HV OpAmp as voltage amplifier. Mihkus Version of opamp driven amp trans amp
Yohoo everyone.
I know that some of us have wanted to use opamp as VAS stage for years...
So I took my brainless head and started this thread, my knowledge so far is sadly poor.
But I managed to design and build something, not 100% my design but who cares
And PCB, You can give it a try but all im able to say... This amp sounds better than many chipamps, just my opinion but thats what Im thinking.
Supply voltage for this design should be around 15-16,5V...
And PCB:
But that is the final design? NO! I will continue to make it better, whenever I have time.
But the road to this started on protoboard...
After that... It sounded good, but I wanted something more :d
I tried with capacitors as diodes to eliminate any switching... Nothing got better, only worse at higher power levels it was VERY UNSTABLE.
After replacing caps for diodes It sounded sweet, but I wasnt happy!
I added driverstage too! and it got better, alot more output power and alot lower distortion. because the opamp had very low output current.
After listening for pop rock for a while...
(Not mine)
And after that...
Yohoo everyone.
I know that some of us have wanted to use opamp as VAS stage for years...
So I took my brainless head and started this thread, my knowledge so far is sadly poor.
But I managed to design and build something, not 100% my design but who cares
And PCB, You can give it a try but all im able to say... This amp sounds better than many chipamps, just my opinion but thats what Im thinking.
Supply voltage for this design should be around 15-16,5V...
An externally hosted image should be here but it was not working when we last tested it.
And PCB:
An externally hosted image should be here but it was not working when we last tested it.
But that is the final design? NO! I will continue to make it better, whenever I have time.
But the road to this started on protoboard...
An externally hosted image should be here but it was not working when we last tested it.
After that... It sounded good, but I wanted something more :d
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
I tried with capacitors as diodes to eliminate any switching... Nothing got better, only worse at higher power levels it was VERY UNSTABLE.
After replacing caps for diodes It sounded sweet, but I wasnt happy!
I added driverstage too! and it got better, alot more output power and alot lower distortion. because the opamp had very low output current.
After listening for pop rock for a while...
An externally hosted image should be here but it was not working when we last tested it.
(Not mine)
And after that...
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
Last edited:
transistor (or "whatever") boosted opamps are quite a common diy amp builds.
actualy the opamp is used only as a voltage amplifier, and the output stage has a voltage gain of 1.
so it supply current.
if you drive the output part with higher supply, the voltage swing of the opamp will not be larger.
so, the output power remains..
actualy a good opamp can bias an output stage, if the output stage is included in the opamp feedback path. can't elliminate the crossover distorsion, but will reduce it to a level where its more than acceptible.
actualy the opamp is used only as a voltage amplifier, and the output stage has a voltage gain of 1.
so it supply current.
if you drive the output part with higher supply, the voltage swing of the opamp will not be larger.
so, the output power remains..
actualy a good opamp can bias an output stage, if the output stage is included in the opamp feedback path. can't elliminate the crossover distorsion, but will reduce it to a level where its more than acceptible.
You''ll never get +-50V from an op amp - at least not one that I know of. The OPA604 or 2604, will do +-24V, which is already amazing. (You might be able to push it a few volts further as well, but be prepared for failures).
If you are looking at simple biasing then how about 3 or 4 diodes? Just to not waste all of that feedback on correcting the crossover distortion. Also, drivers plus o/p will make life easier on the op amp, esp. if you are using it over-voltage. A slightly higher supply on the output stage probably wouldn't go amiss but I suspect the max output of the op amp will already never get above 3 or 4V less that the rails so that's room enough.
A perfectly nice idea.
If you are looking at simple biasing then how about 3 or 4 diodes? Just to not waste all of that feedback on correcting the crossover distortion. Also, drivers plus o/p will make life easier on the op amp, esp. if you are using it over-voltage. A slightly higher supply on the output stage probably wouldn't go amiss but I suspect the max output of the op amp will already never get above 3 or 4V less that the rails so that's room enough.
