[ppl]

Welcome to Head-Fi speaker Guy. I have not built Your circuit. I included the link because of the somewhat different approach taken and because a circuit explanation is given so as people could gain some insight as to how it works.

I am quite sure the readers hear would love your musing on letting the smoke out of components and similar stuff.

 

[The Speaker Guy]

Now I have a fixture designed to survive longer.

1 - I use a MAX400 op amp as the front end. This accurate but slow amp prevents high output slewing, which is what kills this thing

2 - I use a diode biased 2n3904/2n3906 output stage. I ganged four of each transitor types in parallel each with I think about 10 ohms of ballast.

3 - The input is RC filtered to slow down the input pulse

4 - The device was optimized to drive controlled rise time pulses into exactly 0.1µF loads.

WE mostly use this for datasheet scope photos, where we show the device's response to a step change on the input. This has lasted a year without destruction.

 

[Syzygies]

Quote:

Originally Posted by ppl A definitive long time reference of mine regarding op amp and output booster stages from linear technology. This is a long 132 page pdf file However IMHO a Must read Pages 45-47 show Discrete Dimond buffer circuits http://www.linear.com/pdf/an47fa.pdf

I thought this had completely disappeared from the web, until I tried again and found it at:

http://www.stanford.edu/class/ee122/...ets/an47fa.pdf

Note that Figure C11 on page 91 reprises Figure 101, with more information about oscillation.

 

[ppl]

Quote:

Originally Posted by Syzygies I thought this had completely disappeared from the web, until I tried again and found it at: http://www.stanford.edu/class/ee122/...ets/an47fa.pdf Note that Figure C11 on page 91 reprises Figure 101, with more information about oscillation.

Oh Manny thanks for finding that ling other than LTC's site. Ill give a sweeping overeview or the areas yoy talked about.

(1) " Compensation scheme that may work well for one load and output level might be compleatly unstable at another. This is refering to the use of compensation caps as these tend to operate over a small range of capacitence Loads. Anyone thay has had an Amp work perfectly with one set of Transducers and find other transducers result in instability showes the limitation of Phase Leed and Lag compensation.

(2) While on the same subject the tlk turns to ferrite Beeds and the use if series resistors on the input of the Buffer tend to give stability over a wider load range this IMHO indicates a better method of compensation as it adresses the root cause of the problem rather than just put a band Aid on the wound to keep vitial fluids from oozing out. The same alaligy can be used with Frequency compensation as it is better to fix the problem that is causing the instability rather than reduce the syatem bandwidth to obtain seamingly stable operation. Like the Band aid its a nice temperary fix to get the unit working so as to track down the cause, it's nevertheless only a fix and mearly pushes the dirt under the Rug so to speak.

 

[peranders]

Message to the moderator: Could you correct the spelling of "diamond" in the thread title and when you have done this you can delete this post.