|
Components and Circuits For discussions about component types, alternatives and availability, circuit configurations and modifications etc. Discussions here should be of a general nature and not about specific sets. |
|
Thread Tools |
21st May 2020, 8:59 am | #1 |
Dekatron
Join Date: Jul 2009
Location: Sleaford, Lincs. UK.
Posts: 7,667
|
PCB layout - rails to DIP 8 dual opamps.
I'm after making a PCB for this - https://sound-au.com/project97.htm but it won't be printed, I'll route out the traces with a Dremmel. This is just about do-able for dip 8 opamps. With this in mind I thought the best place to start is to figure out how to route the +/- supplies to the opamps and came up with this, see attached pic.
After reading a few application reports on the Texas site I got a few tips, like this one - https://e2e.ti.com/blogs_/archives/b...-for-an-op-amp , but all the relevant articles I could find used two sided boards and SMD's. Particular emphasis was with regards to the non inverting IP. Another issue is where to put the ground plane, I usually lay out all traces and whats left becomes the ground plane but it would obviously help if this in the right places. Any tips would be welcome; for instance do you start with the OP and work backwards or the other way round? how to avoid crosstalk? TFL, Andy.
__________________
Curiosity hasn't killed this cat...so far. |
21st May 2020, 10:58 am | #2 |
Octode
Join Date: Jul 2009
Location: Carmel, Llannerchymedd, Anglesey, UK.
Posts: 1,509
|
Re: PCB layout - rails to DIP 8 dual opamps.
The only issues I have come across is in a mixer where various inputs fed in to a virtual earth mixing node. In this case a ground plane helps, but also if long tracks to the node are involved, it is worth including a ground run between each input line. If this is not done, some capacitive coupling can occur resulting in crostalk at the higher frequencies.
|
21st May 2020, 2:55 pm | #3 |
Moderator
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 22,899
|
Re: PCB layout - rails to DIP 8 dual opamps.
I don't do single sided boards any more. The slots to clear tracks disrupt the grounding and it doesn't function as a plane anyway. I use double sided layout.
Done by hand, you can leave the topside as the ground plane and countersink component leg holes for clearance, or don't countersink and solder those going to ground. With ICs or IC soclets, bend out the pin to be grounded and solder to the plane. If anything, it makes layout easier. David
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done |
21st May 2020, 3:57 pm | #4 |
Dekatron
Join Date: Jul 2009
Location: Sleaford, Lincs. UK.
Posts: 7,667
|
Re: PCB layout - rails to DIP 8 dual opamps.
Thanks both. I keep meaning to try some double sided board and could have sworn i bought some, where it went is anyone's guess. So with double sided board one side is ground plane only? Also is there a better way to feed +/- to the op-amps. i considered jumpering to the pins.
For decoupling I was going to solder direct to the pins underneath, i often mount other components on the underneath too. doesn't look neat but is functional. I find the hardest bit measuring the component distances, tricky to do with 2.5mm pin separation but using the sockets to make holes helps. Andy.
__________________
Curiosity hasn't killed this cat...so far. |
21st May 2020, 6:56 pm | #5 | |
Heptode
Join Date: Jan 2009
Location: West Lothian, UK.
Posts: 762
|
Re: PCB layout - rails to DIP 8 dual opamps.
Quote:
Drawing the layout on graph paper is a good way to start - as a rubber is easier than a Dremel later.
__________________
George |
|
21st May 2020, 7:43 pm | #6 |
Moderator
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 22,899
|
Re: PCB layout - rails to DIP 8 dual opamps.
Definitely a time when an automatic centre punch pays its way.
At work, someone asked to borrow a centre punch to mark where he was going to drill some holes. I fetched my personal toolbox in from the car and handed him an eclipse automatic centre punch. I thought no more about it until I heard the hammer blow. He'd never come across an automatic punch before. Fortunately it survived. I was helpless! David
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done |
22nd May 2020, 5:56 am | #7 |
Dekatron
Join Date: Jul 2009
Location: Sleaford, Lincs. UK.
Posts: 7,667
|
Re: PCB layout - rails to DIP 8 dual opamps.
"I used to use 0.1 inch (no metric parts then) graph paper, taped to the board then a centre punch or pointed bradawl to mark through the paper for drilling." Now that's a good idea, thanks George.
"Definitely a time when an automatic centre punch pays its way." The first time I saw one was when Trevor used one to nick a car in Made in Britain. Been meaning to get one for years but have managed without. I guess it's one of those things where once you get one you realise that should never have been without one. Andy.
__________________
Curiosity hasn't killed this cat...so far. |
22nd May 2020, 2:56 pm | #8 |
Guest
Posts: n/a
|
Re: PCB layout - rails to DIP 8 dual opamps.
I have an old "automatic" punch which is a closed coil spring with a heavy (ish) knob on it, place point on object, pull up knob and release. Works very well for light punches, better than the Eclipse one as it doesn't wander off.
|
29th Jun 2020, 12:53 pm | #9 |
Dekatron
Join Date: Jul 2009
Location: Sleaford, Lincs. UK.
