Morning chaps!
"he missed it"
I didn't miss it, I was just impatient to hear it working with the full setup. And as you both predicted, it does act all 'screwy'
Altho low level listening last night did work rather well, using it today at higher volume levels, it got stupidly loud (not linearly increasing with main level) and at one end of x-over adjustment introduces crackle into the left speaker channel.
I bought 4x TL071 for a reason
I fully intend to knock up a buffer. I'd also then like to add one in front of my volume control in my chip amp, especially as I use baffle step correction here, as per Rod Elliot's pages. This is another potential problem, altho it 'seems' to work ok as is. I'm conceptually against using something ancient like the TL071 for the main signal though. Perhaps you could suggest to me a sensible (ie. won't break the bank or be fussy) alternative for this.
Thanks!
Simon
"he missed it"
I didn't miss it, I was just impatient to hear it working with the full setup. And as you both predicted, it does act all 'screwy'
Altho low level listening last night did work rather well, using it today at higher volume levels, it got stupidly loud (not linearly increasing with main level) and at one end of x-over adjustment introduces crackle into the left speaker channel.I bought 4x TL071 for a reason
I fully intend to knock up a buffer. I'd also then like to add one in front of my volume control in my chip amp, especially as I use baffle step correction here, as per Rod Elliot's pages. This is another potential problem, altho it 'seems' to work ok as is. I'm conceptually against using something ancient like the TL071 for the main signal though. Perhaps you could suggest to me a sensible (ie. won't break the bank or be fussy) alternative for this.
Thanks!
Simon
Curtesy of ESP, quicker than drawing it myself!
http://sound.westhost.com/p09_fig3.gif
edit: Oh and although it is popular to knock the 071, bear in mind that it is easily the most common op amp family used in pro gear, so likely the signal will have run through 100s already.
http://sound.westhost.com/p09_fig3.gif
edit: Oh and although it is popular to knock the 071, bear in mind that it is easily the most common op amp family used in pro gear, so likely the signal will have run through 100s already.
Glad to see your plans are good
Just because TL07x series is old does not mean it's bad Admittedly they can be bettered fairly easily, but for the price they are great. I am coming round to OPA134 (single) and OPA2134 (dual), but relatively speaking they are a big jump in price from the TL07x. I used to use NE5534 (single) and NE5532 (dual), but their BJT input can be a pain sometimes, and it's a hassle to add compensation to the 5534 when running unity gain.
Cheapest place I've found for the OPA's is Farnell if you order 10+
Just because TL07x series is old does not mean it's bad
Admittedly they can be bettered fairly easily, but for the price they are great. I am coming round to OPA134 (single) and OPA2134 (dual), but relatively speaking they are a big jump in price from the TL07x. I used to use NE5534 (single) and NE5532 (dual), but their BJT input can be a pain sometimes, and it's a hassle to add compensation to the 5534 when running unity gain.Cheapest place I've found for the OPA's is Farnell if you order 10+
Thanks PMpinkmouse said:Curtesy of ESP, quicker than drawing it myself!
http://sound.westhost.com/p09_fig3.gif
edit: Oh and although it is popular to knock the 071, bear in mind that it is easily the most common op amp family used in pro gear, so likely the signal will have run through 100s already.
Next question - I see on Elliot's diagram, that the volume control goes *before* the buffer stage. I thought it might have been the other way round.
My main question is - how many buffers do I want, and where do I put them, when I (currently) have:
input > baffle step correction > pot > lm3886 + active lowpass/sub
Should/could I use:
input > pot > buffer > baffle step c. > lm3886 + lpf/sub
Would I not want to somehow isolate the precious low-level signal that goes to the power amp chips from what goes to the x-over circuit?
Could I place one (100R?) resistor in series with the buffer and LM3886, and another in series with the sub x-over? Am I just imagining this would somehow separate their signals?
What I worry about is contaminating one of the LM3886 inputs with noise picked up from the sub x-over wire - which will lead quite far to a separate box....
OPA134 seems like a popular choice. I think there's an unpleasant hard n harsh character to the 2604 I use in my cd player, but I guess different applications make them sound different (and it's not even the same model, I know!). And it offers a lot the original NJM2114 didn't! (clarity, strong bass). Maybe it's all in my head of course, and they all work the same, who knows.richie00boy said:Glad to see your plans are good
Just because TL07x series is old does not mean it's bad
Cheapest place I've found for the OPA's is Farnell if you order 10+
Anyway, I'm sidetracking. The OPA134 is an acceptable price for one or two, no problem. Is it pin compatible with TL071? If so, I can use a socket and compare the little finger stabbers...
SimontY said:
Yes, just like I told you in one of my earlier posts
You already did it with the unused half of the first op-amp, remember? The first place you need to put a buffer is between the passive volume control and the filter input. This buffers the filter from the volume control and stops the volume control's variable and relatively high source impedance messing up the filter.
EDIT: added diagram
Input - volume - buffer - baffle step - chipamp
............................... - filter - subamp
Simon,
If the baffle step circuit is active, it will probably have a buffer function of its own. If so, Richie's scheme is OK to me.
If the radio and CD have very different output impedances, or you have long cables, I would use buffers in other places, too.
Buffers lower the impedance level, and make a circuit less sensitive to electro-magnetic interferrence picked up by cables.
The choice of OP-amp depends on the application. For buffers (x1 gain), you will want a unity gain stable op-amp to make life easier for you. I like the OP-275 dual op-amp, which is unity gain stable.
You can make impedance matching etc., if you really like to, but for many audio application, op-amps can really be used as simple as I illustrated.
Jennice
If the baffle step circuit is active, it will probably have a buffer function of its own. If so, Richie's scheme is OK to me.
If the radio and CD have very different output impedances, or you have long cables, I would use buffers in other places, too.
Buffers lower the impedance level, and make a circuit less sensitive to electro-magnetic interferrence picked up by cables.
The choice of OP-amp depends on the application. For buffers (x1 gain), you will want a unity gain stable op-amp to make life easier for you. I like the OP-275 dual op-amp, which is unity gain stable.
You can make impedance matching etc., if you really like to, but for many audio application, op-amps can really be used as simple as I illustrated.
Jennice
Thanks for the suggestions!
I don't think I could cope with building 6 buffers PM
This amp is destined to find it's eventual home at the office, to prodive background music (from PC, probably no BSC), and I don't want to spend tooo much more time or money on it. It's only temporarily in my main system, for evaluation and experimentation.
The sub x-over is long term though, and will probably be plugged into future chip-amps.
I like the setup suggested by richie00boy - much more simple!
I don't think I could cope with building 6 buffers PM
This amp is destined to find it's eventual home at the office, to prodive background music (from PC, probably no BSC), and I don't want to spend tooo much more time or money on it. It's only temporarily in my main system, for evaluation and experimentation.
The sub x-over is long term though, and will probably be plugged into future chip-amps.
I like the setup suggested by richie00boy - much more simple!
SimontY said:
Ah, yes. My mistake assuming it would be active. IMO passive baffle step circuits are a waste of time. Just tack a buffer on the end of it, no need for an input resistor to ground on this one as the DC bias current will come through the pot and preceding op-amp.
I'd recommend some 22-47 ohm resistors on the outputs of the op-amps driving any interconnects (so the baffle step buffer and filter?), to guard against possible capacitive instability due to the cables.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.