It occurred to preface a pair, or paralleled sets as a composite pair, to an OPA1632. The common-mode bias pin is a handy feature.
It looks like the OPA1611 might be a suitable sub for the 49990. The dual part, the 1612, would be fine except one would have to watch dissipations closely.
Last edited:
Yep, a few good'uns that are a bummer to see disappear. Will be interesting to see what new parts come online in the next few years (using Nationals analog processing IP).
I prefer the OPA1611 over the LME part. 2 main reasons, lower noise and better resistance to EMI/RFI on my test bench.
I done some rudimentary testing with OPA1612 and it is good, but it was beat on distortion by a lowly 5532 (in an AK4490 DAC). On paper the 1612 looks a lot better, but it might take some effort to get that performance out of it.
Some people have this knack of making money. From he cradle to the grave, they make it look easy. The rest of us just struggle and muddle through.
TI is shutting down the whole process and possibly the plant. Don't expect much from them. All the smart guys are gone.
Sent from my Nexus 5 using Tapatalk
Mass Flourishing
I think someone had second thoughts about a post, which showed up on my email notifications 🙂
In the meantime, I recommend a book review in the 19 November 2015 NY Review of Books by Benjamin Friedman, of Mass Flourishing: How Grassroots Innovation Created Jobs, Challenge, and Change, by the Nobelist Edmund Phelps. I confess I was surprised by the discerning, but mostly positive, review. It may be well worth getting the actual book.
It's not overtly political (unusual for NYRB).
I think someone had second thoughts about a post, which showed up on my email notifications 🙂
In the meantime, I recommend a book review in the 19 November 2015 NY Review of Books by Benjamin Friedman, of Mass Flourishing: How Grassroots Innovation Created Jobs, Challenge, and Change, by the Nobelist Edmund Phelps. I confess I was surprised by the discerning, but mostly positive, review. It may be well worth getting the actual book.
It's not overtly political (unusual for NYRB).
For "innovation", I suggest an audio manufacturer etablish a special purpose group which should in effect play around with new designs, to do a very important job which is strictly a try-and-see(hear) the difference. Over the yeras, I hhave heard a number of devices which were beatifully made and showed a degree of seriousness of the intent, and were chock full of prime quality parts, yieldeing no more than a slightly above average sound, and were easy to forget. So, it's not for want of trying and/or cost cutting, it's because it lacked the final stage as I see it.
By far the most interesting part of designing any circuit is the time that comes after it was made to wrok as required electrically, the gross mistakes had been taken out, the usual normal operating ptoblems were sorted out, and only prime quality parts had been used, Yet it fell short of the design objective and sounded just a bit above average. One of the ways to combat this problem is to simply try using other parts, even the unlikely ones, to see what happens. For years, I as many others orientated myself by buying parts which were supposed to be well neight perfect - and costly. A typical example would be Miller capacitance compensation caps. For a long time, the default type of capacitor used here were quality silver mica. I found that in many cases, styroflex would do just as much for a lot less money.
Ceramic caps were frowned upon. Their use was "allowed" only in FM tuners and similar HF devices. Then I literally stumbled on a series of multilayer ceracim caps from Siemens, with an outrageous price for a goddam ceramic cap. Until I replaced the suilver mica caps I was using with them, which left me wondering what the hell was going on, because I was getting a better sound. Not radically so, but audibly better, clearer and subjectively faster. Yet my measurements were essentially the same, off by say 0.01% for THD and IM.
It turned out that these were brand new technology caps just introduced by Siemens, and the prices were like 10 times (!!!) the prices of regular ceramic caps. This sort of thing happens because manufacturers are always introducing new parts of variants of existing parts and are not always loud when doing so.
Now, the fundametal designer of a device has no time to investigate such perhaps oddball variatons on his theme, and that's what the twek group should be doing. Initially, they will cost money, but if they are worth thir salt, they will turn good products into excellent one and could soon become a welcome option. I am, in fact, suggesting, what many tweakers do anyway, but I don't leave it up to them, I have my own group of tweakers who do it right off the bat. Experimeting should include even the not very likely groups of products, simply because you never know what will actualy work best in any particular design until you try it on a live model.
In essence, this is an extension of the product development process which should cover some ground the designer simply had no time for. And they should be freed from cost considerations the designer was bound by. Obviously, they should be able to prove to the designer that they really did some improvement before the product is released.
By far the most interesting part of designing any circuit is the time that comes after it was made to wrok as required electrically, the gross mistakes had been taken out, the usual normal operating ptoblems were sorted out, and only prime quality parts had been used, Yet it fell short of the design objective and sounded just a bit above average. One of the ways to combat this problem is to simply try using other parts, even the unlikely ones, to see what happens. For years, I as many others orientated myself by buying parts which were supposed to be well neight perfect - and costly. A typical example would be Miller capacitance compensation caps. For a long time, the default type of capacitor used here were quality silver mica. I found that in many cases, styroflex would do just as much for a lot less money.
Ceramic caps were frowned upon. Their use was "allowed" only in FM tuners and similar HF devices. Then I literally stumbled on a series of multilayer ceracim caps from Siemens, with an outrageous price for a goddam ceramic cap. Until I replaced the suilver mica caps I was using with them, which left me wondering what the hell was going on, because I was getting a better sound. Not radically so, but audibly better, clearer and subjectively faster. Yet my measurements were essentially the same, off by say 0.01% for THD and IM.
It turned out that these were brand new technology caps just introduced by Siemens, and the prices were like 10 times (!!!) the prices of regular ceramic caps. This sort of thing happens because manufacturers are always introducing new parts of variants of existing parts and are not always loud when doing so.
Now, the fundametal designer of a device has no time to investigate such perhaps oddball variatons on his theme, and that's what the twek group should be doing. Initially, they will cost money, but if they are worth thir salt, they will turn good products into excellent one and could soon become a welcome option. I am, in fact, suggesting, what many tweakers do anyway, but I don't leave it up to them, I have my own group of tweakers who do it right off the bat. Experimeting should include even the not very likely groups of products, simply because you never know what will actualy work best in any particular design until you try it on a live model.
In essence, this is an extension of the product development process which should cover some ground the designer simply had no time for. And they should be freed from cost considerations the designer was bound by. Obviously, they should be able to prove to the designer that they really did some improvement before the product is released.
http://www.diyaudio.com/forums/digi...ng-new-ess-vout-dac-es9022-7.html#post2867664
Just remove the passive filter.
Patrick
Siliconix (IIRC) had published something like that back in the day. It even had cascoded input FETs (again, IIRC). The bypassed output stage is new to me though.
All good fortune,
Chris
All good fortune,
Chris
This is how we used dual jfets with a 5534 to make a very good sounding IC replacement.
We used Toshiba, but the BF862 pair might be even better. Watch out for too much noise from the second stage. This can be minimized by increasing the external load resistors (and proportionally reducing the jfet current) as much as possible to get maximum input stage gain. There is an uneasy compromise that can usually work well. Dave Wilson bought dozens from us for his WAMM ultra speaker frequency equalizer, back in the late '80's. If it was good enough for him, it will work for almost anyone here.
We used Toshiba, but the BF862 pair might be even better. Watch out for too much noise from the second stage. This can be minimized by increasing the external load resistors (and proportionally reducing the jfet current) as much as possible to get maximum input stage gain. There is an uneasy compromise that can usually work well. Dave Wilson bought dozens from us for his WAMM ultra speaker frequency equalizer, back in the late '80's. If it was good enough for him, it will work for almost anyone here.
> We used Toshiba, but the BF862 pair might be even better.
Why not use 2SK2145 for the diff pair, and 2SK209 for the CCS ?
Patrick
Why not use 2SK2145 for the diff pair, and 2SK209 for the CCS ?
Patrick
Attached is one designers jfet/ne5534 implementation?
I posted it one time in a diyaudio thread and one person commented that it can't be any good for audio, it is only good for a THD analyzer use 🙂
One day I will lay it out in smt to give it a try. It so hard to judge the differences between these low THD designs.
I posted it one time in a diyaudio thread and one person commented that it can't be any good for audio, it is only good for a THD analyzer use 🙂
One day I will lay it out in smt to give it a try. It so hard to judge the differences between these low THD designs.
Last edited:
- Status
- Not open for further replies.