FritzS
Been here a while!
Sound of Blue Danube
Sound of Blue Danube
Posts: 1,364
|
Post by FritzS on May 2, 2006 10:50:48 GMT
Dear Mike, really I miss the WNA discussions with you I hope a follower comes back as "phoenix out of ash"! But no-one reply to my suggestion for modification. I think with an J-FET input OP and small modifications it can comes back as a great headphone amp! Another idea I have for the output DC protection - it's a symbiosis from some DIY projects. I will post this ASAP. My experience is a J-FET input OP sounds better than a bipolar input OP - AnalogDevice OP's sounds better than BurrBrown - the AD OP275 sounds better than NE5532N or OPA2134 I hope for feedback
|
|
rickcr42
Fully Modded
Rest in peace my good friend.
Posts: 4,514
|
Post by rickcr42 on May 2, 2006 17:16:49 GMT
I have been using the OP275 for years and I noticed this chip is a love it or hate it part with no nuetral fence sitters.Better was the single opamp version,the OP176 but this has been out of production for years as has (I think) the "clone" SSM chip (SSM is the Analog Devices designator code for "PRO Audio" parts). Will operate at low voltages though better at the full +/- 15VDC and really does need a dicrete "current pump" or monolithic buffer on the output even though alone it specs fairly high in the current delivery area. no speacial bypassing requirements,stable in just about any well layed out circuit (and even poorly layed out circuits ;D ),dynamic,good bass ...... All in all a very civilised easy to use part that is close to "plug-and-play" replacement for any dual opamp using a standard pinout but like I said,hated by many while loved by others. The real strength of a bipolar input opamp is the first stage current delivery to the next stage and input noise levels which better JFET opamps.not as important at low gains but move into the 30-60dB gain range and you will wish you had a bipolar opamp Even better is a discrete Dual JFET front end to a bipolar opamp.The easy way to squeeze one in to a circuit that has zero room on board is to remove the input cap,make a small perf board with the new input circuit which will now have the input cap on board then use the pad positions formaerly occupied by the cap as the in/out to the new "front end" board. A dual jfet is tiny,the size of an 8-Pin DIP opamp so even when you add the rest of the parts you can get the board footprint into the 2x2 inch area for a stereo board-------just another option
|
|
rickcr42
Fully Modded
Rest in peace my good friend.
Posts: 4,514
|
Post by rickcr42 on May 3, 2006 17:07:47 GMT
So how long will this sub forum be able to stay topical Mike ?
seems like with WNA closing the doors and you no longer supporting the products at your previous enthusiastic level the amps will become an asterisk,an "also ran" in the not too distant future.
I must also say it is kind of strange to design then announce a "WNA lll" will be offered then when interest is peaked pretty much say "never mind".
Is there a chance in hell of maybe just purchasing/weasling/begging the rights to the basic amp architecture and pc board layouts ?
that way there is continuity and no one that bought in is left hanging AND would add value to the amps on the used equipment market.
Otherwise two years down the road any attempts to sell one of the amps would bring a "WNA ? Who and what in the hell is that ? anyone ever hear of this obscure amp ?"
|
|
rickcr42
Fully Modded
Rest in peace my good friend.
Posts: 4,514
|
Post by rickcr42 on May 3, 2006 18:53:09 GMT
seems best really otherwise you would have to add a sub forum for every short run amp on the planet and that just does not work.General DIY gets my vote
|
|
|
Post by PinkFloyd on May 3, 2006 19:50:58 GMT
seems best really otherwise you would have to add a sub forum for every short run amp on the planet and that just does not work.General DIY gets my vote OK, will let this forum run for a while and gradually move any technical / interesting parts of it over to the DIY section.
|
|
FritzS
Been here a while!
Sound of Blue Danube
Sound of Blue Danube
Posts: 1,364
|
Post by FritzS on May 4, 2006 7:40:30 GMT
Dear Mike, Rick! I think we should begin a new "Open Source" (!!!) project with name "Phoenix Headphone Amp" and create a new headphone amp reconstructing on WNA and you could rename this forum to this name I will help you to create a new schema - inklusiv a output DC protection and PSU. We could gathering ideas till autumn, winter 2006/07 and then we design a new PCB including all. I think the basics from WNA schema are common knowledge - the most ideas comes from Walt Jung. I found similar ideas in other projects too!!! So we did not have a problem with copyright. I have a subscription from audioXpress too an can send you a pullout-copy of a headphone amp project to your private mail-address (if you tell which).
|
|
rickcr42
Fully Modded
Rest in peace my good friend.
