Hi Granny,
Yep, this is another example of poorly executed balanced circuits that do not meet the balanced specs set by the recording industry but not implemented correctly by home audio. The seeming inability of the home audio and theater industry to conform to those standards is the cause for this problem. And yep, it sure does affect the sound quality of many fine components and has probably caused many bad reviews of products based on the reviewers who don't account for the problem. A true balanced circuit prevents these problems while single ended technology is rife with incompatabilities. Burson Audio makes a single ended buffer available in the USA that would solve the problem and is reported to work very well Unfortunately they don't make the a balanced buffer any more for reasons they only know. Krell with their CAST system is one of the few manufacturers who addressed this problem and essentially implemented a balanced system with their own proprietary connector. As for as I know Atmosphere is the only tube builder that has designed and uses a balanced circuit with tubes, in fact they own a patent for their design. It apparently is much easier to build solid state balanced circuits but the home audio industry does not design for this all the time either. There is apparently much more involved in designing a balanced circuit than just adding a balanced connector to the equipment's design. But if all the manufacturers would conform to the standards set by the music industry impedence mismatches between our equipment would be a thing of the past. |
Sgr, As for balanced tube designs, BAT and Audio Research have been using true differential circuits for many years. I'm sure there are others as well as the aforementioned Atmasphere. |
Thanks for the responses, but looking for an answer to my question which seems to have nothing to do with true balanced designs . |
"My preamp has an output impedance of 5k ohms, making for a possible poor match with the sub (20K input Imp.) if I use the 2nd outputs on my pre into the sub's RCA inputs." You can add a SS Buffer ... 1 to 1 gain amp ... to 2nd output on preamp to lower the preamp's output impedance to better match your sub's 20K input The Buffer should not add any gain but lower the impedance of the preamp from 5k to less than 600 ohms With a 5K output I would guess the pre is Tube .. how is match to main amp ?? "I suppose I could hook the sub up by going from my amp to the high level inputs on my sub?" I wouldn't daisy chain if I didn't have to ... keep the main amp and speakers seperate "I could simply run a second set of speaker cables out from my amp to my 2 subs this way right?" Then the main amp would have to drive both the main speakers and two subs ... may be to much of a load to ask your amp to drive ... |
Grannyring, all of the understandings you expressed your OP are correct. Whether or not significant deep bass rolloff or other frequency response anomalies will result from the non-ideal impedance match depends on whether or not the impedances that are involved vary significantly as a function of frequency. It is very common for a tube preamp to have a coupling capacitor in series with its output. I don't know whether or not the Dude has one, or uses some other design approach, but preamps that do use an output coupling capacitor will usually have an output impedance that rises significantly at deep bass frequencies, relative to its value at higher frequencies (which is already very high in the case of the Dude). That low frequency impedance rise will cause a deep bass rolloff in conjunction with the sub's low input impedance, due to what is known as the voltage divider effect. Also, if the preamp does not drive the two sets of outputs from individual buffer stages, and instead simply wires together the two output jacks for each channel inside the rear panel, then the low input impedance of the sub can also adversely affect the signals going to the main speakers. It is very common for a preamp that provides two sets of output jacks to not buffer them independently. Also, if the preamp does not individually buffer the two sets of outputs the capacitance of the RCA cable to the sub can affect the signals to the main amp and speakers similarly to how the capacitance of the preamp-to-main amp interconnects can. If the total of the lengths of the two sets of interconnects is long, and their capacitance per unit length is not low, given the high output impedance of the Dude the upper treble can be rolled off, at least slightly, as a result of the interaction between that output impedance and the TOTAL of the capacitances of the two sets of cables (if the preamp does not individually buffer the two sets of outputs). Given that your sub has speaker-level inputs, it would make sense to try hooking them up to your power amp outputs, and comparing sonics with the present arrangement. The wires from amp to sub do NOT have to be heavy gauge high quality speaker cables, because they will be conducting an extremely small amount of current (as a result of the high 20K input impedance). Regards, -- Al |
Al, understand all your comments and thanks. Very interesting stuff here. Quite a lot to know and be aware when adding subs to some systems. Yes, my pre has 3.3uf coupling caps. No, my second output is not buffered separately. I use low capacitance Blue Jean cables that are 10 foot long on the subs.
