OKey, I am a bit disappointed.
Somebody else did it before me.
But I should have known... There is nothing new under the Sun.
Don't be
...I'm sure someone else have tried it even before I have.Anyway, you can try something else and be the first one...in Rush cascode there is no limitation for active devices, you don't have to use complementary devices of the same type, e.g. you can use n-jfet/pnp, n-mfet/pnp, triode/pnp...etc
I've been studying Jerald Graeme's "Feedback Plots define Op Amp AC Performance", sboa015.pdf from the TI library, while playing with stability of the FET990.
There are 2 considerations for stability which also lead to 2 definitions of 'Unconditional Stability
I think the advice in the AD797 datasheet is about the limit of complexity. The instructions for uA709 are definitely too much for washed & unwashed masses alike. Hands up those of you who have used this vintage device.
Anyone done this sort of thing recently?
There are 2 considerations for stability which also lead to 2 definitions of 'Unconditional Stability
- Stability with different levels of NFB. This old warhorse is fairly (??!) straightforward. Easy to have a different version for gain less than 3x etc or a single NPO ceramic to change. These are my C4/5 recommendations in post #2060. Graeme helps me formalise much of my experience & thinking and introduces at least one new dodge. Prof. Cherry is excellent if you ken reed, rite & kont.
- Stability with load. This is the important one with Power Amps as you don't often tweak NFB but you do have loadsa different loads. Output Inductors are a big help for this but for a supa dupa GP OPA, SM inductors are truly evil and muck up your 1 ppzillion THD for various reasons. But easy (?!) enough if your noise gain is fixed.
I think the advice in the AD797 datasheet is about the limit of complexity. The instructions for uA709 are definitely too much for washed & unwashed masses alike. Hands up those of you who have used this vintage device.
Anyone done this sort of thing recently?
Last edited:
I've been interested in this stability thing for a while - the thought I would have is that since the device is discrete, what's really the point of making it general purpose, stable no matter what you throw at it? I would prefer to have a base topology, with multiple variations of the feedback and stabilising strategies, so that one could extract maximum performance depending upon what you're using it for ...But a GP OPA might have to cope with combinations of both so the tweaks & recommended C changes can become too complicated. Zobels don't guarantee unconditional.
I think the advice in the AD797 datasheet is about the limit of complexity. The instructions for uA709 are definitely too much for washed & unwashed masses alike. Hands up those of you who have used this vintage device.
Anyone done this sort of thing recently?
Frank
Frank, this is exactly what Scott & I are trying to do. The problem is that it then becomes too complicated to explain even to another guru. (not that I'm claiming guru status
)It took me a long time to be sure about my C4/5 = 15p, 47p & 100p on (mostly) resistive loads at any gain.
I'm still not happy I've got anything sensible I can explain even with polysyllables, when I add different reactive loads. (Dis be double-speak to say I'm all confused by my results
)
Why worry about explaining it? Have one variation that will always work, and then list the conditions that must be met to use the edgier varieties. A simple checklist, a recipe type of thing, that people can refer to: things like, at least so much closed loop gain; and, the load must be resistive, may be so much reactive, etc.Frank, this is exactly what Scott & I are trying to do. The problem is that it then becomes too complicated to explain even to another guru. (not that I'm claiming guru status
Frank
You got it. There several comments here and elsewhere about why bother with discrete today.
So let's list again some bona-fide reasons.
1) High voltage and current operation possibly even as an input to a 50 or 100W PA.
2) Ability to bias output for a specific load (eliminate crossover).
3) Option to custom compensate for any application (very few IC's go above G = 5 decompensation).
This weekend looks good to finish the prototype (cold and dreary). The current thought is the VAS and output on a small card with a daughter card containing the various input stages.
EDIT - I will be having a nice leg of Aussie lamb Saturday. Last week we had a 1955 Belgian bottled Charmes-Chambertin (or so it said on the label) didn't matter it had already joined the parrot.
Last edited:
1955 Charmes-Chambertin is pushing it. Most probably bleedin' deceased. My condolences.
My oldest authentic burg is a 1983 La Tache magnum. When??
1983 was a pretty spotty vintage, lots of rot. The only bottle of 1983 La Tache I have seen did not look that inviting. You better make sure I'm there just in case.
More disappointing was the 1990 Clos de la Bousse d'Or and 2006 Echezeaux Gros Freres both sort of four square and clunky and these are seriously priced wines. I just don't see modern Burgundy.
Last edited:
And since we're in a groove, had this last weekend: 1997 d'Arenberg The Ironstone Pressings, Australia, South Australia, Fleurieu, McLaren Vale - CellarTracker!. Was worried about how it had fared but it was excellent, the comments about the fruit are spot on ...
Frank
Frank
Just the weird stuff - but good stuff left -
How can you keep a fine wine for sooo long? I have to drink it after a few good years of storage. My oldest fine wine now is only from 1996.
So, I thought I would get something made in countries or by makers no longer around -- Like 7 star rated Metaxa from Greece and an old Polish liquer store owner who went to a home for the aged left me with a large bottle from his stash of Polish Vodka made over 50 years ago (from the time he gave it to me) called Krolemska. And, A fine but small Irish whiskey maker who went broke and sold off all its stock of their rare and best single malt Scotch Whiskey -Rose Bank made in 1979, bottled in 1999. If you cant or don't dare drink it, it might as well be unique. -RNM
How can you keep a fine wine for sooo long? I have to drink it after a few good years of storage. My oldest fine wine now is only from 1996.
So, I thought I would get something made in countries or by makers no longer around -- Like 7 star rated Metaxa from Greece
and an old Polish liquer store owner who went to a home for the aged left me with a large bottle from his stash of Polish Vodka made over 50 years ago (from the time he gave it to me) called Krolemska. And, A fine but small Irish whiskey maker who went broke and sold off all its stock of their rare and best single malt Scotch Whiskey -Rose Bank made in 1979, bottled in 1999. If you cant or don't dare drink it, it might as well be unique. -RNM
Last edited:
How can you keep a fine wine for sooo long? I have to drink it after a few good years of storage. My oldest fine wine now is only from 1996.
So, I thought I would get something made in countries or by makers no longer around -- Like 7 star rated Metaxa from Greece
Once bottled spirits tend not to age. I once bought a bottle of Burbon bonded in 1916 and stored to avoid prohibition, lost and found in a warehouse in the 1980's. One taste and Scarface, the Godfather, and Al Capone all make sense.
One of the differences of really old spirits is how they were made and the ingrediants used back then compared to the same process but not quite the same ingrediants. The wood isnt the same, either (if it ever is). So, they taste different/unique and you wont be able to get anything like it again. Even if the brand is still making it. Not unlike wines, in this regard. But, I never drink spirits since my early youth when I learned my lesson on high %age drinks. But, there are those who havent over-indulged yet and they get to drink them and I just stand back and watch.
OTOH the year to year consistency on single malt scotch is amazing, there are golden palates as well as golden ears. The average person could not tell the difference between 15 yr. old Glenlivet across many decades.
consistency - yes, especially when they use the same French oak barrels for storage for decades. When they have to change the wood, then there is a more noticable change which again is consistant for decades in the 'new' wood. I yield to more experience drinkers.
When i lived for three years in Germany, I learned a lot about the wines of France and Germany and the region; Many great wines are fully consumed locally and we never get to taste them over here. Now, I mostly find Calif wines which have become pretty darn good as competitors.
When i lived for three years in Germany, I learned a lot about the wines of France and Germany and the region; Many great wines are fully consumed locally and we never get to taste them over here. Now, I mostly find Calif wines which have become pretty darn good as competitors.
What I meant was take a "The Glenlivet 15" bottled in 1950 and one bottled in 2000 and it would take a 6 sigma palate to tell the difference. They are that careful to maintain consistency all factors not withstanding.
In any case we shall be back OT tomrrow.
Last edited:
Dicey.
Scott holds the ****-you wine record. Sauntered into the restaurant for dinner one night and plopped an 1850 Verdelho on the table. Consensus was that it needed another century. Oldest I have on hand at the moment is a '47 Rioja.