Today, I have seen (have prompted), that the circuit 1 --- 37 times, the circuit very simple 18 times
The big request to all who download circuits, do not collect, read a forum, in circuits it is a lot of discrepancies.
Which I ask to comment on all circuits were collected on breadboard models, and it does not mean that they of 100 % will work for you.
So for example very simple, I today have changed under recommendations ilimzn, many thanks, but all the same is push DC at inclusion and deenergizing
( did not pay attention!!!!!!!!) it is long balanced . Floating feed OPamp provides scope Uout OPamp approximately equal to a Vrail. For this reason it is chosen. I try to result it in 100 % stability.
Very much I ask do not leave(abandon) me, I one am engaged in all it, and there is nobody to communicate nobody understands. Work is far from electronics.
The big request to all who download circuits, do not collect, read a forum, in circuits it is a lot of discrepancies.
Which I ask to comment on all circuits were collected on breadboard models, and it does not mean that they of 100 % will work for you.
So for example very simple, I today have changed under recommendations ilimzn, many thanks, but all the same is push DC at inclusion and deenergizing
( did not pay attention!!!!!!!!) it is long balanced . Floating feed OPamp provides scope Uout OPamp approximately equal to a Vrail. For this reason it is chosen. I try to result it in 100 % stability.
Very much I ask do not leave(abandon) me, I one am engaged in all it, and there is nobody to communicate nobody understands. Work is far from electronics.
All Hi ! I ask the answer to a question, the resistor in base BJT, should transfer(translate) it(him) in a mode of the amplifier of a current
Whether so it?
And how it to be reflected in a sound?
Many thanks deleveld
Many thanks darkfenriz
And many thanks ilimzn
For your comments,
Soon I shall collect final variants of amplifiers. I shall try to provide 100 % stability
Whether so it?
And how it to be reflected in a sound?
Many thanks deleveld
Many thanks darkfenriz
And many thanks ilimzn
For your comments,
Soon I shall collect final variants of amplifiers. I shall try to provide 100 % stability
3F diff
interesting, you probably don't need R60 & R61
I guess you keep output stage in class A?
It may need freq. compensation also, maybe not..
source resistors are pretty high, damping factor will be less than 30, maybe even less than 20, but some don't bother.
I see your head is full of fresh ideas-keep on.
regards
interesting, you probably don't need R60 & R61
I guess you keep output stage in class A?
It may need freq. compensation also, maybe not..
source resistors are pretty high, damping factor will be less than 30, maybe even less than 20, but some don't bother.
I see your head is full of fresh ideas-keep on.
regards
OP_lateral: used this one myself, and stopped precisely for the reasons Darkfenriz mentioned.
In real life TL071 would be completely inadequate for this as it is not possible to reliably drive 330 ohm loads with sufficient current to open up either output transistor. Keep in mind that you need about 6-7V there, across 33 ohms, which means 212mA - well outside what the TL071 can do. You would need to increase the 33 ohm resistors.
Also, take a look at R21 and R22 - these really do nothing except increase the open loop output impedance of the amp and drop voltage from the rails. Since your driving signal from the OPamp is it's output current, the drop on the 33 ohm resistors remains the same regardless of the voltage on the 0.22 ohm resistors. Therefore, these are not inside the control 'loop', so they do not degenerate the output transistor gain, nor do they thermally stabilise the circuit.
It is possible to adjust idle current by means of applying an adjustable current source between Q1 and Q2 emitters. Depending on what MOSFETs you use, this current source also needs to be thermally compensated.
A possible permutation of this circuit results in the output working in current feedback mode. The feedback is then connected from the output of the whole amp, to the output of the OPamp. The voltage feedback for the OPamp is then connected to the output of the OPamp, and sets amplifier input sensitivity. An arrangement where the DC voltage feedback is connected from the output of the whole amp is also possible to reduce output offset voltage.