Posts: 4,514
|
Post by rickcr42 on May 4, 2006 8:10:52 GMT
most ciruits are just that-standard designs,but it is the combination of implementation and layout that makes them uniquely someones baby.
My thoughts are if you want to do an amp that step one should be to go "outside the box" and go straight to a true bipolar power supply of +/-24 to +/-30 VDC 500mA per polarity,use a "bootstrap mode" bipolar opamp (AD826,OP37,AD797) with a discrete matched jfet front end (or not) and a bipolar class-A current pump output.
The high PS voltage along with a bootstrap mode opamp means dynamics will be eye popping and the use of a bipolar opamp (not jfet) with a vanishingly low noise floor means you can up the gain at will to crazy levels and still have a huge drive capability in both the voltage (something around 9 volts audio signal level) and current (100-150 ma class-A sliding to AB for higher current peaks AS IF ! )
Be more like the Hydra than the Pheonix,pure kill power
....just thinking out loud of course ;D
|
|
rickcr42
Fully Modded
Rest in peace my good friend.
Posts: 4,514
|
Post by rickcr42 on May 4, 2006 8:20:21 GMT
|
|
FritzS
Been here a while!
Sound of Blue Danube
Sound of Blue Danube
Posts: 1,364
|
Post by FritzS on May 4, 2006 9:54:39 GMT
both are not really the same - the article from EDN is a true bootstrap but the TNT is a bit other - here is the VCO/VEO fixed. The TNT principle is common used in some HiFi amps far back. I know this principle - I remember it was used with solid state transistors in Threshold power amps .... But both have a drawback - the highspeed OP amps (as used in WNA) will possibly oscillate
|
|
xerxes
Been here a while!
Posts: 1,115
|
Post by xerxes on May 4, 2006 16:31:43 GMT
Between you, you should be able to come up with a killer amp, so get on with it! You'll have at least one customer.
|
|
rickcr42
Fully Modded
Rest in peace my good friend.
Posts: 4,514
|
Post by rickcr42 on May 4, 2006 18:09:22 GMT
for illustration of possibilities only.both ways get you serious voltage drive,maybe enough to even attack the 600-1K ohm AKG cans without going to a "true" power amp or pentodes.Obviously such would require a hig/med/low gain switch if it were to be a universal "any cans" amp or you would lose all possible control of the volume pot and if a potentiometer you always want to keep a dual pot in the "sweet spot" of its range or your inter-cannel tracking will suck big time. Then again,if the output is to the old "standard headphone amp load" there would be 120 ohm resistors inline with the output which would in a way "act" as a gain adjuster by pinching off power to low impedance loads while passing it to high impedance loads ;D And why I purposely use slow opamps when I must use opamps.There truly is not much need for anything over the mid to upper twenties (or at best mid sixties) at audio frequencies which more forten than not can get a person into trouble unless steps are taken to compensate. These ultra-high-speed opamps operate right at the edge of being their own radio transmitter and no matter where RFI enters (power supply leads,either input node,output hence to the feedback loop) it WILL be amplified as if a signal (it is) and mix with the audio frequencies. Where this becomes a serious problem is when the event is not detected (scope does not have a high enough range) and the amp is colored by this event. The reasons for the event being an unknown a reasonable person will usually look to the resistors or caps as the fix thinking better parts will make the amp sound better while all along that "brittle/etched" sound they thought to be detail when the amp was purchased,and what is now causing headaches,is in reality oscillations so high in frequency they are not the dominant sound while still imparting their own tonal "stamp" on everything passing through the stage. I like slow opamps (when I use them, a discrete/valve type person),filtered to the balls on every single potential RFI entry point,TRUE dual power supplies (followed by shunt regs if the part has a high enough PSRR) a fast STOP (settling time) rather than start (rise time) and to be honest bipolar transistor input opamps that idle at a high current (compared to the new breed of class B idlers) Radical/old school would be OP37 bootstraped to a SEPP Class-A output stage AC coupled or with an ultra-low turnover DC servo as an option to lose the cap for the capacitor shy (I HATE servoes but they have a place). No parallel devices,no fet input,no split power buss,no fancy parts,a low impedance feedback network with feedback loop DC shunting (electrolytic to ground in the loop),RF compensation in the loop,at the input and output plus both ps pins......... that or just build a damn Jensen JT-990 discrete opamp (open source design,free to use) with a beefier output stage and call it a day
|
|
FritzS
Been here a while!