Davehrab & Al, going from my Altas amp to the sub's high level (binding posts)does not mean the Atlas amp is driving the two subs? Right? The plate amp in the sub still powers the subwoofer and the Atlas amp is simply sending out the signal for the plate amp to amplify. This is the way I have always understood it? Perhaps I am wrong? |
This Burson buffer may be just the ticket. Thanks for the idea. On paper it sounds good, anything I'm missing here? Seems it allows me to get the best sound out of my subs as well as my main speakers? It costs $320 however and I wonder if just hooking up my subs from my amp's output is the best and no cost answer? As long as the amp is not driving/powering the subs, but simply sending the signal over for plate amp for amplification. |
Going from my Atlas amp to the sub's high level (binding posts) does not mean the Atlas amp is driving the two subs? Right? The plate amp in the sub still powers the subwoofer and the Atlas amp is simply sending out the signal for the plate amp to amplify. This is the way I have always understood it? Yes, again you are completely correct. The amp would see a load impedance equal to the impedance of the main speakers (8 or 4 ohms or whatever) in parallel with the input impedance of the high-level input of the sub, which for all practical purposes will be essentially identical to the impedance of the main speakers. For instance, if the high-level input impedance of the sub is 20,000 ohms, that in parallel with 8 ohms is: (8 x 20000)/(8 + 20000) = 7.997 ohms. My pre has 3.3uf coupling caps. No, my second output is not buffered separately. I use low capacitance Blue Jean cables that are 10 foot long on the subs. Assuming that the length of the cable running to the main amp is also around 10 feet or less, and given the very low capacitance of the Blue Jeans cables (12 pf/ft if you are using the LC-1 cable), you should be ok in the present configuration with respect to the possible effects on the upper treble that I mentioned. With respect to the possibility of deep bass rolloff, 3.3 uf represents a capacitive reactance (the capacitive form of impedance) of about 2413 ohms at 20 Hz. In terms of the magnitude of the overall output impedance of the preamp, that combines with the resistive component of the preamp's output impedance (which to a reasonable approximation is probably about equal to the 5K specified nominal output impedance) as the square root of the sum of the squares of the two numbers. So the 5000 ohm value at higher frequencies would increase to 5552 ohms at 20 Hz. That is not much of a difference in the context of a 20K load, but some amount of phase shift will also occur as a result of the capacitance (I won't take the time to calculate how much, but I believe it will be small), and the slight impedance rise together with the phase shift might, I suppose, be enough to be very slightly perceptible. Also, keep in mind that if you connect the subs to the amplifier outputs instead of the preamp outputs the signals seen by the subs will reflect the sonic effects of the amplifier, which of course are not reflected in the signals seen by the sub in the present configuration. Very conceivably that could be the most significant factor in the sonic differences between the two configurations. In any event, I would definitely give the speaker-level hookup a try, before investing in a buffer (which was an excellent suggestion, but as my math indicates might not be necessary in this case even with the preamp-to-sub configuration). Regards, -- Al |
Al, you are the MAN! thanks so much for your help. |
BTW, my pre to amp ic is only 3 feet long. Yes, I use the LC-1 sub cable. |
Not sure what kind of sub you have, but with the ones that have both line level and speaker level inputs, the sub just 'samples' the input signal (yes, the RCA inputs too). So your preamp isn't really 'driving' the sub's amp(s) per se; so impedance mismatching that might possibly normally occur just isn't a factor. If you're still concerned, contact the sub manufacturer and find out for sure. |
02-07-12: Nsgarch
Not sure what kind of sub you have, but with the ones that have both line level and speaker level inputs, the sub just 'samples' the input signal (yes, the RCA inputs too). So your preamp isn't really 'driving' the sub's amp(s) per se; so impedance mismatching that might possibly normally occur just isn't a factor. Hi Neil, Not sure what you mean by the word "samples" in this context. I think you might mean that the voltage level of the signal being input to the sub is divided down to a lower level which is then processed internally. But that, and any other "sampling" process you may be referring to, are irrelevant. If the sub's line-level inputs are being used, and their input impedance is 20K for example, then the preamp has to drive that 20K load, regardless of whatever "sampling" may occur. If, as in this case, the preamp's two sets of outputs are not individually buffered, then a single output stage in the preamp has to drive the parallel combination of that 20K and the input impedance of the main power amp. That combined impedance will be less than 20K, and will affect the signals seen by both the main amp and the sub. The preamp also has to drive the interconnect cables to both destinations, and the capacitance of both cables will affect the signals seen by the main amp as well as by the sub. Impedance matching considerations are very much a factor in that situation, or in any other situation in which a powered sub is being driven from a line-level signal source having significant output impedance. And the importance of those considerations is increased by the relatively low input impedance that subs commonly have (as Grannyring pointed out), and by the fact that both the main amp and the sub, and their interconnect cables, are driven by a common output stage if the preamp does not individually buffer the two sets of output jacks that it may provide. Best regards, -- Al |
No Al, that's not the way it works (anymore ;--) Yes, in the old days, 'plate amps' were basically full range ss amps that were put on a 'plate' chassis so they could be mounted on the back of a speaker -- quite often a full range monito speaker (not a sub) which (self-powered) was more convenient in studio environments (no long speaker cables; using balanced signal lines of course.) And BTW, do you remember when many subwoofer (drivers) were made with dual voice coils, to keep the preamp's L/R outputs separated?
