ruffrecords
Well-known member
Just for completeness, the original 1073 did not have a phantom supply switch.Phantom power was switched globally across an entire mixer.
Cheers
Ian
Help/advice/guidance translating schematics onto a pcb
- Thread starter djmiggymigz
- Start date
Help Support GroupDIY Audio Forum:
Cheers
Ian
djmiggymigz
Well-known member
Super happy to hear others are learning here too! I sometimes feel bad for my almost infinite questions but there is so much that goes into this that at the very least a little guidance in the right direction feels needed and is immensely appreciated. Thanks to all who've lent a hand already!!
I couldn't recommend KiCad enough. I only downloaded it a week ago and it makes a whole lot of sense to me already. The biggest challenge is learning the ins and outs of the actual electrical engineering side of things. Learning how to use the software itself is cake in comparison. PLUS, you can export you .gbr and drill files, drop them right into JLC and have your custom pcb's in hand within 2 weeks. So very cool to me
djmiggymigz
Well-known member
Whoa meaning if phantom was on one channel it was on all the channels??
I've also noticed that the original 312 design doesnt seem to have phantom power either! I suppose maybe that's because dynamic/ribbon mics were still the majority of mic lockers? Thus 48V wasnt standard or needed until later on?
zamproject
Well-known member
- Joined
- May 11, 2010
- Messages
- 1,600
No, 10k under 48v is 5mA, on the low edge for a LED (depending how bright you want it) 5k may be better
What I mean is that phantom power for one mic is specified around 10mA, let say you have a console with 16 mic input, 48V rail may be scaled to offer around 160+ mA (in fact it is often less as manufacturer cut corners and suppose you won't plug 16 phantom powered mic at once...).
If you add a LED signalling status on each input on the same rail you will overload it as you need 160mA too (@10mA each) to light the LEDs.
Also you better isolate/separate audio supply from signalling supply, usually a 5V rail (modern LED) or 12 to 24V (bulb) is used for all light visual feedback and relay.
API 312 is a part (preamp) of modular console, I never look at API console block diagram, but the P48 is probably handled outside the input module with dedicated supply bus/rail directly wired to previous routing module or mic input connectors.
yes, usual practice back then...
ruffrecords
Well-known member
It is because in those days, phantom power was not often required. Most studios were still using tube based condenser mics which came with their own power supply.
Cheers
Ian
djmiggymigz
Well-known member
okay I totally misunderstood,, 48V is "too low current" for an LED! In that case I'll look into it! I suppose configuring a double pole switch wouldnt be much different than where my layout is right now. However, I dont at all understand how +48v wouldnt provide enough current for an LED but if I use the +16v rail that would?? Don't feel compelled to answer that (unless you want to!!) as I know that's a relatively speaking basic electronics question!
zamproject
Well-known member
- Joined
- May 11, 2010
- Messages
- 1,600
maybe because V=R*I....
Now if you want to use a 30mA P48 for a single preamp it's up to you !
MidnightArrakis
Well-known member
[48V is "too low current" for an LED] -- NO!!!.....48V is -- NOT -- "too low of a current" for an LED!!! It's how the 48VDC power-supply was designed to supply a certain amount of current to its load. BIG DIFFERENCE!!! As an example, condenser microphones typically only draw around 5mA or so. Therefore, the console manufacturers design their phantom power-supply to deliver enough current to supply maybe 10mA to each channel, allowing for enough current to power a microphone that may use around 7mA.
So, let's say you have a 24-channel mixer. Then, the phantom power-supply may be designed to deliver 48VDC @ around 300mA, just for good measure. This works out to be: 24-Channels X 10mA each = 240mA (~1/4-Amp). NOTE: Ian (RuffRecords) could provide you with a much better explanation about this as he once worked at NEVE.
Since you typically don't want to load your power-supply right up to its maximum output capability, you allow for some "headroom" and in my example, I have only allowed for an extra 60mA.
Now, depending upon your LED series resistor and how bright you would like your LED to be, a somewhat typical LED current draw is around 10mA to 20mA or so. And, that's just for a -- RED -- LED. Other colors may draw more current for the same brightness due to how the eye perceives the brightness of different colors.
