Tony, thanks for the reply.
I'll see about making better layout images. Hopefully I can just change the color of the component layer.
As for the current setting resistor or resistors. I did leave room on my board for a dedicated load resistor near the output from rail to ground, after the filter caps of tr regulator. Regarding the Vout to adj resistor, I had seen mentioned a few places on the net mention that this resistor is best at 100-120 ohm, to set the current into the adj pin at 10 mA. That seems to be regardless of current draw of any load. The datasheets for the national and TI both show 240 for that resistor, iirc.
Trust a Texan, Sy is Texan..... That changes everything. Actually, my father in law is Texan and I trust him so......
I'll see about making better layout images. Hopefully I can just change the color of the component layer.
As for the current setting resistor or resistors. I did leave room on my board for a dedicated load resistor near the output from rail to ground, after the filter caps of tr regulator. Regarding the Vout to adj resistor, I had seen mentioned a few places on the net mention that this resistor is best at 100-120 ohm, to set the current into the adj pin at 10 mA. That seems to be regardless of current draw of any load. The datasheets for the national and TI both show 240 for that resistor, iirc.
Trust a Texan, Sy is Texan..... That changes everything. Actually, my father in law is Texan and I trust him so......
What is E$26 and E$27 as they appear to be a short? What software are you using to do this work? Does it generate a netlist? If so are you using the netlist? Where did you show the schematic? I'm not too happy with the ground tree where is necks down. Have you considered a double layer board for this?
I skimmed the thread and didn't stumble on the schematic in there but I've been known to not find the forest amongst all the trees.
G²
I skimmed the thread and didn't stumble on the schematic in there but I've been known to not find the forest amongst all the trees.
G²
G^2, thanks for looking and commenting.
E26 and 27 are filter resistors for a CRC filter. I'm not planning/hoping tonier them, so I shorted the trace. I'll cut it if I need them later.
I am not liking how the ground trace narrowed either, just dis no seem right to me. Looking over the sizes again of my case and such, I realize I can add another 0.5" or so in the height, though no extra width.I'll try to pull that down so the ground is not so narrow there.
I am using the free version of Eagle. I've tried the netlist before but found I really hard to use as I could never find the right part in the net with those little yellow net connections. I have not posted the actual schematic for this, sorry. I'll try to pot one later with my modified board layout.
Also, I did find an error in the phantom section. One of the capacitors in the diode voltage trippler should go from the third diode to ground, not from the second.
Thanks again!
E26 and 27 are filter resistors for a CRC filter. I'm not planning/hoping tonier them, so I shorted the trace. I'll cut it if I need them later.
I am not liking how the ground trace narrowed either, just dis no seem right to me. Looking over the sizes again of my case and such, I realize I can add another 0.5" or so in the height, though no extra width.I'll try to pull that down so the ground is not so narrow there.
I am using the free version of Eagle. I've tried the netlist before but found I really hard to use as I could never find the right part in the net with those little yellow net connections. I have not posted the actual schematic for this, sorry. I'll try to pot one later with my modified board layout.
Also, I did find an error in the phantom section. One of the capacitors in the diode voltage trippler should go from the third diode to ground, not from the second.
Thanks again!
And a quick schematic. Note, the last cap and resistor on the phantom part I have omitted on the board.
Thanks!
psu4schem by Dennis Dietz, on Flickr
Thanks!
psu4schem by Dennis Dietz, on Flickr
Sorry to do this, I'm not normally a bumper, but bump
I ask because I have a chunk of free time tomorrow, but not much after that and I'd like to etch and populate this.
I'm not really expecting anyone to proofread my tracing. I would love to know if the layout seems acceptable or if there are any "Oh god, don't do it that way!" issues
Thanks!
I ask because I have a chunk of free time tomorrow, but not much after that and I'd like to etch and populate this.
I'm not really expecting anyone to proofread my tracing. I would love to know if the layout seems acceptable or if there are any "Oh god, don't do it that way!" issues
Thanks!
Sorry to do this, I'm not normally a bumper, but bump
I ask because I have a chunk of free time tomorrow, but not much after that and I'd like to etch and populate this.
I'm not really expecting anyone to proofread my tracing. I would love to know if the layout seems acceptable or if there are any "Oh god, don't do it that way!" issues
Thanks!
I haven't used Eagle so I don't know how bad the netlist is. I used to run Tango schematic and Tango PCB. The netlist was THE most important part of the task. It guarantees the the PCB is the same as the schematic. Of course if the schematic is wrong....