A perfectly nice idea.
You can using a High Voltage opamp such as Analog Devices ADA4700-1.
Here is a link to the data sheets,
http://www.diyaudio.com/forums/soli...oltage-opamp-based-amplifier.html#post3652631
However it is in a SOIC package and I had just got one but I have not tried it out yet,
http://www.diyaudio.com/forums/soli...oltage-opamp-based-amplifier.html#post3663983
It has great specs and looks like it may do well for such applications as this.
This type of application is the very reason I got one to try out.
FWIW
jer
Here is a link to the data sheets,
http://www.diyaudio.com/forums/soli...oltage-opamp-based-amplifier.html#post3652631
However it is in a SOIC package and I had just got one but I have not tried it out yet,
http://www.diyaudio.com/forums/soli...oltage-opamp-based-amplifier.html#post3663983
It has great specs and looks like it may do well for such applications as this.
This type of application is the very reason I got one to try out.
FWIW
jer
Last edited:
The LTC6090 is the new one from Linear teachnology.
It is rated at 140v (+/-70V),
LTC6090 - 140V CMOS Rail-to-Rail Output, Picoamp Input Current Op Amp - Linear Technology
http://cds.linear.com/docs/en/datasheet/6090fb.pdf
It had just came out earlier this year I think.
These are some of the first of their type since the LM143 that had got discontinued some time ago besides the very expensive ones from Apex.
http://www.ti.com/lit/ds/snvs759/snvs759.pdf
jer
P.S. Yes, it was a link to the links that I had already posted on that part.
It is rated at 140v (+/-70V),
LTC6090 - 140V CMOS Rail-to-Rail Output, Picoamp Input Current Op Amp - Linear Technology
http://cds.linear.com/docs/en/datasheet/6090fb.pdf
It had just came out earlier this year I think.
These are some of the first of their type since the LM143 that had got discontinued some time ago besides the very expensive ones from Apex.
http://www.ti.com/lit/ds/snvs759/snvs759.pdf
jer
P.S. Yes, it was a link to the links that I had already posted on that part.
Last edited:
Here is yet another method of increasing the output voltage swing of opamps,
http://m.eet.com/media/1152270/24127-45890.pdf
I have not tried this myself yet but I have seen it implemented many times.
Typically this circuit doesn't work in simulators but does in real life applications.
I am wondering if this type of configuration could be used up into the >1kv ranges.
FWIW
jer
http://m.eet.com/media/1152270/24127-45890.pdf
I have not tried this myself yet but I have seen it implemented many times.
Typically this circuit doesn't work in simulators but does in real life applications.
I am wondering if this type of configuration could be used up into the >1kv ranges.
FWIW
jer
Anyway... heres how i would apply class A output to HV OpAmp...
Its half done and not sure about components...
The "driver opamp" will be OPA552 will run the whole circuit at +-28V what i mean... 20 0 20 transformer...
Output trans TO3 because I want a bit vintage look... But first I have to make sure everything is correct...
Design isnt finished... Help!
photo sharing
Its half done and not sure about components...
The "driver opamp" will be OPA552 will run the whole circuit at +-28V what i mean... 20 0 20 transformer...
Output trans TO3 because I want a bit vintage look... But first I have to make sure everything is correct...
Design isnt finished... Help!
An externally hosted image should be here but it was not working when we last tested it.
photo sharing
Last edited:
It would make life a bit easier for us if you could post the schematic for this board.
Good find for an op amp. Again, one that I wasn't aware of.
Are you sure you're going to enough current gain to ever warrant a TO-3 case? It'll push those op amps quite hard, especially into 4 ohm speakers. And they can probably only do their 200mA at a pinch. They might even blow up (can they have a near short across them. Though they are horrible things, it might be worth having a Darlington in a single package. Some of the modern ones might not be so bad and you can easily find hfes of up to 5000, though I can't off-hand remember what current they'll take. Or if you want to save a diode drop you could have a CFP.
