Ok. 
In your list of materials you mentioned "a TO220 heatsink", and if you ask a sales guy for that, he will go to a drawer marked "TO220 heatsinks" and give you one ... like the first one I posted.
Which is usually enough for a TO220 transistor, sometimes even for a voltage regulator .... but the LM1875 , although *housed* in a TO220 case, is a real 20W RMS amplifier and with the "regular" one, it will live a hot but short life
It needs something better, like the 2nd example.
As of the other parts, you mentioned "cut and copy " apparently from different schematics, I suggested picking one and sticking to it.
In fact, and specially if you are beginning with this (fine ), it's useful and safe to follow the "learn to walk before you try to run", meaning try to follow the most classic , tried and true path, before trying "improvements" of any sort.
Boring? Maybe so for he who already built a couple dozen projects ... I bet you'll find it quite exciting.
Building a kit is good: it saves many parts sourcing problems, normally uses a classic schematic, many offer PCBs and a few PSU, up to the transformer (or suggest a very specific one).
Some even offer heatsinks !!!
I'd start with something standard, "by the book", and enjoy it.
And later, the Sky is the limit
As a reference, google the LM1875 datasheet, they offer examples and even suggest a PCB.
In your list of materials you mentioned "a TO220 heatsink", and if you ask a sales guy for that, he will go to a drawer marked "TO220 heatsinks" and give you one ... like the first one I posted.
Which is usually enough for a TO220 transistor, sometimes even for a voltage regulator .... but the LM1875 , although *housed* in a TO220 case, is a real 20W RMS amplifier and with the "regular" one, it will live a hot but short life
It needs something better, like the 2nd example.
As of the other parts, you mentioned "cut and copy " apparently from different schematics, I suggested picking one and sticking to it.
In fact, and specially if you are beginning with this (fine
), it's useful and safe to follow the "learn to walk before you try to run", meaning try to follow the most classic , tried and true path, before trying "improvements" of any sort.Boring? Maybe so for he who already built a couple dozen projects ... I bet you'll find it quite exciting.
Building a kit is good: it saves many parts sourcing problems, normally uses a classic schematic, many offer PCBs and a few PSU, up to the transformer (or suggest a very specific one).
Some even offer heatsinks !!!
I'd start with something standard, "by the book", and enjoy it.
And later, the Sky is the limit
As a reference, google the LM1875 datasheet, they offer examples and even suggest a PCB.
Thank you Fahey
I thought the above list was for the same project. First part the basic components and optional components in the last couple posts. After reading your last post I have decided to start with the basic components Daniel listed and forget about the options.
I have built and modified a few D class amps and wanted to move on to a project that would help me learn more about how these things actually work. I am rambling on here
Again- thanks for your help. I'm going to get the order into DigiKey and get going on it
I thought the above list was for the same project. First part the basic components and optional components in the last couple posts. After reading your last post I have decided to start with the basic components Daniel listed and forget about the options.
I have built and modified a few D class amps and wanted to move on to a project that would help me learn more about how these things actually work. I am rambling on here
Again- thanks for your help. I'm going to get the order into DigiKey and get going on it
That's what the post1 design is for. Put together and play. A lot of component variances have been controlled in that design so that the typical "fine tuning" effort are reduced from weeks to about 10 minutes.
Much smaller component count is possible but might be regrettable by taking a lot more time to get good sound.
If someone wants a design that's ready to go the moment it is built, that is at Post#1. I know its funny looking. But did you try it?
I have many other designs, but that one is arranged for very short timeframe from start to finish. It bypasses much of the fine tuning by not putting 34X gain on an RC. Most of the caveats of that omission are patched by adding output caps to restore some protection, some dynamics and some headroom. Those output caps go at the speaker output jack so the amp is driving a bit of cable first. Its really easy to put them on and they do still cost less than speakers. Again, protection utility is only a small portion of their benefits.
