Hi,
Well, imagine it like this.
How about you have six filters all with the same adjustment range and you alter the mechnical stops so the usable range is limited to 24, 12 and 6dB.
Now instead of using mechanical stop put resistors in series with the Pot.
The impedance of your filter will go up, this could be corrected, but instead why not re-adjust the values to take advantage of the larger apparent value of our Pot?
So, the 24dB range filters would use (say) a 10K pot and filters calculated for a 10K Pot with 680R series resistors in the Filter (appx.).
The 12dB range filters would use a 10K Pot and a pair of 5K1 resistors to limit the adjustable range, so the filter would be calculated for 20K Pot and 1.5K series resistors in the Filter (appx.).
Finally the 6dB range filters can use a 10K Pot and 10K series resistors, so the filter would be calculated for a 30K Pot with 2.2K series resistors (appx.).
The other option might be to scale the resistors from the wipers of the filters so that in the summing they contribute appropriately.
Ciao T
Well, imagine it like this.
How about you have six filters all with the same adjustment range and you alter the mechnical stops so the usable range is limited to 24, 12 and 6dB.
Now instead of using mechanical stop put resistors in series with the Pot.
The impedance of your filter will go up, this could be corrected, but instead why not re-adjust the values to take advantage of the larger apparent value of our Pot?
So, the 24dB range filters would use (say) a 10K pot and filters calculated for a 10K Pot with 680R series resistors in the Filter (appx.).
The 12dB range filters would use a 10K Pot and a pair of 5K1 resistors to limit the adjustable range, so the filter would be calculated for 20K Pot and 1.5K series resistors in the Filter (appx.).
Finally the 6dB range filters can use a 10K Pot and 10K series resistors, so the filter would be calculated for a 30K Pot with 2.2K series resistors (appx.).
The other option might be to scale the resistors from the wipers of the filters so that in the summing they contribute appropriately.
Ciao T
I'm thinking. . .
45hz, 120hz, 700hz, 2.2khz, 7khz, 12khz
like: thud, drum, bari, tirade, rasp, hiss
For the amount of lift/cut, 10db seems generous and 6db might do, except for when it didn't.
The LDR idea isn't bad; however, very similar effects can be accomplished with cap//resistor, a cap rigged across a divider set to "almost" max, a resistor series to a resistor, and current drive. . . without having to either purchase or match LDR's. The techniques can be done at input or middle since the conditions are controlled; however, if done at output, it would need a cancel switch or "extent" dial.
Assuming the input signal is quite good but that all tone controls generally reduce resolution somewhat, I think we should put a gain stage at input, prior to the tone controls. There's very few logical spots for gain, but the input jack looks like a good spot to me, since the signal is nice there and we might want to give it a head start, plus generous capacity for current as well.
Check worst case source: Not only the worst, but also the most popular--That's an mp3 player. It is very weak for current, would clip itself for voltage if turned to max, and its 5 fixed "eq" pattern settings are generally selected by choosing the least offensive. Now, what can our proposed tone amp do to help this worst case source?
Second worst quality source and second most popular is a computer, suffering all of the same problems as the worst except that you can choose/upgrade software codecs and choose sound cards to pick more favorable varieties; however, still weak for current, distorts at max voltage output which is still insufficient and has an onboard eq that is very slow to adjust even if an optimal setting could be found. Now, what can our proposed tone amp do to help the second worst source?
Parts check: Generally speaking, op-amps come with different internal compensations, the extra dull Unity Stable, the shouty Never Unity Stable, and the middle ground with sweet sound Near-Unity Stable, which could be run at unity if inverted and still not dull. Considering available op-amp varieties, if we happen to need a non-inverting buffer, I think that a discrete parts alternative is justifiable for obtaining "just right" compensation. As for current, there is so much complex discussion about parallel op-amps but what that does for audio is really simple--sounds like you didn't use cheap chips, even if you did.
I would definitely favor regulated/capmulti power for this device regardless of whatever advertisements there may be as to power rejection figures of op-amps, because the truth is that, where there is gain, you've either got capmulti or noise. Well, I think it important for a hi-fi tone amp to have the cleanest power and ground imaginable--these references are of equal importance to the audio circuit. In regards to the power circuit, we could end up with a very beautiful tone amp, or something else.
Fuzz/texture/reanalog: If using tubes or mosfets to add texture, what that does is actively block droning, reducing resolution in order to make the signal more pleasantly textured and sometimes a lot more useful (especially if loud playback is desired). For example, if one uses a standard "tone" control to reduce the raspy on the Andrews Sisters LP, the result of the tone knob is droning; however, the very next stage can add fuzz to remove the droning, with the end result useful. The same is true of MP4, a codec that drones progressively worse at lower bit rates (many people suffer headaches with "HD" radio); however, many new studio recordings have fuzz/texture added to block this problem, since they won't get a top 40 otherwise; but, unfortunately, they tend to overdo it. It seems that nobody invented a "droning detector" to engage and also govern the extent of fuzz/texture effects. Therefore, need for fuzz/texture is not always, but rather it is sometimes, and so if this (optional) feature is present in a hi-fi tone amp, it will need a cancel switch. Nose gate (so the fuzz isn't nonstop) and mild reverb (obscure the filtering) are other options in this same category and could also need a cancel switch.