But today's powered subs (all the ones I know about, anyway,) including the subwoofer section of certain hybrid speakers (like the MartinLogan models that have a powered woofer section) use Class D ss amps which are fine at low frequencies, but aren't meant to be driven with a full range signal. These modern subs often include an LFE (Low Frequency Effects) input which DOES drive the plate amp directly, from a single (mono) interconnect, since it's assumed that the LFE signal comes 'pre-configured' from a surround processor and only contains mono low frequency information, already with the desired slope and cutoff determined by the processor.
The L/R speaker, and L/R line level inputs on these subs are "connected" to the sub's (Class D) amp through some kind of a buffer which also a.) combines the L/R signals (while keeping the preamp's L and R signals separate) and b.) limits the frequency range going to the subs internal amp. That's why the slope, phase, and often even the level controsl on the sub do not affect the LFE input signal, which it's assumed are configured at the processor. My point is that the "input impedance" at the buffered L/R line inputs (and L/R speaker inputs) of most subs today is 'set' by the sub's designer to accept almost any preamp's L/R main output without doing "damage" to the music signal due to accidental impedance mismatch.
. |
Hi Neil, I recognize and agree with everything in your last post, up until the last sentence: My point is that the "input impedance" at the buffered L/R line inputs (and L/R speaker inputs) of most subs today is 'set' by the sub's designer to accept almost any preamp's L/R main output without doing "damage" to the music signal due to accidental impedance mismatch. My point is that the input impedance of the sub, which is the input impedance of the buffer stage you are referring to, is what it is. And depending on its value, and on the output impedance of the preamp, and on the variation of that output impedance as a function of frequency, it can have significant effects (when driven from a preamp or other line-level source). In the case of Grannyring's sub, it is 20K, which is not an atypical value. The effects on the INPUT SIGNAL to the sub of the interaction of that 20K input impedance with the output impedance of the preamp will be no different than the interaction that would occur if the preamp were driving a main power amp having a 20K input impedance. While the subset of those effects that are relevant to mid and high frequencies will obviously not significantly affect the sound produced by the sub, the subset of those effects that are relevant to deep bass frequencies certainly will affect the sound produced by the sub. And if the preamp does not separately buffer the outputs to the sub and the main power amp, mid and high frequencies can be affected by the sub's input impedance (and also by the capacitance of the cable to the sub) as well. What goes on inside the sub has no relevance to any of that. I would agree with your last sentence, btw, if it were limited to the speaker-level inputs of a powered sub. In that situation the input impedance of the sub is so vastly higher than the output impedance of the power amp that the effects I have described would be completely inconsequential. Best regards, -- Al |
Al, your erudite response notwithstanding, it's unlikely ANY manufacturer these days would leave it to the user to make sure his product performed optimally ;--) I don't think Mr. Grannyring should be concerned in the slightest. As he already said, The system sounds very good as is, just wondering if perhaps I am onto something I have missed with possible bass performance improvement. and so my response is, don't go looking for problems where none exist ;--) . |
Thanks Nsgarch and Almarg for your comments and help. |
Input impedance is stated to be 20K so the preamp sees 20K to the sub and whatever ?? to the amp, but with both in paralell if the amp was 20k as well then the overall input imp. would be only 10k and that could affect freq. response on both the sub and amp. Al is correct, and using the speaker level inputs will negate the paralelling of the signals and offer less problems. Especially if they are using a passive preamp. The only thing to watch out for with using speaker level inputs is if you have a fully differential amp and the sub ties the grounds together to form the mono signal to the woofer, but that is another issue since a good sub will use a floating ground of about 100 Ohms or so. |
