ANY AMPLIFIER WITH SLEW RATE GREATER THAN 50V/ìs?

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Is there any amplifier with voltage slew rate bigger than 50V/ìs? Please not words, only documents like this one quoted below. Also the calculation should be the same as that made by G. Kleinsmidth in below curve. Glen estimated that the slew rate of an amplifier it is 30V/ìs from this curve. Please use the same method.

An externally hosted image should be here but it was not working when we last tested it.
 
sreten said:
Hi,

There are amplifiers with > 50V/uS slew rate, and the way it is
measured does not matter as long as the method is accurate.

Your post seems pointless .....

:)/sreten.

Instead of a vague answer, can you post a curve of output rise time, of an amplifier of those you refer, obtained from a DSO?
So as to be able to calculate the slew rate as did Glen Kleinsmidth? My post it is not so pointless as you suppose, sreten, as it has a specific objective which will see then.
 
Fotios, you are mixing and confusing slew rate with rise time limiting. One adds harmonics, while the other filters out harmonics. This is the source of your problem. Most of us usually deliberately rise time limit our designs to remove RF passthrough, noise, and to make the design 'slew rate proof'. You must REMOVE the bandwidth limitation or increase the drive voltage to really measure the slew rate.
 
john curl said:
Fotios, you are mixing and confusing slew rate with rise time limiting. One adds harmonics, while the other filters out harmonics. This is the source of your problem. Most of us usually deliberately rise time limit our designs to remove RF passthrough, noise, and to make the design 'slew rate proof'. You must REMOVE the bandwidth limitation or increase the drive voltage to really measure the slew rate.

Mr. Curl thanks for your reply which is picking up the meaning of my post with the curve quoted. Without wanting to sound selfish (that i know everything) i know well the difference between them. But these, you must tell to Glen Kleinsmidth, and not to me. He was which made the calculation of slew rate by this curve intended exclusively for measuring the rise time of an amplifier at full power loaded with 8Ù, with feedback closed, with Cdom in place, as well input-output filters in action. As for me, never i tried to measure slew rate. I am only interested for measuring the rise time, and i have as target values between 1,5 to 1,8ìs without load in full output swing, and as much 2,1 to 2,2ìs with 8Ù load in full output swing. And i know that this value it is same for any frequency from 20Hz to 100KHz. During this, i compromise with the value in which the amplifier it is stable (without any overshoot or noise). Moreover the only instruments that i have for measurements are one DSO 50MHZ and one function generator 10MHz both of Hameg. That is my budget.

Fotios
 
one reliable way of measuring slew rate is to apply a sine wave and increase the frequency until slew rate limiting makes the output triangular. the slew rate can then be calculated from the resulting oscope trace. also positive going and negative slew rates may be different. this method doesn't need a pulse generator or a square wave generator with high rise and fall times, once the sine wave is triangularized, the amp no longer swings rail to rail, and the amp doesn't need to be driven to high output swings to see this effect (at lower input voltages, the frequency where slew rate limiting begins will be higher than when the amp is driven rail to rail).
 
Are you trying to drive 80Khz ultrasound transducers?

Both digital and analog recordings are inherently slew rate limited. For example, a 200W@8ohm amplifier connected to a standard 44Khz digital source and playing full volume will never be asked for more than 8V/us. In practice, a music signal will hardly ask for more than half that.

BTW: The required s.r. for a sine wave is 2*pi*V_ampl*freq
 
unclejed613 said:
one reliable way of measuring slew rate is to apply a sine wave and increase the frequency until slew rate limiting makes the output triangular. the slew rate can then be calculated from the resulting oscope trace. also positive going and negative slew rates may be different. this method doesn't need a pulse generator or a square wave generator with high rise and fall times, once the sine wave is triangularized, the amp no longer swings rail to rail, and the amp doesn't need to be driven to high output swings to see this effect (at lower input voltages, the frequency where slew rate limiting begins will be higher than when the amp is driven rail to rail).