Sound of Blue Danube
Sound of Blue Danube
Posts: 1,364
|
Post by FritzS on May 7, 2006 13:14:08 GMT
Where I can found the schema - with google I found only threads I disagree - WNA use high speed OP's - Jan Meier use them too - all amps sounds well I post the modifications I will make on my homepage: www.stockhammer.at/hifi/wna.php#Modification1
|
|
rickcr42
Fully Modded
Rest in peace my good friend.
Posts: 4,514
|
Post by rickcr42 on May 7, 2006 16:51:52 GMT
Jensen Transformers website under "AES Papers" send them an email and they will send out the papers www.jensen-transformers.com/apps_wp.htmlWhile you are at it the other Deane Jensen AES papers are also a very good and very illuminating read,also worthy of a request. the beauty of the JT-990 papers is not just that anyone can build and actually sell identical copies of the design (true public domain) but there is enough blow by blow content of the actual design to go off and do your own version armed with the information. never said it can't sound good,only that additional steps are required to make it so and being a K.I.S.S. type (keep it simple stupid) I would rather strart from a foundation of stability then build off it than I would start out making the foundation work THEN try to build of of that. I regularly use the AD797,OP27/OP37 ands AD744 with output section bypassed and it is a rare day I go with anything over 60V/uS in opamp designs. Why ? Because each succeeding stage needs to be faster than the previous stage or you introduce slewing distortions and that means you power suppply needs also be much faster than the actual amplifying device. This last is easy in a shunt regulator (no in series with the voltage) but a shunt reg can have noise issues many times so off the list for high gain duty (mic/phono/tape head) and will stop regulating if the voltage output exceeds the shunt regulation element voltage (can actually be a good thing ). that the fast opamps can work and can sound fine does not mean additional steps to make them behave has not taken place (in fact says it has) but in the end makes for a more complicated design that in real world use is no faster than the slower device due to the compensation methods. Mate the fast front end to an output section not quite as fast or marginally faster and you introduced MORE problems.Follow this with a slow amplifier (opamp as preamp box) and you need to introduce filters at the upper end on the power amp or you introduce S.I.D. there ! (and why the final stage of a power amp is always the fastest part of the design and uses power transistors operating in the MHz range) So slow and easy is my way though I realise it is not new millenium PC or considered SOTA which many times is just folks trying to blaze their own path to be different and not because it betters what has gone before ;D
|
|
FritzS
Been here a while!
Sound of Blue Danube
Sound of Blue Danube
Posts: 1,364
|
Post by FritzS on May 8, 2006 7:26:11 GMT
I have been using the OP275 for years and I noticed this chip is a love it or hate it part with no nuetral fence sitters.Better was the single opamp version,the OP176 but this has been out of production for years as has (I think) the "clone" SSM chip (SSM is the Analog Devices designator code for "PRO Audio" parts). The OP275 is my favorite - but I did not found a similar single OP in DIL8 - the OP275 sounds very analytic, detailed, down to the last details - so I had never heared my old Philips CD650 before ;D They have the reputation to sound a little dark - but I had not test and heared them ...... I agree with K.I.S.S. (the WNA is it too, or?) - it gives many of headphone amp projects - I had collected some .... www.stockhammer.at/hifi/hifi-audio-tubes.php#HeadphoneAmpsPS: I see the PPA v2 with discrete buffers is similar to my suggestion (of WNA) but use more parts elvencraft.com/ppav2/laroccoaudio.net/discretebuffer.htm
|
|
rickcr42
Fully Modded
Rest in peace my good friend.
Posts: 4,514
|
Post by rickcr42 on May 8, 2006 14:56:56 GMT
just did a fking twenty minute EXTENSIVE response and when I hit "post reply" got the damn white page,hit "back" and all gone !!!!!! Too pissed off to even consider trying to reconstruct my thoughts and posting a response.Maybe later,maybe never,likely the latter since I am not in the habit of repeating myself unless to drive home a point and this is just not that important
|
|
FritzS
Been here a while!
Sound of Blue Danube
Sound of Blue Danube
Posts: 1,364
|
Post by FritzS on May 8, 2006 15:17:54 GMT
Some days ago and NOW I had the same problem Add a smilie - if it don't - quick save the message in a notepad file for later use
|
|
rickcr42
Fully Modded
Rest in peace my good friend.