There might also be something clever you could do with diodes in the f/b loop, though that MAY take a bit of thought. But certainly one thing that does occur to me is that if you have both the transistors switched on you might be loading up your op amps. It might possibly help to have quite large Res, perhaps 0R47. You've got plenty of gain in the op amp to keep the output Z down anyway.
Good find for an op amp. Again, one that I wasn't aware of.
Are you sure you're going to enough current gain to ever warrant a TO-3 case? It'll push those op amps quite hard, especially into 4 ohm speakers. And they can probably only do their 200mA at a pinch. They might even blow up (can they have a near short across them. Though they are horrible things, it might be worth having a Darlington in a single package. Some of the modern ones might not be so bad and you can easily find hfes of up to 5000, though I can't off-hand remember what current they'll take. Or if you want to save a diode drop you could have a CFP.
There might also be something clever you could do with diodes in the f/b loop, though that MAY take a bit of thought. But certainly one thing that does occur to me is that if you have both the transistors switched on you might be loading up your op amps. It might possibly help to have quite large Res, perhaps 0R47. You've got plenty of gain in the op amp to keep the output Z down anyway.
The opamp is feeding a pair or Si- NPN/PNP that drive TO3s
The opamp has to put out only few mw...
This is how I built my first attempt with random not even a pair of output transistors+ opamp, seemed to work super well
There was just a resistor between emmitter and base to switch outputs halfway on it would be crackling sound witout..
Id rather read SCH from PCB and I WILL POST SCH ayway today if I came back from chool.
I dont want to use very high voltages and I dont want more than 10W RMS at 8R.
photo share
I hope you understund, the supply to the opamp hasnt been done and so is for the driver stage trans part just a quick mess, dont worry I will fix this board up very soon.
The opamp has to put out only few mw...
This is how I built my first attempt with random not even a pair of output transistors+ opamp, seemed to work super well
There was just a resistor between emmitter and base to switch outputs halfway on it would be crackling sound witout..
Id rather read SCH from PCB and I WILL POST SCH ayway today if I came back from chool.
I dont want to use very high voltages and I dont want more than 10W RMS at 8R.
An externally hosted image should be here but it was not working when we last tested it.
photo share
I hope you understund, the supply to the opamp hasnt been done and so is for the driver stage trans part just a quick mess, dont worry I will fix this board up very soon.
Hi!!
I just found the schematic of my version of a Current Boosted Opamp power amp circuit.
When I built this I had an ammeter on each leg of the power supply.
And by adjusting the Bias control I could set the bias through each of the output transistor as much as 1.1 amps or so ( the limit of my regulated bench supply) all of the way down to a few milliamp's per device.
At the most they where within about 10 to 20 or less of imbalance between the two halves.
I didn't play with it long enough to make it perfectly balanced but they were very close as it was on the first try.
I believe the offset output voltage was that of the opamp itself of just a few millivolt's or so.
Being that I could set the output bias current into the >1 amp range suggested that the output stage could be run in a class A/AB mode.
It took two or even three diodes to have enough voltage on the base's of the transistors in order to keep them turned on and the resistors set the bias current for each half.
By adjusting just one of the resistors also increases or deceases both of the currents in each halve because the opamp then compensates for the shift of the centering voltage.
I didn't put an ammeter on the opamps output but I will next time to see how much current is demanded from it.
It is not at all much though in this circuit, as I have a lot of current gain of at least 20,000 to 30,000 or so.
About an hfe of 140 to 170 for the small signal transistors and in the order of >220 for the output devices.
This schematic was written as it was before I tore it apart and I did test it in several configuration using a single ended supply as well as bipolar supply.
I am posting this to clarify what I had found as far as biasing goes as well as output voltage swing.
I also shorted the output to observe any distortions with my 1 ohm load and I only observed some flat topping of the output voltage due to the limitations of my bench power supply's self limiting feature.