It should be noted that in current feedback mode, it is possible to cinfigure the OPamp for effective non-inverting gain less than 1. Some OPamps distinctly dislike this and will become unstable. LF351 would be an example - from experience with a very similar design of my own.
3diff:
What Darkfenriz said, plus the following:
As you are running the amp without global feedback, you will need to run the outputs at a fairly high bias - I'm just now playing with a similar output topology, but in my case I have only one output pair and no source resistors. I still need over 100-250mA of idle current.
One thing that REALLY needs to be changed is this: The 10k pot to set output idle current MUST be a Vgs or degenerated Vbe multiplier. The chosen output transistors will have a positive temperature coefficient of idle current even with 1 ohm source resistors, which means that there will always be a measure of thermal runaway potential. Many people suggest that this is not the problem if the ehatsink is large enough, but I'd rather design it properly and independent of heatsink size for runaway potential. You do not want your idle current wandering all over the place, it needs to be stable, or, considering the gm of the MOSFETs increases with temperature for current below some 12-14A, it may even me beneficial if the idle current decreases slightly with temperature.
Possible improvements:
Replacement of R5 and R6 with current sources. This will prevent bias current for M5 and M6 as well as output stage depending on the front end power supply voltage.
BC547/557 may need to be replaced with something with higher Vcemax in real life. It is good practise to plan on the input pairs to withstand the full supply rail to supply rail voltage - otherwise, if there is ever a failure, the input stage can get damaged in rather nasty ways, as in it works, but the transistors leak. I've fixed too many such amps...
You may also want to investigate merging R61/R40 and R60/R39 into a folded cascode configuration. Look at passdiy.com and find the A75 amplifier, for an example.
One more thing - since you are using a simulator, it is a good idea, once the amp simulates at default environment parameters, to try variations in power supply voltage, and especially temperature, and look at what happens. You may be shocked...
In real life TL071 would be completely inadequate for this as it is not possible to reliably drive 330 ohm loads with sufficient current to open up either output transistor. Keep in mind that you need about 6-7V there, across 33 ohms, which means 212mA - well outside what the TL071 can do. You would need to increase the 33 ohm resistors.
Also, take a look at R21 and R22 - these really do nothing except increase the open loop output impedance of the amp and drop voltage from the rails. Since your driving signal from the OPamp is it's output current, the drop on the 33 ohm resistors remains the same regardless of the voltage on the 0.22 ohm resistors. Therefore, these are not inside the control 'loop', so they do not degenerate the output transistor gain, nor do they thermally stabilise the circuit.
It is possible to adjust idle current by means of applying an adjustable current source between Q1 and Q2 emitters. Depending on what MOSFETs you use, this current source also needs to be thermally compensated.
A possible permutation of this circuit results in the output working in current feedback mode. The feedback is then connected from the output of the whole amp, to the output of the OPamp. The voltage feedback for the OPamp is then connected to the output of the OPamp, and sets amplifier input sensitivity. An arrangement where the DC voltage feedback is connected from the output of the whole amp is also possible to reduce output offset voltage.
It should be noted that in current feedback mode, it is possible to cinfigure the OPamp for effective non-inverting gain less than 1. Some OPamps distinctly dislike this and will become unstable. LF351 would be an example - from experience with a very similar design of my own.
3diff:
What Darkfenriz said, plus the following:
As you are running the amp without global feedback, you will need to run the outputs at a fairly high bias - I'm just now playing with a similar output topology, but in my case I have only one output pair and no source resistors. I still need over 100-250mA of idle current.
One thing that REALLY needs to be changed is this: The 10k pot to set output idle current MUST be a Vgs or degenerated Vbe multiplier. The chosen output transistors will have a positive temperature coefficient of idle current even with 1 ohm source resistors, which means that there will always be a measure of thermal runaway potential. Many people suggest that this is not the problem if the ehatsink is large enough, but I'd rather design it properly and independent of heatsink size for runaway potential. You do not want your idle current wandering all over the place, it needs to be stable, or, considering the gm of the MOSFETs increases with temperature for current below some 12-14A, it may even me beneficial if the idle current decreases slightly with temperature.
