Ski Bum Posted January 25, 2015 Share Posted January 25, 2015 Can I assume that if I play the offending passages at lower volume and I dont hear the same grain / distortion that it then becomes an issue somewhere else in the chain. and if I do still hear the grain / distortion it becomes part of the recording ?? To the first part, that would indicate less clipping, somewhere in the chain, most likely the amp (assuming you followed Mark's suggestion on pre and amp gain settings, you're simply driving the amp less hard). If the grain/distortion is in the source, you'll hear it at all volume levels. Quote Link to comment Share on other sites More sharing options...
TJIann Posted January 25, 2015 Share Posted January 25, 2015 Awesome thread! Thanks guys. I'm pouring myself a bourbon and then going down to my lounge and re-adjust the gain knobs on my McIntosh 2105 properly. Before doing that, let me find out where the gain adjustment lies in a Mc2105.Page 6 of your manual give detailed instructions o7n setting the gain controls! Yes, but I don't have the CBS test record to run through their setup. I followed the steps you outlined and came close to where I had the gains set. A bit lower for both and then adjust the right gain up a bit and the imaging is great. Thanks. Quote Link to comment Share on other sites More sharing options...
joessportster Posted January 25, 2015 Author Share Posted January 25, 2015 Can I assume that if I play the offending passages at lower volume and I dont hear the same grain / distortion that it then becomes an issue somewhere else in the chain. and if I do still hear the grain / distortion it becomes part of the recording ?? To the first part, that would indicate less clipping, somewhere in the chain, most likely the amp (assuming you followed Mark's suggestion on pre and amp gain settings, you're simply driving the amp less hard). If the grain/distortion is in the source, you'll hear it at all volume levels. Great, I just tried a SS amp and the grain / distortion remains so I am guessing it is either the dac breaking up or the source. Either way I feel better about my Kenzie Headphone, and I am a little better armed to discuss amps when it comes to power "Mark, now that Joe seems to be satified with the intial questions he had, do you have time to provide a layman's type of description to the "transfer function" of an amplifier?" as the song goes "I aint ever satisfied" I will follow your inquiry as well with great enthusiasm Quote Link to comment Share on other sites More sharing options...
Don Richard Posted January 25, 2015 Share Posted January 25, 2015 (edited) When it comes to audio power amplifiers the case is exacerbated by the speaker itself which is a non-linear load being applied, while a randomly changing signal (music) is applied to the input. Since "frequency" is a circuit variable throughout the entire amplifier including it's output, the non-linear speaker load means each design of amplifier will interact slightly differently with this load (depending on the input source) and thus make different sounds. Loved the old quote, still on the money. Regarding this part, I would add that these load dependent freq response variations are forms of linear distortion, and typically on the order of a db or two even for high output impedance SETs paired with widely varying impedance speakers. Sure, it's audible, wide band changes generally are that way, but it's not objectionable non-linear distortion, and could even be compensated via eq upstream. Here's a graph of a Primaluna PP amp with a high output impedance: Here's one for a Pass Labs amp with a very low output impedance: The first amp is an EL34 tube amp, the second one uses MOSFETs, IIRC. In general, tube amps have a higher output impedance, but there are exceptions to the general rule. Both graphs were run using Stereophile's simulated loudspeaker load. NOTE: The first two graphs are the Primaluna, the last one is the Pass Labs. This was my first attempt at posting Pics, don't know how to delete a picture. Edited January 25, 2015 by Don Richard Quote Link to comment Share on other sites More sharing options...