I think that I saw some crossover distortions but only at a low bias current under such conditions.
I didn't mess with it long enough to do any real testing taking notes on such matters as it was all bread boarded on my protoboard.
I will be re-assembling it again sometime soon using a PCB and a power supply that is worthy of such a power amp.
I will also increase the number of output transistors to 2 or 3 per half as well in the next version.
The BD911/912's are good for a Pd of 90 watts and are very closely matched as are the small signal drivers too.
I also matched the forward voltages of the diodes strings as close as I could as well.
It powered a speaker quite nicely with test tones.
But, I didn't take the time to listen to any music through it.
Everything looked good though, including some square waves.
I did have to compensate it for High Frequency oscillations on the positive halves of the output signal and I think that is what I used the .01uf capacitor for.
This only happened when I swapped out the TL072 for the LME49860.
As I had mentioned it was just a quick project to see if it worked and it worked so well that it warrants a closer look and tweaking.
I think that the new ADA7400-1 HV opamp will work nicely in the next version.
According to my calculations only about 2 to 5 ma. was being demanded from the opamp.
Thanks !!!
jer
P.S. Forum won't allow me to upload the JPG file I will try again later!!!
I just found the schematic of my version of a Current Boosted Opamp power amp circuit.
When I built this I had an ammeter on each leg of the power supply.
And by adjusting the Bias control I could set the bias through each of the output transistor as much as 1.1 amps or so ( the limit of my regulated bench supply) all of the way down to a few milliamp's per device.
At the most they where within about 10 to 20 or less of imbalance between the two halves.
I didn't play with it long enough to make it perfectly balanced but they were very close as it was on the first try.
I believe the offset output voltage was that of the opamp itself of just a few millivolt's or so.
Being that I could set the output bias current into the >1 amp range suggested that the output stage could be run in a class A/AB mode.
It took two or even three diodes to have enough voltage on the base's of the transistors in order to keep them turned on and the resistors set the bias current for each half.
By adjusting just one of the resistors also increases or deceases both of the currents in each halve because the opamp then compensates for the shift of the centering voltage.
I didn't put an ammeter on the opamps output but I will next time to see how much current is demanded from it.
It is not at all much though in this circuit, as I have a lot of current gain of at least 20,000 to 30,000 or so.
About an hfe of 140 to 170 for the small signal transistors and in the order of >220 for the output devices.
This schematic was written as it was before I tore it apart and I did test it in several configuration using a single ended supply as well as bipolar supply.
I am posting this to clarify what I had found as far as biasing goes as well as output voltage swing.
I also shorted the output to observe any distortions with my 1 ohm load and I only observed some flat topping of the output voltage due to the limitations of my bench power supply's self limiting feature.
I think that I saw some crossover distortions but only at a low bias current under such conditions.
I didn't mess with it long enough to do any real testing taking notes on such matters as it was all bread boarded on my protoboard.
I will be re-assembling it again sometime soon using a PCB and a power supply that is worthy of such a power amp.
I will also increase the number of output transistors to 2 or 3 per half as well in the next version.
The BD911/912's are good for a Pd of 90 watts and are very closely matched as are the small signal drivers too.
I also matched the forward voltages of the diodes strings as close as I could as well.
It powered a speaker quite nicely with test tones.
But, I didn't take the time to listen to any music through it.
Everything looked good though, including some square waves.
I did have to compensate it for High Frequency oscillations on the positive halves of the output signal and I think that is what I used the .01uf capacitor for.
This only happened when I swapped out the TL072 for the LME49860.
As I had mentioned it was just a quick project to see if it worked and it worked so well that it warrants a closer look and tweaking.
I think that the new ADA7400-1 HV opamp will work nicely in the next version.
According to my calculations only about 2 to 5 ma. was being demanded from the opamp.
Thanks !!!
jer
P.S. Forum won't allow me to upload the JPG file I will try again later!!!
Last edited:
That's Cool.