A problem with tracing manually is it's possible to see something that is in error and consistently miss it. The computer doesn't make that mistake.
So you really DO wnat others to check your work.
G²
Just at a quick glance, I see an awful lot of junctions, (the blobs on the wires), in your schematic that I wouldn't expect to see in Eagle. I'd try dragging all the components around and make sure the traces "rubber band" with them in the way you would expect, especially around the rectifiers.
Hi Dennis, the new colours are much better! As I said before I'm not experienced with layout, however if It were me I'd consider moving the reg closer to the output, I think you could put the components on the left, instead on the right of the reg. I'd also be careful about using a 100nF cap on the output as a bypass cap. In my sims with a 1000uF main output cap I had to go to 470nF to avoid ringing problems. edit: sorry just realised it is 100uF not 100nf, you probably should consider putting series resistance with that one as well. Someone else chime in if I'm off the mark here please!
Additionally I'd suggest putting in provision for series resistance between the output cap and the 0V pad. You will quite probably get ringing if you don't put in a small amount of series resistance, and with a 1000uF cap this will almost certainly be in the audio band. I'm using 0.33 ohms. There will be very little current flowing so it doesn't have to be anything big. I'm actually using SMD resistors for this purpose.
Tony.
As I said before I'm not experienced with layout, however if It were me I'd consider moving the reg closer to the output, I think you could put the components on the left, instead on the right of the reg. I'd also be careful about using a 100nF cap on the output as a bypass cap. In my sims with a 1000uF main output cap I had to go to 470nF to avoid ringing problems. edit: sorry just realised it is 100uF not 100nf, you probably should consider putting series resistance with that one as well. Someone else chime in if I'm off the mark here please! Additionally I'd suggest putting in provision for series resistance between the output cap and the 0V pad. You will quite probably get ringing if you don't put in a small amount of series resistance, and with a 1000uF cap this will almost certainly be in the audio band. I'm using 0.33 ohms. There will be very little current flowing so it doesn't have to be anything big. I'm actually using SMD resistors for this purpose.
Tony.
Last edited:
Thanks everyone for the replies, I appreciate it.
Tony, I'll see about moving the regs. I wanted to make sure there was space for a heatsink behind the regulators.
Series output cap resistors: I'll see about making space on the bottom for that also as you suggest. I though the smaller caps after the big one filtered that out but I really don't know.
Tony, I'll see about moving the regs. I wanted to make sure there was space for a heatsink behind the regulators.
Series output cap resistors: I'll see about making space on the bottom for that also as you suggest. I though the smaller caps after the big one filtered that out but I really don't know.
So I went ahead and etched the board. I also populated it.
I've not hooked it up to see if there is any audible noise, and I don't have a scope or anything, but it works pretty well.
There is one problem though. The positive and phenom rails are at the expected voltage but the negative rail is ~2V high. I did check everything as I populated the board. The resistors are matched and the supply rails before the regulator are within 0.2v of each other and with the voltage expected after rectifying a 15vac supply (around 21vdc).
The positive rail is at 17.5, negative at 19.2vdv. I did hand match the resistors so I know they are right, and I've double-checked to make sure I did not drop and pick up the wrong bit. I've switched out the LM337 from TI to a new one for an identical final voltage. I added a load to the circuit which only made the regulator get hot (I switched regulators afterwards just in case). I even tried switching which secondary wires where connected to form the center tap. No change. The LM317t used are Fairfield (or is it fairchild?) and the 337 is TI.
The negative regulator part is exactly as shown in my board layout above with the exception the the protection diode from In to Out is reversed in the drawing. On the real board, the strip points to out, as does the protection diode from Adj to Out. I am using 330uF/25 electrolytics from adj/our divider to ground. I only used one set of 1000uf/25 filter caps from rail to ground after the regulator.
I'd love some ideas on what to check. Does the 337 behave a little differently? Could one of the protection diodes be wrong, backwards, or bad. I'm pretty sure I checked each diode with the diode fxn on my meter before using them.
Thanks!
I've not hooked it up to see if there is any audible noise, and I don't have a scope or anything, but it works pretty well.
There is one problem though. The positive and phenom rails are at the expected voltage but the negative rail is ~2V high. I did check everything as I populated the board. The resistors are matched and the supply rails before the regulator are within 0.2v of each other and with the voltage expected after rectifying a 15vac supply (around 21vdc).