Limiter: These are as useful as clipping/x-max isn't. Of course such a thing is optional; however, the useful feature isn't present in the majority of power amplifiers. Most power amplifiers that do have limiters have unpleasant limiters. Well, the matter is really as simple as a signal big enough to trip a voltage divider (engaged as needed), resulting in a soft clip limiter (since the extent big enough to make a hard clip doesn't occur). I do not know how to make one adjustable, but that might not be important. The important aspect is, if we're going to crank up the bass, we might want to disallow exceeding the rails of our tone amp. This signal with a pre-set ceiling could be fed into any power amp that has its own volume control (or adjustable gain), with the result that it could be dialed in so that clipping/x-max would never occur regardless of the large varieties amongst different music tracks. Since the extent of the rails of our tone amp is known, a "soft limiter" does not need a cancel switch.
I hope that some of these notes are helpful.
45hz, 120hz, 700hz, 2.2khz, 7khz, 12khz
like: thud, drum, bari, tirade, rasp, hiss
For the amount of lift/cut, 10db seems generous and 6db might do, except for when it didn't.
The LDR idea isn't bad; however, very similar effects can be accomplished with cap//resistor, a cap rigged across a divider set to "almost" max, a resistor series to a resistor, and current drive. . . without having to either purchase or match LDR's. The techniques can be done at input or middle since the conditions are controlled; however, if done at output, it would need a cancel switch or "extent" dial.
Assuming the input signal is quite good but that all tone controls generally reduce resolution somewhat, I think we should put a gain stage at input, prior to the tone controls. There's very few logical spots for gain, but the input jack looks like a good spot to me, since the signal is nice there and we might want to give it a head start, plus generous capacity for current as well.
Check worst case source: Not only the worst, but also the most popular--That's an mp3 player. It is very weak for current, would clip itself for voltage if turned to max, and its 5 fixed "eq" pattern settings are generally selected by choosing the least offensive. Now, what can our proposed tone amp do to help this worst case source?
Second worst quality source and second most popular is a computer, suffering all of the same problems as the worst except that you can choose/upgrade software codecs and choose sound cards to pick more favorable varieties; however, still weak for current, distorts at max voltage output which is still insufficient and has an onboard eq that is very slow to adjust even if an optimal setting could be found. Now, what can our proposed tone amp do to help the second worst source?
Parts check: Generally speaking, op-amps come with different internal compensations, the extra dull Unity Stable, the shouty Never Unity Stable, and the middle ground with sweet sound Near-Unity Stable, which could be run at unity if inverted and still not dull. Considering available op-amp varieties, if we happen to need a non-inverting buffer, I think that a discrete parts alternative is justifiable for obtaining "just right" compensation. As for current, there is so much complex discussion about parallel op-amps but what that does for audio is really simple--sounds like you didn't use cheap chips, even if you did.
I would definitely favor regulated/capmulti power for this device regardless of whatever advertisements there may be as to power rejection figures of op-amps, because the truth is that, where there is gain, you've either got capmulti or noise. Well, I think it important for a hi-fi tone amp to have the cleanest power and ground imaginable--these references are of equal importance to the audio circuit. In regards to the power circuit, we could end up with a very beautiful tone amp, or something else.
Fuzz/texture/reanalog: If using tubes or mosfets to add texture, what that does is actively block droning, reducing resolution in order to make the signal more pleasantly textured and sometimes a lot more useful (especially if loud playback is desired). For example, if one uses a standard "tone" control to reduce the raspy on the Andrews Sisters LP, the result of the tone knob is droning; however, the very next stage can add fuzz to remove the droning, with the end result useful. The same is true of MP4, a codec that drones progressively worse at lower bit rates (many people suffer headaches with "HD" radio); however, many new studio recordings have fuzz/texture added to block this problem, since they won't get a top 40 otherwise; but, unfortunately, they tend to overdo it. It seems that nobody invented a "droning detector" to engage and also govern the extent of fuzz/texture effects. Therefore, need for fuzz/texture is not always, but rather it is sometimes, and so if this (optional) feature is present in a hi-fi tone amp, it will need a cancel switch. Nose gate (so the fuzz isn't nonstop) and mild reverb (obscure the filtering) are other options in this same category and could also need a cancel switch.
Limiter: These are as useful as clipping/x-max isn't. Of course such a thing is optional; however, the useful feature isn't present in the majority of power amplifiers. Most power amplifiers that do have limiters have unpleasant limiters. Well, the matter is really as simple as a signal big enough to trip a voltage divider (engaged as needed), resulting in a soft clip limiter (since the extent big enough to make a hard clip doesn't occur). I do not know how to make one adjustable, but that might not be important. The important aspect is, if we're going to crank up the bass, we might want to disallow exceeding the rails of our tone amp. This signal with a pre-set ceiling could be fed into any power amp that has its own volume control (or adjustable gain), with the result that it could be dialed in so that clipping/x-max would never occur regardless of the large varieties amongst different music tracks. Since the extent of the rails of our tone amp is known, a "soft limiter" does not need a cancel switch.
I hope that some of these notes are helpful.
Last edited:
Hi,
The frequencies of the original palette and the range of boost/cut are well chosen. They come close to the ones you list as well. It is worth looking closely.