So, now.....you go tacking on the current load of IC chips for mic-preamps and LED's to show that the circuitry is "ON" and.....you end up exceeding the allowed phantom-power current load of the microphone channel!!! So.....how could you fix this problem???
EASY!!! ..... You just design your phantom power-supply to be able to deliver 1-Amp to each mic-channel!!! Then, your phantom power-supply becomes a 48VDC @ 25-Amp (allowing for a smidgen of "headroom") power-supply that's the size of a brick!!!
Now.....you can add-on all of the LED's and peripheral circuitry to your hearts content and probably -- STILL -- have enough current left-over to light-up an LED Christmas Tree for holiday events!!! Do you see how things go here???
[I dont at all understand how +48v wouldnt provide enough current for an LED] -- It's all in the power-supply design.
[the +16v rail that would??] -- It's all in the power-supply design.
As I pointed out above, the console manufacturers -- COULD -- design their phantom power-supplies to provide 25-Amps if they wanted to!!! But, why would they??? Their -- intended -- total current load is anticipated/expected to be: 24 X 10mA. The console companies aren't allowing, nor are they expecting, for you to increase the current-load demand by adding-on additional circuitry. They (apparently "foolishly") designed their phantom power-supplies to -- ONLY -- provide enough current to power a microphone. The console manufacturers should have anticipated that you would want to plug a toaster-oven into an XLR!!! Right???
In any case.....here's a handy little calculator that will assist you in calculating what value an LED series resistor needs to be:
https://www.digikey.com/en/resource...ors/conversion-calculator-led-series-resistor
[maybe because V=R*I] -- More below:
Hope this helps!!!
/
That all depends on your power supply design. Zam seems to be assuming that your 48V supply will be sized to just barely provide enough current for maximum draw with microphones connected to all inputs.
It is of course perfectly possibly to design in a power supply with enough current capability to supply all the microphones and all the LEDs.
Keep in mind that the voltage dropped from 48V to the LED forward voltage (so typically around 46.5V for a red LED) is dissipating power in the resistor. So for your 10k example that is V^2/R, or just over 0.2W. If you drop the resistor to 5K as suggested for higher brightness that increases to around 0.45W dissipated in the resistor.
Were you accounting for half a watt of power dissipation when selecting which resistor to use? You would want a safety margin, so for long life I would probably use 1W resistors. And that heat goes somewhere, so think about what the resistor will be heating up (i.e. what is near and above the resistors, is there an air vent, etc.).
Check your LED specs as well, a lot of modern LEDs are plenty bright for use indoors at 5mA. If you want them visible in sunlight then make sure the maximum forward current is plenty higher than what you want to use (in other words if you want to run 10mA for higher brightness, don't use an LED with a maximum current of 10mA, use one with 15mA or 20mA max current spec).
djmiggymigz
Well-known member
Wow just got around to reading these latest replies and thank you for such detailed explanations! @MidnightArrakis you've brought to my attention some things I've yet to and need to learn re: the chart you attached! Currently taking an into to electric engineering class but it skipped past the super basic things I.e current vs. voltage vs watts, etc. Guess I'm going to youtube for that one!
I suppose that is one thing that I'm maybe confused on. In the above replies lots of y'all keep mentioning "how the power supply is designed" and other things in regards to a seperate PSU circuit(ry) for phantom power. This makes sense to me in regards to (for example) a 1U, etc, rack mounted Mic Pre build where you're using 120V from the wall that must be converted, etc. BUT, I'm building this in the 500 series format. Which means that the +48v is simply coming from the 500 series chassis PSU and is simply routed to the individual cards via pin 16 on the cards. Which looking at the manual for my chassis (Fredenstein Bento 8), they state "Bento 8 outputs +/-16V DC 2Aand +48V DC 0.25A, 500mA per slot"
So am I missing something here. Do people/companies sometimes build PSU circuits onto/into their 500 series PCB's? Or are the chassis made with the possible additions of LED's, etc. in mind?
To answer this, no. After analyzing Capi's VP312 schematic, I thought it would be easiest to do what they did. Which is 500 Series Pin 15 (+48v) -> SPDT switch that when on sends phantom to both the Resistor (10k in their schem) and LED as well as splitting off towards some resistors in the In Cold path. To sum up, it seemed the simple and easiest way to go about it since it's essentially just hardwired straight in!