The positive rail is at 17.5, negative at 19.2vdv. I did hand match the resistors so I know they are right, and I've double-checked to make sure I did not drop and pick up the wrong bit. I've switched out the LM337 from TI to a new one for an identical final voltage. I added a load to the circuit which only made the regulator get hot (I switched regulators afterwards just in case). I even tried switching which secondary wires where connected to form the center tap. No change. The LM317t used are Fairfield (or is it fairchild?) and the 337 is TI.
The negative regulator part is exactly as shown in my board layout above with the exception the the protection diode from In to Out is reversed in the drawing. On the real board, the strip points to out, as does the protection diode from Adj to Out. I am using 330uF/25 electrolytics from adj/our divider to ground. I only used one set of 1000uf/25 filter caps from rail to ground after the regulator.
I'd love some ideas on what to check. Does the 337 behave a little differently? Could one of the protection diodes be wrong, backwards, or bad. I'm pretty sure I checked each diode with the diode fxn on my meter before using them.
Thanks!
Thanks for the reply Sy.
The trafo is 15.5 each secondary, not connected to anything.
After rectification and filtering with 1x1000uf each rail and 1.0uf each rail, voltage is 21.1 for each rail to ground.
After the regulator voltage is 17.5 positive to ground, 19.2 negative to ground. Adj to ground is 16.2 positive, 18.8 negative rail.
The trafo is 15.5 each secondary, not connected to anything.
After rectification and filtering with 1x1000uf each rail and 1.0uf each rail, voltage is 21.1 for each rail to ground.
After the regulator voltage is 17.5 positive to ground, 19.2 negative to ground. Adj to ground is 16.2 positive, 18.8 negative rail.
Yup, backward diode from the adj to out pin. Rookie!
To double check, the diode between the in and out pin should have the stripe toward the out pin for he negative rail and toward the in pin on the positive?
Thanks for the direction!
Ripple voltage, I'm not sure what that is. Do you mean how much the VAC fluctuates under a load?
To double check, the diode between the in and out pin should have the stripe toward the out pin for he negative rail and toward the in pin on the positive?
Thanks for the direction!
Ripple voltage, I'm not sure what that is. Do you mean how much the VAC fluctuates under a load?
Thanks Tony. I actually had checked that and foundry all seemed ok. I have a hard time visualizing the fxn of the protection diodes on the negative rail, so I messes that up.
The voltage matches perfectly now. 17.4vdc for each rail both without a load and with a ~100mA load (resistor from rail to ground).
Thanks again all! A learning experience indeed.
The voltage matches perfectly now. 17.4vdc for each rail both without a load and with a ~100mA load (resistor from rail to ground).
Thanks again all! A learning experience indeed.
It can be confusing, especially the way the schematics are drawn, You need to think about the fact that everything is relative to 0V and which way the current flows. So for your -ve reg the adj pin is more positive than the output pin. You don't want current flowing through these diodes normally so they need to be "backwards" These diodes ensure an easy path for current to flow in the case of a short on the input or output of the reg, which would otherwise go through the reg itself.
Tony.
Tony.
Thanks Tony for the helpful explanation. It makes more sense to me thinking in more positive.
Anyone have an opinion on the actual phantom voltage? How critical is 48vdv? I kmow this will depend on the mic mostly, but I only have one mic right now requiring phantom so I don't have much to go on. I've seen some sources suggest that less is usually ok, others say that you will loose a lot of dynamic range with lower voltage. Right now I have it at 42vdc because that is what I could get with the resistors I had on hand.
Anyone have an opinion on the actual phantom voltage? How critical is 48vdv? I kmow this will depend on the mic mostly, but I only have one mic right now requiring phantom so I don't have much to go on. I've seen some sources suggest that less is usually ok, others say that you will loose a lot of dynamic range with lower voltage. Right now I have it at 42vdc because that is what I could get with the resistors I had on hand.
Couple more questions now
My antec 50VA 15vac transformer has a purple static shield wire. What do
I do with that?
For switches and fuses, how much leeway should I allow. The transformer is 1.5A rating. I don't think I would ever draw more than 0.5A at most. How should I size my switch and fuses based on that?
Thanks as always!!!
My antec 50VA 15vac transformer has a purple static shield wire. What do
I do with that?
For switches and fuses, how much leeway should I allow. The transformer is 1.5A rating. I don't think I would ever draw more than 0.5A at most. How should I size my switch and fuses based on that?
Thanks as always!!!
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.