HI!
I've just finished a buffer by Pavel Macura using it as a preamp instead. I think his design may have been discussed before - couldn't find any threads though.
It utilizes OPA627 with a BUF634T in it's feedback loop. BUF643T can handle current of uptil 250ma, so it should have what it takes to drive a power amp.
It sound nice, stereo perspective is wide and noise is all most nonexisting. Putting my ear to the speaker I can only hear plenty of "nothing"......
. BUF634T runs in class-A with an Iq of 15mA. I gets a little varm and seems to sound better after running an hour or so.
I have done a few mods that I would like to have some feedback on:
I have increased power to +/- 18 volt to have more dynamics.
Is this a good idea?
I would like to have more gain. I have changed R4 to 22K. Can I just raise it until I get the gain I want?
Can I remove C1 (1uf) and the 470p cap sitting over R2?
Can I omit C10 (18p) going from pin 2 to 6 on OPA637T?
I've just finished a buffer by Pavel Macura using it as a preamp instead. I think his design may have been discussed before - couldn't find any threads though.
It utilizes OPA627 with a BUF634T in it's feedback loop. BUF643T can handle current of uptil 250ma, so it should have what it takes to drive a power amp.
It sound nice, stereo perspective is wide and noise is all most nonexisting. Putting my ear to the speaker I can only hear plenty of "nothing"......
. BUF634T runs in class-A with an Iq of 15mA. I gets a little varm and seems to sound better after running an hour or so.I have done a few mods that I would like to have some feedback on:
I have increased power to +/- 18 volt to have more dynamics.
Is this a good idea?
I would like to have more gain. I have changed R4 to 22K. Can I just raise it until I get the gain I want?
Can I remove C1 (1uf) and the 470p cap sitting over R2?
Can I omit C10 (18p) going from pin 2 to 6 on OPA637T?
Attachments
Hi jazzpeter,
you can use supply voltage of +/-18V. In any case I would recommend to mount small heatsink onto BUF634T when operated with Iq=15mA.
Voltage gain can be reasonably increased up to +20, in case that you need it.
Input cap (1uF) prevents the buffer from possible DC shift due to preceding circuits, you can omitt the cap, but probably it is safer to keep it. Foil cap would do no sound harm, IMHO.
The 470 pF cap creates a HF filter with R1 and it is better to keep it, it also decreases possible intermodulations with HF.
18pF cap is not needed for high gain as you have, it is connected for buffer application (G=+2 unloaded).
OPA637 can be used for gain higher than 5, as is your case. Not for buffer application (G=+2).
Pavel Macura
you can use supply voltage of +/-18V. In any case I would recommend to mount small heatsink onto BUF634T when operated with Iq=15mA.
Voltage gain can be reasonably increased up to +20, in case that you need it.
Input cap (1uF) prevents the buffer from possible DC shift due to preceding circuits, you can omitt the cap, but probably it is safer to keep it. Foil cap would do no sound harm, IMHO.
The 470 pF cap creates a HF filter with R1 and it is better to keep it, it also decreases possible intermodulations with HF.
18pF cap is not needed for high gain as you have, it is connected for buffer application (G=+2 unloaded).
OPA637 can be used for gain higher than 5, as is your case. Not for buffer application (G=+2).
Pavel Macura
More gain....
HI Pavel,
How do I increase gain? Raise the value of R4? By how much? Now it's 22K which doesn't seem enogh to give the "swing" I need to feed my poweramp.
Is there a formel to calculate exact gain?
PS: I forgot to mention: This circuit was built in two hours a sounds great . . Highly recommended for a compact preamp. Noise is incredible low using OPA627AP.
HI Pavel,
How do I increase gain? Raise the value of R4? By how much? Now it's 22K which doesn't seem enogh to give the "swing" I need to feed my poweramp.
Is there a formel to calculate exact gain?
PS: I forgot to mention: This circuit was built in two hours a sounds great .
. Highly recommended for a compact preamp. Noise is incredible low using OPA627AP.
Hi Peter,
the gain can be calculated as G = 1 + (R4/R3).
Now (for R4=22k) you have voltage gain of 7,67 (17.7 dB). For voltage gain of 20 (i.e. 26dB) I would recommend to change R3 to 1k2 and R4 to 22k (as you have it now). The gain will be
G = 1 + 22/1.2 = 19.33, i.e. 25.7 dB.
P.S. The calculations hereabove are valid without terminating resistor R12 50 Ohm (that should be placed at the end of output cable). R12 creates a divider with R10 and reduces calculated gain to one halve compared to calculations. With R12 the gain is:
G = 0.5 x (1 + (R4/R3))
the gain can be calculated as G = 1 + (R4/R3).
Now (for R4=22k) you have voltage gain of 7,67 (17.7 dB). For voltage gain of 20 (i.e. 26dB) I would recommend to change R3 to 1k2 and R4 to 22k (as you have it now). The gain will be
G = 1 + 22/1.2 = 19.33, i.e. 25.7 dB.
P.S. The calculations hereabove are valid without terminating resistor R12 50 Ohm (that should be placed at the end of output cable). R12 creates a divider with R10 and reduces calculated gain to one halve compared to calculations. With R12 the gain is:
G = 0.5 x (1 + (R4/R3))
Per-Anders just sent me a link to his page. He says that Your design is very similar to his.
