I have built a few amps using discrete devices in the past. Now, I am thinking about trying out an op amp based mosfet. I have a few devices in my mind (from the good old days), like tl082 or ad835. The first is a little too slow (in terms of slew rate) and the 2nd might be too fast (50v/us) so it could be a source of parasitic oscilation.
The amp will be an op amp driving a complementary-feedback based output stage (so I don't have to have a voltage amplification stage).
I like to get about 100w out of this amp so I am looking for an op amp with slew rate roughtly in the 35-40v/us range.
Any suggestions? thanks in advance.
The amp will be an op amp driving a complementary-feedback based output stage (so I don't have to have a voltage amplification stage).
I like to get about 100w out of this amp so I am looking for an op amp with slew rate roughtly in the 35-40v/us range.
Any suggestions? thanks in advance.
Slew rate is certainly a consideration, but to judge your choice specifically on slew is silly.
You will get a lot of opinions here...my own being one based on 'bang-for-the-buck'. In this catagory, the OPA2604 rules (25V/µS). I also really like the OPA627, but it is hardly the bargan that the 2604 is.
I realize I need to give some of the Analog Devices opamps a try, but all the ones I want to check out are SMD. Too much trouble and money for so little gain.
You will get a lot of opinions here...my own being one based on 'bang-for-the-buck'. In this catagory, the OPA2604 rules (25V/µS). I also really like the OPA627, but it is hardly the bargan that the 2604 is.
I realize I need to give some of the Analog Devices opamps a try, but all the ones I want to check out are SMD. Too much trouble and money for so little gain.
I've built an amp with the NTE equilivant of the TL072 and FETs for current drive. It sounds very good, clean and VERY little noise. I'm sure that there are better OP amps and I plan on trying some out. But one thing to consider is how much voltage will the OP amp swing (before clipping). Take the OPA227 as an example. Operation is +- 18, working V at +- 15. Even if this is what you do swing, this is your limiting factor in output power. SO, I=V/R=15/8=1.875 and Wattage = V*I= 28.125 peak power assuming 100% effiency. Realistcly this would be 18-20 Watts. I guess that my point is that voltage of the OP amp should also be taken into consideration if you are trying to get 100 Watts. Check the TI site for Power Operational Amps for additional options. But, I do have to say that I run my OPA based amp in my studio with a pair of '70s Advents and a seperate Sub and it is quite loud enough for the front rooms of my house, even during parties.
-Dozuki
-Dozuki
I did try the amp with a TL082. I don't remember the reason, but I felt that the TL072 worked better with the circuit I had put together. Either it didn't sound right to me, or I couldn't get it to swing enough voltage.
Sorry, I forgot that a CF OPS has voltage gain. This will definately help. And using BJTs as opposed to FETs in the OPS will bump the power up a bit too. I should actuall try using a CF OPS as I've recently come into a bunch of different OP amps to play with this same type of amplifier. When I was trying different OP amps I bread boarded the whole circuit on a small BB. It was easy to tweak and a fun little amp to play with. I could post my schematic if there is interest and I can get my scanner working. I probably made some mistakes, but I've been using it for 6 months or so now and I'm quite pleased with the sound.
-Dozuki
Sorry, I forgot that a CF OPS has voltage gain. This will definately help. And using BJTs as opposed to FETs in the OPS will bump the power up a bit too. I should actuall try using a CF OPS as I've recently come into a bunch of different OP amps to play with this same type of amplifier. When I was trying different OP amps I bread boarded the whole circuit on a small BB. It was easy to tweak and a fun little amp to play with. I could post my schematic if there is interest and I can get my scanner working. I probably made some mistakes, but I've been using it for 6 months or so now and I'm quite pleased with the sound.
-Dozuki
Your standard op amp can be bade to operate at higher Rail voltages by a method called bootstraping. see the AN-211 from ADI http://www.analog.com/UploadedFiles/Application_Notes/21917334AN211.pdf
this is a basic bootstraped op amp. you must be sure your chosen device is free from phase reversal if the Input is overdriven as this will cause seriose damage to this circuit is that happens. Most modern op amps meet this requirement however some older FET input Op amps are prone to this. check data sheet.
Attachments
Current feedback by a patented design could be something? This Mark Alexander amp is not very well regarded here but I think it's quite intersting. I help a friend many years ago with this amp and it worked alright. I have tried this concept myself with excellent results. Check here
Original, but I have a better copy of the scanned document. Send me a message if you want the better one.
http://www.analog.com/library/applicationNotes/AdAudio/AN211.pdf
http://homepages.strath.ac.uk/~cnbp111/amp1.html
Original, but I have a better copy of the scanned document. Send me a message if you want the better one.
http://www.analog.com/library/applicationNotes/AdAudio/AN211.pdf
http://homepages.strath.ac.uk/~cnbp111/amp1.html
during the tube days, people would put a small (and high wattage) resistor between the speaker and the ground and use that for negative feedback. I have seen some SS amps with that design and tried one myself. Not much sonic improvement other than for a subwoofer.
Anyone remember those designs?
Anyone remember those designs?
millwood> Ah... Yes i do Remember those days and also alot of old Kenwood Tube Amps and receivers had a switch that allow different damping factors using a similar technique.
peranders> Hi glad you droped in. The alxender circuit was intended only as an illistration of the Bootstraped Technique i talked about. I was not intending to indicate it was any Kind of sonic marval.
I have used Just plain BUF-634 Buffers Bootstraped to High Voltage transistors to follow a tube output stage in a topology similar to my posted op amp schematic above with splendid results. sure true Tube-O-Files could detect the solid state eliment in the sound But over all no one complained about the sound and alot of local Tube Equipment owners have since adopted this technique into there equipment in order to gain low impedance drive for stuff like Headphones.
peranders> Hi glad you droped in. The alxender circuit was intended only as an illistration of the Bootstraped Technique i talked about. I was not intending to indicate it was any Kind of sonic marval.
I have used Just plain BUF-634 Buffers Bootstraped to High Voltage transistors to follow a tube output stage in a topology similar to my posted op amp schematic above with splendid results. sure true Tube-O-Files could detect the solid state eliment in the sound But over all no one complained about the sound and alot of local Tube Equipment owners have since adopted this technique into there equipment in order to gain low impedance drive for stuff like Headphones.
...to continue...
The hypothetical op-amp., G1, (see attached), with a slew rate of 40V/uS, is connected to an output stage, G2, with a gain of 10, and the whole is enclosed in global feedback loop, giving an overall closed loop gain of 40.
It is straightfowardly demonstrated that the output slews at approx. 400V/us. This obviously assumes that the output stage possess a slew rate well in excess of this figure, otherwise the overall slew rate is then determined by the output stage.
This is almost invariably the case, and should therefore not be a concern.
The hypothetical op-amp., G1, (see attached), with a slew rate of 40V/uS, is connected to an output stage, G2, with a gain of 10, and the whole is enclosed in global feedback loop, giving an overall closed loop gain of 40.
It is straightfowardly demonstrated that the output slews at approx. 400V/us. This obviously assumes that the output stage possess a slew rate well in excess of this figure, otherwise the overall slew rate is then determined by the output stage.
This is almost invariably the case, and should therefore not be a concern.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.