The question wasn't dumb. I did answer. See schematic at Post#1.Were my questions above to dumb to warrant a response from anyone
If you want the solutions "all together in a single page", take a hard long look at this:
ELECTRONICA / components | projects | applications
Good circuit implementation, "by the book", layout, PCB, finished product pictures, transformer suggestions, even an auxiliary (optional) +/-15V PSU to feed your preamp, if so needed.
The finished PCBs looks like this:
EDIT: If you don't want to mess with making your own PCB, this one looks very good:
seems to be available at:
http://diyhifishop.com/lm1875-tda2030-amplifier-amp-board-pcb-p-18.html
It's for two LM1875/TDA2030 (I guess TDA2050 should fit there too) and includes the PSU PCB.
All for a claimed $9.90 which sounds reasonable.
ELECTRONICA / components | projects | applications
Good circuit implementation, "by the book", layout, PCB, finished product pictures, transformer suggestions, even an auxiliary (optional) +/-15V PSU to feed your preamp, if so needed.
The finished PCBs looks like this:
An externally hosted image should be here but it was not working when we last tested it.
EDIT: If you don't want to mess with making your own PCB, this one looks very good:
An externally hosted image should be here but it was not working when we last tested it.
seems to be available at:
http://diyhifishop.com/lm1875-tda2030-amplifier-amp-board-pcb-p-18.html
It's for two LM1875/TDA2030 (I guess TDA2050 should fit there too) and includes the PSU PCB.
All for a claimed $9.90 which sounds reasonable.
Last edited:
Thanks for the information!
Two-chip boards can be transformed into Parallel LM1875 amplifiers which are more sturdy and more elegant. Also, one could choose to upgrade to an offboard power supply.
P.S.
For parallel amplifier on a stereo board, it is possible to jumper the NFB (feedback shunt) caps together at the point where feedback-shunt resistor meets capacitor, so that the capacitor differences (which have gain on them) don't cause excess heat. What I'm saying results in only 1 feedback-shunt cap (it is two caps parallel) so that results are same for both chips. Likewise the input caps can be jumpered together. Parallel amplifiers share all of the caps, none of the resistors. That's easy to remember. Lastly, we'd just add ballast of 0.33R (or similar value) 5w to each + speaker output as is customary for parallel amplifiers.
Two-chip boards can be transformed into Parallel LM1875 amplifiers which are more sturdy and more elegant. Also, one could choose to upgrade to an offboard power supply.
P.S.
For parallel amplifier on a stereo board, it is possible to jumper the NFB (feedback shunt) caps together at the point where feedback-shunt resistor meets capacitor, so that the capacitor differences (which have gain on them) don't cause excess heat. What I'm saying results in only 1 feedback-shunt cap (it is two caps parallel) so that results are same for both chips. Likewise the input caps can be jumpered together. Parallel amplifiers share all of the caps, none of the resistors. That's easy to remember.
Lastly, we'd just add ballast of 0.33R (or similar value) 5w to each + speaker output as is customary for parallel amplifiers.
Last edited:
Thanks Daniel and Fahey. I really appreciate your help and your suggestions I've been reading so much and looking at so many options it's starting to make my head spin. I was dreaming about caps last knight and that can't be good
The one at DiyHiFi looks interesting. I just read the thread that Daniel mentioned somewhere along the way about dual parallel 3886 boards. I have also been considering ChipAmp's 1875 which is quite inexpensive. I wish he provided more info on his site. Click on manual or photos and there is nothing there O well
I've been reading so much and looking at so many options it's starting to make my head spin. I was dreaming about caps last knight and that can't be goodThe one at DiyHiFi looks interesting. I just read the thread that Daniel mentioned somewhere along the way about dual parallel 3886 boards. I have also been considering ChipAmp's 1875 which is quite inexpensive. I wish he provided more info on his site. Click on manual or photos and there is nothing there
O well
One thing. When comparing LM3886, with LM1875, you'll see LM3886's with onboard power supply and large caps right at the chip to muffle their shout, but if you do exactly the same circuit with LM1875, that would be as dull as any cheap ebay amp kit. The LM1875 favors some 220u~470u caps at the amplifier board, and an offboard power supply.
