I would make the 90-degree tracks 45-degree camfer. Makes them shorter.
Looks fine otherwise. What is that open connection from L-in all the way to the left? Is it an error in the schematic?
Jan
Looks fine otherwise. What is that open connection from L-in all the way to the left? Is it an error in the schematic?
Jan
It is the analog ground. I was thinking to leave it open between left and right channels or is it better to connect them in PCB?
Is there any particular tracks you recommend ,that must be made shorter. not about this particular pcb but in your design.
In general which is more harmful routing between upper and lower surfaces vs a longer track?
Is there any particular tracks you recommend ,that must be made shorter. not about this particular pcb but in your design.
In general which is more harmful routing between upper and lower surfaces vs a longer track?
Last edited:
Run both the hot and screen from each RCA plug to the points on the PCB.
The same with the connections to the output connectors. Run them as if they are two separate channels just passing over the PCB.
Then, at one point only, connect the analog grounds to the supply ground. This should work well here since the supply is on the same PCB, you are effectively creating a star ground at the PCB.
I am a fan of shorter is better so I would always use a double sided PCB, but here it's just a personal preference; this relatively simple circuitry will surely be fine on a single sided PCB.
Jan
Forgot to mentiuon, ground the ref and feedback dividers of the regulator also to that central ground point. That point now becomes by definition 'clean ground/zero volts'.
Jan
Jan
Thanks Jan.That was what I was looking for. Seperate analog and power grounds seperate until somewhere they meet outside the PCB ( before the power supply bypass caps?)
Also the zener and Zener noise is it important to look for that in picking D2?i picked this one randomly.
BZX84J-B7V5,115 Nexperia | Mouser
Also the zener and Zener noise is it important to look for that in picking D2?i picked this one randomly.
BZX84J-B7V5,115 Nexperia | Mouser
Sorry Jan, when I read more , I think I may not have understood it right. So two channels with the screen of cable coming to PCB in , then connect them using seperate traces on PCB with power supply ground?
Or is it to keep all grounds seperate and connect them using cable To a point outside PCB?
Or is it to keep all grounds seperate and connect them using cable To a point outside PCB?
The first. Just pick a spot on the PCB, probably somewhere near the center, and route all grounds to that. RCA screens, power supply output ground, reference ground (the ref diode bottom), the bottom of the feedback resistor divider that sets the supply output voltage.
The input to the power supply is just two wires coming from the rectifier/caps, and are not grounded anywhere except on the PCB.
Edit: you may want to put an extra hole on the central PCB ground point so you can use that to connect to the chassis.
Jan
Last edited:
Jan, sorry with my silly questions and appreciate your patience with me. The points you mentioned above are not the trace forming the positive return sense trace in your Schematic ? What I understand is LM329 is the voltage reference diode and R6 and R7 in positive regulators forms the resistor divider. So instead of taking the trace closest to the circuit/ part it powers along with ground trace as it shows in schematic route it to the star ground at the central part of PCB? So there will be only three traces going to the circuit it powers instead of 4 ( kelvin connection)
Last edited:
Try Vishay BZX series for higher voltage outputs , use it both for v ref and output series zener. I used 15 volt Vishays for 30 volt output Super regulators.
