theAnonymous1 said:I know I plug this stuff way too much, but J-B KWIK WELD epoxy works wonders for holding heatsinks on. It's a whole lot cheaper than "real" thermal epoxy too.
http://jbweld.net/products/jbkwik.php
And yes, those little OPA1632 get blazing hot. I cut a fin off of an old CPU heatsink and used some KWIK WELD to hold it on.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
In a pinch,I've used J-B Weld,and a few coins (pennies) glued to various chips and stuff (small bridge rects.)
You can even stack a couple pennies,or 'stagger' them (half overlapping),to make a sorta finned stack.
It works well.
@alexw88
Have you detected that your circuit in the schematic gives out inverted output of a signal ? I mean if signal is positive on input the output will be negative. I found this by chance while making measurements with RMAA. The Message from "Polarity" was allways detected as "inverted". When i changed the polarity it reported "correct"
I wondered that time, why the left channel report had only Stereo crosstalk between 20Hz-20kHz of approximately linear -78dB and the right channel was meanwhile @ -90dB.
After i changed the polarity the crosstalk lowered down, while noise level result was nearly constant as before changing polarity.
(remark that my first crosstalk diagramm i have posted here was allready the corrected one.)
So i think it could be useful for you and all @ here.
So these are just my 2 cents ...
Best regards
artQuake
Have you detected that your circuit in the schematic gives out inverted output of a signal ? I mean if signal is positive on input the output will be negative. I found this by chance while making measurements with RMAA. The Message from "Polarity" was allways detected as "inverted". When i changed the polarity it reported "correct"
I wondered that time, why the left channel report had only Stereo crosstalk between 20Hz-20kHz of approximately linear -78dB and the right channel was meanwhile @ -90dB.
After i changed the polarity the crosstalk lowered down, while noise level result was nearly constant as before changing polarity.
An externally hosted image should be here but it was not working when we last tested it.
(remark that my first crosstalk diagramm i have posted here was allready the corrected one.)
So i think it could be useful for you and all @ here.
So these are just my 2 cents ...
Best regards
artQuake
Finally got the boards.An externally hosted image should be here but it was not working when we last tested it.
Thank u once more alexw88 for support.
Will assemble them soon ... i wonder if this boards will give better results ...
Best regards
artQuake
Alex, the thermal pad is really easy to solder, capillary forces helping the solder to flow under the chip, while heating from the bottom side of the pcb. Only the heat during soldering made me thinking.
But soldering the legs of the OPA is pure surgery.
Here some results with RMAA from p2p soldering board (before) and the new pcb board.
Noise with p2p soldered
Noise with new pcb
Crosstalk with p2p soldered
Crosstalk with new pcb
Summary p2p soldered
Frequency response (from 40 Hz to 15 kHz), dB +0.05, -0.01 Excellent
Noise level, dB (A) -98.3 Excellent
Dynamic range, dB (A) 92.5 Very good
THD, % 0.012 Good
THD + Noise, dB (A) -75.4 Average
IMD + Noise, % 0.015 Very good
Stereo crosstalk, dB -95.4 Excellent
IMD at 10 kHz, % 0.015 Very good
General performance Very good
Summary new pcb
Frequency response (from 40 Hz to 15 kHz), dB +0.07, -0.15 Very good
Noise level, dB (A) -97.3 Excellent
Dynamic range, dB (A) 92.1 Very good
THD, % 0.019 Good
THD + Noise, dB (A) -71.7 Average
IMD + Noise, % 0.023 Good
Stereo crosstalk, dB -96.2 Excellent
IMD at 10 kHz, % 0.022 Good
General performance Very good
Interesting that the general performance of the p2p soldered got tiny better results ...
I'll test it once more when i get the better remote controlled stepped attenuator inside the circuit ... @ the moment i'm using two 1 $ stereo 10 k Pots ... but result is good.
Have you / or maybee others got about same results ? Or is there something questionable in the data ?