If it was going to be as easy as using an onboard power supply I would have done it--pair of big caps and a KBU6D would be so easy, and sound dreadful. Hi-fi is not that easy.
If it was going to be as easy as using an onboard power supply I would have done it--pair of big caps and a KBU6D would be so easy, and sound dreadful. Hi-fi is not that easy.
Last edited:
I actually disagree with this. I've built two 1875s just in the last few weeks. One had 470u caps at chip, the other 2200u. Same off board psu. There wasn't much in it but the 2200u had the edge at volume: purer, more lifelike. They were both going into nominally 4 ohm, which may have made the difference.
If I had some 8 ohm speakers in my workshop I would be interested to hear any difference there.
I don't find a reason why having higher filter capacitance (= less ripple - stiffer voltage - softer current delivery - higher power reserve - etc. etc. etc.), even better stsying close by, connected with thick tracks ..... might "muffle" ??????????? the sound ???
Now the amplifier will be *dull* ????
No logic to it.
It's quite the other way around.
PS: there's a reason for typical 0.1uF decoupling caps being applied "as close as possible to power pins, ideally on the same pads".
Same here.
??????????? the sound ???Now the amplifier will be *dull* ????
No logic to it.
It's quite the other way around.
PS: there's a reason for typical 0.1uF decoupling caps being applied "as close as possible to power pins, ideally on the same pads".
Same here.
Last edited:
My stereo LM1875 board has two 10,000uF supply caps ( shared by both channels) about 2.5 inches away from the chips and 100uF caps about 0.5 to 1 inch away from the chips on each of their supply rails. Supply is +/- 21V unloaded.
I drove a 4 ohm speaker ( Behringer 2031P) with this amp and it goes very loud and sounds great. Great HF AND LF with unbelievable extension for such a simple set up ! But yes, I think the circuit does benefit from better quality parts. My first board had 10,000uF caps made by Lelon ( who's that ? ) ! The second board used 10,000uF from Samsung. Very clearly the LF extension is noticeable better in the Samsung cap version.
I'm trying to make a version using all film caps ( input and NFB ) except the supply caps. Wonder if it will sound good. I'm expecting it because of pp film caps in the signal path !
I drove a 4 ohm speaker ( Behringer 2031P) with this amp and it goes very loud and sounds great. Great HF AND LF with unbelievable extension for such a simple set up ! But yes, I think the circuit does benefit from better quality parts. My first board had 10,000uF caps made by Lelon ( who's that ? ) ! The second board used 10,000uF from Samsung. Very clearly the LF extension is noticeable better in the Samsung cap version.
I'm trying to make a version using all film caps ( input and NFB ) except the supply caps. Wonder if it will sound good. I'm expecting it because of pp film caps in the signal path !
Big caps somewhat close, but the little caps doing the high speed work. Excellent.
Thanks! What gauge is the 2-1/2" interconnect between the 10,000uF to 100uF?
Last edited:
FWIW, each 125 mil wide track , of 2.5" long standard 1 ounce (35uM thick), will have a resistance of: 0.00964 ohms, rounded to: 0.01 ohms.
Not bad.
From: The CircuitCalculator.com Blog PCB Trace Resistance Calculator
Not bad.
From: The CircuitCalculator.com Blog PCB Trace Resistance Calculator
I think this is one of those super value chips that hasn't won any awards on this forum and maybe elsewhere. That said , it's better to just use it and not say too much so that it's price doesn't shoot up !
For anyone with a tight budget this is probably their last ( best!) stop. Their next jump will be VERY expensive . There are better amps but we aren't talking about chalk and cheese here ! Of course 20 watts a channel isn't huge but it does go quite loud , enough for most people. Make it an active system with a powered sub and it really takes off !