Coytee Posted January 26, 2015 Share Posted January 26, 2015 Hey Joe, Maybe this will help. Amplifiers take small voltage signals and make them bigger and more powerful by automatically multiplying the input signal voltage and current. Different amp designs have different amounts of multiplication - which is called gain. They also have a factor called sensitivity, which is a measure of how much voltage it takes at the input to make them go to full power at the output. A sensitivity of 2V means that a 2V input signal will cause the amp to put out full rated power. There are no real standards for this. Some amps have a 0.5V sensitivity, some are 3V, most are around 2V for power amps. Integrated amps, which have a preamp built in, will have lower sensitivity. The knob at the input panel is there to limit the input voltage. Suppose your CDP puts out a 2V signal, and your amp has a 2V sensitivity. It would mean that the CDP would always drive the amp to full power. You don't want that. So, the knob is connected to a voltage divider, called a potentiometer. It divides the input voltage so that some of it is wasted into ground, and some is passed on to the amplifier. When the knob is very low, let's say 1.8V goes to ground and 0.2V goes to the amp and produces 5W of output. When the knob is turned higher, 1V goes to ground and 1V goes to the amp, and now you have 30W of power. When the knob is all the way up, 0V goes to ground and 2V goes to the amp producing full power - however much that is. By dividing up the input voltage, you can control how much power the amp puts out. Usually, that knob is located on the preamp. So, the preamp determines how much power the amp puts out. But, that knob can also be put on the amp too. A point about the knobs. The potentiometers behind those knobs are not usually linear. Linear would mean half a turn creates half the power. These pots are usually shaped so that it takes about 90% of the turn to create 50% of the power. It's called log taper. It allows much finer control of the lower power levels where most listening is done and people want to be very precise. Now, if you combine the taper of the pot, with the gain of the amp, and the sensitivity of the amp, you can see why 50% rotation on one amp might blast you, while 75% on some other amp, might not. You have three variables, and each designer may implement them differently. There's no standard. Mdeneen, that is the most lucid explanation I have ever read, you should be writing books........... Write books or perhaps take up engineering.... Quote Link to comment Share on other sites More sharing options...
Don Richard Posted January 26, 2015 Share Posted January 26, 2015 And the lesson is this: next time you decide to listen to a simulated loudspeaker load, do so with a Nelson Pass amplifier. It will produce a smoother curve. Wrong lesson. What is significant is that the higher the output impedance of an amplifier the more closely it's output tracks the loudspeaker's impedance curve. Generally, tube amps have higher output impedances than solid state amps, but not always. For example, ARC's Reference 75 looks more like the Pass Lab amp's graph than it looks like the Primaluna's. 1 Quote Link to comment Share on other sites More sharing options...
WMcD Posted January 26, 2015 Share Posted January 26, 2015 (edited) If it helps: The amplifier is not a magnifying glass. The British call vacuum tubes, "valves." That is an accurate description of transistors too. The analogy is a supply of water at pressure which goes though a water valve, and then flows out. You control the flow rate by twisting the handle. In a very simple analogy, the grid of a vacuum tube or the base of a transistor, or the gate of an FET, is the handle (this depends on the topology of the amp but it is generally true). The input signal twists the handle, or at least allow current to flow from the amplifier's power supply to the output and to the speaker. This does mean that you are listening to the power supply as the valve allows the flow to change in accordance with the input signal. This means that the output voltage can never be higher than the max voltage of the amp's power supply. But generally we are reducing the input level (twist amount) with the volume control. But in any event, when the input level is low (say in quiet passages), the valve is not turned on very much. So the output is not loud. This might address a misconception. If an amp is capable of delivering 40 watts, it almost never delivering 40 watts. Certainly when the recording has silence, the amp is putting out zero watts. When the music is soft, the amp may be delivering, for example 0.1 watts. You can look at this in "real time" which I the output level versus time, on an oscilloscope. If you input a sine wave, and use the volume control to reduce the reduction (increase) the sine wave as a whole gets larger in voltage, but in a perfect world, it will scale up, perfectly. We can run up the volume control but eventually, we are trying to output a sine wave which has a peak larger than the peak voltage from the power supply. The top of the sine wave levels out or "clips." This distortion of the top of the sine wave introduces additional frequencies. This is a little difficult to imaging but it is because the output "sine" is no longer a sine, but a sine which is warped. The warpedness looks like other sine waves which have -- surprise -- other frequencies, usually multiples (or harmonics) of the original sine. So that clipping is distortion version of the input, even if the valve is otherwise perfect at scaling up the signal. But, valves are not perfectly linear in the scaling up. If linear the waveform would be perfectly scaled up. Different devices and designs have different imperfections. There are similar problems when the current (rather than voltage) available from the power supply can not keep up. This is typically when the resistance of the speaker is low and the given voltage would like to supply current but the power supply can not. Remember if we have a pure resistance in the speaker, voltage across it and current through it, scale perfectly. This is what Ohm's law describes when we write it in the form of resistance = volt / amps. If we have an 8 ohm resistor and apply 8 volts, the current is going to be 1 amp. If we apply 4 volts, there will be 1/2 amp. We're used to that calculation in d.c. but it works for a.c. too. If you have the proper equipment you can see an 8 volt peak sine wave applied across the resistor, and a 1 amp sine wave of current through the resistor. Voltage and current are perfectly scaled versions of each other. There is no distortion. Nonetheless, if we increase the input, it is the power supply which can't produce voltage and current in a scaled fashion, and we get a distorted sine wave in the output of the amp. WMcD Edited January 26, 2015 by William F. Gil McDermott 2 Quote Link to comment Share on other sites More sharing options...