It is bad. Very bad, actually. The distortion from LDR's makes anything else look harmless. This can be offset using complex bridge arrangements, but complexity is drastic.
Gain before the tone control presents an overload margin issue. Even if we do not expect sources to be higher than 4V RMS (many pro-ones are higher) usual rail voltages severely limit the gain we can place here.
Also, the whole point of this thread is to precisely avoid the effects of common tone controls, while retaining their functions.
I do not think this type of source would fall under consideration here. It is below lo-fi.
My main Audio source is a PC and trust me, it is nothing like an MP3 player... What you state is a rather common misconception. Again, the kind of computer source you describe would be unlikely to be considered by most here on this thread.
I would definitely favour something a bit more elaborate and high performance, we are talking about a HIGH END capable tone control, not about a generic "Lo-Fi" EQ (you can buy these for next to nothing anyway).
Tubes or other options may be available if we make the EQ fully passive. In this case the debate about which to use and which is better becomes moot.
I would however add that correctly implemented Tubes or Fet do not "add texture", I would rather say that using them correctly avoids the gritty "texture" common bipolar solid state circuits tend to impart.
Limiters have their place in low grade Semi-Pro-Gear (where we expect the operators to be incapable to correctly operate controls and to back off if they hear distortion). They are rather inappropriate for high quality sound equipment and even for mastering calibre studio gear.
They are very helpful if the goal is to produce a low cost, low-end semi-pro or pro-sumer effect box (incidentally, such are available in large numbers, if not precisely to your desired spec).
Even the Circuit Carl_Huff shows (with appropriate additions) would probably be overkill for what you want. Maybe even the NightPro EQ3D is OTT, however i would suggest you closely investigate it as the schematic can easily be adjusted to change the Bands to those you desire.
Details of a DIY NitePro clone are here:
Nite - EQ3D
However, your suggestions in my view entirely miss the point of the discussion here and to take much of them on board would meant to accept significantly lower available fidelity for the result than desired. In fact much of what you suggest would have been unacceptable in the 80's when I designed studio electronics for the east german state radio/TV.
And the designs we used and created (they fell under the RFZ banner but where mostly created by small companies) have as I have noticed recently acquired cult following among recordists. it seems someone lucked into a fairly large pile of the stuff. Seem I was wrong when I thought the gear that was ripped out of the studio very early on in the german re-unification was scrapped and dropped into landfills, seems someone rescued it and stored it...
I would suggest to re-read the whole thread and possibly also the originating one about the "Blowtorch" preamplifier by John Curl, Bob Crump and Carl Thompson and about the original Cello Palette as part of the Audio Suite, to get a feel where this is aimed.
The original palette for example used 61-Position rotary switches with individual resistors for each control (including EQ) and fully discrete, largely J-Fet based Transconductance Operational Amplifiers.
A picture of the insides of the original is here:
The aim to keep as much of the aspirations of high quality of these types of products, but to provide a way to actually make these "DIY productable" and more accessible, as true clones are not doable.
Ciao T
I'm thinking. . .
45hz, 120hz, 700hz, 2.2khz, 7khz, 12khz
like: thud, drum, bari, tirade, rasp, hiss
For the amount of lift/cut, 10db seems generous and 6db might do, except for when it didn't.
The frequencies of the original palette and the range of boost/cut are well chosen. They come close to the ones you list as well. It is worth looking closely.
It is bad. Very bad, actually. The distortion from LDR's makes anything else look harmless. This can be offset using complex bridge arrangements, but complexity is drastic.
Gain before the tone control presents an overload margin issue. Even if we do not expect sources to be higher than 4V RMS (many pro-ones are higher) usual rail voltages severely limit the gain we can place here.
Also, the whole point of this thread is to precisely avoid the effects of common tone controls, while retaining their functions.
I do not think this type of source would fall under consideration here. It is below lo-fi.
My main Audio source is a PC and trust me, it is nothing like an MP3 player...
What you state is a rather common misconception. Again, the kind of computer source you describe would be unlikely to be considered by most here on this thread.I would definitely favour something a bit more elaborate and high performance, we are talking about a HIGH END capable tone control, not about a generic "Lo-Fi" EQ (you can buy these for next to nothing anyway).
Tubes or other options may be available if we make the EQ fully passive. In this case the debate about which to use and which is better becomes moot.
I would however add that correctly implemented Tubes or Fet do not "add texture", I would rather say that using them correctly avoids the gritty "texture" common bipolar solid state circuits tend to impart.
Limiters have their place in low grade Semi-Pro-Gear (where we expect the operators to be incapable to correctly operate controls and to back off if they hear distortion). They are rather inappropriate for high quality sound equipment and even for mastering calibre studio gear.
They are very helpful if the goal is to produce a low cost, low-end semi-pro or pro-sumer effect box (incidentally, such are available in large numbers, if not precisely to your desired spec).
Even the Circuit Carl_Huff shows (with appropriate additions) would probably be overkill for what you want. Maybe even the NightPro EQ3D is OTT, however i would suggest you closely investigate it as the schematic can easily be adjusted to change the Bands to those you desire.
Details of a DIY NitePro clone are here:
Nite - EQ3D
However, your suggestions in my view entirely miss the point of the discussion here and to take much of them on board would meant to accept significantly lower available fidelity for the result than desired. In fact much of what you suggest would have been unacceptable in the 80's when I designed studio electronics for the east german state radio/TV.