Koestner, thanks. Grannyring, Koestner makes a very important point about using a fully balanced amp with a sub, and your Aesthetix Atlas is indeed fully balanced. The negative speaker-level input of the sub should be connected to a ground point on the amp, and NOT to either of the amp's negative output terminals (which put out signals, as opposed to being grounded). Connecting the negative speaker-level input of the sub to one of the amp's negative output terminals could short the signal on that terminal to ground, assuming the amp and the sub both have three-prong power plugs, and depending on their internal grounding configuration. A suitable ground point will often (but not always) be a chassis screw. You can verify that by using your multimeter to check for continuity between a chassis screw and either the ground sleeve of an rca connector or pin 1 of an xlr connector (while nothing is connected to the amp). Concerning the input impedance of the amp, according to the manual it is an extremely high 530K, and is therefore insignificant with respect to the issues we have been discussing. Regards, -- Al |
Al, the builder told me if the plate amp has the negative high level input grounded or in common with another connector (left minus connected to right minus, then I would need to do a special connection from the Atlas. Connect only one phase of each channel ( presemably the plus) to the high level input.
I will look at the back plates tonight. |
Yes, but in addition to connecting the + output of the Atlas to the + high level input of the sub (for each channel; I believe you are using two subs), you also need a ground or signal return connection between the sub(s) and the amp. Otherwise you'll most likely get a BIG hum.
Best regards,
-- Al |
Ha, ya a BIG hum is the last thing I need1 |
Ok, one last question so I can make my final decision. I am not sure I want to go the high level hook up route as I simply don't know how the plate amps are wired. Can't see anything as all the sub binding posts are hooked together on a circuit board with the traces hidden from view. I would have to take boards off etc and I don't want to do all that. My amp is fully balanced and I fear causing an issue due to the way the sub high level inputs may well be wired -Keiga KG 5150.
Bottom line. Is the line level hook up going to cause an issue in terms of impedance mismatch or not? Al comments as well as Nsgarch suggest I will not have an issue in terms of sound degregation. Here are the facts on my gear.
Preamps output impedance is 5k ohms
Plate amps input impedance is 22k ohms
My stereo Atlas amp has an input impedance of 470k ohms
My sub cable is very, very low capacitance -Al referred to it above and they 10 feet long. I run two subs.
My ic's between the pre and amp are 3 feet long
While the music sounds good I do wonder if I am rolling off in the bass area and possibly compromising the sound of my main speakers. Again, both Al and Nsgarch seem to suggest no issue. But some other comments suggest otherwise? I would buy a Buson buffer to place between my pre and subs if that is deemed as best. I understand that such a buffer would reduce what the subs see as output impedance down to under 600 ohms and perhaps as low as 100 ohms.
I would rather not spend the money if there is nothing here to be concerned with. I set my subs to play 40 hertz and below only. Since I have 2 I set the volume very low - near the minimum really.
What do you expert say. |
I would add to the key facts that you listed for your gear that the output coupling capacitor in your preamp is 3.3 uf. Many tube preamps use somewhat smaller coupling capacitors, which would increase the likelihood of deep bass rolloff. Many tube preamps also have much lower output impedance at mid and high frequencies, which would further increase deep bass rolloff when used with a smaller coupling capacitor, because it would result in more of a DIFFERENCE between the low frequency and mid/hi frequency output impedances.
Per my earlier analysis, my feeling is that you are certainly not experiencing any major deep bass rolloff, and at most any deep bass rolloff would be extremely slight and would only occur on material containing spectral components approaching 20 Hz.
As I indicated earlier, if it were practicable to go to the high-level sub connections probably the most significant difference would be due to putting the sonic effects of the amplifier in the path to the sub. And that difference, while standing a good chance of being significant, is probably unpredictable in terms of being better or worse.
