First LM3875 impressions

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
There is a guy in the swap meet that is selling those 4700uf Rubycon 63v USR caps.

He sells a box of 20 for around $20 or $25 plus shipping. Pretty good deal.

They are one inch in diameter, not sure what your board will take. Also note they are the "Snap in" kind so you might have to drill the board etc, or make a new setup.

No affiliation http://www.diyaudio.com/forums/swap-meet/184977-rubycon-4700uf-x-63v-rohs-1-25-each.html

That is great, thanks!
I will check into those caps.
 

6L6

Moderator
Joined 2010
Paid Member
so I had to replace them with six 2,200 uF capacitors that I had laying around. The LM kits had an addition 2 x 1,000 uF caps right on board. So I have a total of about 8,800 uF per amp.

Question is, do I tear things up and kick those caps up to 4,700-6,600 uF for a noticeable gain in bass?.

Looking at your amp you have significantly more than you think... Each amp board has four 2200uf caps, for a total of 8, and the six 2200uf on the PSU board add up to 30,800uf total, or 15,400uf per rail.

You are squarely in the realm of diminishing returns if you want to add more. This isn't like an Audiosector build where the stock amount is 1/5 of what you already have...
 
Looking at your amp you have significantly more than you think... Each amp board has four 2200uf caps, for a total of 8, and the six 2200uf on the PSU board add up to 30,800uf total, or 15,400uf per rail.

You are squarely in the realm of diminishing returns if you want to add more. This isn't like an Audiosector build where the stock amount is 1/5 of what you already have...

Thanks 6L6, sometimes you need an extra pair of eyes. I had missed that fact (8 x 2,200 uF on the LM3875 boards) though it has been in front of me. Think you may be right, I am approaching the diminishing returns line if not on it already. The amp is not bass shy.
 
That's hard to tell, I can't say that I have.

Whaleman; Discrete amps are OK, if your going to a project with more parts and effort, tubes are the way go.

You have 7,700uf per rail right now, not 15400uf when you divide by two amplifiers. You should an improvement with a little more.
 
That's hard to tell, I can't say that I have.

Whaleman; Discrete amps are OK, if your going to a project with more parts and effort, tubes are the way go.

You have 7,700uf per rail right now, not 15400uf when you divide by two amplifiers. You should an improvement with a little more.

Yes, thanks. I emailed the guy with the Rubycons for sale, I have not heard back from him.

I just built an Audio Note Kits tube amp, it mainly sits on the shelf. Sounds great but my hurried current lifestyle does not allow a lot of warm up tube time. Up next is a discrete amp, but not sure what is out there and most stuff is out of China which at the moment, is a bit dicey with all the fakes.
 
Actually he is on jury duty, you might have to find some different ones if he does not get back to you.

I could not wait any longer, the Nichicon FW 10,000 uF x6 caps arrived today from Mouser and went right in. I think the overall sound is much improved, not just necessarily the bass. I think the mids and treble benefited the most.

So now I have 77,600 uF total capacitance and the Gainclone sounds terrific.
 

Attachments

  • DSC_0587.jpg
    DSC_0587.jpg
    270.9 KB · Views: 324
  • DSC_0571.jpg
    DSC_0571.jpg
    291.3 KB · Views: 322
  • DSC_0582.jpg
    DSC_0582.jpg
    369.2 KB · Views: 311
Thanks folks for the pos. comments, I hope to let it do its thing later on when no one is around.

Those adhesive stand offs came from mouser, I will post a part no.

I ended up getting a box of those rubycon caps and I've yet to test them all but the ones I have tested (about 4) show about 4100uf - tested with a fluke and an atlas ESR70 showing consistent between the two - .04ohm ESR. I've yet to use them in a project.

Is that low for a 6,800 uF cap?
 
They are 4700uf caps +/- 20%. So anything between 3760uf and 5640uf would be within spec. (From the swap meet here, not yours.)

Mine also measures around 4100uf, within ~12% of spec.
When I measured them in a "rail" consisting of 5 x 4700uf (23,500uf), they did a little better measuring within around 6%.
 
Since there was a talk about GC grounding I feel a need to ask all of you about my specific problem. I have built a lot of gainclones with different ICs (LM3875, LM1875, TDAQ2050, THD2030 and OPA549). With all of them I had a problem with a ground layout. Everyone are suggesting that signal and power grounds should be wired separately. I have tried different wirings from this forum but this just didn't work for me. I allways got a hum and a big DC offset (90 to 120 mV without a cap Ci on resistor to ground). The only way I had no hum and a reasonable DC offset was to wire all the grounds separately to ground plane like on attached photo. It's not the best resolution but in the middle there is a cooper foil as grounding plane and everything that should go to ground is wired directly to that piece of cooper.
As I understand this is not the best solution so I would like to ask if anyone of you might have an idea why only that kind of groundung works for me. Recently I tried to rewire this GC on the photo as mrVetz did but I got hum and about 120mV of DC offset. I was moving and changign wires for three days then I gave up and just dissasembled the damn thing.
I would appreciate any oppinion about this because I have no idea what seems to be the problem.
Regards
 

Attachments

  • LM1875 ground.jpg
    LM1875 ground.jpg
    88.9 KB · Views: 230
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.