buildafriend
Well-known member
deleted
slow blow pre PCB
- Thread starter buildafriend
- Start date
Help Support GroupDIY Audio Forum:
ruffrecords
Well-known member
A couple of comments on your PCB layout.
1. Try to get the first tube as close to the input transformer as possible. The grid circuit is very sensitive so its track must be kept as short as possible.
2. You don't really need the grid stopper for the 12AU7 input. All it does is worsen the noise.
3. Try to get your heater tracks as short and fat as possible. It is worth arranging the orientation of the tubes to achieve this. It is also worth routing the heaters first.
4. You have a lot of right angled turns in your tracks. I am no PCB expert but 45 degree turns are more normal.
Cheers
Ian
1. Try to get the first tube as close to the input transformer as possible. The grid circuit is very sensitive so its track must be kept as short as possible.
2. You don't really need the grid stopper for the 12AU7 input. All it does is worsen the noise.
3. Try to get your heater tracks as short and fat as possible. It is worth arranging the orientation of the tubes to achieve this. It is also worth routing the heaters first.
4. You have a lot of right angled turns in your tracks. I am no PCB expert but 45 degree turns are more normal.
Cheers
Ian
mattamatta
Well-known member
What did you end up doing with the grounds in question in the schematic?
Looks like that blue section should definitely be ground (which would schematically attach it to the same line where you circled in black whether it's explicitly drawn or not).
The current running through the tubes on the right side needs to eventually make its way back to the supply return from where it started (B+ in this case).
If you connect those nets together like drawn but not to the B+ return/ground, there isn't really a DC return path, and you may end up forcing one somewhere unexpected or not getting the tube bias you expect and only having current flow through various leakages (tubes, caps, etc).
Looks like that blue section should definitely be ground (which would schematically attach it to the same line where you circled in black whether it's explicitly drawn or not).
The current running through the tubes on the right side needs to eventually make its way back to the supply return from where it started (B+ in this case).
If you connect those nets together like drawn but not to the B+ return/ground, there isn't really a DC return path, and you may end up forcing one somewhere unexpected or not getting the tube bias you expect and only having current flow through various leakages (tubes, caps, etc).
I always use 45º angles but I think they only worth (in audio applications) for self etched boards (without solder mask to protect them from lifting) But I also like more the design like that, 90º seems too old school for me 
In many cases the center tap is very useful, if you are working in the same project with tubes that only have 6.3V heaters you end using 6.3V for everyone, so you short the ends together (4 and 5 in the 12AU7) and apply the 6.3 between there and pin 9.
The star distribution of B+ isn't optimal at all, you should take the trace from the input to the caps and from there to the tubes, same thing for the ground of the PS, to the caps, then to the tube. (of course when I say tube is the anode and cathode resistors respectively if there are any)
Then the signal ground goes from the input transformer, to the first tube, to the second tube, to the output, following the path. Here is where it gets tricky since you also have to connect the caps and PS ground, just connect the PS caps close to the tubes and then use the cap on the output tube to go to the PS GND. In any case, in tubes circuit current is pretty low, so you can take some more freedoms than working on SS circuits, but it's always good practice to optimize this things.
As Ian said, place your first tube close to the I/P transformer, the other tube close so heaters run short traces. Then the GND from input to output, at the output tube you make a split going to PS GND. Place the caps close to each tube, add the PS rails. Route everything else as tidy as you can. The GND traces at the input are particularly strange, that cap at the lower left corner runs a trace all the way just for it, you should probably move that cap, it makes the GND make that huge loop from top to bottom.
I draw my boards quite a few times before I'm happy with them, don't worry, keep routing! Good luck with that!
JS
In many cases the center tap is very useful, if you are working in the same project with tubes that only have 6.3V heaters you end using 6.3V for everyone, so you short the ends together (4 and 5 in the 12AU7) and apply the 6.3 between there and pin 9.
The star distribution of B+ isn't optimal at all, you should take the trace from the input to the caps and from there to the tubes, same thing for the ground of the PS, to the caps, then to the tube. (of course when I say tube is the anode and cathode resistors respectively if there are any)
Then the signal ground goes from the input transformer, to the first tube, to the second tube, to the output, following the path. Here is where it gets tricky since you also have to connect the caps and PS ground, just connect the PS caps close to the tubes and then use the cap on the output tube to go to the PS GND. In any case, in tubes circuit current is pretty low, so you can take some more freedoms than working on SS circuits, but it's always good practice to optimize this things.
As Ian said, place your first tube close to the I/P transformer, the other tube close so heaters run short traces. Then the GND from input to output, at the output tube you make a split going to PS GND. Place the caps close to each tube, add the PS rails. Route everything else as tidy as you can. The GND traces at the input are particularly strange, that cap at the lower left corner runs a trace all the way just for it, you should probably move that cap, it makes the GND make that huge loop from top to bottom.
I draw my boards quite a few times before I'm happy with them, don't worry, keep routing! Good luck with that!
JS
thanks;
trx are LL5402 and LL1538XL
heater and relay supply doesnt have connection to GRD plane. 2 trace for each (+/-)
GRD plane is for audio or 0 V only. ( which ever is the correct term)
47uf/400V for on board filtering cap :
on prototype it didnt even needed one.
i am using un_regulated but filtered PSU, C-R-C-R-C and same for heaters too. relays regulated.
when 4 micpre in one case, with all power connections, i might need extra caps as in schematic,
but, i can always slap one extra on each board, thats pretty easy...
trx are LL5402 and LL1538XL
heater and relay supply doesnt have connection to GRD plane. 2 trace for each (+/-)
GRD plane is for audio or 0 V only. ( which ever is the correct term)
47uf/400V for on board filtering cap :
on prototype it didnt even needed one.
i am using un_regulated but filtered PSU, C-R-C-R-C and same for heaters too. relays regulated.
when 4 micpre in one case, with all power connections, i might need extra caps as in schematic,
but, i can always slap one extra on each board, thats pretty easy...
Upacesky
Well-known member
What kind of power supply, power transformer do you use? Did you also design something for it?
Andy Peters
Well-known member
kambo said:
Most standard PCB libraries do not have such things.
buildafriend said:
We had discussion about big vs small core OTs here somewhere, can't find it now. I was reminded about the fact that pronounced lows with some trafos of this kind are not needed on every application, even if they sound impressive when you hear them at first. Is your OT XSM series? IIRC NYD found out WSM series have a bit better response. You might be used to big cores and now find lacking bass. Well, my suggestion would be something like CM-27101 with 50Ni:50Fe core, i'm pretty sure Cinemag has something similar with lower ratio, maybe from CMOX series...
It would also be worth checking if 6CG7 sounds better as input tube, this paper compares it to 6SN7/6CG7:
http://www.ianbell.ukfsn.org/data/6CG7mufollowerdistortion.pdf
audiophreak
Well-known member
not sure if this helps , but check this thread - http://groupdiy.com/index.php?topic=49279.msg623266#msg623266
This seems similar to a situation I was having with a project and some wonderful people here helped me out
This seems similar to a situation I was having with a project and some wonderful people here helped me out
Similar threads
Latest posts
-
-
-
Built custom channel for Soundcraft Boards (SSL like)- Latest: MidnightArrakis
-
-
-