The info I posted I found by experimenting and I have never seen it explained anywhere and I have been in electronics for many many years (self taught).
I have seen many questions about this type of configuration many times (as well as myself) but never any solid answers about the biasing of the output devices.
One particular thread that uses this type of configuration uses the TDA7293 Power Chipamp and some heavy duty output transistors here,
http://www.diyaudio.com/forums/chip...istors-amp-tda7293-come-also.html#post2904001
IMHO I think that it is a waste of the power amp chip unless one is driving at least 4 pairs of devices but it does work.
The questions of crossover distortions have come up many times in that thread but no one has ever done any proper tests to show the effects with a scope and very low impedance loads.
I would do it myself but I don't have any of the output transistors that they are using yet.
I hope the info was of use to you and that your circuit gives you much success as mine did, as this is a wonderful hobby!!!
I will upload my schematic for your reference as soon as they get the forum problems fixed.
Most of the HV opamps that I have run across didn't have very good specs for use in audio circuits until recently.
And your are correct, I think that they were designed mostly for low speed control circuits.
For many years I have been trying to come up with a circuit that just used a dual opamp and a pair of output transistors but not using a Current Boosted configuration.
I found out the hard way that it was the limited voltage output swing of the opamps that was hendering the performance of my past designs.
Cheers!!
jer
The info I posted I found by experimenting and I have never seen it explained anywhere and I have been in electronics for many many years (self taught).
I have seen many questions about this type of configuration many times (as well as myself) but never any solid answers about the biasing of the output devices.
One particular thread that uses this type of configuration uses the TDA7293 Power Chipamp and some heavy duty output transistors here,
http://www.diyaudio.com/forums/chip...istors-amp-tda7293-come-also.html#post2904001
IMHO I think that it is a waste of the power amp chip unless one is driving at least 4 pairs of devices but it does work.
The questions of crossover distortions have come up many times in that thread but no one has ever done any proper tests to show the effects with a scope and very low impedance loads.
I would do it myself but I don't have any of the output transistors that they are using yet.
I hope the info was of use to you and that your circuit gives you much success as mine did, as this is a wonderful hobby!!!
I will upload my schematic for your reference as soon as they get the forum problems fixed.
Most of the HV opamps that I have run across didn't have very good specs for use in audio circuits until recently.
And your are correct, I think that they were designed mostly for low speed control circuits.
For many years I have been trying to come up with a circuit that just used a dual opamp and a pair of output transistors but not using a Current Boosted configuration.
I found out the hard way that it was the limited voltage output swing of the opamps that was hendering the performance of my past designs.
Cheers!!
jer
Last edited:
Hi Mihkus, I didn't notice that you had some drivers in there - or I couldn't be bothered to look. I expect that what you have will be ample for more than 10W.
I look forward to seeing Gerald's schematic. 1A of Iq could be immense fun in a low powered amp. It must be 30+Watts of dissipation on each side but it could be very cool (or hot!). It also might be worth dividing that between two sets of paralleled output transistors, or the same thing in one package - which can be found, but I don't know where off-hand.
I look forward to seeing Gerald's schematic. 1A of Iq could be immense fun in a low powered amp. It must be 30+Watts of dissipation on each side but it could be very cool (or hot!). It also might be worth dividing that between two sets of paralleled output transistors, or the same thing in one package - which can be found, but I don't know where off-hand.
Don't forget to use a small bypass cap from each opamp power pin to ground. They must connect right at the pins, not a few mm away.
WILL DO or else... oscillations may occur.. Will use SMDs as i like them alot more than Thru-hole
The SCH i promised
I think i made a mistake somewhere on the output transistors side, I cant remember how I built it all up at home...
An externally hosted image should be here but it was not working when we last tested it.
screen capture program
Last edited:
Noooooooo! Im a bit high so i made a lot mistakes, then fixed them and did them again
image hosting websites
An externally hosted image should be here but it was not working when we last tested it.
image hosting websites
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.