Possible improvements:
Replacement of R5 and R6 with current sources. This will prevent bias current for M5 and M6 as well as output stage depending on the front end power supply voltage.
BC547/557 may need to be replaced with something with higher Vcemax in real life. It is good practise to plan on the input pairs to withstand the full supply rail to supply rail voltage - otherwise, if there is ever a failure, the input stage can get damaged in rather nasty ways, as in it works, but the transistors leak. I've fixed too many such amps...
You may also want to investigate merging R61/R40 and R60/R39 into a folded cascode configuration. Look at passdiy.com and find the A75 amplifier, for an example.
One more thing - since you are using a simulator, it is a good idea, once the amp simulates at default environment parameters, to try variations in power supply voltage, and especially temperature, and look at what happens. You may be shocked...
ilimzn said:
Really. Both circuits OP_lateral and 3diff are made only on a simulator, but 3diff was collected and worked 1 year, but with a global feedback,
Yes in real circuits are very far from a simulator, but the simulator gives an approximate picture it works whether or not.
The simulator basically is used for this purpose, and also for drawing circuits (conveniently).
Vgs multiplier - laziness to draw (excuse)
Op_lateral What OPamp is better for using.
Hi ! One of the first projects, please your comments,
the project up to this time has remained only in a simulator, laziness to collect,
though should work and very much it is not bad, OP_amp on an input(entrance)
it is possible to replace on discrete?))) Detailed face values if are necessary write.
Excuse, that not having completed the first projects I offer new, but in internet recently, in the theory very bad (when learned many saws of beer), therefore is very glad that my projects were though are a little interesting! Thank.
the project up to this time has remained only in a simulator, laziness to collect,
though should work and very much it is not bad, OP_amp on an input(entrance)
it is possible to replace on discrete?))) Detailed face values if are necessary write.
Excuse, that not having completed the first projects I offer new, but in internet recently, in the theory very bad (when learned many saws of beer), therefore is very glad that my projects were though are a little interesting! Thank.
Interesting configuration!
R55 and 56 are not really necessary, and Q7 will need a small resistor in the emitter or it will overcompensate the FETs.
The problem that i see is a large rail voltage loss to the output - at a glance at least some 8V, posibly more. The folded cascode drivers need a set of higher supply rails...
Regarding your earlyer question about the 'lateral' amp, which OPamp to use - it's unlikely you'll find a good one for the other components being what they are. you need to increase the current sensing resistors and the OPamp load resistor until your OPamp output current can drive it all properly. The simple choice would be a NE5534, as long as you remember to set it's gain >5, or compensate it externally for lower CL gain. Keep in mind that the NE5534 has a rather high quiescent current so you may not be able to find the ideal combination of resistors. On it's own, the NE5534 will drive a 600 ohm load within a couple of V of it's power rails, which means you can get about +-20mA out of it with low distortion (less distortion for less loading!).
R55 and 56 are not really necessary, and Q7 will need a small resistor in the emitter or it will overcompensate the FETs.
The problem that i see is a large rail voltage loss to the output - at a glance at least some 8V, posibly more. The folded cascode drivers need a set of higher supply rails...
Regarding your earlyer question about the 'lateral' amp, which OPamp to use - it's unlikely you'll find a good one for the other components being what they are. you need to increase the current sensing resistors and the OPamp load resistor until your OPamp output current can drive it all properly. The simple choice would be a NE5534, as long as you remember to set it's gain >5, or compensate it externally for lower CL gain. Keep in mind that the NE5534 has a rather high quiescent current so you may not be able to find the ideal combination of resistors. On it's own, the NE5534 will drive a 600 ohm load within a couple of V of it's power rails, which means you can get about +-20mA out of it with low distortion (less distortion for less loading!).
ilimzn said:
The majority of circuits are collected, really, from that that was (on a knee, from devices there is only a multimeter?))), therefore it is limited Urail, and components, in a simulator I can give anyone Urail, but I first of all do(make) achieving 100 % of repeatability. Therefore I ask your comments to improve.