And the designs we used and created (they fell under the RFZ banner but where mostly created by small companies) have as I have noticed recently acquired cult following among recordists. it seems someone lucked into a fairly large pile of the stuff. Seem I was wrong when I thought the gear that was ripped out of the studio very early on in the german re-unification was scrapped and dropped into landfills, seems someone rescued it and stored it...
I would suggest to re-read the whole thread and possibly also the originating one about the "Blowtorch" preamplifier by John Curl, Bob Crump and Carl Thompson and about the original Cello Palette as part of the Audio Suite, to get a feel where this is aimed.
The original palette for example used 61-Position rotary switches with individual resistors for each control (including EQ) and fully discrete, largely J-Fet based Transconductance Operational Amplifiers.
A picture of the insides of the original is here:
The aim to keep as much of the aspirations of high quality of these types of products, but to provide a way to actually make these "DIY productable" and more accessible, as true clones are not doable.
Ciao T
Hi,
Okay, I think we should consider this boiled down.
1) Everyone wants different active electronics and I doubt we can ever get agreement. This suggests a purely passive circuit.
2) I am sure if we can conjure up something, however I am not 100% sure that we can rely on using potentiometers.
I am sure we can implement the needed filters with T-Circuits, Cinema/Langevin style, but something a bit less complex without needing inductors and ton's of decks would be better.
If we stick to parallel passive James/Baxandall Filters then we can retain the option for 10K log (type B not type A) potentiometers.
For some estimates, this would produce around 1.5K input impedance at mid/high frequencies, controls flat with some variation when controls are adjusted. We could include an optional buffer (op-amp or other) for the very low LF bands where most of the impedance drop happens. Output impedance would be even lower, < 1K.
So the circuits self-noise would probably be around -125dBV (unweighted) with an insertion loss of around 24dB, so noise cannot be better -101dBV if the "re-gain" Amp is noiseless. This is not great for SNR, however in the other thread I argued that "extreme" SNR figures have little real-world benefit, so it is probably okay.
Given the impedances and levels etc. the input needs buffering capable of driving a few 100 ohm with little concern of very low noise low noise.
The output needs a suitable low noise Amp with a gain of around 24dB and less than 4nV|/Hz or -125dBV Ein noise and greater than or equal to 10K input impedance and output drive capabilities as needed for the application. This will be a touch difficult to do with tubes, no problems is using hybrid or solid state.
So we may still want to retain a few Op-Amp Footprints for DIP and SMD (use optional) on the PCB and a few buffer components (use optional) anyway, plus perhaps the "Paris Hilton/Nicole Richie enhanced 3-Pin Regulator" for powersupply, CRC filtering before the regs and discrete diodes..
So a PCB can be used with these positions filled and a choice of fidelity levels by actually selecting Regulator IC's, Rectifier diodes, Capacitors and Op-Amp's (plus buffer/no buffer) will be up to the builder, with a wide range of options.
We could even include options for a level (Volume) control around the output Op-Amp (see my post in the other thread) and optional gain and even attenuation for the input section.
Then the whole PCB could be bolted into a case to produce something akin to the Palette Preamp, single-ended in, single-ended out and with similar noise levels to Douglas Self's latest contraptions at similar gain levels (e.g. theoretically -118dBV output noise for -20dB gain) and have line input and output transformers added for balanced/pro options.
Using the right quality of components should give excellent fidelity for most uses, as only two stages are needed, one at notional unity and one to offer gain up to 40dB (if +16dB maximum gain are desired). Using an OPA827 0.001% THD20 for 3V into 600 Ohm can be attained.
Special requirements could be met using off-board electronics (Tubes etc.).
Ciao T
I would love to have a Palette style eq!
A cheap opamp one would be great, but a tube one that is linear with lower levels and adding nice distortion when pushed would be killer.
Okay, I think we should consider this boiled down.
1) Everyone wants different active electronics and I doubt we can ever get agreement. This suggests a purely passive circuit.
2) I am sure if we can conjure up something, however I am not 100% sure that we can rely on using potentiometers.
I am sure we can implement the needed filters with T-Circuits, Cinema/Langevin style, but something a bit less complex without needing inductors and ton's of decks would be better.
If we stick to parallel passive James/Baxandall Filters then we can retain the option for 10K log (type B not type A) potentiometers.
For some estimates, this would produce around 1.5K input impedance at mid/high frequencies, controls flat with some variation when controls are adjusted. We could include an optional buffer (op-amp or other) for the very low LF bands where most of the impedance drop happens. Output impedance would be even lower, < 1K.
So the circuits self-noise would probably be around -125dBV (unweighted) with an insertion loss of around 24dB, so noise cannot be better -101dBV if the "re-gain" Amp is noiseless. This is not great for SNR, however in the other thread I argued that "extreme" SNR figures have little real-world benefit, so it is probably okay.
Given the impedances and levels etc. the input needs buffering capable of driving a few 100 ohm with little concern of very low noise low noise.
The output needs a suitable low noise Amp with a gain of around 24dB and less than 4nV|/Hz or -125dBV Ein noise and greater than or equal to 10K input impedance and output drive capabilities as needed for the application. This will be a touch difficult to do with tubes, no problems is using hybrid or solid state.
