This is not just another gainclone

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
And that's the PS case.
 

Attachments

  • ps.jpg
    ps.jpg
    40.5 KB · Views: 2,388
Hey, so how critical do you guys think that the RC combination on the non-inverting input is? If you take another look at my gainclone pcb, you'll see that I tied the non-inverting input directly to ground.. Any thoughts?

Also, what is the best place to get pcb's made? My school will only make me so many of these boards. :)

--Jordan
 
Jordan,
I initially used the RC on the pos side then took it off then tried a few different combinations going back and forth and I couldn't hear any difference. I really gave it a good listen and honestly ended up shruging my shoulders and left it the last way I had it which was shorted to ground. My speakers have good detail but are getting long in the tooth (12 years), but I consider myself to have good hearing (being in sonar and all). I don't know if technically this isn't an optimum thing to do but it works well for me.
 
"baby GC" using the LM1876

I have just - as for curiousities sake - made a kind of "baby GC" using the LM1876.
As you all know this is a stereo version, with 15W max. or so.

The main disadvantages are: The distances for the feedback are different from one channel to the other and, most of all, there is only one negative supply pin for both channels, which prevents it from using two separated psus.

I - this time - followed more or less the application schematic from National (non-inverted) but without the cap in the feedback path and without the zobel network. The psu are two 22200yF caps by Roederstein (the two pos. pins connected to each other) and a .47yF HQ MKT from + to -. I used a single 10A regular bridge and a 2 x 16V / 250VA transformer. All wiring in done in p2p, of course.

After firing up the tiny beast, it quickly showed up that there was too much DC offset (0.3V), which was corrected by two 47yF panasonic caps in the FB path.

To my very great astonishment even this amp sounds surprisingly good. It does not have the control of its "bigger" brothers, but is not bad nevertheless. And although I only had it mounted on a small heatsink it was capable to play to a fairly high level.

But I think that it still gets "into some kind of oscillation" after some listening, because after it is getting hot it stays a little warm, even with no signal. Maybe I should add the zobel to get out the last instabilities.

Voila ! This is maybe the cheapest possible DIY amp with reasonable sound quality you can make, and it is surely good enough for secondary duties, like driving surround speakers etc. And it is so small ! Without thinking about the heatsink (and the transformer) you could fit one of these into a box in the size of a compact cassette.

Klaus / - likes tiny amps
 
Gain clone problems.

Ok, after seeing the gain clones built I had to try one.
Things are working well enough, no occ and good power.
However, I have a 60 cycle rep as shown in the picture.

It's not quite hum (not sin signal) but it is 60 hz. The peak-2-peak
is just under .02V. What do others get?

The volume has no effect on the signal. Touching the case,
vol control, and other things have no effect.

I have tried moving the xfmer with jumpers 12 inchs from the
chip with the same hum.

I checks the +/- power rails and is's a small sawtooth at 60 cycles.
Any ideas.....

An externally hosted image should be here but it was not working when we last tested it.
 
Re: Gain clone problems.

ted said:
Ok, after seeing the gain clones built I had to try one.
Things are working well enough, no occ and good power.
However, I have a 60 cycle rep as shown in the picture.

It's not quite hum (not sin signal) but it is 60 hz. The peak-2-peak
is just under .02V. What do others get?

The volume has no effect on the signal. Touching the case,
vol control, and other things have no effect.

I have tried moving the xfmer with jumpers 12 inchs from the
chip with the same hum.

I checks the +/- power rails and is's a small sawtooth at 60 cycles.
Any ideas.....

An externally hosted image should be here but it was not working when we last tested it.

Hi,

It seems like it must be mains induced, due to the frequency, and I wonder if you have some unwanted 'stray' coupling from one of the power rails to the output, possibly caused by some capacitive or inductive coupling here.

If it was an artifact which was being 'amplified' by the chip, I would have thought that this would have changed in level with the vol control which you say it doesn't.

Therefore, an unwanted 'signal' to the chip, itself, or to the output, from the PS seems more likely.

There is very little smoothing in these GCs and when you say "small sawtooth" on the power rails, how big is "small" ?

I am not ultra-familiar with the circuit as I haven't yet built one, nor all of the variations which I have seen from time to time, but if you are using the resistor from output down towards the neg supply (the blue one as in post #485) I would look at this resistor (which is a direct DC connection between the [not very smooth] power supply and the output) to begin with.

I would certainly try without this resistor to see whether it affects the issue, asuming you are using one here, as it might give you a clue about where this problem stems from.

Good luck. :)

Edit. As an afterthought, it seems that the circuit relies mainly on the very high PS rejection of the chip to keep things quiet, and maybe you have a 'rogue' chip, which I have come across before, which may amplify OK, but has reduced PSRR. If this affects two chips in the same way, this is rather (very!) unlikely, though.
 
Peter Daniel said:
I imagine that depending on pot's resistance, the circuit will be more or less prone to oscillatintg. I used both 100k and 50k at the input, and 50k version sounded better, but other factors had influence as well. Also it was my impression that 100k version had more tendency to oscillate.
I meant that I was going to use a relay-based attenuator *instead* of a pot. But I guess all I need to do is minimize the attenuator's impedance, right?

--Jordan
 
grimberg said:
I am curious as to how you attach the front and back plates to the two extrusion sections that make the body. Would you be willing to share that secret?

:)


As you see, there are 5/8" bars attached to front and rear panels. The size and placing of the bars is such that the square tubing fits tightly over them. There is actually no need for extra fastners, but I'm placing 4 screws at the bottom to make sure it stays together.

Jam,

What do you think it's worth?;)
 
I didn't want to give a customer a reason for a doubt if I used BG or not.;)

BTW, my acrylic GC has the caps skinned, but the first one also didn't and it sounded better. The bit to drill the holes is 7/8" and when I skin the cap it is loose, so I have to put the tape around it anyway. Since the first amp sounds so good, I decided to do this one exactly the same way. The only difference is a resistor in the feedback loop. First one had Holco, this one is using Rikken, I can't get Holcos anymore.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.