muel Posted January 26, 2015 Share Posted January 26, 2015 ...Great, I just tried a SS amp and the grain / distortion remains so I am guessing it is either the dac breaking up or the source... We kind of touched on this grain/distortion in another thread. I'm thinking you have a decent ear and you are hearing a weakness of many DAC's that I've used. To a degree most of them can't seem to completely hold it all together when playing complicated passages. I don't know what this is or why but I know it when I hear it. When I replaced the V-dac with the Grant Fidelity with the HRT with the DacMagic with the Havana with the Bel Canto I finally got away from that "touch of grain" sound. All of these (among others) have it to a small degree at least but the Bel Canto is sublime as is the Ayre QB-9. It would be nice to know of a DAC around a 1k price point that could accomplish this. The DacMagic probably came the closest of the sub 1k dacs but also had a tiny sibilance. Sorry to detract from the amp discussion. Very interesting Quote Link to comment Share on other sites More sharing options...
russ69 Posted January 26, 2015 Share Posted January 26, 2015 I thought the volume knob was the smoke controller....live and learn. Quote Link to comment Share on other sites More sharing options...
derrickdj1 Posted January 26, 2015 Share Posted January 26, 2015 Great information, Thanks Mark and all So let me ask this, At times on more complicated passages of music I will get some grain / poor sound, and at the same volume on slower say acoustic passages it will sound sublime. Now I dont think the amount of music in a passage can affect how the amp performs but I am by no means a professional so why not ask one (No such thing as a dumb ?, or so I have been told) Most likely what is happening is that if you adjust your volume for average musical signal a big dynamic peak will drive your amp into clipping. You must be aware that the dynamics in music can make huge demands on the amp. If you have a 50W amp, and you are playing it at 10W for average musical passages, it will be clipping during even modestly dynamic passages. So one of the important parameters to KNOW about my amp would be the Sensitivity and then match that with say my Pre-amp In other words if my amp is 2 V, Try and find a pre that has an output of 2V............Correct ?? The preamp should be AT LEAST 2V. Yes. More is no problem. Now I must assume my integrated was designed in a way that the pre section matches (or closely matches) my amp section, what about source materials into an integrated, does it follow the same basic principle, IE... have source output closely match integrated input sensitivity ??? It rarely matters what the source output is. Your preamp will have enough gain to make the signal big enough. The exception is "passive preamps" (an oxymoron, actually) which have zero gain. When I setup my preamp, I have two gain decisions to make. First is the individual channel settings. Second is the master volume setting. What do you recommend as the best setting for the individual channel setting to avoid clipping? An amp of adquate power and reserve is the best approach. Transient clipping is not a problem for the most part. Prolong clipping can damage the speaker. A lot of tube amp will deliver the same power at 8 or 4 ohms. This is not the case with SS amps. Knowing your gear and it's limitation is a key point in having an enjoyable experience and protecting your gear. Quote Link to comment Share on other sites More sharing options...
seanwalsh Posted January 26, 2015 Share Posted January 26, 2015 I have hotrodded La Scalas and most of the time I prefer my ASL Tulip 2A3. I use 360 on the Sylvania 45 but 630 bias sounds best on 2a3c Shuguang. What is the effect of various bias points on set amps? Quote Link to comment Share on other sites More sharing options...
joessportster Posted January 26, 2015 Author Share Posted January 26, 2015 ...Great, I just tried a SS amp and the grain / distortion remains so I am guessing it is either the dac breaking up or the source... We kind of touched on this grain/distortion in another thread. I'm thinking you have a decent ear and you are hearing a weakness of many DAC's that I've used. To a degree most of them can't seem to completely hold it all together when playing complicated passages. I don't know what this is or why but I know it when I hear it. When I replaced the V-dac with the Grant Fidelity with the HRT with the DacMagic with the Havana with the Bel Canto I finally got away from that "touch of grain" sound. All of these (among others) have it to a small degree at least but the Bel Canto is sublime as is the Ayre QB-9. It would be nice to know of a DAC around a 1k price point that could accomplish this. The DacMagic probably came the closest of the sub 1k dacs but also had a tiny sibilance. Sorry to detract from the amp discussion. Very interesting Hey Muel, I know it may be the Dac but I am still not convinced, I feel loke it may lie elsewhere replacing tubes seemed to make for better synergy and some of the grain is better. much more listenable, I need to try my reference for this phenom from other sources and see if it resides as part of the source, Its weird because some higher energy music does fine. If I ever get my injury settled a Bel Canto is on my A list. There are some great explanations on here that are easy to understand Quote Link to comment Share on other sites More sharing options...