So we may still want to retain a few Op-Amp Footprints for DIP and SMD (use optional) on the PCB and a few buffer components (use optional) anyway, plus perhaps the "Paris Hilton/Nicole Richie enhanced 3-Pin Regulator" for powersupply, CRC filtering before the regs and discrete diodes..
So a PCB can be used with these positions filled and a choice of fidelity levels by actually selecting Regulator IC's, Rectifier diodes, Capacitors and Op-Amp's (plus buffer/no buffer) will be up to the builder, with a wide range of options.
We could even include options for a level (Volume) control around the output Op-Amp (see my post in the other thread) and optional gain and even attenuation for the input section.
Then the whole PCB could be bolted into a case to produce something akin to the Palette Preamp, single-ended in, single-ended out and with similar noise levels to Douglas Self's latest contraptions at similar gain levels (e.g. theoretically -118dBV output noise for -20dB gain) and have line input and output transformers added for balanced/pro options.
Using the right quality of components should give excellent fidelity for most uses, as only two stages are needed, one at notional unity and one to offer gain up to 40dB (if +16dB maximum gain are desired). Using an OPA827 0.001% THD20 for 3V into 600 Ohm can be attained.
Special requirements could be met using off-board electronics (Tubes etc.).
Ciao T
We approximately concur on the crossover points and using nicely clear op-amps and that super high quality power is important. I'm especially grateful for high quality power--that helps the "High End" part.
I think the main concentration should be on the Pallete style tone amp. Potentiometers are fine although Log and Lawfake (also Log) are nicest. Softer sound effect LDRs and similar effects from passive devices can be added externally to output or input and are not needed inside of the tone amp, so don't worry about it.
I still have 3 areas of concern, which are resource management for home audio.
1. The home is not assumed full of studio audio equipment.
Maximum applicability: Because of our clean power, Adjustable gain is much more appropriate running *somewhere* in with our tone amp instead of having gain relying on (afflicted by) the unregulated power of a power amp. Simply ensure enough capacity for a variety of sources.
2. Exclusivity also excludes the need of the tone amp; however. . .
Current at the input: When we also add an input buffer for the tone amp, then you have a wider variety of source devices is supported and that increases the usefulness of the tone amp.
3. The operator may wander to a different chair or room.
Sensible bass knob: With adding just two resistors for a voltage divider, two inversely series zeners connected to the ground leg of the voltage divider, as to "switch on" the voltage divider rarely and this device ONLY at the lowest pitched bass control. Tracks do vary and I don't wish to dash for the knob per each during loud playback. I wasn't suggesting overkill limiter operated by a discrete mainframe, but rather a super simple and slight mitigation for unmanned potentiometer since we probably don't want to maintain a grip on the volume/bass knob(s) during playback at home.
So, if those 3 things made sense, most sources work even if they're not completely gold plated, and you didn't need either LDR or tubes.
P.S.
The deal with an op-amp per each tone knob looked more favorable for performance than passive options.
I think the main concentration should be on the Pallete style tone amp. Potentiometers are fine although Log and Lawfake (also Log) are nicest. Softer sound effect LDRs and similar effects from passive devices can be added externally to output or input and are not needed inside of the tone amp, so don't worry about it.
I still have 3 areas of concern, which are resource management for home audio.
1. The home is not assumed full of studio audio equipment.
Maximum applicability: Because of our clean power, Adjustable gain is much more appropriate running *somewhere* in with our tone amp instead of having gain relying on (afflicted by) the unregulated power of a power amp. Simply ensure enough capacity for a variety of sources.
2. Exclusivity also excludes the need of the tone amp; however. . .
Current at the input: When we also add an input buffer for the tone amp, then you have a wider variety of source devices is supported and that increases the usefulness of the tone amp.
3. The operator may wander to a different chair or room.
Sensible bass knob: With adding just two resistors for a voltage divider, two inversely series zeners connected to the ground leg of the voltage divider, as to "switch on" the voltage divider rarely and this device ONLY at the lowest pitched bass control. Tracks do vary and I don't wish to dash for the knob per each during loud playback. I wasn't suggesting overkill limiter operated by a discrete mainframe, but rather a super simple and slight mitigation for unmanned potentiometer since we probably don't want to maintain a grip on the volume/bass knob(s) during playback at home.
So, if those 3 things made sense, most sources work even if they're not completely gold plated, and you didn't need either LDR or tubes.
P.S.
The deal with an op-amp per each tone knob looked more favorable for performance than passive options.
Hi TL,
Why aren't you considering dual linear potentiometers with law
faking resistors added ? Surely tracking / matching is better ?
rgds, sreten.
Last edited:
Linear law fake. I like that.
Let us assume for the moment that the unit is to be monophonic and that we will build one unit per each channel.
Do we always use 2 channels? Not always. Do we always have an equal amount of stuffing and/or reflective surfaces on both sides of every given room? Not always. Is the tracking of dual potentiometers always exact? Not always. Are speaker drivers made precisely the same every time? Not always.
These distracting problems can be dodged with monophonic individual units. . . aka monobloc tone amps.
Either dual channel or single channel knobs could be used and that is unimportant. The main interest is not in the distractions but rather the main interest is in a reasonably low loss and somewhat authentic Palette tone amp.