Specially for THD + Noise.
artQuake
But soldering the legs of the OPA is pure surgery
. Here some results with RMAA from p2p soldering board (before) and the new pcb board.
Noise with p2p soldered
An externally hosted image should be here but it was not working when we last tested it.
Noise with new pcb
An externally hosted image should be here but it was not working when we last tested it.
Crosstalk with p2p soldered
An externally hosted image should be here but it was not working when we last tested it.
Crosstalk with new pcb
An externally hosted image should be here but it was not working when we last tested it.
Summary p2p soldered
Frequency response (from 40 Hz to 15 kHz), dB +0.05, -0.01 Excellent
Noise level, dB (A) -98.3 Excellent
Dynamic range, dB (A) 92.5 Very good
THD, % 0.012 Good
THD + Noise, dB (A) -75.4 Average
IMD + Noise, % 0.015 Very good
Stereo crosstalk, dB -95.4 Excellent
IMD at 10 kHz, % 0.015 Very good
General performance Very good
Summary new pcb
Frequency response (from 40 Hz to 15 kHz), dB +0.07, -0.15 Very good
Noise level, dB (A) -97.3 Excellent
Dynamic range, dB (A) 92.1 Very good
THD, % 0.019 Good
THD + Noise, dB (A) -71.7 Average
IMD + Noise, % 0.023 Good
Stereo crosstalk, dB -96.2 Excellent
IMD at 10 kHz, % 0.022 Good
General performance Very good
Interesting that the general performance of the p2p soldered got tiny better results ...
I'll test it once more when i get the better remote controlled stepped attenuator inside the circuit ... @ the moment i'm using two 1 $ stereo 10 k Pots ... but result is good.
Have you / or maybee others got about same results ? Or is there something questionable in the data ?
Specially for THD + Noise.
artQuake
Here is my THD results, it is very low at 0.0002%.
Measurement is taken with the EMU 1212m PCI sound card, using balanced input and output. Output voltage from preamp is about 1.7V p-p.
Basically this is about the same as what I would get if I do a soundcard external loopback test. So I guess my test result is soundcard limited.
Can you post your THD graph? What would you get if you do an external loopback test?
Alex
Measurement is taken with the EMU 1212m PCI sound card, using balanced input and output. Output voltage from preamp is about 1.7V p-p.
Basically this is about the same as what I would get if I do a soundcard external loopback test. So I guess my test result is soundcard limited.
Can you post your THD graph? What would you get if you do an external loopback test?
Alex
The results that RMAA attested @ my pcb board above was a THD of 0.019 %
(Noise level = -97.3 dB)
Using math with the same noise level results in THD = 10EXP(-97.3/20) x100 = 0.00136 %
Alexw88’ THD of 0.0002 % should give a noise level of 20 x lg (0.0002 /100) = -113.97 dB
What result of THD have you got with RMAA alexw88 ?
Is there something wrong with calculation or RMAA ?
I made a external loopback with my soundcard some time ago (M-Audio Transit / unsymmetric operation) I’ll post that results/pics this evening … but what I remember that the noisefloor was near – 99dB and the THD of 0.005%.
Since the values of the preamp must be better then my soundcard, and when I look to alexw88’s results of his preamp, my RMAA results should at least also be soundcard limited.
I do not think that a better soundcard will/should attest this preamp better results. Or should I think so ?
I’m a little now … think it could be misinterpretation of RMAA Values or calculation ? The sound it delivers anyhow makes me really confident.
But for conscientiousness I wonder also why THD+Noise is so low …
(Noise level = -97.3 dB)
Using math with the same noise level results in THD = 10EXP(-97.3/20) x100 = 0.00136 %
Alexw88’ THD of 0.0002 % should give a noise level of 20 x lg (0.0002 /100) = -113.97 dB
What result of THD have you got with RMAA alexw88 ?
Is there something wrong with calculation or RMAA ?
