Try with mixing one blue Led and I will try too so to see if it has any real negative impact on noise or not.
I don't know about buffalo to expertly tell you. You will see in practice.
I am looking on some noise measurements, and mixing one 3V9 0.5W Zener or a blue Led with the other two, looks safe by the way. Better the blue, the Zener will need reverse orientation and we don't know if it will be that stable without strong current.
Thanks for the input. 3xblue can be too much voltage though, I would mix just one firstly.
To Salas and people concerned in Buffalo ii+ Tridents and Legato. I´ve never measured the current draw, because my first exerience with Trident was the my PS then,(Placid 350mA) gave not enough current to lite the leds in local AVCC and Tridents. Then I gave as good what I read in forums: idle current for Buffalo+tridents 450-475mA and 300 mA for Legato.
Today I have measured the idle current with the following results:
Buffalo + Tridents: 410mA
Legato 2 and 3.1 (without Buffers):Between 190 and 200 mA.
So that´s the departure point (for me)
Today I have measured the idle current with the following results:
Buffalo + Tridents: 410mA
Legato 2 and 3.1 (without Buffers):Between 190 and 200 mA.
So that´s the departure point (for me)
Hi Salas
I'm riding a version for 5.25 V output with ccs 470 ma. I think I've managed with R1 3R3 and R305 1k trimmer (jumper r303). Through r1 I have 1.55v and 5.25v in the output .. good!
My guess is that with R1 3R9, the maximum voltage at the outputs is 4.9 v. Although r305 (trimer) rises above 300 ohms. However, with R1 3R3 I can get with the trimer (1k) the 5.25 volts I want.
This ok?
Thanks!
I'm riding a version for 5.25 V output with ccs 470 ma. I think I've managed with R1 3R3 and R305 1k trimmer (jumper r303). Through r1 I have 1.55v and 5.25v in the output .. good!
My guess is that with R1 3R9, the maximum voltage at the outputs is 4.9 v. Although r305 (trimer) rises above 300 ohms. However, with R1 3R3 I can get with the trimer (1k) the 5.25 volts I want.
This ok?
Thanks!
Use BJT output as in the guide if you want the highest bandwidth of flat Zo it can give. It goes higher Zo mOhm in total is the flip side.
At 100MHz even if we had an ideal voltage source I doubt that any layout and wiring would not rise impedance either on the reg or in the receiving circuit anyway. Radios, communication, and computers succeed to work on LM317 kHz bandwidth flat Zo and then shooting through the roof style, and SMPS to keep the world going, so we are in better hope with this reg I would dare guess.
At 100MHz even if we had an ideal voltage source I doubt that any layout and wiring would not rise impedance either on the reg or in the receiving circuit anyway. Radios, communication, and computers succeed to work on LM317 kHz bandwidth flat Zo and then shooting through the roof style, and SMPS to keep the world going, so we are in better hope with this reg I would dare guess.
thanks Salas, yeah i had already decided that the bjt is the one for the job, thats what the singular lv bjt+ kit I ordered was for, its a better compromise to have flatter although a touch higher z output for this situation. i just wondered if you had seen any sims or plotted what happens at much higher frequency, as the graph only goes to 100khz, a factor of 1000x different to my load and the mosfet sim started a pretty steep upward trend towards the edge of the plot, i wondered if the bjt did the same, but later. i guess a really nice teflon cap is in order and a well bypassed XO directly across the pins with an np0
I was considering using the new lt ultrafast transient response mosfet controller LT1575, which basically does what it sounds like and drives an external mosfet as a source follower, its designed for exactly this type of load, primarily high speed CPUs, considerably better than the ol' 317
I was considering using the new lt ultrafast transient response mosfet controller LT1575, which basically does what it sounds like and drives an external mosfet as a source follower, its designed for exactly this type of load, primarily high speed CPUs, considerably better than the ol' 317
Last edited:
It should be about 1 Ohm at 100MHz with the BJT at the output. At those frequencies, if your digital circuit isn't best decoupled and awesome layout, nothing can save it. I doubt there aren't traces of the clock and multiples on rails, gnd, and on field if poking around with a 1GHz oscilloscope even on air defense grade high clock digital equipment.
oh the cct is well designed and well decoupled (ackodac AKD12P PTFE version sabre dac), but it usually uses a lifepo4 battery powered AD797 based reg that is able to be placed considerably closer than i will be able to put the shunt due to physical size. I guess we'll see how i go. it may turn out better to use the shunts for the AVCC L/R and reference supplies, but use the ad797 reg for the clock.
Last edited: