HURZ I FIXED IT MOSTLY

Very theoretical circuit with 3.3V supply. Its easy to make this working in EC with perfect current sources. Check the voltage at center current source 100uA which goes even above 3.4V! Probably increase the supply significant and try to replace the current sources with real ones then it gets a challange! If you managed this we are talking about temperatur issues. All in all this will be a quite complicated circuit in discret ;-)

Long trip to work. Might give it a go lol.

I will follow you ;-)

Yes, now it gets closer to a discret solution. You see 3.3V power supply was a bit to low. There is a good reason why discrect bipolar ICs have a supply voltage of more than 3V. The last design i have been worked long time ago, we had 5V as supply voltage and it was already a optimised version, we come from 12V supply years before. The competition at that time still had 9V. Suppose a quadrature mixer were you have stacked 4 or even 5 bipolar transistors! 5*0.7 is alread 3.5V and does not leave much headroom for current mirrors and the signal itself. However, you can simplify you circuit, it does not need for all current sources again a reference source. You can reuse the 42.6kOhm reference current as current source bank with some emitter resistor to control the current multiplication. And as long as the currents are that low you dont run into temperatur problems. Anyway, discret transistors are not matching, they are on separated silicon and have different charateristics. So some emitter tuning resistor are needed anyway. Sorry for late response, I've been on vacation :-)

Thanks for sharing anyway. This was mostly a curiosity thing (always wondered how these simplified schematics actually implement current sources). Might come back to it and improve things when I'm bored.