Best regards,
-- Al |
That is what I thought and thanks Al. |
If you hook the sub up in parallel to the preamp you will go from an output to input impedance ratio of 94 to only 4. I feel that will certainly result in high frequency sparkle loss. You should look into a buffer between the preamp and sub for the line level inputs. The Musical Fidelity X10 v3 has a high input impedance around 400K that will mate well with your amp. I think the ratio would then be about 44:1 and certainly acceptable compared to 4. I am speaking from experience here, not just textbook. I noticed a definate drop in the highs when I hooked up my sub the same way without the buffer. These buffers don't cost that much. Just get one and try it both ways... you will hear a difference. |
Koestner, I'm curious as to what preamp you were using. Perhaps it has an output impedance that rises in the upper treble region.
Also, what were the lengths of the cables to both the sub and the power amp, and what was their capacitance per unit length (or alternatively, what was their make and model)? As I mentioned earlier in the thread, the capacitance of the cable from preamp to sub, as well as the capacitance of the cable from preamp to power amp, can roll off the upper treble (as seen by the power amp) if it is high in relation to the preamp's output impedance.
Best regards,
-- Al |
Koestner, I have only seen these placed between the preamp and cd or dac. Can it be placed between my preamp and sub amps and work fine?
I assume the second set of preamp outs go into the x10 inputs and then I output to the sub's plate amps right? Is this better or the same as placing the x10 between the cd player and pre? Will my preamps output impedance be lowered by placing the x10 between cd player and preamp? Lowered in the same way as placing it between the pre and subs...... |
Grannyring, the answers to the four questions in your last post are:
1)Yes
2)Yes
3)Different
4)No
Also, I would be cautious in extrapolating from the results a buffer provides in one system to the results it would provide in another system, unless there is a good understanding of the reasons for the improvements it provided in the first system.
Best regards,
-- Al |
First off, I must say that Al definately knows more about electronics/stereos than me from my reading of these forums for so long. So with a bit of confusion I am asking - "Why aren't we seeing this the same way?" This is what I think:
A preamp to amp connection should have an impedance ratio of at least 40:1 from the amps input to the pre out. So if the OP's pre has an output impedance of 5K, and his amp has an input imp of 450K, then his ratio is 90:1
Now if he hooks up the sub using the 2nd line level output on his preamp and the sub has an input inp of only 20K, then the 20K and the 450K are seen to be in parallel by the preamp and thus the preamp is now seeing an input impedance of only 19K, and a ratio of 4.
Won't this cause rolloffs of the frequency lows and highs? It did when I hooked my passive preamp to my amp and sub by splitting the signals, but when I used a buffer to the sub it sounded much better through my mains.
I see this as a necessity for the OP to at least try as a ratio of 4:1 is a serious mismatch compare to his original value of 90.
Al, what are your thoughts here? |
Hi Koestner, A preamp to amp connection should have an impedance ratio of at least 40:1 from the amps input to the pre out. So if the OP's pre has an output impedance of 5K, and his amp has an input imp of 450K, then his ratio is 90:1
Now if he hooks up the sub using the 2nd line level output on his preamp and the sub has an input inp of only 20K, then the 20K and the 450K are seen to be in parallel by the preamp and thus the preamp is now seeing an input impedance of only 19K, and a ratio of 4. Agreed 100%, assuming that only a nominal preamp output impedance (5K in this case) is known. If the worst case (maximum) preamp output impedance across the 20 Hz to 20 kHz frequency range is known, the 40:1 figure could be comfortably reduced to 10:1. Won't this cause rolloffs of the frequency lows and highs? Sometimes. Achieving a high ratio of load impedance to output impedance will definitely eliminate the possibility of rolloffs or other frequency response irregularities due to impedance incompatibilities. However, that does not NECESSARILY mean that a low ratio will cause problems. If all of the impedances that are involved do not vary significantly as a function of frequency (meaning that they are essentially resistive), and if the sum of the capacitances of the two interconnect cables is kept appropriately low in relation to the output impedance of the preamp at high frequencies, the only consequence of the low ratio would be a slight attenuation that would affect all frequencies equally, and that would be easily compensated for by turning the volume control up slightly. Earlier in the thread we established that cable capacitance is a non-issue in Grannyring's case. The very high nominal output impedance of the preamp, and the relatively high value of its output coupling capacitor, as I see it suggest that the output impedance of the preamp is likely to vary relatively little as a function of frequency. My analysis earlier in the thread