??? The RS+ & RS- connectors are also on the PCB from ikoflexer. ???
I thought, if the load is not near by the PCB, you must connect the RS+ & RS- additionally between the PCB & the load....
The two points are on the pcb, but they don't connect on the pcb to the shunt mosfet. They are supposed to be connected to the load via wires. If you keep the distance between the regulator and load very small then don't bother with wires for remote sensing, just connect them on the pcb to the shunt mosfet.
Do yourself a favour and add a couple of small resistors (10ohms will do) between RS-out's and RS-in's. This will protect the load if by chance you forget to connect the sense wires at the load terminals.
And please do something about that self advertising. It looks whorish.
I'm lucky, I'm on his ignore list 
Oops, just realized the Scorpions are playing "Can't live without you" ... a mere coincidence? Oh no, and they're from Germany too :O
P.S. Ok, it's been year since I've listened to the Scorpions. Came across the Blackout album in a second hand store the other day and nostalgia won, couldn't pass it for $1.49. It also happens I don't believe in coincidences. After I wrote this message (jokingly of course), I was washing dishes and there comes the next song on the disc... http://www.youtube.com/watch?v=-YNUu7w-WpU which I dedicate respectfully to syn08 There, now you know... the rest of the story.
Oops, just realized the Scorpions are playing "Can't live without you" ... a mere coincidence? Oh no, and they're from Germany too :O
P.S. Ok, it's been year since I've listened to the Scorpions. Came across the Blackout album in a second hand store the other day and nostalgia won, couldn't pass it for $1.49. It also happens I don't believe in coincidences. After I wrote this message (jokingly of course), I was washing dishes and there comes the next song on the disc... http://www.youtube.com/watch?v=-YNUu7w-WpU which I dedicate respectfully to syn08
There, now you know... the rest of the story.
Last edited:
Thank´s for your advice. But hey, what self advertising?? The dvb-projekt writing on the PCB is on ikoflexers request
An update should be in order. First, I'd like to apologize for being so slow. Although I'm doing terribly as far as spare time goes, I managed to move the project forward a bit. I think the pcb layout is final for both positive and negative rails. The changes were minor, mostly aesthetic. None of the pcb layout software that I tried (both commercial and free) makes the component footprints to my liking, so I've spent quite a bit of time to develop custom footprints for all. Have you noticed the tiny pads that most software makes? IMHO those pads are for robot assembly. We people, need thicker pads. So I spent quite some time to redo component footprints. Nice thick pads, pads that aren't necessarily aligned as a straight line should be placed accordingly, plenty of room between traces and pads, allowance for both to-220 and to-247 power mosfets, for most flexibility, and so on.
I've run further tests on the positive regulator, it works well. Lately people have been talking about oscillations. No oscillations observed with a Tek scope that can do up to 60MHz. As far as I can say the positive rail pcb is done and tested, ready for pcb fabrication. The negative pcb is being etched right now, so I'll report on it soon.
I've run further tests on the positive regulator, it works well. Lately people have been talking about oscillations. No oscillations observed with a Tek scope that can do up to 60MHz. As far as I can say the positive rail pcb is done and tested, ready for pcb fabrication. The negative pcb is being etched right now, so I'll report on it soon.
About shielded cable for the sense wire.
I thought shielded cable is only needed for the voltage sense but not the ground sense, as the voltage potential of the ground sense would be too low to pick up any noise. Correct me if I am wrong.
An interesting thing I discovered today was the extended ground (shield) wire of the oscilloscope actually picked up a lot of noise. The ground wire was only extended for about 20cm and it made such a difference (2mV?). So I am no longer sure hence the above question.
I thought shielded cable is only needed for the voltage sense but not the ground sense, as the voltage potential of the ground sense would be too low to pick up any noise. Correct me if I am wrong.
An interesting thing I discovered today was the extended ground (shield) wire of the oscilloscope actually picked up a lot of noise. The ground wire was only extended for about 20cm and it made such a difference (2mV?). So I am no longer sure hence the above question.