Posts: 7,667
|
Re: PCB layout - rails to DIP 8 dual opamps.
I've had this preamp built now for a month, and it works well, apart from a few issue's where my work wasn't 100%, EG not quite soldered components, SC's where solder had blobbed over two traces etc all was well. See attached schematic.
Yesterday whilst correcting one such I noticed HF oscillation on the inverting IP's of the final OP stage, these are singing away at 357,142khz. When I did the layout I paid attention to the traces on the inverting IP's as advised in the Texas article linked in post #1. They are very short, about 1.5cm. The opamp is decoupled as advised by Rod Elliot, EG 100n multilayer ceramic caps bang on the +Vcc & -Vcc ins, there are also two 10u caps across the rails to ground where the supply connects to the board, as well as a 10u near to the said pins on the OP opamp. These work ok on the other two opamps, it's just this one that is giving me gyp. I tried a 12p cap from one of the inverting IP's to ground, this being the first low value cap I pulled out the drawer, no go, I then tried a 100p ceramic in parallel with R113, which is a 15k, not the 33k as per the schematic, which I calculated would reduce gain at 100khz approx, this resulted in the OP oscillating, this turned out to be a short on the - IP. I then tried a 47p ceramic but finally settled on a 330p multilayer ceramic, this stops the HF on the - IP, but reduces the OP slightly at 20khz, only 25mV RMS, which I can live with. So, have I done right here or is there a better way? Swapping the 5532 for a TL072CP solves the issue too, because AFAIK it's not such a fast opamp, I think. Andy.
__________________
Curiosity hasn't killed this cat...so far. |
30th Jun 2020, 10:46 am | #10 |
Dekatron
Join Date: Feb 2007
Location: Leominster, Herefordshire, UK.
Posts: 16,536
|
Re: PCB layout - rails to DIP 8 dual opamps.
Bunging those 100R inside the loop (between output pin and FB resistor might help?
__________________
....__________ ....|____||__|__\_____ .=.| _---\__|__|_---_|. .........O..Chris....O |
30th Jun 2020, 12:30 pm | #11 |
Rest in Peace
Join Date: Jun 2006
Location: Chard, South Somerset, UK.
Posts: 7,457
|
Re: PCB layout - rails to DIP 8 dual opamps.
Your cct. diag. does not show any decoupling capacitors fitted adjacent to the PSU connections of the op-amp. Are those caps actually installed?
Al. |
30th Jun 2020, 3:05 pm | #12 |
Dekatron
Join Date: Jul 2009
Location: Sleaford, Lincs. UK.
Posts: 7,667
|
Re: PCB layout - rails to DIP 8 dual opamps.
"Bunging those 100R inside the loop (between output pin and FB resistor might help? " You mean move the junction of R113/pin 7 to junction of R114/C103 Chris?
Re decoupling caps...."Are those caps actually installed?" Yes Al, see attached. Attached are also the other schematics so we know what were talking about. C1 is on across U1, C2 across U2 and C3 & C6 across U3 where the issue is. C1 & 2 are within spitting distance of the Vcc pins on U1 & 2, but on U3 I soldered the cap bang on the pins. C6 is across the rails adjacent to U3 as described in post #9. Andy.
__________________
Curiosity hasn't killed this cat...so far. |
1st Jul 2020, 1:30 am | #13 |
Nonode
Join Date: Oct 2010
Location: Tintinara, South Australia, Australia
Posts: 2,340
|
Re: PCB layout - rails to DIP 8 dual opamps.
Your arrangement of the decoupling caps is not a common one that I have come across.
Normal practise for split rail supplies is to put the caps from each supply rail to ground, not across the + & - rails. Remember, they are two seperate supplies referenced to ground. Most op-amp application notes only give the most minimum mention of decoupling of power rails and if you got that from the TI note I think you did, it was talking about single supply op-amps even though it is in the part of the app note that was dealing with split supply circuits. |
1st Jul 2020, 4:26 pm | #14 | |
Rest in Peace
Join Date: Jun 2006
Location: Chard, South Somerset, UK.
Posts: 7,457
|
Re: PCB layout - rails to DIP 8 dual opamps.
Quote:
Al. Last edited by Skywave; 1st Jul 2020 at 4:29 pm. Reason: General re-write |
|
2nd Jul 2020, 7:13 am | #15 |
Dekatron
Join Date: Jul 2009
Location: Sleaford, Lincs. UK.
Posts: 7,667
|
Re: PCB layout - rails to DIP 8 dual opamps.
Re coupling caps, see original article - https://sound-au.com/project97.htm . Davids mentioned coupling caps across the rails before, see this thread - https://www.vintage-radio.net/forum/...d.php?t=162698, post #12. The caps I used have very short leads, bang on the pins.
Andy.
__________________
Curiosity hasn't killed this cat...so far. |