Schu Posted January 26, 2015 Share Posted January 26, 2015 (edited) One of the main gains I've experienced in going with a better dac was a significant reduction in granularity... even between dacs that used the same chipset... More liquidity and reduced granularity. I am not saying this is your issue, it may be a combination... but I bet it wouldn't hurt either. Dac tech has moved quickly over the last year or two with chips introduced and timing refined, I am not sure how much further it can go. Edited January 26, 2015 by Schu Quote Link to comment Share on other sites More sharing options...
NOSValves Posted January 26, 2015 Share Posted January 26, 2015 And the lesson is this: next time you decide to listen to a simulated loudspeaker load, do so with a Nelson Pass amplifier. It will produce a smoother curve. Wrong lesson. What is significant is that the higher the output impedance of an amplifier the more closely it's output tracks the loudspeaker's impedance curve. Generally, tube amps have higher output impedances than solid state amps, but not always. For example, ARC's Reference 75 looks more like the Pass Lab amp's graph than it looks like the Primaluna's. To me the lesson is the Primaluna is not nearly as good a tube amp as Up Scale audio has paid in advertising to portray it as.... 1 Quote Link to comment Share on other sites More sharing options...
seanwalsh Posted January 26, 2015 Share Posted January 26, 2015 ASL recommends 600 for the 2a3 and online 360 for the 45 tube along with dropping down a transformer tap. I believe there is an extra decimal place in the digital readout provided, I am amazed that this ancient technology even makes sound let alone become preferred over modern designs. Thank you for the concise explanation. Quote Link to comment Share on other sites More sharing options...
joessportster Posted January 26, 2015 Author Share Posted January 26, 2015 One of the main gains I've experienced in going with a better dac was a significant reduction in granularity... even between dacs that used the same chipset... More liquidity and reduced granularity. I am not saying this is your issue, it may be a combination... but I bet it wouldn't hurt either. Dac tech has moved quickly over the last year or two with chips introduced and timing refined, I am not sure how much further it can go. I understand unfortunately any dac upgrades will be in the distant future rather than near Quote Link to comment Share on other sites More sharing options...
joessportster Posted January 26, 2015 Author Share Posted January 26, 2015 One of the main gains I've experienced in going with a better dac was a significant reduction in granularity... even between dacs that used the same chipset... More liquidity and reduced granularity. I am not saying this is your issue, it may be a combination... but I bet it wouldn't hurt either. Dac tech has moved quickly over the last year or two with chips introduced and timing refined, I am not sure how much further it can go. I understand unfortunately any dac upgrades will be in the distant future rather than near Guess I will suffer with my Havana (wish all suffering were this good) Quote Link to comment Share on other sites More sharing options...
Schu Posted January 26, 2015 Share Posted January 26, 2015 suffering so good is a curse of the well informed and knowledgable... walking blindly is bliss. 1 Quote Link to comment Share on other sites More sharing options...
tube fanatic Posted January 26, 2015 Share Posted January 26, 2015 Let's suppose our triode has a maximum power of 25W. We don't want to use the whole 25W for BIAS. We want to bias the tube so that the tube is carrying 25W when we have applied our full signal. Otherwise, all we have is a 25W heater. So we must choose a BIAS that keeps the tube in the linear transfer function without exceeding the power capability when full signal is applied. Mark, if I understand your point correctly, I don't fully agree. If the tube has a maximum plate dissipation of 25 watts, obviously we don't want to exceed that under zero signal conditions. However, if the load line swings a bit above that with signal applied it should not be an issue since the power is being transferred to the load and is not being dissipated in the tube itself. I can draw out a representation tomorrow if you want to see what I'm saying graphically. In actuality, this occurs all the time with a speaker load when the impedance goes way up with frequency. If the bias point is brought down far enough to avoid this, the distortion is often too high. Love these technical discussions! I agree with the forum member who suggested that we have a dedicated tube amp section. With the growing interest in them I think the participation would be very good. Maynard Quote Link to comment Share on other sites More sharing options...
CECAA850 Posted January 26, 2015 Share Posted January 26, 2015 Typing on a tablet makes me try to conserve words. You'd never know it 2 Quote Link to comment Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.