Thanks for the advice

Unfortunatelly, my new amplifier project does not triangularize a sine wave of small level applied in its input up to 5MHz!, the only that i have seen in my DSO it is the reduction of output signal level. For the record, the input sensitivity it is 0,92Vrms for full output swing which is 42,42Vrms (supply rails = +/-60Vdc). So, to triangularized a sine wave - from my tests right now - the output must be 90Vpp at least (or 31,82Vrms which is the 75% of full output swing) and the frequency must be from 160KHz and above. To this, i had removed first the zobel from output to not burned the resistor, and after i did again the measurement which proved finally NOT reliable. From experience, an amplifier like this with a rise time of 1,8ìs under a load of 8Ù in output in full output swing (measured by aplying a square in input to obtained a 60Vpp square in output from 20Hz to 100KHz) and without removed any filter or feedback, it is possible to limited in a 40V/ìs slew rate? So much was the slew rate that i measured according to your suggestion.
Finally, why is there the need for measuring the slew rate? The measuring of rise time it is not enough? I am stable in my place to estimate only the rise time.
In real audition tests, usually we are satisfied from the mid-low to low frequencies reproduced which is the difficult work for an amplifier. In this point i am focused mainly, in the steep bass reproduction. I quote a curve taken not to measure the rise time, but the tilt of a square duty portion. From this, it is obvious the high performance of amplifier to manage bass frequencies.

An externally hosted image should be here but it was not working when we last tested it.

Full output swing in 100Hz square under 8Ù load in output. The measured tilt of duty portion it is only 2,4V

Thanks again
Fotios
 
Fotios said:

"Unfortunatelly, my new amplifier project does not triangularize a sine wave of small level applied in its input up to 5MHz!, the only that i have seen in my DSO it is the reduction of output signal level."

You don't expect to see the effects of slew-rate limiting on low-level signals. It is a non-linear effect seen on high-amplitude signals. Are you sure you understand the mechanisms that cause slew-rate limiting?
 
Ouroboros said:
Fotios said:

"Unfortunatelly, my new amplifier project does not triangularize a sine wave of small level applied in its input up to 5MHz!, the only that i have seen in my DSO it is the reduction of output signal level."

You don't expect to see the effects of slew-rate limiting on low-level signals. It is a non-linear effect seen on high-amplitude signals. Are you sure you understand the mechanisms that cause slew-rate limiting?

I am sure that i understand VERY WELL the mechanisms that cause an inadmissible high RISE TIME! Let us to say above 2,5ìs from 1Hz to 150KHz. Am i wrong in this way of estimation?
Also when i say small level i mean a sine wave of 60Vpp in output at least in 20KHz.
Thanks
Fotios
 
This is from J. Curl post

"Most of us usually deliberately rise time limit our designs to remove RF passthrough, noise, and to make the design 'slew rate proof'."

That is exactly the correct answeer. If i understood well, mr. Curl says that we must examine first how much it is the slew rate of the amplifier, and then by keeping in mind this value, we must place a limit in the rise time of amplifier so as it can control its voltage slewing possibility to not produced harmonics etc. As for me, i only examine from clear curves (square) how i can obtain the minimum rise time in these by changing caps and resistors in whole the feedback loop or in the nested feedback loops.
 
Administrator
Joined 2007
Paid Member
Bit late with this one. There was a design by Giavani Stochino ( sorry if I have spelt it incorrectly ) with a claimed 300 volt's/microsecond slew rate. The amp is rated 100 watts 8 ohm/4ohm if I remember and it used 2 pairs of HEXFET outputs.
Has any one ever built it I wonder ? and what does it sound like.
 

GK

Disabled Account
Joined 2006
fotios said:
Is there any amplifier with voltage slew rate bigger than 50V/ìs? Please not words, only documents like this one quoted below. Also the calculation should be the same as that made by G. Kleinsmidth in below curve. Glen estimated that the slew rate of an amplifier it is 30V/ìs from this curve. Please use the same method.

An externally hosted image should be here but it was not working when we last tested it.


Gee wiz. I'm begining to wish I kept quiet.

To be more accurate, I did not calculate your amplifiers slew rate from the graph shown. I calculated it as 30V/us from the tail current of the LTP and the value of Cdom.

Following a little protest from you I then pointed out that your step response measurement confirms the simple calculation.

I have attached (again :rolleyes: ) an annotated version of your step response graph below. The slew rate is indeed 30V/us - and please don’t mix up the non-linear, exponential R-C response as the voltage approaches full amplitude into your slew rate limiting estimations, as here the input stage has recovered from overload and the rate of change is no longer slew rate limited


Cheers,
Glen
 

Attachments

  • slew.jpg
    slew.jpg
    39.3 KB · Views: 385
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.