FS: powered bus board PCB

From circuitbending to homebrew stompboxes & synths, keep the DIY spirit alive!

Moderators: lisa, luketeaford, Kent, Joe.

User avatar
EATyourGUITAR
has no life
Posts: 4811
Joined: Tue Aug 31, 2010 12:24 am
Location: Providence, RI, USA

Re: interest check small powered bus board

Post by EATyourGUITAR » Sun Apr 26, 2020 2:32 pm

all working good. the dummy loads should be built with much higher value resistor to reduce heat. 100mA + 100mA is probably the limit for heat. my 300mA*2 per dummy load is getting too hot to touch in 10 seconds! use 120 or 122 Ohms instead.
IMG_20200426_143531865.jpg
PCB $10
PCB + caps $14
dummy load PCB $1 each
+ shipping $8 USA $20 world $24 UK (thats what they charge me for UK)

shipping terms, warranty, liability
I am not liable for customs fees in your country. you assume all liability for fees or lost shipments outside of the USA. International tracking numbers will be given but they are not a guarantee that your package will arrive and I do not reimburse people because they do not like the tracking number I gave them. international sales are final. defective international merchandise can not be returned. DIY kits in the united states are warrantied for significant manufacturer defects of the PCB, or caps, or shipping damage that is reported within 7 days of receipt of the goods. assembling and installing power supplies is done every day by many people without injury or damage but that does not ensure the safe installation of power systems by people who are not properly trained and licensed and insured. you assume all liability if you attempt to do it yourself. you have been warned of the risk here in.
You do not have the required permissions to view the files attached to this post.
WWW.EATYOURGUITAR.COM <---- MY DIY STUFF

User avatar
EATyourGUITAR
has no life
Posts: 4811
Joined: Tue Aug 31, 2010 12:24 am
Location: Providence, RI, USA

Re: FS: powered bus board PCB

Post by EATyourGUITAR » Tue Apr 28, 2020 9:58 pm

BUSBOARD.jpg
BUSTANK.jpg
You do not have the required permissions to view the files attached to this post.
WWW.EATYOURGUITAR.COM <---- MY DIY STUFF

tele_player
Common Wiggler
Posts: 191
Joined: Fri Feb 15, 2019 10:50 am
Location: Sacramento

Re: FS: powered bus board PCB

Post by tele_player » Tue Apr 28, 2020 10:20 pm

What was the reasoning behind the choice of filtering?
Why did you choose adjustable regulators, when +12 and -12 parts are available?
What are your specs for line regulation, load regulation, ripple?

User avatar
EATyourGUITAR
has no life
Posts: 4811
Joined: Tue Aug 31, 2010 12:24 am
Location: Providence, RI, USA

Re: FS: powered bus board PCB

Post by EATyourGUITAR » Wed Apr 29, 2020 7:27 am

tele_player wrote:
Tue Apr 28, 2020 10:20 pm
What was the reasoning behind the choice of filtering?
Why did you choose adjustable regulators, when +12 and -12 parts are available?
What are your specs for line regulation, load regulation, ripple?
I am going to take this as a joke since most experienced people would recognize that this is the same setup used by %90 of modular synth power supplies. there is nothing significantly different about my design compared to other manufacturers or the example on the LM317 LM337 datasheets. I am selling a PCB which means the components used will vary depending on the builder. the PCB without components has no specs. it is just a PCB. specs like what you are asking for are generally given for assembled and tested power supplies sold with warranty on the specs. I have no yet invested the time and money into this type of market or this type of product development. I do not own an oscilloscope right now. I lost everything 2 years ago. I am rebuilding. hopefully with sales from the bus board I can buy what I need to provide more testing as value added for my supporters.

if you want to see the difference between the LM317 vs 7812 just compare the datasheets. there is really no point in me typing the datasheets here. in some situations, depending on a lot of details, LM317 can handle larger loads and can be calibrated where as the 7812 has precision based on silicon lottery. in my bus board I only claim 750mA. I need to build a better dummy load to confirm the heat sinks are sufficient at sustained 750mA. right now that 750mA is based on the manual calculations using simple formulas for power dissipation in watts based on thermal conductivity, mass, surface area, estimated efficiency according to the datasheets of all components used in my personal build.

you generally don't have to worry about ripple with linear supplies. most people stop fussing after they switch to a linear regulator. the advantage is low noise power rails the only disadvantage is efficiency. I have designed this with 15v laptop bricks which are super efficient if you live in the EU. because the voltage drop through the LM317 (or LM337) is exactly 3v (see datasheet) and the load never gets to 1.5A, you are operating in the conditions that will absolutely minimize inefficiency and heat generated in the LM317 without sacrificing regulation.
WWW.EATYOURGUITAR.COM <---- MY DIY STUFF

tele_player
Common Wiggler
Posts: 191
Joined: Fri Feb 15, 2019 10:50 am
Location: Sacramento

Re: FS: powered bus board PCB

Post by tele_player » Wed Apr 29, 2020 9:08 am

I wasn't exactly joking, just wondering if you'd done any of the arithmetic.

User avatar
EATyourGUITAR
has no life
Posts: 4811
Joined: Tue Aug 31, 2010 12:24 am
Location: Providence, RI, USA

Re: FS: powered bus board PCB

Post by EATyourGUITAR » Wed Apr 29, 2020 9:45 am

I will try to get you some real specs after I upgrade my test equipment and build the 2.0 of my eurorack dummy loads. I will be honest I have never performed a test for load regulation and line regulation. I have zero test equipment. I have a super cheap transistor tester and a borrowed garbage multimeter. I used 100uF everywhere except I used 10uF for the trimmer RC lowpass filter to make it more quick to respond. there is no reason to put 100uF on the trimmer. I also didn't want to put 10,000uF caps on the output of the bus board because that can cause problems with ripple, in rush current, latching, over-current protection etc..I'm sure there are people on this forum with more experience than me that can extrapolate the most probably results given the circuit and bill of materials which is public. that might be a better short term solution while you are waiting on me to prove it. you can also use some other brand of power supply with almost identical specs to make an educated guess.

there is also the datasheet
TheLM317deviceis an adjustablethree-terminalpositive-voltageregulatorcapableof supplyingmorethan1.5 A overan output-voltagerangeof 1.25V to37 V. It requiresonly two externalresistorsto set theoutputvoltage.Thedevicefeaturesa typicallineregulationof 0.01%andtypicalloadregulationof0.1%.It includescurrentlimiting,thermaloverloadprotection,andsafeoperatingareaprotection.Overloadprotectionremainsfunctionalevenif theADJUSTterminalis disconnected
WWW.EATYOURGUITAR.COM <---- MY DIY STUFF

User avatar
latigid on
Ultra Wiggler
Posts: 931
Joined: Fri Oct 25, 2013 11:05 am

Re: FS: powered bus board PCB

Post by latigid on » Wed Apr 29, 2020 9:56 am

EATyourGUITAR wrote:
Wed Apr 29, 2020 7:27 am

you generally don't have to worry about ripple with linear supplies. most people stop fussing after they switch to a linear regulator. the advantage is low noise power rails the only disadvantage is efficiency. I have designed this with 15v laptop bricks which are super efficient if you live in the EU. because the voltage drop through the LM317 (or LM337) is exactly 3v (see datasheet) and the load never gets to 1.5A, you are operating in the conditions that will absolutely minimize inefficiency and heat generated in the LM317 without sacrificing regulation.
Interestingly enough, the ripple characteristics are specced for 50/60Hz, not for what typical switchers use. The ripple rejection decreases with increasing frequency (page 7, TI datasheet).

edit: "decreases with increasing"

User avatar
EATyourGUITAR
has no life
Posts: 4811
Joined: Tue Aug 31, 2010 12:24 am
Location: Providence, RI, USA

Re: FS: powered bus board PCB

Post by EATyourGUITAR » Wed Apr 29, 2020 10:45 am

latigid on wrote:
Wed Apr 29, 2020 9:56 am
Interestingly enough, the ripple characteristics are specced for 50/60Hz, not for what typical switchers use. The ripple rejection decreases with increasing frequency (page 7, TI datasheet).

