Re: i felt induced to reply
That explains why you seem to have a certain reluctance to lending
your name to a new unit.
Anyway, as I hope everybody understand, my oriignal suggestion
was just meant as a joke since the unit seemed so uncommon.
I don't think it is the task of this forum to invent new units.
Besides, thinking a bit more about it, I found that 1 Dieckmann
would actually simplify to the inverse of 1 second, however
strange that may sound.
Fred Dieckmann said:"I vote for Dieckmanns. This would be expressed in:
Amps per Farad per Volt. I / C / V"
You what makes this truly funny is my middle name is Henry( who would make such a terrible thing up)! That would make it too confusing even for me.Why does this remind me of that useful conversion of the units for velocity into Furlongs per Fortnight?
That explains why you seem to have a certain reluctance to lending
your name to a new unit.
Anyway, as I hope everybody understand, my oriignal suggestion
was just meant as a joke since the unit seemed so uncommon.
I don't think it is the task of this forum to invent new units.
Besides, thinking a bit more about it, I found that 1 Dieckmann
would actually simplify to the inverse of 1 second, however
strange that may sound.
Hi subwo,
That Vas to differential feedback C is where I place the 2xC+R network. Splitting the C with the R to output maintains stability without degenerating so much output stage drive speed; though obviously you have phenominal speed already.
I always found that instability was more likely to arise with lower values of capacitive loading. Larger Cs sharply reduce the hf gain in the phase change region and this proves nothing, while values around 5 to 50nF can trigger additional undesirable responses in the phase changed region.
Has your distortion improved at lower frequencies, or is the simulation still reading phase ?
Hi Chris,
I'm waiting to see whether your valve topology uses a transformer, or is a non-complementary push-pull.
Cheers ........ Graham.
That Vas to differential feedback C is where I place the 2xC+R network. Splitting the C with the R to output maintains stability without degenerating so much output stage drive speed; though obviously you have phenominal speed already.
I always found that instability was more likely to arise with lower values of capacitive loading. Larger Cs sharply reduce the hf gain in the phase change region and this proves nothing, while values around 5 to 50nF can trigger additional undesirable responses in the phase changed region.
Has your distortion improved at lower frequencies, or is the simulation still reading phase ?
Hi Chris,
I'm waiting to see whether your valve topology uses a transformer, or is a non-complementary push-pull.
Cheers ........ Graham.
Hello Graham,
I too realized that it is the same principle as the 2C+R method. It trades some overall feedback for some from the more local loop.
The size of the capacitor in simulations has reacted as you mention. I later noticed that at 100nF the circuit still became unstable on the LTspice (but not on Microcap) simulator with the more local
compensation capacitor in place. But I could also see how drastic the benefit is.
The (phase) distortion at low frequencies is dependent on the value of the DC blocking capacitors as you were saying. Making those capacitors large or omitting them during distortion measurements eliminates that problem.
Best regards
I too realized that it is the same principle as the 2C+R method. It trades some overall feedback for some from the more local loop.
The size of the capacitor in simulations has reacted as you mention. I later noticed that at 100nF the circuit still became unstable on the LTspice (but not on Microcap) simulator with the more local
compensation capacitor in place. But I could also see how drastic the benefit is.
The (phase) distortion at low frequencies is dependent on the value of the DC blocking capacitors as you were saying. Making those capacitors large or omitting them during distortion measurements eliminates that problem.
Best regards
Yup. I've got a book that so indicates. I've noticed some ambiguities in what people mean when that use the phrase "Lin topology". Some mean just what you describe plus the quasi complementary OS. For others the key item is the absence of an output cap. Some mean more generally anything laid out IS(diff pair)-->VAS-->OS. Some would call Self's "blamless" Lin topology but a symetric variation of it.
I took tmblack's question to refere to anything laid out IS(diff pair)-->VAS-->OS.
If fear a discussion of what is "Lin" vs "not-Lin" could end up as tedious as : "When is Class B really Class AB" or visa versa.
Lin Topology
The Lin topology was developed and patented in the mid 1950s by Dr. H. C. Lin at RCA laboratories in Princeton, N. J. An article was published in Electronics Magazine, September 1956.
It had a single ended front end, capacitively coupled input and output, used germanium transistors, was quasi-complementary and used bootstrapping on the gain stage.
Here are some links:
http://www.semiconductormuseum.com/Transistors/RCA/OralHistories/Lin/Lin_Page7.htm
http://www.semiconductormuseum.com/Transistors/RCA/OralHistories/Lin/Lin_Index.htm
Here's a zip file with the article:
http://www.alectralink.com/pc/QuasiComp/Lin.zip
And some history here: I don't agree with his opinions but he does have links to some interesting technical papers by Lin, Shaw, and Baxandall:
http://www.alphalink.com.au/~cambie/
The Lin topology was developed and patented in the mid 1950s by Dr. H. C. Lin at RCA laboratories in Princeton, N. J. An article was published in Electronics Magazine, September 1956.
It had a single ended front end, capacitively coupled input and output, used germanium transistors, was quasi-complementary and used bootstrapping on the gain stage.
Here are some links:
http://www.semiconductormuseum.com/Transistors/RCA/OralHistories/Lin/Lin_Page7.htm
http://www.semiconductormuseum.com/Transistors/RCA/OralHistories/Lin/Lin_Index.htm
Here's a zip file with the article:
http://www.alectralink.com/pc/QuasiComp/Lin.zip
And some history here: I don't agree with his opinions but he does have links to some interesting technical papers by Lin, Shaw, and Baxandall:
http://www.alphalink.com.au/~cambie/
The information forwarded is greatly appreciated. Lots of interesting soundbites of Lin.
Still I am wonderful who made first DC coupled amp like the one in ETI-480 article @1974, (well inverted input. has one capacitor but it could be eliiminated).
I doubt Ge transistors would work with thier high leakage and heat sensitivity.
Tom
Still I am wonderful who made first DC coupled amp like the one in ETI-480 article @1974, (well inverted input. has one capacitor but it could be eliiminated).
I doubt Ge transistors would work with thier high leakage and heat sensitivity.
Tom
tmblack said:
I've never researched the question of first diff pair, and first direct coupled amp. I believe diff pairs had been used with tubes, here's a tube OP amp:
http://www.uoguelph.ca/~antoon/gadgets/741/741.html
I just noticed that Lin's article shows a split supply, direct coupled output stage but he didn't use it in his final implementation. I would give him credit anyway. I'd guess that the Germanium (Ge) transistors were prone to failure and that was risky with direct coupling.
I started out with tubes when it was hard to get transistors. I remember the CK722 Ge transistor and built my first preamp from an article in Popular Electronics with 2N404 Ge transistors - it was NOISY. Silicon transistors had come out by this time, I just don't think we had a schematic for a preamp. I don't remember ever getting a power amp to work (for any length of time) with Ge power transistors but I was very young at the time. Our first RCA transistor manual had power amps with Ge devices, the 1969 printing still had Ge based preamps but all the power amps now used Si transistors. The RCA direct coupled, diff amp input, quasi-complementary 70W power amp design was in this 1969 printing. Power amps were borrowing from OP amp topologies that had been around for several years.
Pete B.
Dan Meyer actually called the Tiger amp an operational amplifier circuit. I purchased a Tiger Amp kit at a yard sale around 1982 and it came with the 10 year old copy of Popular Electronics magazine which explained how the circuit worked. By studying the explanation and experimenting with the amp, I learned much of the theory and got a grounding in the field.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.