Don´do that, very inefficient way to run a transformer, you are way better with the current 15+15VAC one, which "should" give you around +/- 20V rails, not far rom what you have today.
Your amp seems to be working fine, and is ruler flat within the 20Hz to 20kHz audio band, so if you hear bass shy and slightly trebly sound, attribute it fully to your speakers ... or personal preferences.
You have an excellent setup, just not designed for chest thumping Bass or rattling windows.
Sounds like a pretty rewarding yet very economical system, I like it.
The speakers will likely benefit from some poly fill or whatever, probably never had any I would guess. Maybe the crossover would benefit from some better capacitors, but that would be more costly and take more trial and error potentially.
The speakers will likely benefit from some poly fill or whatever, probably never had any I would guess. Maybe the crossover would benefit from some better capacitors, but that would be more costly and take more trial and error potentially.
Unfortunately, these "hum breakers" usually degrade performance. I'd consider it a bandaid. You're far better off with 0 Ω between the signal ground and the power ground, but you do have to ensure that you only route small signal currents through the signal ground if you do that.
Tom
I made a thread sort of asking about the 10 ohm signal ground resistor a couple weeks ago, for the same schematic. I was told to include it on both channels, for a stereo build using a single 3.5mm jack as input. And I was also told to include back to back diodes for protection.
But in my head it still doesn't feel right. I'm tempted to wire the input ground sections of both channels to each other first and then back to the main ground with a single wire. That just seems like it wouldn't have any loops and wouldn't need those resistors.
I'm still waiting on parts and haven't been able to play around yet. If I can find any difference I'll probably post about it.
Anyway, to answer the original question, some possibilities for the poor bass could be,
A. It's just the flatness of the frequency response
B. If you used the included chips. They usually look terrible compared to real ones. Who knows what could go wrong.
C. The power supply capacitors can be increased in size. It's a different chip, but I remember the datasheet for the lm3886 saying that the on-board power capacitors should be minimum 470 uf. That's the size I plan on using for my lm1875 project, but just because I have them already and they fit. It probably won't make a difference in this case.
D. Try the classic output resistor + inductor combo and see if you can hear a difference. You shouldn't be able to. If you do it could mean oscillation. But that's probably not it if the amp sounds fine.
If you can't find any problems other than lack of bass, consider building a preamp for your next project.
But in my head it still doesn't feel right. I'm tempted to wire the input ground sections of both channels to each other first and then back to the main ground with a single wire. That just seems like it wouldn't have any loops and wouldn't need those resistors.
I'm still waiting on parts and haven't been able to play around yet. If I can find any difference I'll probably post about it.
Anyway, to answer the original question, some possibilities for the poor bass could be,
A. It's just the flatness of the frequency response
B. If you used the included chips. They usually look terrible compared to real ones. Who knows what could go wrong.
C. The power supply capacitors can be increased in size. It's a different chip, but I remember the datasheet for the lm3886 saying that the on-board power capacitors should be minimum 470 uf. That's the size I plan on using for my lm1875 project, but just because I have them already and they fit. It probably won't make a difference in this case.
D. Try the classic output resistor + inductor combo and see if you can hear a difference. You shouldn't be able to. If you do it could mean oscillation. But that's probably not it if the amp sounds fine.
If you can't find any problems other than lack of bass, consider building a preamp for your next project.
"seemed to lack a little bass"
JUSS10 -
Post #1: "seemed to lack a little bass"
If you used a receiver with your same speakers before you built this diy amp,
and the receiver had tone controls so you could adjust the bass volume relative to the overall volume, that could account for some difference in what you hear.
JUSS10 -
Post #1: "seemed to lack a little bass"
If you used a receiver with your same speakers before you built this diy amp,
and the receiver had tone controls so you could adjust the bass volume relative to the overall volume, that could account for some difference in what you hear.
Hmm, wasn't getting updates on this thread.
well I went ahead and ordered a few more these kits on ebay. They were going for $2 a kit so if all else, I figured I could used the boards. Thats cheaper than I could get the boards made at OSHpark (though I still may design my own board)
Anyway, I also made a digikey order and based the components on the data sheet for the LM1875. Went with film caps and wire film resistors (not knowing what the stock resistors were) Only components I kept from the kit were the power supply caps (100uf) and the LM1875 it came with.
I jumpered the 1K resistor that was in the signal path going to the input on the IC. I left the 330pf cap out and then jumpered the 10ohm resistor so the signal ground and power ground are on the same lined (that appears to be how the schematic is)
haven't tried it yet but plan to do a side my side comparison with this new board and the ones I already had in. I'll replace one and then pan back and forth to hear the difference. Hoping the film caps will make the most difference but we'll see. Still very suspect of the chinese LM1875 chip. I do have of the real ones laying around if this is a bust.
Justin
well I went ahead and ordered a few more these kits on ebay. They were going for $2 a kit so if all else, I figured I could used the boards. Thats cheaper than I could get the boards made at OSHpark (though I still may design my own board)
Anyway, I also made a digikey order and based the components on the data sheet for the LM1875. Went with film caps and wire film resistors (not knowing what the stock resistors were) Only components I kept from the kit were the power supply caps (100uf) and the LM1875 it came with.
I jumpered the 1K resistor that was in the signal path going to the input on the IC. I left the 330pf cap out and then jumpered the 10ohm resistor so the signal ground and power ground are on the same lined (that appears to be how the schematic is)
haven't tried it yet but plan to do a side my side comparison with this new board and the ones I already had in. I'll replace one and then pan back and forth to hear the difference. Hoping the film caps will make the most difference but we'll see. Still very suspect of the chinese LM1875 chip. I do have of the real ones laying around if this is a bust.
Justin
Change the 100u per each rail to parallel 220u (220u//220u aka 440u with two caps each rail).
That should resolve the tone problem.
If not, then try 470u//470u per each rail, at most.
Otherwise, resolve layout foobar issue.
There's a fair list of other ways to do it, but none either as straightforth nor as entertaining.
Stability though, is still worth a mention, because if it runs more stable then the tone is more neutral.
Both, yes, I vote for both a goodly power circuit with a stable amp. That's because it is so much fun when you crank it up.
I have actually cranked up the LM1875, but never beyond 3300 watts (an array amp for digital organ). For sure, enjoying it loud requires 2 things: Attention to technical stability (square wave test must actually look like squares), and finessing the power circuit a bit.
I noticed missing earth GND too. Not safe without it but case is made of wood so all you can do is connect earth GND to PSU power GND via a ground loop breaker consisting of large 3W 10R resistor plus two large back to back 10A10 diodes in parallel between earth GND and power ground. If case was metal, connect chassis to earth GND side of ground loop breaker.
That 10ohm ground lift resistor on the board is a ground loop breaker and decreases hum if you have any. I would not bypass it with jumper. But have a listen - maybe it sounds fine.
That 10ohm ground lift resistor on the board is a ground loop breaker and decreases hum if you have any. I would not bypass it with jumper. But have a listen - maybe it sounds fine.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.