indicated that the output impedance rise at low frequencies is likely to affect the deep bass very minimally, if at all. And I see no particular reason to suspect in this case that the output impedance will rise significantly at high frequencies (which would cause an upper treble rolloff, given the 19K loading), although there are at least a few preamps for which that does occur. The usually unspecified input capacitance of the sub and the power amp could also affect the highs, in conjunction with the preamp's output impedance, but the total cable capacitance of only 159 pf (13 feet x 12.2 pf/ft) would seem low enough to provide a reasonable margin for that. I see this as a necessity for the OP to at least try as a ratio of 4:1 is a serious mismatch compare to his original value of 90. So given the foregoing, I see it as something to consider, but I would keep in mind what I consider to be a significant chance that there may not be a problem at present that it would solve (although it could conceivably affect the sound produced by the sub just as a result of its own intrinsic sonic character). I would also suggest that further consideration be given to trying the speaker-level connections. After thinking further about the issues Grannyring cited concerning uncertainty about the grounding configuration in the sub, I don't see that that has any relevance. What is needed is identification of a connection point to the circuit ground of the Atlas amp (perhaps Aesthetix can confirm whether or not a chassis screw would be such a point, or it could be checked with a multimeter). The negative high-level input terminal of the sub would be connected to that point, with the positive high-level terminal of the sub being connected to the positive output terminal of the amp. If that is done, the internal grounding configuration of the sub would be irrelevant, as I see it, aside from the usual possibility that a slight ground loop hum could conceivably occur. Best regards, -- Al |
Thanks Al for taking the time to write such an informing response. In my case I noticed a distinct reduction of the sparkle in the highs. So I got a much better sound using the buffer, then I just concluded that it's that impedence issue that I had read about on these forums. Not knowing that it is much more complex then that, I figured the ratio thing was written in stone. So now what comes to mind is the Seierra Audio amps and preamps that have reportably excellent sound that have identical output to input impedence (75 Ohms) and use a BNC coaxial wire to connect them. This effectively is a ratio of only 1:1 yet people say it sounds great. So as you say there is more to it.
Thank you for sticking with this thread and I'm sure that others as well as myself have benefited from this discussion. |
Hi All,
While it is true that many manufacturers including ARC do make a balanced designed preamp, they don't conform to the balanced spec. (I'm sure this has a name, I'm not aware of it.) But I was having a similar problem as Grannyring between my ARC REf 3 and my Xilica Crossover. I made a quick call to ARC to see is my suspicions of incompatability were correct. The tech at ARC said, there is no point in even trying to get the two to work together that the ARC preamps could not be configured to work with low impedence input and outputs because of the circuit design. His quick less than two minute explanation was sell your ARC unit and don't purchase another unless the equipment you use are compatible with the ARC design.
A subsequent conversation with the guys from Atamasphere provided me with much of the explanations I outlined above but I did leave out all the technical reasons why tube preamp designs do not conform to the balanced standard that the music industry uses today which prevents impedence matches between equipment. The audio industry has adopted the balanced connector part of the equation and the cables are built correctly by the cable guys, it is the electronics guys who are at fault.
If all audio equipment was built to the balanced specs the music industry uses, we would not have these issues of impedence mismatching. And, I believe the music industry does successfully mix tube gear with solid state with no problems.
Not to sound like I'm picking on the tube manufacturers alone, but shame on the subwoofer guys for not dealing with this problem in their designs as well. If one is going to design products with class D amplifier circuits, then care should be taken to raise the impedence so that it matches so that it conforms with an industry standard as well. One wonders how many subwoofers have received bad reviews or that have had unhappy customers because of an impedence mismatch, and not the design and construction of the woofer itself. |
Wanted to update folks on my progress. Well I have two subs now hooked up to my system. I have a/b'd the sound with and without the subs. When I turn the subs off I also remove the ic's from my preamp to compare the sound.
No loss of highs. No loss of bass. No bad anything. So what I feared as an impedance mismatch did not materialize. Most interesting as on paper one would think I could have an issue with impedance between my preamp and sub plate amps.
Al, you certainly got this one correct. |