I made a external loopback with my soundcard some time ago (M-Audio Transit / unsymmetric operation) I’ll post that results/pics this evening … but what I remember that the noisefloor was near – 99dB and the THD of 0.005%.
Since the values of the preamp must be better then my soundcard, and when I look to alexw88’s results of his preamp, my RMAA results should at least also be soundcard limited.
I do not think that a better soundcard will/should attest this preamp better results. Or should I think so ?
I’m a little
now … think it could be misinterpretation of RMAA Values or calculation ? The sound it delivers anyhow makes me really confident.But for conscientiousness I wonder also why THD+Noise is so low …
In your noise graph, I see the 50Hz and 150Hz hum is gone on the PCB version. The noise level is low enough. About your THD result, it is better to look at the graph to see what happened.
To know if it is soundcard limited or not, you need to do a soundcard loopback test. If the loopback test result is very close to your preamp result, then it may be soundcard limited. If the loopback test is a lot better, then it is not.
To know if it is soundcard limited or not, you need to do a soundcard loopback test. If the loopback test result is very close to your preamp result, then it may be soundcard limited. If the loopback test is a lot better, then it is not.
yes indeed, my soundcard as i mentioned has noise of -99dB and the pcb version -97.3dB, but THD is with 0.019% much higher than Soundcard 0.005%. Specially THD+Noise @ -71.7dB !
Or should i not attach that much importance on the values and only look to relative changes?
Have you noticed that you have also the same peak @ 150 Hz ?
Or should i not attach that much importance on the values and only look to relative changes?
Have you noticed that you have also the same peak @ 150 Hz ?
OK further in text ...
Summary M-Audio Transit USB Loopback
Frequency response (from 40 Hz to 15 kHz), dB +0.03, -0.01 Excellent
Noise level, dB (A) -99.7 Excellent
Dynamic range, dB (A) 99.7 Excellent
THD, % 0.0011 Excellent
THD + Noise, dB (A) -91.8 Very good
IMD + Noise, % 0.0034 Excellent
Stereo crosstalk, dB -97.6 Excellent
IMD at 10 kHz, % 0.0033 Excellent
General performance Excellent
Summary Preamp_pcb
Frequency response (from 40 Hz to 15 kHz), dB +0.06, -0.03 Excellent
Noise level, dB (A) -99.2 Excellent
Dynamic range, dB (A) 99.1 Excellent
THD, % 0.018 Good
THD + Noise, dB (A) -72.5 Average
IMD + Noise, % 0.021 Good
Stereo crosstalk, dB -91.1 Excellent
IMD at 10 kHz, % 0.020 Very good
General performance Very good
Measurementa @ Sampling mode 24-bit, 48 kHz
The better the sampling rate resolution, the better the results. In fact my first results where @ 16-bit, 44.1 kHz.
What Sampling mode did you use ?
Noise level, Dynamic range are about the same but THD & THD+Noise and especially IMD+Noise differ very clearly.
Look / compare the following graphs (specially between 0.1 - 1kHz)
M-Audio Transit USB Loopback - IMD + Noise %
Preamp_pcb - IMD + Noise %
What are the reasons ?
Any hints or suggestions are highly appreciated ...
artQuake
Summary M-Audio Transit USB Loopback
Frequency response (from 40 Hz to 15 kHz), dB +0.03, -0.01 Excellent
Noise level, dB (A) -99.7 Excellent
Dynamic range, dB (A) 99.7 Excellent
THD, % 0.0011 Excellent
THD + Noise, dB (A) -91.8 Very good
IMD + Noise, % 0.0034 Excellent
Stereo crosstalk, dB -97.6 Excellent
IMD at 10 kHz, % 0.0033 Excellent
General performance Excellent
Summary Preamp_pcb
Frequency response (from 40 Hz to 15 kHz), dB +0.06, -0.03 Excellent
Noise level, dB (A) -99.2 Excellent
Dynamic range, dB (A) 99.1 Excellent
THD, % 0.018 Good
THD + Noise, dB (A) -72.5 Average
IMD + Noise, % 0.021 Good
Stereo crosstalk, dB -91.1 Excellent
IMD at 10 kHz, % 0.020 Very good
General performance Very good
Measurementa @ Sampling mode 24-bit, 48 kHz
The better the sampling rate resolution, the better the results. In fact my first results where @ 16-bit, 44.1 kHz.
