Bob, almost all amps and IC's slew rate limit, because they use feedback. Only SOME amps and IC's bandwidth limit or have slew rate limits above 50V/us.
Even today, 40 years since Otala first stuck his head out, amps and IC's are made with little regard for the distortion leading back to the slew rate limit.
However, 30 years ago, or more, Matti Otala moved to DIM in order to segregate the slew rate generating TIM and adding other dynamic distiortions like PIM and non-linear capacitance in active components.
I am completely aware of what is possible and not possible. I use feedback in most of my designs, even today. Please show me something that I don't know, and I will acknowledge it.
Even today, 40 years since Otala first stuck his head out, amps and IC's are made with little regard for the distortion leading back to the slew rate limit.
However, 30 years ago, or more, Matti Otala moved to DIM in order to segregate the slew rate generating TIM and adding other dynamic distiortions like PIM and non-linear capacitance in active components.
I am completely aware of what is possible and not possible. I use feedback in most of my designs, even today. Please show me something that I don't know, and I will acknowledge it.
Demian,
I've been probing for ideas/suggestions regarding spectrum analyzers. John suggested that I ask you, so...
What's your thinking on spectrum analyzers?
I once bought an HP off of e-bay, but got clobbered on the transaction. The thing never worked, was shipped rattling loose in an oversized cardboard box, you name it. Bad mess. I've been leaning towards ever-wider bandwidth circuits (currently .25 to .5MHz throughput) anyway and have come to think that a 50kHz (if I remember correctly) analyzer wouldn't necessarily show me everything that I want to see.
Grey
I've been probing for ideas/suggestions regarding spectrum analyzers. John suggested that I ask you, so...
What's your thinking on spectrum analyzers?
I once bought an HP off of e-bay, but got clobbered on the transaction. The thing never worked, was shipped rattling loose in an oversized cardboard box, you name it. Bad mess. I've been leaning towards ever-wider bandwidth circuits (currently .25 to .5MHz throughput) anyway and have come to think that a 50kHz (if I remember correctly) analyzer wouldn't necessarily show me everything that I want to see.
Grey
john curl said:
Are you speaking of the original Electrocompaniet? I used to sell those, way back when. They were head and shoulders above any other solid state amps I heard in that time period in terms of sound quality, but hard to move because they were something like $1400, but only 40 or 50W. Customers expected more power for that amount of money.
I'd like to track down a used one, just for fun. I think I've got the schematics somewhere and could presumably build one, but that would take the fun out of it.
Grey
I had the first prototype, and used it for about 15 years, and it was destroyed in the firestorm. It beat virtually everything that I could get ahold of in the 1970's, especially with STAX headphones. The AR D-150 was actually smoother with STAX headphones, but it was not so good with loudspeakers, go figure.
The Electro--- actually ran WATT puppies, in an OK way, once I added a .5 ohm resistor on the output. I bought a second one from a hi fi store for $200 about 10 years ago. I was just in that very same hi fi store today. I DID have to re-adjust it to get it to perform well, as it does use cheap pots. Still, it still sounds great, within its limitations.
The Electro--- actually ran WATT puppies, in an OK way, once I added a .5 ohm resistor on the output. I bought a second one from a hi fi store for $200 about 10 years ago. I was just in that very same hi fi store today. I DID have to re-adjust it to get it to perform well, as it does use cheap pots. Still, it still sounds great, within its limitations.
The founder of EC, Abrahamsen (ABE) is still in business, with his new company, Abrahamsen.
BTW the firs amp was a 25W.
Cheers
Stinius
http://www.abrahamsenaudio.com/
BTW the firs amp was a 25W.
Cheers
Stinius
http://www.abrahamsenaudio.com/
Gerhard, while your input is interesting to me, I hope it doesn't confuse many others. You are of course, correct. You might call your TIM, secondary TIM. For example what is the slew rate limit of a Dyna Pas-3 tube phono stage? Think it through. Measure it, if you can. Hint: It is related to what you just mentioned.
However, MOST TIM is generated by the dominant pole that is added to make the feedback loop stable.
However, MOST TIM is generated by the dominant pole that is added to make the feedback loop stable.
Joshua_G said:
Joshua,
a proper digital scope should have a sampling rate at least 5 times the bandwidth. 100 Msps is enough only for 20 MHz scope. If I see a "250 MHz scope" with only 100 Msps real-time sampling I would not bother even to look at it in detail. At the moment there are some $500 scopes with a reasonable performance, like Rigol 1052E.