The LM1875 and TDA2050 sound similar but there are some small differences. Noticeable on some musical tracks. I prefer the 1875 ( in my implementation ) but several people prefer the 2050. Both are acceptable for a low cost ( not really inexpensive !) decent amp.
It's finally the music that matters ! Enjoy !
For anyone with a tight budget this is probably their last ( best!) stop. Their next jump will be VERY expensive . There are better amps but we aren't talking about chalk and cheese here ! Of course 20 watts a channel isn't huge but it does go quite loud , enough for most people. Make it an active system with a powered sub and it really takes off !
The LM1875 and TDA2050 sound similar but there are some small differences. Noticeable on some musical tracks. I prefer the 1875 ( in my implementation ) but several people prefer the 2050. Both are acceptable for a low cost ( not really inexpensive !) decent amp.
It's finally the music that matters ! Enjoy !
My post got truncated. I've been looking for the board layout and haven't found it yet. It's a few years old.
I changed some parts values.
Cin = 1uF polyester film. Rin = 47K. NFB R = 47 K. -ve input to gnd is 1.8K and decoupling cap for the nfb is 47uF Samwah. Two supply decoupling 100uF Samwah caps and 10,000uF Samsung supply caps. BR1010 bridge rectifier.
I changed some parts values.
Cin = 1uF polyester film. Rin = 47K. NFB R = 47 K. -ve input to gnd is 1.8K and decoupling cap for the nfb is 47uF Samwah. Two supply decoupling 100uF Samwah caps and 10,000uF Samsung supply caps. BR1010 bridge rectifier.
Looks great!
Power circuit decoupling caps with 100u on a stereo build do achieve the minimum 200u per rail needed. The power circuit format of BR, 10,000u, series element, 100u, LM1875, does look like it should be quite successful and very clear. The feedback resistor of 47k is "in the ballpark" although for feedback-shunt the 1.8k+47u is a "warm" setting cleaned up by smallish size 1u input cap which would need to be polyester and looks like a good idea. No problem except for making the caps somewhat more difficult to select (brands, esr pattern, etc). Seems like there would be really good efficiency/headroom with this setup. It looks like it should play very well, highly dynamic and fairly powerful.
At a gain of 27, some MP3 players wouldn't be impressed, but it is enough if one has an M-Audio Revo PC sound card, or stronger source.
Question, does the power circuit look like:
100u, L, 10,000u, L, 100u? or
100u||100u, L, 10,000u?
P.S.
Suggestion for easier cap selection:
That 47u? Try 47u||47u identical models in parallel. The bass would be slightly less warm (94u+1.8k is still a warm setting), but lower notes clearer which could benefit some speakers. It could be a good and quick little experiment if you have more of the same model cap available.
Power circuit decoupling caps with 100u on a stereo build do achieve the minimum 200u per rail needed. The power circuit format of BR, 10,000u, series element, 100u, LM1875, does look like it should be quite successful and very clear. The feedback resistor of 47k is "in the ballpark" although for feedback-shunt the 1.8k+47u is a "warm" setting cleaned up by smallish size 1u input cap which would need to be polyester and looks like a good idea. No problem except for making the caps somewhat more difficult to select (brands, esr pattern, etc). Seems like there would be really good efficiency/headroom with this setup. It looks like it should play very well, highly dynamic and fairly powerful.
At a gain of 27, some MP3 players wouldn't be impressed, but it is enough if one has an M-Audio Revo PC sound card, or stronger source.
Question, does the power circuit look like:
100u, L, 10,000u, L, 100u? or
100u||100u, L, 10,000u?
P.S.
Suggestion for easier cap selection:
That 47u? Try 47u||47u identical models in parallel. The bass would be slightly less warm (94u+1.8k is still a warm setting), but lower notes clearer which could benefit some speakers. It could be a good and quick little experiment if you have more of the same model cap available.