)

Comments:

SLA v. unregulated power

I've previously used SLAs on preamps and dacs, and have been very pleased with the results. For the Amp3 the linear supply seems to be a much better fit. It's unlikely that the extra 0.6-0.9V would have done much, so perhaps its the ability to refeed the caps in time? SLA power sounded a little clinical compared to ac.

Amp3 (installed) v. T-amp

Compared to a modded T-amp (extra supply capacitance; better input caps) the Amp3 installed with ac power is a winner by a huge margin. The Amp3's clarity is different -- it's like viewing a 8-12 megapixel image shot with a quality lens. Everything's clear and natural, with nothing sticks out. With the T-amp it's as if the image was 3-5 mpixel, and heavy sharpening was applied to make the image look sharp.

Ultimate power delivery: haven't not tested yet. Using very inefficient, tough Gale 401As from the 70s (84dB or worse, 4ohms). At moderate levels in a medium room, orchestral music, the peak voltage measured at speakers was still ~3V, so I probably had 3-6dB to spare, if the supply, etc, are built correctly. I suspect that the Amp3 with SLA would have clipped at this point. The T-amp would have clipped long before this...

Great kit. SMD work is not for everyone, and the tendency to induce noise if not properly enclosed, grounded, etc means that builders need to be more careful than with simpler designs (eg. chip amps), but my Amp3 already beats my best TDA7294 attempt.

[stuorguk]

quote:Originally posted by mb1a

Great kit. SMD work is not for everyone, and the tendency to induce noise if not properly enclosed, grounded, etc means that builders need to be more careful than with simpler designs (eg. chip amps), but my Amp3 already beats my best TDA7294 attempt.

I hope there will be a non-SMD version of this kit sometime. It's not as if the PCB is going to be much bigger. I'm sure a lot more people would be interested. It's the perfect amp for many Squeezebox owners, which is where I heard about this site. If it's anything like what is happening with Nixie Clock kits, we may see some interest from suppliers of bespoke cases.

[Jan]

Hi,

the problem with a non-SMT version is that the chip itself is SMT. Mounting the chip only is IMO a problem, as it could not be tested until the other components are in place....

[stuorguk]

Hmmm... take your point. How about manufacturing a small PCB with just the Tripath and a pin header pre-installed on it, that can be plugged in, or soldered onto another PCB?

Stuart.

[bearman]

I think the SMD components are the way to go. You have done a good job making space for optional critical components, and if it were to be all conventional components, the board would easily be 3 times larger. With a decent soldering station, the SMD just isn't that tough, and this is coming from an electrical knucklehead.

One thing I would like to see is a PSU board for the AMP 1 available.

Out to make the premier backpacking sound system.