Alex
Syn08, Bob, Edmond, can I nevertheless return to my previous question? I understand your reasoning, allow the proper input dynamic range in the stages preceding lag compensation, allow the proper output current capability to recharge lag compensation networks, make the good recovery from overload, put the proper filters for out-of band signals,… Again, what is your criteria?
john curl said:
Hi John,
If you increase the impedance at the VAS collector node, and do not change the gain crossover frequency, you increase the amount of low-frequency negative feedback, and decrease the dominant pole frequency, but do not decrease slew rate or increase TIM.
That is what you don't know.
Cheers,
Bob
Electrocompaniet
Hello John
Are you talking about Electrocompanient, yes I have heard these amps and their biggest weakness is that they are soft and flabby in the bass, and consequently their sound is pleasant due to omision of information that they are not reproducing. Their sound that is not true to the source they are playing . Put another way they sound a bit boring.
Interestingly though Electrocompanient (today) have a new higher performing amp in their range and its called the DP line, which has a design philosophy at odds with "low feedback and wide open loop bandwidth" Have a look for your self the standard amps are called the "classic line" and they sit below the new DP series have a look at the power amps specs.
http://www.electrocompaniet.com/products_cat_DP.html
Regards
Arthur
john curl said:
Hello John
Are you talking about Electrocompanient, yes I have heard these amps and their biggest weakness is that they are soft and flabby in the bass, and consequently their sound is pleasant due to omision of information that they are not reproducing. Their sound that is not true to the source they are playing . Put another way they sound a bit boring.
Interestingly though Electrocompanient (today) have a new higher performing amp in their range and its called the DP line, which has a design philosophy at odds with "low feedback and wide open loop bandwidth" Have a look for your self the standard amps are called the "classic line" and they sit below the new DP series have a look at the power amps specs.
http://www.electrocompaniet.com/products_cat_DP.html
Regards
Arthur
Re: Electrocompaniet
The DP line is very good, it was a brand called Dynamic Precision (DP)
The new owners of EC bought the DP Company and made it an EC brand.
I think I was (I know I was) the first one that tested and used the DP amps, back in the 80’.
Cheers
Stinius
PHEONIX said:
The DP line is very good, it was a brand called Dynamic Precision (DP)
The new owners of EC bought the DP Company and made it an EC brand.
I think I was (I know I was) the first one that tested and used the DP amps, back in the 80’.
Cheers
Stinius
Of course I know that, I am not Dr. Greiner. I just like high open loop feedback to reduce 'imaginary' distortions, like Barrie Gilbert.
Hi Joshua,
I have to agree with Alex here. His 5 X your highest frequency of interest = listed bandwidth is correct. If you don't need to see sharp edges on square waves (you are looking at sine type wave forms), you can relax that to 3 X. For your 200 KHz bandwidth, this would be fine.
I did test a number of smaller DSOs and a USB 'scope. The Rigol does offer good value for the money and is pretty easy to use. I did test units up to the Tek DPO7000 series and the Agilent 6000 series. The lower end products are getting much better for performance. With the decreasing cost of DSP semiconductors and memory, I expect this trend to continue. We are just starting to see the effects of reduced technology costs.
USB products use your keyboard and mouse and screen as the interface. So using these is more difficult simply because the controls are not intuitive and are not manipulated as easily. Still, you can get good performance from these products. One advantage is that the software is upgradeable easily. That means if you are with a better company, your instrument may improve with more features as time goes on. That brings you back to companies such as Picoscope who have enough corporate mass to ensure updates and a good initial interface.
I bought a model "DSO-2150" recently. You see these advertised in many places on the 'net. I finally figured out how to use it (the software has "issues", or should we call them features?), then channel 1 failed. The importer didn't want to deal with it, so here I am shipping it back to China for service. They would not release any service information at all. Well, at least they are willing to look at it. They have also updated the software once - a good sign.
-Chris
I have to agree with Alex here. His 5 X your highest frequency of interest = listed bandwidth is correct. If you don't need to see sharp edges on square waves (you are looking at sine type wave forms), you can relax that to 3 X. For your 200 KHz bandwidth, this would be fine.
