New ChipAmp Project Thread

[slipperybidness]

Some updates.

*I finished 1st channel of the LM3886
*Hooked it up to my O-scope, signal generator, and benchtop PS for measurements, and to try to snoop out any fatal flaws
*Well, I was getting some severe clipping on the negative portion of the sine wave output, starting at ~80Hz and continuing on down in freq.
*However, this seems to be that my bench PS is not robust enough. I noted that as I start clipping, my bench supply droops from the set-point of 25V down to about 20V. I did not see the same droop on the +V supply, and did not see the same clipping on the positive portion of the output waveform. D'oh!
*I'm just about done with my DIY PS. Just need to wire up the power transformer and measure the PS.

So, I hit what is hopefully a small hurdle. All the evidence right now suggests that my bench PS just runs out of gas. I'll have to get my DIY supply done and tested, then come back and repeat this work.

 

[slipperybidness]

Quick update.

I have decided to go with a separate PS for the LM4780 build (at least for now until I need my PS in a bigger badder amp). The PS is done!

PS specs:

*EMI filter on the mains input, IEC connection
*Safety fuses on the mains and on the secondaries.
*625KVA Toroid Transformer (Avel)
*Fast Switching Diodes (On-Semiconductor) that are heat-sinked and snubber cap'd (you don't often see proper snubbing in commercial designs)
*36,000 uF filter caps PER Rail for a total of 72,000uF! (Panasonic)
*Bypass caps on the output (Vishay)
*16AWG rail outputs
*Total cost was about $260, but I also had some stuff in stock already like the power switch and connectors. So, say total build in the neighborhood of $275 just for the PS. While that doesn't "seem cheap", this thing is a beast! It should be able to handle 400W continuous without a hiccup, and likely even more. I would put it up against most any PS and any consumer product. I should be set for robust and reliable power for decades now!

I have 1 channel of the LM4780 parallel amp completed and tested. I tested and measured it on my o-scope as 2 separate channels, then tied the outputs together in parallel config. After that, I tested it on my DT SM350, sounded great! The bass was amazing! I need to go back and scope it to be sure that nothing odd turned up when I did the parallel connection. It is installed in the chassis.

The 2nd channel is complete, but untested. My biggest hold-up right now--I ran out of thermal compound, D'oh. I should have a new 12g tube next week. I could possibly be done with this next weekend!

 

[slipperybidness]

Oh yeah.....the LM3886 channels and PS are completed and tested too. I realized that I'm approaching the point of "more amps than I can possibly use". I know, a tough spot to be in, guess I "need" more speakers.

So, I've kind of put that project on hold. I think I'm going to hold off until I either find a free case for it (case is typically one of the most expensive parts), or I get some inspiration to make some really cool/artistic amp out of it. Maybe I'll rummage around some thrift shops and see if inspiration hits.

 

[slipperybidness]

 

[slipperybidness]

 

[TheWarrior]

Like constructing your first light saber...

 

[MrBoat]

Nice to see this thread. I had bought some other series of amp chips for the Gain Clone project and forgot that I bought them.

Nice work here!

 

[MrBoat]

I did indeed build the gain clone with the LM3886 chips and it sounds pretty darned spectacular to my ears. Yet more evidence that we likely pay too much for audio. I built it and installed it into a box I made at work out of .125 x 4" anodized aluminum and I hid it when the whole family was living here temporarily and now I can't find it without digging deep. I know that in the box with it is my new set of Sandvik diag cutters and other wiring tools, along with my Klein crimpers.

Building, and then listening to these little amp projects sure adds some satisfying fun to this hobby. Hopefully, you are able to use what you have built. I am hoping that class D keeps evolving as it seems to be.

 

[slipperybidness]

I did indeed build the gain clone with the LM3886 chips and it sounds pretty darned spectacular to my ears. Yet more evidence that we likely pay too much for audio. I built it and installed it into a box I made at work out of .125 x 4" anodized aluminum and I hid it when the whole family was living here temporarily and now I can't find it without digging deep. I know that in the box with it is my new set of Sandvik diag cutters and other wiring tools, along with my Klein crimpers.

Building, and then listening to these little amp projects sure adds some satisfying fun to this hobby. Hopefully, you are able to use what you have built. I am hoping that class D keeps evolving as it seems to be.

Take a good look at the neurochrome variants for the LM3886. I have never built nor heard one. But, the guy looks to have some serious engineering chops, and claims to have the best performing chip amp variant.

LM3886 Done Right

This DIY LM3886 amplifier board offers the best performance of any LM3886 chipamp implementation due to the optimized PCB layout. Enjoy the cleanest sound with the LM3886.

neurochrome.com

 

[MrBoat]

Take a good look at the neurochrome variants for the LM3886. I have never built nor heard one. But, the guy looks to have some serious engineering chops, and claims to have the best performing chip amp variant.

LM3886 Done Right

This DIY LM3886 amplifier board offers the best performance of any LM3886 chipamp implementation due to the optimized PCB layout. Enjoy the cleanest sound with the LM3886.

neurochrome.com

Thank you for that information. Have not followed the chipamp scene for awhile and would not have noticed this, most likely.