Hi kaloyan,
I am very happy to read your posting. You tested in practice what we fantasized about in the thread. Neither did I use genuine LM1875, nor do I have distortion measurement gear of that quality. -125dB should be 0.00006%! That is in the order of what Neurochrome mentions. My 3x2 “fake” LM1875 managed to play very well for a good hour when I had a fatal avalanche-failure phenomena.
Chris and I have been discussing a PCB for a 2-parallel LM1875 (+ composite OP-AMP loop) stereo amplifier that leaves some 45W in 8 Ohm, 65W in 6 Ohm and almost 90W in 4Ohm. That should do at home for most music. I must admit that the 3x2 LM1875 version (six per channel) became rather bulky and even the idle current level was a bit annoying.
The increase in CM distortion with low gains, is that not countered by using the LM1875 in inverting configuration?
I am very happy to read your posting. You tested in practice what we fantasized about in the thread. Neither did I use genuine LM1875, nor do I have distortion measurement gear of that quality. -125dB should be 0.00006%! That is in the order of what Neurochrome mentions. My 3x2 “fake” LM1875 managed to play very well for a good hour when I had a fatal avalanche-failure phenomena.
Chris and I have been discussing a PCB for a 2-parallel LM1875 (+ composite OP-AMP loop) stereo amplifier that leaves some 45W in 8 Ohm, 65W in 6 Ohm and almost 90W in 4Ohm. That should do at home for most music. I must admit that the 3x2 LM1875 version (six per channel) became rather bulky and even the idle current level was a bit annoying.
The increase in CM distortion with low gains, is that not countered by using the LM1875 in inverting configuration?
Last edited:
Hi Joe
yes you are right. one TDA7293 cost the same as 1875. yes its cost 3 euro at reichelt now.!!
but this is still not composite amp! the thermal split with 2 chips might be better.
the trigger was to have a modular design to parallel and bridge the amps like we think/what..but i learnt that this is not easy to do with THT components, offset trimmer...etc.
chris
yes you are right. one TDA7293 cost the same as 1875. yes its cost 3 euro at reichelt now.!!
but this is still not composite amp! the thermal split with 2 chips might be better.
the trigger was to have a modular design to parallel and bridge the amps like we think/what..but i learnt that this is not easy to do with THT components, offset trimmer...etc.
chris
TDA7293 is a good choice as well. It is for higher power levels than the LM1875 and should be more simple to connect in parallel. I have not tried parallel-coupling yet. The TDA7293 I tested for use in composite amplifiers (mainly postings #37 and #46): Chip-amps suited as power stage in a composite amplifier, LM1875/TDA2050 excluded.
We know that the LM1875 has a particularly nice sound and sufficient power for those who are not into heavy metal music and rave parties
. With LM1875 we can make small amplifier building blocks that can be put in parallel or BTL-coupled according to individual needs. This way members can tailor their own amplifier.
We know that the LM1875 has a particularly nice sound and sufficient power for those who are not into heavy metal music and rave parties
. With LM1875 we can make small amplifier building blocks that can be put in parallel or BTL-coupled according to individual needs. This way members can tailor their own amplifier.I asked kaloyan for the schematic which will be usefull for my bookshelf speakers. He said that he will open a seperate topic for it.
But then i read his post again;
THD and noise bellow -125dB (the limit of my measuring gadgets) 0.1 to 40W - 2 parallel LM1875 amps
Even AP555B can not able to meause that low SINAD. So i guess there is a big miss-understanding here. Or he is just "trolling".
But then i read his post again;
THD and noise bellow -125dB (the limit of my measuring gadgets) 0.1 to 40W - 2 parallel LM1875 amps
Even AP555B can not able to meause that low SINAD. So i guess there is a big miss-understanding here. Or he is just "trolling".
Last edited:
Thanks for the link but I don´t see a schematic of the whole composite you used with the TDA7293.TDA7293 is a good choice as well. It is for higher power levels than the LM1875 and should be more simple to connect in parallel. I have not tried parallel-coupling yet. The TDA7293 I tested for use in composite amplifiers (mainly postings #37 and #46): Chip-amps suited as power stage in a composite amplifier, LM1875/TDA2050 excluded.
We know that the LM1875 has a particularly nice sound and sufficient power for those who are not into heavy metal music and rave parties
Yes, I can understand the point of the LM1875 if you really like the sound.
But the effort to arrive at a stable composite design and then bridge+parallel is quite an effort.
With the TDA one chip would be sufficient. And if you don´t like the sound you can always put a H2-generator in front ;-)
That link is only concerned with the TDA7293 as buffer amplifier. The OP-AMP coupling was discussed and disclosed in the other thread using LM1875 as buffer amplifier in a composite amplifier.
