I'd say the current mirror would work equally well with 470ohms emitter vs 4.7k ohms emitter.
The main thing you want for best operation is perfect matching.
I do not understand where that guy was comming from, where emitter resistance higher then your transistors are not a real current mirror anny more.
The main thing you want for best operation is perfect matching.
I do not understand where that guy was comming from, where emitter resistance higher then your transistors are not a real current mirror anny more.
Why are you attack me, if You don't understand the main benefit of using current mirror????
Sajti
You said that if you make the emitter reistance too high, then the transistor pair isn't a "true current mirror any more".
I do not see where your comming from, and I do not think this is correct
If I actually have allot of time which I do not, I would make my own version of this high freq op amp as a test model.
However due to lack of time, I don't even have time to build any more productive amp schematics which I designed.
Last edited:
You can always build a amp on a prototype board if your confident and exprienced.
It takes me normally 1 day to make a test model on a prototype board, if everything goes right first time I may finish a channel in a day if not takes longer.
I normally perfer to save money so I can buy more components. Sometimes things don't work out with pcb then the money is wasted.
It takes me normally 1 day to make a test model on a prototype board, if everything goes right first time I may finish a channel in a day if not takes longer.
I normally perfer to save money so I can buy more components. Sometimes things don't work out with pcb then the money is wasted.
Last edited:
you are right, I must buy some for these things ...Prototype board is a trade off, your circuit takes up more space, and your much more likeley to make connetion error when buiding. It is also more time consuming using prototype. Depends on personall prefrence
I normally use this one for most projects, more than enough for a channel with relay. installed.
20Pcs 9 x 15 cm DIY Prototype Paper PCB Universal Board Good quality-in Integrated Circuits from Electronic Components & Supplies on Aliexpress.com | Alibaba Group
You can also get 10x22 and 12x18 and 5x7 boards
I normally use this one for most projects, more than enough for a channel with relay. installed.
20Pcs 9 x 15 cm DIY Prototype Paper PCB Universal Board Good quality-in Integrated Circuits from Electronic Components & Supplies on Aliexpress.com | Alibaba Group
You can also get 10x22 and 12x18 and 5x7 boards
these boards you think, I thought you mean these solderless:
Mb-102 Mb102 Breadboard 830 Tie Point Solderless PCB Bread Board for Arduino | eBay
... I look into that what you sent ...
Mb-102 Mb102 Breadboard 830 Tie Point Solderless PCB Bread Board for Arduino | eBay
... I look into that what you sent ...
The basic idea behind the current mirror, to keep the voltage change low (few mV only!), to drive the output bjt. This small change means high speed, because it don't have to charge/discharge the internal capacitances.
If You check the schematic provided by kct, using 5.76k/3.92k, You will see that the voltages are very high. And what happening, if You simply shorting the C-E of Q4, and Q8? The output current of Q3 and Q9 will be lower with some 5%, but the amplifier will work. What does it mean? It means, that Q3, and Q9 are simple common emitter stage, driven from 5.76k source impedance, which slowing down the whole amplifier. So Q4, Q8 does practically nothing, just add 0,6V voltage shift, and some thermal compensation. This is not a true current mirror any more.
Sajti
10pcs of them costs only 5USD, if You send the gerbers to any chinese pcb manufaturer...
Sajti
yes Sajti, I sent them, with shipping it cost about 8USD, but I ordered only 5PCS (because this is only small test), if it pass, I post gerbers here ... there can be used any good SMD input transistors where current resistors values are small (for current mirrors) ... do you know some SMD replacement for KSC1845/KSA992 ?
Sorry but I nowhere near agreeing with you, nor am I convinced.
The basic idea behind the current mirror is to keep the current of two transistors or more equal to each other.
Having larger emitter resistor means nothing but less ce voltage beacsue more voltage is being dropped through the resistor.
Driven from a higher source impedance dosen't mean its slower. Driven from a source with higher capcaitance means its slower.
I wounder how many amps you have designed from scratch.
You obviously won't understand why a current mirror is not the solution to best audio.
This is the main reason why this high freq amp is not considered to be built as one of my project.
Speed of a amp means nothing, human hears up to only 20khz. HI FI is not Raido Frequency going faster means nothing.
I think your understanding of emitter resistance and current mirror is wrong and you should consider re studying it.
You obviously won't understand why a current mirror is not the solution to best audio.
This is the main reason why this high freq amp is not considered to be built as one of my project.
Speed of a amp means nothing, human hears up to only 20khz. HI FI is not Raido Frequency going faster means nothing.
I think your understanding of emitter resistance and current mirror is wrong and you should consider re studying it.
The only reason I can think of in why your saying emitter resistance higher means slower speed is that it takes more time for current to travel into next stage.
(becasue the bias current is less, then the ammount of base current next stage draws becomes a bigger fraction of bias current of first stage).
I'd also say these effects are negligible.
But if you assume that the transistor is perfectlly buffed or next stage dosen't draw any base current then I do not see how speed would be deducted.
(becasue the bias current is less, then the ammount of base current next stage draws becomes a bigger fraction of bias current of first stage).
I'd also say these effects are negligible.
But if you assume that the transistor is perfectlly buffed or next stage dosen't draw any base current then I do not see how speed would be deducted.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.