What Sampling mode did you use ?
Noise level, Dynamic range are about the same but THD & THD+Noise and especially IMD+Noise differ very clearly.
Look / compare the following graphs (specially between 0.1 - 1kHz)
M-Audio Transit USB Loopback - IMD + Noise %
An externally hosted image should be here but it was not working when we last tested it.
Preamp_pcb - IMD + Noise %
An externally hosted image should be here but it was not working when we last tested it.
What are the reasons ?
Any hints or suggestions are highly appreciated ...
artQuake
Thanks alex,
i ordered already the E-MU 0404 USB for better measurements, as the M-Audio Transit now is only good enough to listen music on my workstation. I'll get it on thuesday, then i’ll be able to make also measurements @ 24bit / 96 kHz or even in higher resolution. The Transit has only USB 1.0 so it's not possible to get in to 24bit / 96 kHz.
As voltage regulators i use two LT1083 for + & -. See details of the used PSU-Circuit at the following schematic. For this purpose they are for sure a bit to large, but i don't think that this will be the reason. Or is larger in this case badder ? I could give a pair of LT1085's also a try, but only when it's really making sense.
Shall i add additional capacitors the output of the PSU ? For example 1000uF//100uF//100nF. As you see i only have the Tantalum 10 uF. I did not put additional caps because i allready have 680uF on the preamp pcb board.
BTW: As PSU filtering caps on the pcb board i mounted :
680uF (Rubicon ZL) (http://www.farnell.com/datasheets/75224.pdf) paralelled with 47uF Sanyo OS-Con SP (http://www.farnell.com/datasheets/78196.pdf) and 100 nF Ceramic Kemet Golden Max C320 - X7R.
On pcb board :
SMD Resistors : 0.1%, 10 ppm/°C , 0.125 W, Tyco Holsworthy RP73-Series (http://www.farnell.com/datasheets/102809.pdf)
SMD Capacitors : Metallized pps film chip cap, Panasonic ECHU(X) (http://www.panasonic.com/industrial/components/pdf/abd0000ce13.pdf)
I think solving the problem is to aim on the psu, because my p2p soldering version has shown the same behaviour …
Would it be better in using a negative and positive Voltage regulator ?
I really have no answer to this.
i ordered already the E-MU 0404 USB for better measurements, as the M-Audio Transit now is only good enough to listen music on my workstation. I'll get it on thuesday, then i’ll be able to make also measurements @ 24bit / 96 kHz or even in higher resolution. The Transit has only USB 1.0 so it's not possible to get in to 24bit / 96 kHz.
As voltage regulators i use two LT1083 for + & -. See details of the used PSU-Circuit at the following schematic. For this purpose they are for sure a bit to large, but i don't think that this will be the reason. Or is larger in this case badder ? I could give a pair of LT1085's also a try, but only when it's really making sense.
Shall i add additional capacitors the output of the PSU ? For example 1000uF//100uF//100nF. As you see i only have the Tantalum 10 uF. I did not put additional caps because i allready have 680uF on the preamp pcb board.
An externally hosted image should be here but it was not working when we last tested it.
BTW: As PSU filtering caps on the pcb board i mounted :
680uF (Rubicon ZL) (http://www.farnell.com/datasheets/75224.pdf) paralelled with 47uF Sanyo OS-Con SP (http://www.farnell.com/datasheets/78196.pdf) and 100 nF Ceramic Kemet Golden Max C320 - X7R.
On pcb board :
SMD Resistors : 0.1%, 10 ppm/°C , 0.125 W, Tyco Holsworthy RP73-Series (http://www.farnell.com/datasheets/102809.pdf)
SMD Capacitors : Metallized pps film chip cap, Panasonic ECHU(X) (http://www.panasonic.com/industrial/components/pdf/abd0000ce13.pdf)
I think solving the problem is to aim on the psu, because my p2p soldering version has shown the same behaviour …
Would it be better in using a negative and positive Voltage regulator ?
I really have no answer to this.
Now it's accomplished 
:
Results of measurement with RMAA 6.1.0 @ 24 bit / 96 kHz with E-MU 0404 USB 2.0
Here are the results and graphs of the pcb version of the preamp
Frequency response (from 40 Hz to 15 kHz), dB +0.01, -0.04
Noise level, dB (A) -109.8 (E-MU is 3 dB lower)
Dynamic range, dB (A) 110.3
THD, % 0.0006
THD + Noise, dB (A) -98.6
IMD + Noise, % 0.0013
Stereo crosstalk, dB -102.7
Noise level
THD + Noise (at -3 dB FS)
Intermodulation distortion IMD + Noise
The both peaks between 100 and 200 Hz are already existent in the graphs from the soundcard. --> Anyhow i'll make a new psu for the E-MU, as it seems for me that it's very small and get's also very warm. (1000mA@5V ...
"Eliminate the wall warts !" .
Definitely wonderfull results, at least in my eyes. I'll measure again when all the components are mounted, connected and definetely boxed ... I wonder what results my p2p soldered one will deliver. I guess about nearly the same ...
I think the visual result will even be better when soundcard is better. Whatever this is all measurement up to here. Now it's time for listening to my music and get some wine from the cellar.
BTW: I have about min. 50 boards last. (the learning by burning and nice price reserve boards) So if there exists some demand for them just mail me. If there are interested parties from the same country then i'll desireabely gather the boards and send them in one parcel. average cost price 4$ per Board + shipment. But i think, since alex put the gerbers in here, you'll probably get them less expensive in your neighbourhood.
Thanks to all of you again for helping me along this progress. It's always nice to get some useful echos in here.
Best regards n cheers
artQuake
:Results of measurement with RMAA 6.1.0 @ 24 bit / 96 kHz with E-MU 0404 USB 2.0
Here are the results and graphs of the pcb version of the preamp
Frequency response (from 40 Hz to 15 kHz), dB +0.01, -0.04
Noise level, dB (A) -109.8 (E-MU is 3 dB lower)
Dynamic range, dB (A) 110.3
THD, % 0.0006
THD + Noise, dB (A) -98.6
IMD + Noise, % 0.0013
Stereo crosstalk, dB -102.7
Noise level
An externally hosted image should be here but it was not working when we last tested it.
THD + Noise (at -3 dB FS)
An externally hosted image should be here but it was not working when we last tested it.
Intermodulation distortion IMD + Noise
An externally hosted image should be here but it was not working when we last tested it.
The both peaks between 100 and 200 Hz are already existent in the graphs from the soundcard. --> Anyhow i'll make a new psu for the E-MU, as it seems for me that it's very small and get's also very warm. (1000mA@5V ...
"Eliminate the wall warts !" .Definitely wonderfull results, at least in my eyes. I'll measure again when all the components are mounted, connected and definetely boxed ... I wonder what results my p2p soldered one will deliver. I guess about nearly the same ...
I think the visual result will even be better when soundcard is better. Whatever this is all measurement up to here. Now it's time for listening to my music and get some wine from the cellar.
BTW: I have about min. 50 boards last. (the learning by burning and nice price reserve boards) So if there exists some demand for them just mail me. If there are interested parties from the same country then i'll desireabely gather the boards and send them in one parcel. average cost price 4$ per Board + shipment. But i think, since alex put the gerbers in here, you'll probably get them less expensive in your neighbourhood.
Thanks to all of you again for helping me along this progress. It's always nice to get some useful echos in here.
Best regards n cheers
artQuake
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.