Follow this: Copy the Reflektor-D schematic from its guide, but instead of R6 put a 2SK117GR about 4mA strong with D toward R4 and G,S shorted together toward S0. Use in the Vref string (all parts that C2 encompasses) four LM329 in series with cathodes looking up like Zeners, a green LED with cathode opposite to the LMs and a 2K trimmer in rheostat mode.
I hope your oscilloscope is faster even, because you will have to check that it will be stable at those voltages and parts too in your own layout.😀
Oscillation issues. I had to get the rig up and running, so I defaulted to the successful BiB. I will work out the bugs and retry. You could hear it. Wrong kind of audio.
No. One channel did not work. The other did, but. When turned off, the LED would flicker, diminishing in both speed and intensity as the current drained. I will have a look at it again this weekend. Need to check layout first.
I guess the BD140 are to give higher Vce, what's the benefit of using a CCS to replace R6, would it be helpful in lower voltage application also?
He needs a 2K trimmer also bcs he asks for smoothly changing 32-36V range of Vo. So having a CCS on one leg, steadier current is mirrored to the other leg over the trimmer not to thermally drift much the Vo as a whole system by I/V across that largish value passive ref part. Side-effects are more open loop gain but in less open loop bandwidth and less phase margin and an extra active noise source component also. What is preferred it depends on the client app. Analog preamp could appreciate the lower Zo due to higher OLG for its swings and stages but digital chip with local LC filter, final SMD local reg etc. already has Z between and could appreciate the less reactive nature and less 1/F noise of passive part R6. For those curious in lower Vo or even in digital apps powering they can experiment the same in the 5mm official pcb's space for VAR R6 exchanging with G+S prepared two legs to solder GR JFET and decide. But careful, it will boost the LEDS output a little, plus it must be rechecked that the rail is still stable when having the real application as load.
Is seems reasonable that if the specs are more defined, then other options present themselves. What I will most likely do is get the RIAA up and running. Its needs will be more well defined and then I can be more accurate with my needs. I am going to look into a current mirrors operation to understand it a bit better.
Small variations in the Reflektor config will better suite one or other reg+app system nicer, but better have a basic working system first as you say. And then its only fun to taste or tune.
Hi Salas,
I want to use reflektor D to supply my SDTRANS 384 but i need 5V at > 0.5 A (at the moment it's running on a BIB).
This is a smidgen over recommended max , do you recommenced changing power transistors ?
I'm using a separate PSU box with ample external heatsinking for this and associated BIII dac and IV ( EUVL ), so getting rid of the heat should not be a problem.
I want to use reflektor D to supply my SDTRANS 384 but i need 5V at > 0.5 A (at the moment it's running on a BIB).
This is a smidgen over recommended max , do you recommenced changing power transistors ?
I'm using a separate PSU box with ample external heatsinking for this and associated BIII dac and IV ( EUVL ), so getting rid of the heat should not be a problem.
Set to 0.73A with 0.82R R1 and no more than 9-10VDC across the main filter cap and it will take it even on 50mm tall board level sinks.
- Status
- Not open for further replies.