Posts: 4,514
|
Post by rickcr42 on May 8, 2006 15:50:38 GMT
I was in the habit of always doing my responses in notepad in another forum for this very reason but since it has never happened here before I got totally Pear Harbored (sneak attack) on this was and was not prepared at all. BUT !The old "screw me once shame on you screw me twice shame on ME" expression means the rickmonster is ready for the bastards on Round-ll and it will NOT happen again on my watch ! i may even try to recronstruct my screwy thoughts/post later when i calm down a bit more and find that patience thing so many talk about but has so far eluded me in life ;D
|
|
rickcr42
Fully Modded
Rest in peace my good friend.
Posts: 4,514
|
Post by rickcr42 on May 8, 2006 16:49:55 GMT
___________________________________________________ PART l:
The Op-Amps-"One mans opinion"___________________________________________________ The AD797www.physics.ucdavis.edu/Classes/Physics116/P116C_lab/AD797.pdfThis really is all the Op-amp most will ever need at gains of x10 or more (have used it all the way to +60dB gain !!!) but it needs a low noise power source or you are pissing away the superior noise floor of the chip.The 50ma output drive says even as a long line driver it is hard to beat with a stick without any additional help (discrete stage) though for real kill power/dynamics with low-Z cans a discrete class-A stage is the icing on the cake. Want a warmer tone ? Use MosfetsThe 50 ma current out means it has enough muscle to overcome the gate capacitance/drive requirements of the mosfets which needs to be low impedance and why so many screw up on the matchup (a 300-600 ohm series opamp load/mosfet gate resistor is ideal). Notice the temp stability spec,the input noise spec at AUDIO relevant frequencies and slew rate is only 20V/uS.This is a prime time audio part and it even has a DC "null" connection,something missing from 99% of most "modern" parts that claim to be a step forward The OP37www.mit.edu/~6.301/OP37b.pdfanother stellar performer in audio circuits.This baby is maybe a better choice in the "common for headphone amps" gain of x5 (the AD797 is unity gain stable,the OP37 is not). again the noise level is measured against AUDIO frequencies unlike the modern "superchips" that meaure in the mhz range but you would be hard pressed to find ANY information on frequencies that are relevant for our purposes and there many times is a reason for this-they suck ! again we have the option for a DC balance control,again we have what appears to be a lacking slew rate at 17V/uS but which is more than good for the purpose,again we have a low freq noise corner,good temp stability and again a bipolar (not jfet) input topology. This baby is OLD and dates all the way back to the pre-ADI days of PMI as does....... The OP27www.mit.edu/~6.301/OP27c.pdfPretty much the same part as the OP37 with a couple of significant differences : the slew rate is only 2.8V/uS folks,+/- 10V into 600 ohm drive is no slouch ! and if you think the slew rate is a limit on the sonics you need to be aware that most triode amps,amps that done right will kick the shit out of a solid state equivelant slew in the very same range ! If this is an area of concern then just limit the use to microphone/tape head/phono pickup preamplifier duty where the input signal will be in the microvolt range anyway. what do all the above have in common ? 1-slow as f**k 2-great input noise specs for audibly significant frequencies 3-they like to be run right at the high range of the power supply spec (=/-16.5 VDC) 4-have a DC balance 5-stable over a wide range of temps 6-bipolar input (only way to true low noise) 7-all settle fast which means increased low level detail 8-all sound good in the "classic" sense as opposed to the modern "artificially etched highs" most seem to favor at first but find tiring over time and why so many are always looking for the next amp. Add a discrete Class-A output stage to any of the above and you have an amp that will hold up over time.The OP27/37 rule at voltage swing,the AD797 adds enough current for most apps,all can be helped with either a "classic" buffer (not CFB amp but open loop unity gain monolithic buffer in the LH0033 mold or LT1010 in pure class-A up to 70mA or so) The AD744www.analog.com/UploadedFiles/Data_Sheets/871938516AD744_c.pdfyou want jfet inputs ? You got it ! you looking for fast settling times ? In spades ! Unity gain stable ? No problem ! Internal or external compensation,good Ac performance (audio is an AC signal),PLENTY fast for the speed whores,low noise,DC blance pins,has it all. Weaknesses ? Ouput current delivery. So that means stand alone use is off the list though a close call for driving a proper output stage.So why is this part so "cool" ? look at fig.28 you see Pin #5 ? That is a total output section bypass ! That means you can add whatever output stage you want while using the front end of the AD744 purely as a gain stage but by being on a single die with far better P/N device matching than we could do and WAAAAY better temperature tracking so the parts stay in sync over a wider range for better CMMR and PSSR than any dicrete opamp. Consider it a preamp on a chip waiting for an output section Part #2 will take a look at my personal opinion of powering these bastards.Not the last word but just an opinion from an old tinkerer of things that make a racket ;D .
|
|
FritzS
Been here a while!