Let us assume for the moment that the unit is to be monophonic and that we will build one unit per each channel.
Do we always use 2 channels? Not always. Do we always have an equal amount of stuffing and/or reflective surfaces on both sides of every given room? Not always. Is the tracking of dual potentiometers always exact? Not always. Are speaker drivers made precisely the same every time? Not always.
These distracting problems can be dodged with monophonic individual units. . . aka monobloc tone amps.
Either dual channel or single channel knobs could be used and that is unimportant. The main interest is not in the distractions but rather the main interest is in a reasonably low loss and somewhat authentic Palette tone amp.
Last edited:
Thorsten and Daniel - Very cool project you are starting here .
My preferences, if I may, would be:
- stereo
- the same six frequencies as the Palette
- ability to use opamps in 8 pin DIP sockets
- enough room on the PCB to plug into the same 8 pin DIP sockets some discrete jfet-based opamps (for midrange and highs) instead of ICs
- I would probably use switches instead of pots, and give up some min/max range
- as Thorsten suggested, single-ended filtering circuit, with transformer (or jfet discrete opamp) conversion for balanced inputs and outputs.
Hopefully this will turn out to be compatible with others' and your interests.
I'll keep an eye on this thread.
Pierre
.My preferences, if I may, would be:
- stereo
- the same six frequencies as the Palette
- ability to use opamps in 8 pin DIP sockets
- enough room on the PCB to plug into the same 8 pin DIP sockets some discrete jfet-based opamps (for midrange and highs) instead of ICs
- I would probably use switches instead of pots, and give up some min/max range
- as Thorsten suggested, single-ended filtering circuit, with transformer (or jfet discrete opamp) conversion for balanced inputs and outputs.
Hopefully this will turn out to be compatible with others' and your interests.
I'll keep an eye on this thread
.Pierre
It is Thorsten's project, but I'd probably want to build one. I'm hoping for low loss even if that means a few more op-amps to solder. He and I agreed on most of the critical aspects, and for that I'm grateful.
P.S. If you favor balanced, here's a good sounding little thing with a fet input:
P.S. If you favor balanced, here's a good sounding little thing with a fet input:
Attachments
![SSphaseinverter2[1].gif](/community/data/attachments/253/253232-5493d28aa4f4a08131a4b7c994661fd5.jpg)
Well, best possible quality within diy abilities with variable quality componentsis the aim and if a "dual mono" and 14 freq knobs plus vol, etc is required, then "so be it" -
IMO, the "pots" are one of the most difficult components - I do use a bridged LDR attenuator with quite reasonable results, but doubt it could be applied here.
With only 2 gain stages, including discrete jfet/IC option allows increased pcb density/flexibility, particularly if pots designed for "offbrd" connection rather than single large pcb (not sure about this!), and still offers options of external gain stages and power supplies
What is a "Paris Hilton/Nicole Richie enhanced 3-Pin Regulator" for powersupply"?? That's Interesting! Cct, please?
IMO, the "pots" are one of the most difficult components - I do use a bridged LDR attenuator with quite reasonable results, but doubt it could be applied here.
With only 2 gain stages, including discrete jfet/IC option allows increased pcb density/flexibility, particularly if pots designed for "offbrd" connection rather than single large pcb (not sure about this!), and still offers options of external gain stages and power supplies
What is a "Paris Hilton/Nicole Richie enhanced 3-Pin Regulator" for powersupply"?? That's Interesting! Cct, please?
Hi,
Lawfaked pot's change impedance with adjustment, thus they would change filter behaviour. Traditionally I am a great advocate of lawfaked linears, but one needs to know their drawbacks.
If we want to use lawfaked linears we need to redesign the circuit to accommodate them and then a completely passive design may be off the table.
Ciao T
Lawfaked pot's change impedance with adjustment, thus they would change filter behaviour. Traditionally I am a great advocate of lawfaked linears, but one needs to know their drawbacks.
If we want to use lawfaked linears we need to redesign the circuit to accommodate them and then a completely passive design may be off the table.
Ciao T
Hi,
This will not work for the discussed "massive passive parallel" james tonestack. We can use any number of options that have linear pots, but these mean Op-Amp's are unavoidable (though you can build Op-Amp's with tubes as well).
A purely passive system means we can use any gain stage and buffer stage we ruddy well care and there is no need to debate which op-amp etc.
Lets not.
We ARE always using at least 2 channels or multiples or 2.
The object of this EQ is program re-equalisation, strictly analogue, at the highest possible levels of objective and subjective performance. Given the wide bandwidth of the circuit is singulary unsuited to equalising rooms and speakers.
If this sort of thing is what you need get a Behringer DEQ2496 and modify the analogue stages and power supplies. It has all the functions you want including room Auto-EQ etc. (and some more you did not mention but which can come in handy, such as the frequency dependent stereo width control) all for pittance in money.
It is not possible to beat that unit using generic parts and circuits and attempting to create the functions or even something close in the analogue domain.
Finally, I personally seriously object to any "dual-mono" controls in any stereo system, they are a nightmare to set correctly.
The whole point of the Palette is to allow the user to correct tonal aberrations in the recording in a way that is both intuitive, easy and quick with minimal loss of fidelity. If any of the group ""intuitive, easy, quick, minimal loss of fidelity" are missing the whole project becomes pointless from my viewpoint.