I suppose that is one thing that I'm maybe confused on. In the above replies lots of y'all keep mentioning "how the power supply is designed" and other things in regards to a seperate PSU circuit(ry) for phantom power. This makes sense to me in regards to (for example) a 1U, etc, rack mounted Mic Pre build where you're using 120V from the wall that must be converted, etc. BUT, I'm building this in the 500 series format. Which means that the +48v is simply coming from the 500 series chassis PSU and is simply routed to the individual cards via pin 16 on the cards. Which looking at the manual for my chassis (Fredenstein Bento 8), they state "Bento 8 outputs +/-16V DC 2Aand +48V DC 0.25A, 500mA per slot"
So am I missing something here. Do people/companies sometimes build PSU circuits onto/into their 500 series PCB's? Or are the chassis made with the possible additions of LED's, etc. in mind?
djmiggymigz
Well-known member
Question about this: Does this apply to the relationship/"intersections" of top and bottom layer tracks as well? Running into a few situations at the moment where top and bottom traces (when routed most directly) are creating the acute angles between the top and bottom layers. My guess is that it is maybe not a problem (or at least not the same "acip trap" problem) since the tracks aren't on the same plane. Screenshot for reference (red=top, blue=bottom).
Attachments
Brian Roth
Well-known member
MidnightArrakis
Well-known member
[Does this apply to the relationship/"intersections" of top and bottom layer tracks as well?] -- NO!!! Acid-Traps are only formed when the angled tracks are intersecting together on the same layer. WHEW!!!
[Screenshot for reference (red=top, blue=bottom)] -- I am so glad that you are using -- RED = TOP / BLUE = BOTTOM -- as those are the original layer color-codes that were used back when we designed multi-layer PCBs by hand. Tracks were routed on a piece of 22" X 34" Mylar-film (at either a 2:1 or 4:1 scale. However, tiny "RF" PCBs were designed at a 10:1 scale!!!) using adhesive-backed RED & BLUE crepe or plastic tape. Because nobody could look at a hand-taped PCB layout and know which color was which layer, designating RED = TOP / BLUE = BOTTOM became an "unofficial standard" only primarily because people could remember that BLUE & BOTTOM both began with the letter "B"!!! Crazy, huh???
Now.....get this!!! Just try to imagine designing a 16-layer PCB -- BY HAND!!! -- using only BLACK crepe tape using >> 20-SHEETS!!!* -- of Mylar-film stacked one upon another, that are held in alignment with one another by guide-holes that were punched into the Mylar sheets that were placed over a set of guide-pins that were affixed to a metal bar!!! I had to do this when working for a defense contractor that was designing PCBs for U.S. F-16 fighter jets back during the mid-1980s. Back then, computers and any CAD programs were >> JUST << beginning to be introduced into the marketplace. All of you "today's kids" entirely missed out on making the transition from designing PCBs manually on Mylar-film to designing PCBs with these newfangled computers!!!
>> * -- 16-sheets for circuit routing, 2-sheets for TOP & BOTTOM soldermask, 1-sheet for the silkscreen and 1-sheet for the mechanically-dimensioned detail drawing.
/
There is more than one vendor building 500 series racks, and sometimes more than one model per vendor.
According to the Wikipedia article on 500-series:
"In 2006 API established the VPR Alliance, which established 500-series specification standards to ensure compatibility with API 500-series rack systems."
"VPR Alliance standards dictate voltage at +/- 16 volts, with +48 volts for phantom power, with maximum current draw of 130, 130 and 5 mA respectively per slot."
5mA per slot is significantly less then you are planning to use for phantom power and LED current combined.
500 series VPR specs for reference:
API VPR 500 spec
djmiggymigz
Well-known member
Wow okay that's awesome that's going to clean up my layout for sure!!
WOW that's insane that must've been such a process!! Is the reason for all the layers simply that the circuits you were working on were so dense and complex that you needed 16 layers to be able efficiently route everything?? If my 500 series board was 16 layers almost everything would have a direct path to where its going!! lol. However, doing all that work by hand using physical layers seems like such a time consuming task. Tho, I'm sure that in this day and age (KiCad and similar programs) you can probably design and layout a PCB in a heartbeat!