I did test a number of smaller DSOs and a USB 'scope. The Rigol does offer good value for the money and is pretty easy to use. I did test units up to the Tek DPO7000 series and the Agilent 6000 series. The lower end products are getting much better for performance. With the decreasing cost of DSP semiconductors and memory, I expect this trend to continue. We are just starting to see the effects of reduced technology costs.
USB products use your keyboard and mouse and screen as the interface. So using these is more difficult simply because the controls are not intuitive and are not manipulated as easily. Still, you can get good performance from these products. One advantage is that the software is upgradeable easily. That means if you are with a better company, your instrument may improve with more features as time goes on. That brings you back to companies such as Picoscope who have enough corporate mass to ensure updates and a good initial interface.
I bought a model "DSO-2150" recently. You see these advertised in many places on the 'net. I finally figured out how to use it (the software has "issues", or should we call them features?), then channel 1 failed. The importer didn't want to deal with it, so here I am shipping it back to China for service. They would not release any service information at all. Well, at least they are willing to look at it. They have also updated the software once - a good sign.
-Chris
Hi Grey,
Almost every single thing I ever bought in the way of test instruments off Eeeekbay has been damaged or non-functional. Usually in obvious ways. Honesty is not that common amongst Ebay equipment sellers.
One company I did deal with was "Teletek Enterprises". I had a problem with an HP lab meter and received it with "Calibration constants lost" error. Of course, that makes the meter completely useless without paying for a full calibration (it came with a certification) on top of the repair. The owner, Joe McCauley, was quick to correct the problem and simply replaced the instrument with a good one.
I would highly recommend dealing with Joe. Turns out he is a very reasonable and nice individual. Since he is a US business, much easier and cheaper for you to deal with him than it was for me.
I bought an HP 3580A, working perfectly. Yeah, right. Didn't respond to switching. Once the switches were cleaned, the top half of the display was not working. I saw this before doing anything, but the vendor, in Quebec, left me high and dry. $650.00 worth of junk. That hurt, and still does as I am still paying that off.
-Chris
I'm sorry to hear that. What was it that you bought?
Almost every single thing I ever bought in the way of test instruments off Eeeekbay has been damaged or non-functional. Usually in obvious ways. Honesty is not that common amongst Ebay equipment sellers.
One company I did deal with was "Teletek Enterprises". I had a problem with an HP lab meter and received it with "Calibration constants lost" error. Of course, that makes the meter completely useless without paying for a full calibration (it came with a certification) on top of the repair. The owner, Joe McCauley, was quick to correct the problem and simply replaced the instrument with a good one.
I would highly recommend dealing with Joe. Turns out he is a very reasonable and nice individual. Since he is a US business, much easier and cheaper for you to deal with him than it was for me.
I bought an HP 3580A, working perfectly. Yeah, right. Didn't respond to switching. Once the switches were cleaned, the top half of the display was not working. I saw this before doing anything, but the vendor, in Quebec, left me high and dry. $650.00 worth of junk. That hurt, and still does as I am still paying that off.
-Chris
I had remembered it as a 3580A, but I happened to run across a note from back during that era earlier today and discovered that it was actually a 3582A...assuming that I'm reading the note correctly. This was five or six years ago and I don't remember a lot of the details. Hadn't even remembered having any notes about the stupid thing, but there it was.
Grey
EDIT: Joshua_G, I'd suggest something more like a couple of MHz, absolute bare minimum, even if you only anticipate circuits with a few hundred kHz bandwidth. The reason being that if you've got an oscillation that's way up there--and they usually are--you'll want to be able to see it. I've got two Tektronix 7603s, one bench, the other rack mount, and another, older, Tektronix that I can't remember the model number. The 7600 mainframes were (I think) 100MHz with the right plug-ins. I like having the bandwidth...just in case.
Grey
EDIT: Joshua_G, I'd suggest something more like a couple of MHz, absolute bare minimum, even if you only anticipate circuits with a few hundred kHz bandwidth. The reason being that if you've got an oscillation that's way up there--and they usually are--you'll want to be able to see it. I've got two Tektronix 7603s, one bench, the other rack mount, and another, older, Tektronix that I can't remember the model number. The 7600 mainframes were (I think) 100MHz with the right plug-ins. I like having the bandwidth...just in case.
Hi Grey,
Please tell me you fixed it .... It's got some nice features you might find useful. I know I would!
-Chris
Please tell me you fixed it .... It's got some nice features you might find useful. I know I would!
-Chris
- Status
- Not open for further replies.