For "JAEQ" (just any other EQ) we have no need to laboriously design new stuff. There are many designs for flexible pro-EQ's that may be found on line and we may take any that we like the looks of (the Studer 169 EQ would probably suit me) or buy some flexible Pro-Audio EQ and modify it.
Ciao T
This will not work for the discussed "massive passive parallel" james tonestack. We can use any number of options that have linear pots, but these mean Op-Amp's are unavoidable (though you can build Op-Amp's with tubes as well).
A purely passive system means we can use any gain stage and buffer stage we ruddy well care and there is no need to debate which op-amp etc.
Lets not.
Do we always use 2 channels? Not always. Do we always have an equal amount of stuffing and/or reflective surfaces on both sides of every given room? Not always. Is the tracking of dual potentiometers always exact? Not always. Are speaker drivers made precisely the same every time? Not always.
These distracting problems can be dodged with monophonic individual units. . . aka monobloc tone amps.
We ARE always using at least 2 channels or multiples or 2.
The object of this EQ is program re-equalisation, strictly analogue, at the highest possible levels of objective and subjective performance. Given the wide bandwidth of the circuit is singulary unsuited to equalising rooms and speakers.
If this sort of thing is what you need get a Behringer DEQ2496 and modify the analogue stages and power supplies. It has all the functions you want including room Auto-EQ etc. (and some more you did not mention but which can come in handy, such as the frequency dependent stereo width control) all for pittance in money.
It is not possible to beat that unit using generic parts and circuits and attempting to create the functions or even something close in the analogue domain.
Finally, I personally seriously object to any "dual-mono" controls in any stereo system, they are a nightmare to set correctly.
The whole point of the Palette is to allow the user to correct tonal aberrations in the recording in a way that is both intuitive, easy and quick with minimal loss of fidelity. If any of the group ""intuitive, easy, quick, minimal loss of fidelity" are missing the whole project becomes pointless from my viewpoint.
For "JAEQ" (just any other EQ) we have no need to laboriously design new stuff. There are many designs for flexible pro-EQ's that may be found on line and we may take any that we like the looks of (the Studer 169 EQ would probably suit me) or buy some flexible Pro-Audio EQ and modify it.
Ciao T
Hi,
To me low loss requires fewer Op-Amp's and smarter circuit design, rather than more op-amp's. More circuitry can never improve the signal. One reason that most people in High end and HiFi are biased against Tone Controls and Equalisers is that their impact on fidelity is usually dramatic and not positive. This is at least in part due to the large (active) parts count of many.
Back in East Germany one of the products we had was a graphic 10-Band Equaliser, fully discrete and with only 3 Transistors per channel. This EQ is still justly famous and more than 20 Years after it ceased production is greatly sought after.
So everything should be as simple as possible but no simpler.
Ciao T
PS, attached is the 9-band version of the east german EQ I referenced above... It should be noted also that this EQ is "zero feedback as well, not just minimal and LC based...
The inductors for the lower bands may be replaced by Gyrators (the HF ones are available) and of course the old Bipolar transistors (read BC for SC) can be in part at least replaced by J-Fets.
This could be called "ZenEQ" aka ZEQ and should sit well with the crowd over at the Pass section.
It should be equally clear that the circuit may be directly translated to zero feedback tube circuitry, however it is imperative that it be designed as either Tube or Solid state (or Hybrid) circuit ground-up, so it means choices for circutry must be made or the project must be forked into a Tube and a Solid State Fork.
To me low loss requires fewer Op-Amp's and smarter circuit design, rather than more op-amp's. More circuitry can never improve the signal. One reason that most people in High end and HiFi are biased against Tone Controls and Equalisers is that their impact on fidelity is usually dramatic and not positive. This is at least in part due to the large (active) parts count of many.
Back in East Germany one of the products we had was a graphic 10-Band Equaliser, fully discrete and with only 3 Transistors per channel. This EQ is still justly famous and more than 20 Years after it ceased production is greatly sought after.
So everything should be as simple as possible but no simpler.
Ciao T
PS, attached is the 9-band version of the east german EQ I referenced above... It should be noted also that this EQ is "zero feedback as well, not just minimal and LC based...
The inductors for the lower bands may be replaced by Gyrators (the HF ones are available) and of course the old Bipolar transistors (read BC for SC) can be in part at least replaced by J-Fets.
This could be called "ZenEQ" aka ZEQ and should sit well with the crowd over at the Pass section.
It should be equally clear that the circuit may be directly translated to zero feedback tube circuitry, however it is imperative that it be designed as either Tube or Solid state (or Hybrid) circuit ground-up, so it means choices for circutry must be made or the project must be forked into a Tube and a Solid State Fork.
Attachments
Last edited:
Hi,
Actually, even if the pots are designed onto the PCB (make space for Alps Blue) you can easily wire them with wire to the pads, so that would be the easy part. As is external Gain stages.
That is if we use the "Massive Parallel Passive James/Baxandal Tone Stack" (MPPJBTS).
I think the two came up with this when they where hanging with Beau de Plaz who drew some other very controversial and (in)famous schematics...
Ciao T
Actually, even if the pots are designed onto the PCB (make space for Alps Blue) you can easily wire them with wire to the pads, so that would be the easy part. As is external Gain stages.