Yeah I'm understanding now (having looked into the current draw of LED's amongst other things) that 5mA wouldn't quite work with what I have laid out. However, I checked the manual for the specific rack that I have and am going to be using and it states that it has 500mA of current at the +48V supply per slot. (Check post #50 in this thread. I quoted directly from the manual where it says this.) That should be plenty??
Now I'm starting to wonder, with everyone saying that I may have problems with the 48V/LED aspect of the circuit, why did CAPI design their 312 this way? It seems as though some racks have enough current to accommodate this design but maybe others don't.
Cqwet Dbdfte
Well-known member
- Joined
- Oct 2, 2023
- Messages
- 628
I would not lose sleep over "acid traps" in this day and age.
If professionally made by a PCB mfg. this is nothing to be concerned about.
Use a mfg. recommended Design Rules file (DRU) and CAM file
and you are good to go.
Pay attention to insulation distances not included in the above.
Sometimes cuts in boards are needed to increase tracking distances.
These may be imposed by safety agencies, UL, CSA, TUV etc.
If professionally made by a PCB mfg. this is nothing to be concerned about.
Use a mfg. recommended Design Rules file (DRU) and CAM file
and you are good to go.
Pay attention to insulation distances not included in the above.
Sometimes cuts in boards are needed to increase tracking distances.
These may be imposed by safety agencies, UL, CSA, TUV etc.
Last edited:
One manufacturer makes a tape simulator 500 series unit that draws something like 250mA out of only the positive rail. I wound up having to make some pretty hefty +16v only PSU's for friends who wanted to use 8 of them in one lunchrack.
When using 2 of these units in one of the 1U 500 series racks that come from China the supplied wall wart PSU caught fire ...
That should just be a concern with high voltages, either line voltage coming directly onto a PCB, or something like a tube power supply routed on a PCB.
Just for the OP benefit: not a concern at all when working with 500 series cards.
Yes, that is common with things like processors, where a memory bus might need over 180 pins routed between the processor and DRAM devices with strict requirements for length matching, which power supply is directly adjacent to the signal lines, multiple different high current power supply voltages, etc.
Two layer is very common for line level analog audio designs.
Right, but your question was why everyone kept mentioning "how the power supply is designed" as a consideration.
I looked back through your early posts and I did not see any statement to the effect of "I am using model X chassis from vendor Blah, and I only care if my design works in that chassis, general interoperability is not a consideration at all."
Without a disclaimer that specific then experienced commenters are going to point out that you have to be aware of the limitations caused by pieces outside of your PCB design, and it is helpful to be aware of how that fits into the bigger picture, in the sense of how the limitations of your specific chassis compare to minimum required by spec, and common de-facto limits of chassis vendors.
djmiggymigz
Well-known member
Aha yupp that's because I didn't know at the time that that had to be taken into consideration! Thanks for all the explanation and context! It makes a lot of sense that originally, the API 500 series format would have different specs for power and such vs. current day racks. I've been looking around at specs tho to try and see how much of a difference there is between racks! It seems silly to me that everyone would be making racks with different specs. Guess that's just part of it tho!
For the sake of my own learning process: Can anyone take a look at the CAPI Schematic (linked here) and help me discern their +48V plan. Like I said in a previous post, I used the same layout in my schematic as it seemed to be the easiest way to go about it. However, now knowing that the current on the +48v pin might not be enough for both phantom AND the LED, I'm confused as to why they designed it like this? Seems to me like anyone could buy the kit, build it, and depending on what 500 rack they have, the LED might not work OR the draw would be too much and the phantom power going to a mic would suffer the consequences. Is that correct? Are they maybe assuming that anyone building the kit would have a rack that could handle it?
Cqwet Dbdfte
Well-known member
- Joined
- Oct 2, 2023
- Messages
- 628
There are LEDs that require very little current, look at Digikey for 2mA types, I found a green one (3mm, thru hole, $1.00) that is really bright with 2mA, even 0.1mA should suffice. A higher current standard LED in series with the 48V to the load may be OK if 46V is OK.
