| drfrink24 |
Hello all.
I've finished my GC using BrianGTs standard rev 2 PCB kit.
Firstly, I'd like to say I was very impressed with the documentation provided, the PCB layout and simple steps.
I am also very impressed with the sound. I think I've avoided all the major problems people have seen. I've got ZERO hum in the speakers, no turn-on thump. Just excellent, clean, detailed, rich sound.
Only one problem remains. Its not loud enough. I'm hoping its not the amp, but my pre-amp setup. Here are the details:
Speakers: Atlantic tech 371LCR 91db sensitivity, 8ohms.
I'm using a Yamaha mid-range CD player. It has a volume output control. I've used it on my other amp (Sony TA-N55ES), it works great for that.
For the amp, I used a 330VA trans, at the measured output voltages are 36.8 volts.
Using my trusty RadioShack SPL meter, at about 1 meter, with the CD player at full output, I'm seeing peaks of 82 db, on a CD that seemed the loudest biased I had on hand.
The bass is great, its just a tad out of the "sweet" volume. Not exactly room filling. I would think with 91db sens speakers, I could at least hit 90db in room?
I've also used a CD player, with headphone output, as another pre-amp. Again, less than 82 db peaks, with the volume all the way up.
Any ideas? I mean, my clock radio can put out more db than this!
I think I'm being realistic too. Not like I'm looking for 110 db in room at 40hz... |
|
|
| DonoMan |
If it's not clipping then increase the gain. Gain = Rf / Rin (add 1 if non-inverting)
Also, buffer it. A buffer is very easy to build on perfboard.
Also, which chip are you using? |
|
|
| drfrink24 |
whoops!
LM3875 non-inverted basic kit.
Its not clipping at all.
Why do you recommend buffering? I'm a newb, so, be gentle...
Oh, one more thing.
The sound is great. Noise floor is amazing. I would like to avoid adding complexity, assuming that there is a chance I'd lose something in the process. |
|
|
| DonoMan |
| Because if your source can't provide the necessary power the op amp requires, it will not be as loud... May not be your problem, but worth a try, I think. |
|
|
| drfrink24 |
Ok, I'll start looking into it then. The only problem I have is that my chassis layout would require longer input/output runs.
Although, this may not be an issue, I'll just use shielded cable this time around... |
|
|
| joensd |
| quote: | | Any ideas? I mean, my clock radio can put out more db than this! |
Hi,
I actually donīt think a buffer is what you need in the first place.
You should get plenty of volume with your setup/speakers.
Any CD-player is able to drive the input impedance of a GC easily (a buffer might still sound better but thatīs another story)
Have you double-checked everything, resistors sitting in the right place?
Good luck! |
|
|
| kropf |
You mention that you have no hum. How about hiss? With 91dB speakers, even a dead quiet GC should have some hiss when you put your ear close to the speaker. If you don't hear anything at about 3-4", your gain is probably set wrong - check your input and feedback resistors.
A surefire way to check is with an oscilloscope. However, without that, you could play some bass heavy music and get an approximate idea of the gain using a multimeter.
Jeremy |
|
|
| drfrink24 |
| quote: | Originally posted by joensd
Hi,
I actually donīt think a buffer is what you need in the first place.
You should get plenty of volume with your setup/speakers.
Any CD-player is able to drive the input impedance of a GC easily (a buffer might still sound better but thatīs another story)
Have you double-checked everything, resistors sitting in the right place?
Good luck! |
I can take another look tonite.
But, lets go hypothetical here:
1. Wrong values used. If you look at the basic kit components, there is only one set of resitors packaged in with quantity of 4, the 22 kohm .5w. If I didn't use these right, I'd have leftover parts, or, clearly mismatched size differences. So, we can rule that out for R1 and R2.
That leaves R3 ( I didn't use any Zobel components). The picture in the .pdf file is pretty obvious which one to use. I think we can rule out that.
I think I can pretty much rule out using the wrong values in general for the input.