That is if we use the "Massive Parallel Passive James/Baxandal Tone Stack" (MPPJBTS).
I think the two came up with this when they where hanging with Beau de Plaz who drew some other very controversial and (in)famous schematics...
Ciao T
Hi,
So you would vote for an active EQ circuit, say derived from the NitePro EQ3D (guys, I gave the link to this schematic earlier, please look it up), which embeds Op-Amp's as essential (and un-replacable) part of the circuit?
Or would you consider using Op-Amp's more of a "last resort of the incompetent" (Salvor Hardin originally talking about Violence)?
Ciao T
So you would vote for an active EQ circuit, say derived from the NitePro EQ3D (guys, I gave the link to this schematic earlier, please look it up), which embeds Op-Amp's as essential (and un-replacable) part of the circuit?
Or would you consider using Op-Amp's more of a "last resort of the incompetent" (Salvor Hardin originally talking about Violence)?
Ciao T
on the subject of pots, these were all the rage a while ago:
http://www.diyaudio.com/forums/chip...ot-ex-tawan-photo-ignore-my-ugly-pinky-2.html
I used them in gainclones and really sounded good. Not sure how accurately they track.
http://www.diyaudio.com/forums/chip...ot-ex-tawan-photo-ignore-my-ugly-pinky-2.html
I used them in gainclones and really sounded good. Not sure how accurately they track.
I would favor active EQ utilizing an op amp per each frequency control knob, only if the signal DIDN'T have to run through every one of those in series.
See, it would be just fine for audio if the op-amp assigned to a frequency control knob only had a major affect during the assigned frequency range of that particular control knob.
This could let you use the strengths of op-amps, for example bias the NE5532 to class A for the 12k knob (so you can dial up 12k without crossover distortion) or parallel the JRC4580 for the 2.2k knob where the emitter resistors reduce hard sound compression (don't laugh before you check the output of your M-audio), or something really forward from NatSemi at the 120hz knob (where its harmonics are clear, tame, and probably beneficial to low frequency audio). . . all of which are much different that what they'd do if used for the entire audio band.
Hi,
The best way in my for something like that is an LC style EQ with the inductors partially of fully replaced by gyrators. If using switches instead of pots the controls can be even made completely "centre-neutral", that is completely disconnected out of the circuit.
In most cases a single NPN BJT (or N-Channel J-Fet) can perform splendidly as virtual inductor. Op-Amp's can be used for those who absolutely must bodge an NE553X into everything, even where it is neither beneficial or appropriate...
I doubt I would design in any of Op-Amp's you mention into any commercial product for audio use, except the lowest grade of gear. The closest they come to doing audio in any of my designs is as power supply regulator. They are inappropriate in my personal view for high quality audio.
Moreover, I would suggest that avoiding wherever possible may be a better policy, unless we are concerned about delivering ppm or sub ppm distortion levels, which gives bragging rights but little genuine benefit in the real world.
However, I think the NitePro EQ3D clone that I have linked to previously would probably suit your requirements quite well and let you indulge in op-amp rolling and tweaking. It includes PCB design for mono channels and all other needed details, the OpAmp's are indeed used in parallel, a reasonable amount of effort has been applied to the power-supply, so if you simply adjust the circuit to match the frequencies you want you can just build it, use whatever op-amp you like and be done.
Here an actual, build-up example:
click for more
I personally would not consider this approach however, or if at all then only as the absolute last resort. Chances are I'd rather put up with Waves REQ6 loaded onto my PC...
Ciao T
The best way in my for something like that is an LC style EQ with the inductors partially of fully replaced by gyrators. If using switches instead of pots the controls can be even made completely "centre-neutral", that is completely disconnected out of the circuit.
In most cases a single NPN BJT (or N-Channel J-Fet) can perform splendidly as virtual inductor. Op-Amp's can be used for those who absolutely must bodge an NE553X into everything, even where it is neither beneficial or appropriate...
I doubt I would design in any of Op-Amp's you mention into any commercial product for audio use, except the lowest grade of gear. The closest they come to doing audio in any of my designs is as power supply regulator. They are inappropriate in my personal view for high quality audio.
Moreover, I would suggest that avoiding wherever possible may be a better policy, unless we are concerned about delivering ppm or sub ppm distortion levels, which gives bragging rights but little genuine benefit in the real world.
However, I think the NitePro EQ3D clone that I have linked to previously would probably suit your requirements quite well and let you indulge in op-amp rolling and tweaking. It includes PCB design for mono channels and all other needed details, the OpAmp's are indeed used in parallel, a reasonable amount of effort has been applied to the power-supply, so if you simply adjust the circuit to match the frequencies you want you can just build it, use whatever op-amp you like and be done.
Here an actual, build-up example:
click for more
I personally would not consider this approach however, or if at all then only as the absolute last resort. Chances are I'd rather put up with Waves REQ6 loaded onto my PC...
Ciao T
Hi,
Looks interesting but noisy. And here I was worrying about the around -100dB for my MPPJBTS.
Curves look suspiciously like the Palette, but less range.
Ciao T
Looks interesting but noisy. And here I was worrying about the around -100dB for my MPPJBTS.
Curves look suspiciously like the Palette, but less range.
Ciao T