Sound of Blue Danube
Sound of Blue Danube
Posts: 1,364
|
Post by FritzS on May 8, 2006 18:29:23 GMT
My Stock of OP AMP's NS - LM6171, LM6172, LM6181, BB / TI - OPA655P, OPA134, OPA2132, OPA2227, OPA2228, OPA627, OPA637, OPA2134, OPA227, OPA228, OPA604, OPA2604, BUF634P, BUF634T, DRV134 AD - AD711, AD811, AD797, AD9631, AD9632, AD812, AD823, AD8610, AD797, AD843, AD847, AD8065, OP275 PMI - OP279, OP113, SSM2135 ON Semiconductor - SE5532, SA5534 / SE5534 Maxim - MAX410, MAX412, MAX427, MAX437 LT - LT1364CN8, LT1365CN, LT1363CN8, LT1360CN8
|
|
|
Post by spendorspain on May 8, 2006 18:39:47 GMT
Hi
I've just installed in my WNA MkII the modified AD744 op-amp, with the output stage bypassed (output taken from pin 5, not 6). The mod is very easy: solder in a 8-pin DIL socket a bridge between pins 5 and 6 and bend slightly the pin 6 of op-amp to avoid any contact with the socket (I've also isolated it with fine heatshrink). Thus, the output from op-amp pin 5 goes through the bridge to the socket pin 6, and the modified assembly can be used like any normal op-amp.
Now the AD744 is in "IC1" position instead of LM6171 and the sound is very good, even with the output caps needed with LM6171 and without any burn-in time. The detail is similar (and without doubt it will improve as it doesn't need any output cap in signal path) although it seems a little more distant sounding (which I prefer). The most improvement is, to my ears, a bass much more tuneful and detailed.
Using output caps, I've not yet measured the possible DC offset, but there is absolutely no thumps when powering on and off (with AD8065 there was a loud and frightening thump at turn-on and even more at turn-off).
As listening goes on, I´ll post more impressions, but it looks very promising.
Regards Jose
PS.: I'd like to attach some photos, but I don't know how... this is my first post here. How can this be done?
|
|
FritzS
Been here a while!
Sound of Blue Danube
Sound of Blue Danube
Posts: 1,364
|
Post by FritzS on May 8, 2006 19:00:37 GMT
I've just installed in my WNA MkII the modified AD744 op-amp, with the output stage bypassed (output taken from pin 5, not 6). The mod is very easy: solder in a 8-pin DIL socket a bridge between pins 5 and 6 and bend slightly the pin 6 of op-amp to avoid any contact with the socket (I've also isolated it with fine heatshrink). Thus, the output from op-amp pin 5 goes through the bridge to the socket pin 6, and the modified assembly can be used like any normal op-amp. How is the sound with no modifications - only put the AD744 op-amp in the socket instead of the LM6171? I will test AD843 and AD8065 too
|
|
rickcr42
Fully Modded
Rest in peace my good friend.
Posts: 4,514
|
Post by rickcr42 on May 8, 2006 19:05:50 GMT
one time when the data sheets actually give some clue as to "why". Fig.6,Fig.16 and Fig.17 from the data sheet.Low frequency distortion is a tough nut many opamps can not crack and why they many times sound leaned out even though the measurements "say' you are flat to below 20 hz. The ability to swing large voltage with low distortion in the lower regions is not usually an opamp strength I personally consider that more accurate and the other "forwrd" which translates to most of the power being in the mid to upper octaves as would be expected for an opamp designed to operate in the mhz range (most modern chips) If you have a website it is a simple matter of right clicking the pictures,choosing properties then "highlighting" the web address found there,right click "copy" then past here,"highlight" the address again,go to the top (above the emoticons) when you post and click "insert image" which will add the meta tags to the picture. If you have no web page just park the pictures at one of the free hosting sites and follow the above instructions ************************************************** someone whack that guy ! .
|
|
rickcr42
Fully Modded
Rest in peace my good friend.