I reported earlier that after 2 days run-in the sound was a bit smoother. Since the only electrolytic capacitor I use is C3, which serves the function of improving the line regulation. So a "better" C3 might have improved the sound.
With that in mind, today I experimented shunting it with a 0.1uF MKP to provide a low impedance at high frequencies. The sound is slightly different, possibly to the better. I need to do comparisons for a few times before I can make a conclusion.
I normally have high suspicion on bypassing electrolytic capacitors. If not done properly it introduces more problems than it solves, usually causing ringing. However, in that position, the wire length may be sufficiently small (i.e. low inductance) for ringing to be a concern.
Anyone wants to experiment please report back.
With that in mind, today I experimented shunting it with a 0.1uF MKP to provide a low impedance at high frequencies. The sound is slightly different, possibly to the better. I need to do comparisons for a few times before I can make a conclusion.
I normally have high suspicion on bypassing electrolytic capacitors. If not done properly it introduces more problems than it solves, usually causing ringing. However, in that position, the wire length may be sufficiently small (i.e. low inductance) for ringing to be a concern.
Anyone wants to experiment please report back.
Last edited:
Andrew,
Your advice is always welcome.
Sorry for my ignorance as I don't know exactly the terms of "bridge" and "terminate". I understand that the top sense output needs to connect to the voltage input of the load, and the bottom sense output needs to connect to the ground of the load. Were you saying that both sense wires must be shielded? Of course, the shield will be connected to the ground. I think the shield should be connected to the ground on the regulator side, not the load side.
Regards,
Bill
Your advice is always welcome.
Sorry for my ignorance as I don't know exactly the terms of "bridge" and "terminate". I understand that the top sense output needs to connect to the voltage input of the load, and the bottom sense output needs to connect to the ground of the load. Were you saying that both sense wires must be shielded? Of course, the shield will be connected to the ground. I think the shield should be connected to the ground on the regulator side, not the load side.
Regards,
Bill
Last edited:
excellent to hear the PCBs are nearing completion. for sure I will grab a few when you have sorted out the negative version. I will use a few for the higher current I/V section of a PCM1794A dac
is there even a vague idea of price? I have registered interest for 3 positive and 2 negative, but may increase this depending on the damage to the wallet
is there even a vague idea of price? I have registered interest for 3 positive and 2 negative, but may increase this depending on the damage to the wallet
Last edited:
Finished stuffing the neg pcb last night, but it was too late and was too tired to start testing.
@merlin, thanks for the encouragement!
@hifinutnut, the pcb has changed a little since post 446. I see you found out about the length of the grounding cable on the scope probe. There are accessories that address this problem, which replace the scope probe grounding wire with a very short (less than 2cm) connection. It is indeed difficult to make correct low noise measurements. Your scope should have two channels most likely, try to measure on one channel the regulator ground. There will be quite a bit of noise there too.
Regarding noise pickup in the sense wires, yes, that's definitely a concern and it's best to have the whole construction boxed. But your build might not allow that, so you can try a shielded cable. If you do that, do not connect the shield at both ends to gnd.
@qusp, about the price. Don't know exactly, but it will be a bit less than $10 per piece, with pos and neg being same price, and shipping extra.
@merlin, thanks for the encouragement!
@hifinutnut, the pcb has changed a little since post 446. I see you found out about the length of the grounding cable on the scope probe. There are accessories that address this problem, which replace the scope probe grounding wire with a very short (less than 2cm) connection. It is indeed difficult to make correct low noise measurements. Your scope should have two channels most likely, try to measure on one channel the regulator ground. There will be quite a bit of noise there too.
Regarding noise pickup in the sense wires, yes, that's definitely a concern and it's best to have the whole construction boxed. But your build might not allow that, so you can try a shielded cable. If you do that, do not connect the shield at both ends to gnd.
@qusp, about the price. Don't know exactly, but it will be a bit less than $10 per piece, with pos and neg being same price, and shipping extra.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.