Look at: http://home.swipnet.se/~w-50719/hifi/qrv04
Look at: http://home.swipnet.se/~w-50719/hifi/qrv04
Meta42, qrv04 and Pavel's design is more or less the same and also what the datasheet of BUF634 says.jazzpeter65 said:
Peter the reason that I emailed you was to avoid attention from somebody.... and the reason was just technical talk, no spam.
Pavel, I think we have discussed it before but I don't understand your cable interface. Why not just add so much resistance so you will get the output impedance equal to the cable? If the cable has 50 ohms (just an example) you chose 51 ohms as series resistor and also as load in the other end. Isn't this the normal procedure?
Per, that's exactly how it is. Please visit my www (just click the button below), and see Audio Buffer. There is a column about placement of the resistors, R10 is in series with the cable and R12 is at the end as a terminating resitor.
From the engineering point of view, R12 seems to be useless for audio frequency band. In fact this resistor changes the sound impression in a very positive way. It is worth trying
From the engineering point of view, R12 seems to be useless for audio frequency band. In fact this resistor changes the sound impression in a very positive way. It is worth trying
Jazzy Peter,
I would also lower R2. With the gain of one you have it is OK, but raising the gain to 22 means you will also amplify any thermal noise of the 100k R2 22 times. 10K would still be high enough to cause no trouble to your sources. But you would want to increase the input coupling cap to keep the same lf cut-off point.
Jan Didden
I would also lower R2. With the gain of one you have it is OK, but raising the gain to 22 means you will also amplify any thermal noise of the 100k R2 22 times. 10K would still be high enough to cause no trouble to your sources. But you would want to increase the input coupling cap to keep the same lf cut-off point.
Jan Didden
gain and impedance...
Hi Jan and PMA,
Per-Anders wrote me something interesting: That I should keep the value of R4 to a value of max. 10K. It should give a "cleaner" resistance in the feedbackloop, he says. I'm not following him on that one.
Can you exsplain?
If R4 is a constant then I have to change R3 I guess. But will I get a gain of fx. +25db by changing R3 only?
I need this gain to drive my poweramp.....
Jan suggest to change R2 to a lower value. I have an ALPS 2 x 50K potmeter infront of R1. I guess that means that I can't go lower on R2 than 50K?
Also this means that I have to recalculate the filter R1 and the 470p going to ground at the input.
Best,
Peter
Hi Jan and PMA,
Per-Anders wrote me something interesting: That I should keep the value of R4 to a value of max. 10K. It should give a "cleaner" resistance in the feedbackloop, he says. I'm not following him on that one.
Can you exsplain?
If R4 is a constant then I have to change R3 I guess. But will I get a gain of fx. +25db by changing R3 only?
I need this gain to drive my poweramp.....
Jan suggest to change R2 to a lower value. I have an ALPS 2 x 50K potmeter infront of R1. I guess that means that I can't go lower on R2 than 50K?
Also this means that I have to recalculate the filter R1 and the 470p going to ground at the input.
Best,
Peter
Pavel,
Before reading your last pst I came to the same conclusion: R10 and R12 is a voltage divider and halfs the output gain.
I'll try this out to see what happens:
R4=10K
R3=510 Ohm
R12=omit
There is still R2 at the input. Wouldn't is be wrom to have me gain pot of 50K followed by R2 of 10K?
Before reading your last pst I came to the same conclusion: R10 and R12 is a voltage divider and halfs the output gain.
I'll try this out to see what happens:
R4=10K
R3=510 Ohm
R12=omit
There is still R2 at the input. Wouldn't is be wrom to have me gain pot of 50K followed by R2 of 10K?
Peter,
do not follow 50k pot with 10k resistor, keep original 100k.
Omitting R12 wil result in a little bit different sound, try it before you make a final decision. You might also have R12 connected as it is and furthermore reduce R3 to 250 Ohm, then you will get gain of 21 (circuit loaded by R12).
Also, in case of 50k pot reduce the 470pF input filter capacitor to 47pF.
do not follow 50k pot with 10k resistor, keep original 100k.
Omitting R12 wil result in a little bit different sound, try it before you make a final decision. You might also have R12 connected as it is and furthermore reduce R3 to 250 Ohm, then you will get gain of 21 (circuit loaded by R12).
Also, in case of 50k pot reduce the 470pF input filter capacitor to 47pF.
In case that you omitt R12: The bass is not so steady, it is less defined. More grainy mids. The better audio chain, the more it is pronounced. I would recommend you to make a simple try. It is easy to disconnect and connect R12.
I have another small objection. I would rather use 10k potentiometer instead of 50k potentiometer, smaller value of the pot always brings sonic improvement.
I have another small objection. I would rather use 10k potentiometer instead of 50k potentiometer, smaller value of the pot always brings sonic improvement.
Jan, not quite right about the noise: You must add all resistances including the signal source and this will in parallel so the noise contribution of R2 is none (almost) if the signal source is of low impedance type, 1k or less. I agree though with Jan that I also strive to have as low input impedance as possible, just for "pureness" reasons. Every signal source to put up with 10 k inpedance but the problem is to get a good input coupling cap. In my qrv04 case I search for a SMD capacitor and above 100 nF is hard to find.janneman said:
Peter why do you have to have so much gain. More than 10 is much for preamp.
Let me also add that it's good to have low resistor values if the design can take it. The problem with high values and high speed opamps is the input capacitance of the opamp. This (may) cause excessive phase shifts = oscillations
Pavel, I see what you mean with R12 but I had written and *) and "see text" a little more clearly in the schematics, just a point of view only.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.