Posts: 4,514
|
Post by rickcr42 on May 8, 2006 20:22:02 GMT
____________________________________________________ PART ll:
Powering the Bastards while maintaining the essence of the active ciruitry sonics____________________________________________________ First off,this page needs to be opened and all files read to get an understanding of just how audible the power supply is.It is also very cool to see the prosses span thirty one years ! waltjung.org/Regs.htmlNow for my personal opinion I like the shunt regulator for all but the most noise sensitive circuits and that means anything with a gain over X5 though even here i sometimes go off the ranch and into "outlaw" territory just because I can MY reasons are power supply bandwidth and a "less is more" general philosophy. Because a shunt regulator is not asctually in the circuit path between the "raw" DC and the power input to the actual circuit being powered the bandwidth is limited only by the power transformer bandwidth with the "regulator" having ZERO effect other than to shunt excess voltages to ground (the regulation part).For the same reasons this regulator is lightening fast ! Because it is out of circuit mostly (by not being a series connected element but a shunt connection) it also has the ability to DO NOTHING if the voltage demands of the circuit during a dynamic passage of music. That means if the voltage requested is over and above what the regulator is set at it just stops regulating then allows this "requested" voltage to pass through to the circuit "asking" which in music terms translates into a dynamic voltage output with the speed of it being limited solely by the topology of the active circuit and NOT by the speed of the power supply which MUST be faster than the audio circuitry or you introduce slewing distortions no matter how fast the op-amp/circuit is (and why light speed opamps mated to a typical three terminal voltage regulator are not a good match). So my V-regs are pretty much there for a bit,not there for a bit,there again,not............................. The downside is they are not only not the most quiet regulator choice but are not the lowest in output impedance AND being in a shunt position means they will pass along any noise already on the power line unlike a "series" regulator which performs filtration/cancellation of this noise. I have personally found this to be more of a "measuring" problem than i have found it to be limiting in any way if SQ is the test until we start to get up in gain then it will be the limiting factor in how much self noise the overall circuit produces. Kinda defeats the purpose to use a seriously low noise part like an AD797 at a gain of X10 or better then mate it to a "noisy" shunt regulator but for low gains and/or buffer/follower circuits not just my first but only choice. I overbuild the AC/DC section to x2 the peak current delivery requirements of the circuit worst case then filter the fk out of both sides of the transformer (primary and secondary),add slow start resistors inline with the first stage capacitor so the rectifier is not stressed at startup then use either CLC or CRC filtration (again bypassed,now up to several poles of RFI bypassing ) leading to the shunt reg itself which in turn is followed by the final "storage" cap which in turn is decoupled to both channels and the final filters right at the active devices a damn lot of capaitors and resistors for such a simple regulator but worth it to me in the end when the music flows unrestrained and noise free.no "uh ohs" no "oopsies" no "gotchas" just free flowing music the way it was meant to be Series regulators ? Yup.I use 'em also and a LOT because of all the high gain stages I screw with but these babies I don't fk around with either and taking a page from Walt Jung go straight to ultra-high bandwidth ultra-low noise discrete Super-Regulators ! Three terminal "plug and play" monolithic regulator chips ? HAH ! No fkn way ! Not in MY audio gear ;D
|
|
FritzS
Been here a while!
Sound of Blue Danube
Sound of Blue Danube
Posts: 1,364
|
Post by FritzS on May 9, 2006 7:46:33 GMT
About the preferred use of HighSpeed OP-Amps ..... I think the speed (high upper frequency limit) of an OP amp cannot be high enough. The open loop gain of an OP is very high, with an high speed design of the OP the phase shift is lower too. I an real design the gain of the amp is very low (in circumstance to open loop) by use of a high NFB. To limit influences in sound the upper roll off frequency should be greater with a factor 10 to 100 or more than 20kHz. This at the highest output power. To avoid the real amp to send RF you should use an low pass filter at the input. If you use low speed amps with high NFB (and significant phase shift) some problems are act. This is one of the advantage of tubes - the preferred tube family E88CC (6DJ8 / 6922 / ECC88 / E88CC / CV2492 / etc. - www.tubedepot.com/69226dj8.html) are used as PCC88 in older VHF/UHF TV tuner till to 1 GHz. If you don't want high NFB you should use an designs looks like as Pass ZEN (http://www.passdiy.com/amps.htm). I want to go the way with higher speed OP amps - LM6171, AD797, AD843, AD8065, etc. - and higher fT transistors in the power stage. The ideal amp looks like "a straight wire with gain" - this is my goal.
|
|
FritzS
Been here a while!
Sound of Blue Danube
Sound of Blue Danube
Posts: 1,364
|
Post by FritzS on May 9, 2006 7:53:56 GMT
Why???
|
|