2. Poor solder connections.
Lets assume that either R1, R2 or R3 are poorly soldered in some combination or another. This would have to be a consistent mistake on both channels, since both channels measure the same volume.
While not entirely out of the question, probably highly unlikely. But, lets just say I made the same soldering mistake, or, overheated the components or something messed up.
What would the effects be? Would the sound be great, but, at a lower volume? |
|
|
| drfrink24 |
| quote: | Originally posted by kropf
You mention that you have no hum. How about hiss? With 91dB speakers, even a dead quiet GC should have some hiss when you put your ear close to the speaker.
Jeremy |
Nope, no hiss. It's absolutely dead-quiet.
I must have something wrong on the input. I'll measure the r values on all my components, then consult the values in the documentation tonite for R1, R2, R3 and Rf. |
|
|
| HDTVman |
I agree with kropf. The feedback resistors set the gain of the amp and they should not all be the same value. The value of the resistor from the output to - input / - input to signal ground sets the gain. If both are 22k the 22,000/22,000=1 that will double the signal voltage but is only 3 dB gain. You want more like 20 - 30 dB gain. 30 dB would be 1000 times voltage gain, 20 dB would be 100 times.
Later BZ |
|
|
| Hanzwillem |
| quote: | Originally posted by drfrink24
Nope, no hiss. It's absolutely dead-quiet.
I must have something wrong on the input. I'll measure the r values on all my components, then consult the values in the documentation tonite for R1, R2, R3 and Rf. |
odd. My 3875 from brain are hissing and humming lot out of my 92dB sensative speakers. But they do have 'infinite' power. I don't know how loud these can play, but the levels I tried were already uncomfortable to listen to but still not clipping. I use Pass's (un)balanced line stage. |
|
|
| drfrink24 |
| quote: | Originally posted by HDTVman
I agree with kropf. The feedback resistors set the gain of the amp and they should not all be the same value. The value of the resistor from the output to - input / - input to signal ground sets the gain. If both are 22k the 22,000/22,000=1 that will double the signal voltage but is only 3 dB gain. You want more like 20 - 30 dB gain. 30 dB would be 1000 times voltage gain, 20 dB would be 100 times.
Later BZ |
Yeah, I think you're exactly on track. I went back over the the instructions, I found that R1 and R2 aren't supposed to be the same value...
So, yeah, I've got some de-soldering to do.
This would exactly explain what I'm seeing... err.. hearing, I guess
:xeye:
Thanks to all, I'll verify this tonite!
Update:
Actually, looking at the instructions, its pretty clear what I've done wrong.
R1 and R2 on both channels are 22k ohm, and 220 ohm on Rnfb.
But hey, if anyone wants an amp thats dead quiet, and IMHO sounds fantastic, but doesn't have the ability to even TEMPT you to play loud, this is the ticket! You all can thank my incompetence later! |
|
|
| cjd |
With no zobel, you should have installed four resistors.
3 of the same size, one bigger.
The big one is pretty obvious.
Of the three small resistors, there're two values included. (4) 20kohm and (2) 220ohm.
The 20k resistors go in Rf and R2
The 220 goes in R1
The 680ohm goes in R3.
I think. :)
C |
|
|
| drfrink24 |
| quote: | Originally posted by cjd
With no zobel, you should have installed four resistors.
3 of the same size, one bigger.
The big one is pretty obvious.
Of the three small resistors, there're two values included. (4) 20kohm and (2) 220ohm.
The 20k resistors go in Rf and R2
The 220 goes in R1
The 680ohm goes in R3.
I think. :)
C |
Yep, I'm 99.99999% positive that I put 20k on both R1 and R2, and 220 on Rnfb.
Don't think I messed up 680 though. :D
Brian is sending me new resistors for R1, R2 and Rf.
I also found that he is using 1k ohm for R1, instead of 220, so, maybe a silver lining in the whole messup. |
|
|
|