Are you using translation software? If yes, it is not doing a very good job, it is often very difficult to understand what you mean. I immagine it is the same for you when my answers are translated back to Russian...
It is actually very easy to get extra higher voltage rails (again, look at the power supply for the A75 on passdiy.com). A few extra diodes and capacitaors will not break the bank and can offer better results. It is especially easy because you do not reeally need a lot of current, so voltage stabilization is simplified.
It is actually very easy to get extra higher voltage rails (again, look at the power supply for the A75 on passdiy.com). A few extra diodes and capacitaors will not break the bank and can offer better results. It is especially easy because you do not reeally need a lot of current, so voltage stabilization is simplified.
ilimzn said:
Yes, using translation software. It is awful
ilimzn said:
( the Multiplier of a voltage)----it is interesting, did not think. At me old disassembled (it is a pity, and it was Øóøóðèí---LAMM, but such has got), amplifier Amfiton, on his(its) base and are done(made) these circuits.
(((((( (Óìíîæèòåëü íàïðÿæåíèÿ) ---- î÷åíü èíòåðåñíî , íå äóìàë( íî êàê ýòî ïîâëèÿåò íà çâóê). Ó ìåíÿ ñòàðûé ðàçîáðàííûé
( æàëü, à ýòî áûë Øóøóðèí ---LAMM, íî òàêîé äîñòàëñÿ), óñèëèòåëü Àìôèòîí, íà åãî áàçå è äåëàþòñÿ ýòè ñõåìû.
Îò óñèëèòåëÿ îñòàëñÿ òîëüêî êîðïóñ è òðàíñôîðìàòîð, â í¸ì ñîáðàë ïåðâûì Zen ( ñ 3 âåíòèëÿòîðàìè, íà êàíàë,) âûñîòà óñèëèòåëÿ ïðèìåðíî 8 ñì. Ðàáîòàë ïîëãîäà . Ðàçîáðàí. È ïîíåñëîñü, êàæäûé ìåñÿö ñîáèðàåòñÿ íîâàÿ ñõåìà, ðàçáèðàåòñÿ è âñ¸ ñíîâà, íî íàäîåëî êîïèðîâàíèå èçîáðåòåíèé 70-õ ãîäîâ, òîëüêî Ïàññ âí¸ñ ÷òî-òî íîâîå, âñ¸ îñòàëüíîå òîò æå âåëîñèïåä èçîáðåò¸ííûé ïðåäêàìè, äàæå ìîé âîïðîñ ïðî âûõîäíîé êàñêàä âêëþ÷åííûé ñ îáùåé áàçîé(çàòâîðîì) ýòî ïîïûòêà ñêîïèðîâàòü êîãî-òî , íàäî ÷òî òî êàðäèíàëüíî íîâîå , Âû òàê íå äóìàåòå .
À ýòèõ "êîïèé " ó ìåíÿ áîëüøå ñîòíè , è êàæäàÿ ïî ñâîåìó èíäèâèäóàëüíà, íî íå èíòåðåñíî êîïèðîâàòü. ß ïðàâäà î÷åíü ïðèçíàòåëåí çà Âàøè êîììåíòàðèè, òåîðèÿ ìîÿ óòåêëà â ïèâî , à æàëü, íî âñ¸ æå ìåíÿ ÷åìó-òî ó÷èëè , â ÑÑÑÐ , è ÿ õîòü ÷òî-òî âñïîìèíàþ. Ñïàñèáî. Î÷åíü ðàä îáùåíèþ. Ýòî òàêîå íîâîå äëÿ ìåíÿ.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.