A lot easier than trying to find optimal resistor values on the voltage divider. Most zeners aren't this quiet
Low noise as per Christers tests below
BZX55/C12 (0.5W 12V):
#1 @ 1mA: 0.35 0.37 0.37 0.39 0.39 uV
#1 @ 5mA: 0.30 0.28 0.28 0.28 0.30 uV
#1 @ 20mA: 0.24 0.25 0.25 0.26 0.25 uV
#2 @ 1mA: 0.32 0.33 0.32 0.33 0.32 uV (Vr = 11.32 V)
#2 @ 5mA: 0.26 0.26 0.27 0.32 0.26 uV (Vr = 11.37 V)
#2 @ 20mA: 0.25 0.26 0.28 0.24 0.30 uV (vr = 11.42 V)
A lot easier than trying to find optimal resistor values on the voltage divider. Most zeners aren't this quiet
Low noise as per Christers tests below
BZX55/C12 (0.5W 12V):
#1 @ 1mA: 0.35 0.37 0.37 0.39 0.39 uV
#1 @ 5mA: 0.30 0.28 0.28 0.28 0.30 uV
#1 @ 20mA: 0.24 0.25 0.25 0.26 0.25 uV
#2 @ 1mA: 0.32 0.33 0.32 0.33 0.32 uV (Vr = 11.32 V)
#2 @ 5mA: 0.26 0.26 0.27 0.32 0.26 uV (Vr = 11.37 V)
#2 @ 20mA: 0.25 0.26 0.28 0.24 0.30 uV (vr = 11.42 V)
Sorry no, I didn't see your post until now.
I wasn't aware you are using the remote connection from the superreg. That is really for cases where load and supply have appreciable separation. Since you have everything on one PCB, there's no need.
Everything that is connected to supply ground and remote bround can all be connected to a central point.
Jan
Really? I can easily find 1% or even 0.1% resistors of all kind of funny values. Try that with zeners - you're lucky to find 5% values.
Jan
Sigh!I thought I tested your patience.
Thanks.so I will avoid sensing trace for ground !
Does that means I can use one V+ connection for load too, or you recommend seperate traces /connections as load and sense ?
Jan,I am not sure if this circuit uses low ESR cap(must) everywhere ? I read output cap should not be LOW ESR ,but which one are the output CAP?
Any cap connected to the output ;-)
If there are film caps further on as part of the preamp that's probably OK. But be sure to have an electrolytic as output cap for the regulator.
Jan
Jan, sorry. You answered yes to both my questions in# 1355. So I can skip sense and load traces for positive traces ?
Also are the electrolytic cap used in regulator all are Low ESR ? I read in earlier post to use to avoid low ESR caps as output caps, but not sure which on in the schematic is an output cap? C4 ?
Also are the electrolytic cap used in regulator all are Low ESR ? I read in earlier post to use to avoid low ESR caps as output caps, but not sure which on in the schematic is an output cap? C4 ?
Last edited:
Also I was reading the original Article which used the supply to opamp taken before the pass Transistor with filter caps.Did you change that in new boards to avoid oscillation?
Juan, the tolerance of the LM 329 output voltage means having to measure each one for it's voltage and calculate the resistors for each one, and buy 1 or 2 of each value., since I am using the super regulator as a +\- 30 volt pre -regulators for a shunt regulator I'm not too worried about the extact output voltage and so keeping the 1 k resistors and using the Vishay BZX -15 as V ref works for me.
Note: when using the Super regulator at higher output voltages you really want to use 1/2 watt resistors for the voltage setting resistors as they get warm due to the low circuit impedance Walt used to keep the noise low. At 30 volts output they dissipate a 1/4 watt each if you use a 15 volt reference as I did.
Ideally the series resistor for setting the current for the reference voltage would be optimized for each voltage etc. in reality while you can scale just voltage setting resistors or change the voltage reference, optimally each output voltage requires tweaking of several resistors, depending on the V reference used which opens up a new can of worms.
Note: when using the Super regulator at higher output voltages you really want to use 1/2 watt resistors for the voltage setting resistors as they get warm due to the low circuit impedance Walt used to keep the noise low. At 30 volts output they dissipate a 1/4 watt each if you use a 15 volt reference as I did.
Ideally the series resistor for setting the current for the reference voltage would be optimized for each voltage etc. in reality while you can scale just voltage setting resistors or change the voltage reference, optimally each output voltage requires tweaking of several resistors, depending on the V reference used which opens up a new can of worms.
Last edited:
You could also use an LM4040-5V reference, you can get those at 0.1% if that is your worry.
I fail to see how +/- 0.5V at the Vout could be relevant.
Jan
I fail to see how +/- 0.5V at the Vout could be relevant.
Jan