Bridge+parallel is not an enormous effort because it is just the same block being repeated. It is (much) more components and some adjustment to make the blocks match.
H2 generator?
Bridge+parallel is not an enormous effort because it is just the same block being repeated. It is (much) more components and some adjustment to make the blocks match.
H2 generator?
Ah, OK. So that was a pre-evaluation of the chip without composite circuitry.
I think I saw that 6xLM1875 schematic.It is probably simpler to do than one might think, especially when it´s in blocks. Like a 3xparallel-PCB and then via bridge-adapter connected with another module.
Still a single TDA7293 in composite with another opamp can be a really small thing.
Talking about composite, bridge, parallel and small: have you seen this design in post #31?
Bridged LM3886 + LME49720 + remote FB
That´s compact!
The H2-generator is from Nelson Pass. Probably a nice add-on if your amp doesn´t have a "nice harmonics" profile. You can add some H2 yourself!:
FIRST WATT
I think I saw that 6xLM1875 schematic.It is probably simpler to do than one might think, especially when it´s in blocks. Like a 3xparallel-PCB and then via bridge-adapter connected with another module.
Still a single TDA7293 in composite with another opamp can be a really small thing.
Talking about composite, bridge, parallel and small: have you seen this design in post #31?
Bridged LM3886 + LME49720 + remote FB
That´s compact!
The H2-generator is from Nelson Pass. Probably a nice add-on if your amp doesn´t have a "nice harmonics" profile. You can add some H2 yourself!:
FIRST WATT
Last edited:
Triggered by the conservation I did some googleing (despite having enough DIY projects on the table ;-)) and found this very neat layout for the LM1875.
Kicad-files provided!
Recommendations for paralleling are also given.
At JLCPCB it probably costs not even 5$ to get 10 boards made.
GitHub - nradulovic/hw_minima: LM1875 based audio power amplifier
Kicad-files provided!
Recommendations for paralleling are also given.
At JLCPCB it probably costs not even 5$ to get 10 boards made.
GitHub - nradulovic/hw_minima: LM1875 based audio power amplifier
Many thanks for the hints. I have seen more composite designs based on the LM3886 and some based on TDA7293. I am sure they are all good. The idea with this thread was to try to explain to members how simple it can be to make pretty good amplifier constructions. So they can go from implementing designs of others or modifying "China-boards" to designing own constructions.
At least the LM1875 has the advantage that it only has 5 pins and accordingly is easy to use in a Vero-board.
$5 for 10 PCBs is much less than I thought. Would you happen to know the order of sending costs? Sometimes the Chineese "revenge" is in the sending costs.
I was educated just before PCB-CAD became accessible in price. I have made many with manual track layout and Ferro-Chloride etching. Later I was the engineer in the group that had to take care of plannings, coordination and other boring activities while somebody else did the PCB-layout. I never really learned PCB-layout on PCs. But if needed, I will learn,
At least the LM1875 has the advantage that it only has 5 pins and accordingly is easy to use in a Vero-board.
$5 for 10 PCBs is much less than I thought. Would you happen to know the order of sending costs? Sometimes the Chineese "revenge" is in the sending costs.
I was educated just before PCB-CAD became accessible in price. I have made many with manual track layout and Ferro-Chloride etching. Later I was the engineer in the group that had to take care of plannings, coordination and other boring activities while somebody else did the PCB-layout. I never really learned PCB-layout on PCs. But if needed, I will learn,
Can you give me a link or hint where to find one?
You can send a PM also if that´s too offtopic.
I´d really like to try a composite TDA7293 although the xmax-amp from X is really tempting. It´s quite a big PCB but very well done and apparently low noise as it seems. It´s just more exciting to make your own amp.
That´s nice although I think a composite amp is always lots of trial and error as the simulations never tell the whole story.
Just go to the JLCPCB website and quote for 10PCBs (or 5 etc.) and put in your destination afterwards. You don´t have to input the gerber files to get the quote.
10PCBs, 80x80mm, 2-layer to france is 5$ + 7.46$ shipping.
If you order just 5PCBs it´s 2$ + shipping!
I already used them for making a "Baby Aksa" amplifier and the quality is very very nice!
Even 4-layer boards aren´t much more expensive if any.
You could alternatively start with a Kicad-project (open source) and modify it to your needs.
We could also do a project together if there is some interest. Slowly getting back into PCB-design and Kicad is quite a nice SW.
That said I have a couple of amps on the workbench.
Last edited:
Hi joensd,
Many thanks for your quick reply. Sorry for my slower response.
I looked through my schematic downloads and had another look on the Internet. I found some very basic diagrams with composite regulation and TDA7293 on homecinema-fr.com but nothing that was ready for implementation. Sorry, I remembered wrong. It is time to make one such TDA7293 design. TDA7293 has nice regular clipping.
At present I am busy with some other activities but eventually afterwards.
Two TDA7293 in parallel for some 200W in 4 Ohm, is that sufficient?
The PCB costs are much less than I expected. I will have to learn KiCAD one day.
Making composite amplifiers need not to be very difficult as long as you do not expect to get the very last dB’s in distortion. I learned empiric stability trimming of regulation loops by some very skilled technicians many years ago when simulation tools were expensive. You do most likely not reach the very optimum but it is pretty fast and it works.
What about if I make a first draft of a schematics - > we discuss it and improve the schematics - > when we agree, I make the test-version (mono) on Veroboard and test that it works - > when I am happy with it, I put it in a box and send it to you in Germany - > you test it - > we make the last adjustments to the schematics - > we start making a PCB?
If we prepare the PCB for two TDA7293 in parallel, it is easy to leave one TDA7293 out (per channel) if it is only used with 8 Ohm speakers. Honestly, I do not need 200W in 4 Ohm.
What-do-you-say?
Many thanks for your quick reply. Sorry for my slower response.
I looked through my schematic downloads and had another look on the Internet. I found some very basic diagrams with composite regulation and TDA7293 on homecinema-fr.com but nothing that was ready for implementation. Sorry, I remembered wrong. It is time to make one such TDA7293 design. TDA7293 has nice regular clipping.
At present I am busy with some other activities but eventually afterwards.
Two TDA7293 in parallel for some 200W in 4 Ohm, is that sufficient?
The PCB costs are much less than I expected. I will have to learn KiCAD one day.
Making composite amplifiers need not to be very difficult as long as you do not expect to get the very last dB’s in distortion. I learned empiric stability trimming of regulation loops by some very skilled technicians many years ago when simulation tools were expensive. You do most likely not reach the very optimum but it is pretty fast and it works.
What about if I make a first draft of a schematics - > we discuss it and improve the schematics - > when we agree, I make the test-version (mono) on Veroboard and test that it works - > when I am happy with it, I put it in a box and send it to you in Germany - > you test it - > we make the last adjustments to the schematics - > we start making a PCB?
If we prepare the PCB for two TDA7293 in parallel, it is easy to leave one TDA7293 out (per channel) if it is only used with 8 Ohm speakers. Honestly, I do not need 200W in 4 Ohm.
What-do-you-say?
Hi FF,
don´t worry, we are not on a schedule here and don´t want to be, right?
As super-easy and eventually practical it would be to parallel, I really don´t need that kind of power and am absolutely fine with 100W peaks.
Compact would be my priority and thus a single TDA7293 with its controlling opamp. But I´m very open to other suggestions.
I personally would rather bridge two amps with +-24V rails or so for my 8Ohm woofers.
Great to hear you already got some experience. Getting it right on veroboard sounds tough but I guess if it finally works, it also will on a PCB.
If we could come up with a schematic that includes all eventual needs (like additional RC-poles for stability or similar), we could also directly try a PCB if that´s easier but I´d go with your suggestion in any case.
If you want to try the open-source-LM3886 design in the process for testing, I have two populated PCBs left over that will end up in a stereo amp at a later point. Also, if you want to try a discrete design, I have a couple of Baksa-PCB left over. (couple of SMD parts though)
I suppose we should start a new thread or continue via PM for the time being.
Maybe we should also go the "open-source" route. I´m sure others with more experience will chime in.
Best regards
Jens
don´t worry, we are not on a schedule here and don´t want to be, right?
As super-easy and eventually practical it would be to parallel, I really don´t need that kind of power and am absolutely fine with 100W peaks.
Compact would be my priority and thus a single TDA7293 with its controlling opamp. But I´m very open to other suggestions.
I personally would rather bridge two amps with +-24V rails or so for my 8Ohm woofers.
Great to hear you already got some experience. Getting it right on veroboard sounds tough but I guess if it finally works, it also will on a PCB.
If we could come up with a schematic that includes all eventual needs (like additional RC-poles for stability or similar), we could also directly try a PCB if that´s easier but I´d go with your suggestion in any case.
If you want to try the open-source-LM3886 design in the process for testing, I have two populated PCBs left over that will end up in a stereo amp at a later point. Also, if you want to try a discrete design, I have a couple of Baksa-PCB left over. (couple of SMD parts though)
I suppose we should start a new thread or continue via PM for the time being.
Maybe we should also go the "open-source" route. I´m sure others with more experience will chime in.
Best regards
Jens