edit: "decreases with increasing"
this is very interesting. thank you for bringing that up. now I will go back and read the datasheet some more. I see that troggotronic has been doing this for years. the tiptop bus board does this on the +15v +12v side only using the exact same parts pretty much. the reviews on sweetwater say that the tiptop bus board is the noiseless perfection that cures cancer. also interesting is that the tiptop -12v side is being supplied by a -15v output switching around 40KHz or 100KHz based on what the most likely DC to DC converter may be.

https://www.sweetwater.com/store/detail ... studio-bus

my personal theory is that this works because the amplitude of the 40KHz switch noise is already pretty low after the RC filter or RCL pi filter that most people have on the output of a SMPS.
WWW.EATYOURGUITAR.COM <---- MY DIY STUFF

User avatar
EATyourGUITAR
has no life
Posts: 4811
Joined: Tue Aug 31, 2010 12:24 am
Location: Providence, RI, USA

Re: FS: powered bus board PCB

Post by EATyourGUITAR » Sat May 02, 2020 9:25 pm

latigid on wrote:
Wed Apr 29, 2020 9:56 am
EATyourGUITAR wrote:
Wed Apr 29, 2020 7:27 am

you generally don't have to worry about ripple with linear supplies. most people stop fussing after they switch to a linear regulator. the advantage is low noise power rails the only disadvantage is efficiency. I have designed this with 15v laptop bricks which are super efficient if you live in the EU. because the voltage drop through the LM317 (or LM337) is exactly 3v (see datasheet) and the load never gets to 1.5A, you are operating in the conditions that will absolutely minimize inefficiency and heat generated in the LM317 without sacrificing regulation.
Interestingly enough, the ripple characteristics are specced for 50/60Hz, not for what typical switchers use. The ripple rejection decreases with increasing frequency (page 7, TI datasheet).

edit: "decreases with increasing"
so I was digging into this some more. the specs are actually good up to 10KHz. I am testing with a 65KHz SMPS module that is specific to my needs. the ripple rejection at 10KHz is -60dB. at 65KHz it drops to -45dB. I used an online calculator that tells me that line regulation drops from max 0.01% @ 10KHz to 0.05% @ 65KHz @ 25C expect it to get as bad as 0.1% across the entire temperature range of the LM317. however, the max capacitance on that same AC to DC converter is 7000uF. I have found some nice inductors to make a pi filter with 2000uF. -45db is not so bad if I can get the ripple down with passives after the SMPS. people building as kits will have slightly different results but I think this is definitely better than a lot of other cheap solutions. you can fit a reverb tank in 6U. that makes it worth it.
WWW.EATYOURGUITAR.COM <---- MY DIY STUFF

tele_player
Common Wiggler
Posts: 191
Joined: Fri Feb 15, 2019 10:50 am
Location: Sacramento

Re: FS: powered bus board PCB

Post by tele_player » Sat May 02, 2020 11:52 pm

My earlier questions about specs came from it looking to me, with largish filter caps, like this design was intended to use unregulated diode rectified AC as input, not a regulated SMPS which delivers a very different ripple on its output. Line regulation will be handled by the SMPS.

I’d just look at it on a scope, to see what kind of hash gets past the linear regulators.

User avatar
EATyourGUITAR
has no life
Posts: 4811
Joined: Tue Aug 31, 2010 12:24 am
Location: Providence, RI, USA

Re: FS: powered bus board PCB

Post by EATyourGUITAR » Sun May 03, 2020 6:47 am

I definitely want to. I spent a lot of money stocking up on parts to offer full kits. I just purchased an air compressor for my solder paste dispenser. I will test with the frequency counter and also volts AC on the +12v DC line. I might not be able to buy an oscilloscope right now. I need to make some money first. I spent money on the matrix mixer prototypes. Spent money on my case. I'm not working right now so buying a scope is sketch.
WWW.EATYOURGUITAR.COM <---- MY DIY STUFF

User avatar
mrand
Common Wiggler
Posts: 164
Joined: Thu May 23, 2013 11:26 pm
Location: Whitehorse, Yukon
Contact:

Re: FS: powered bus board PCB

Post by mrand » Sun May 03, 2020 2:51 pm

Sorry if my caffeine isn't working, but this looks to me like it's designed for DC input from a power brick, etc.
But I do see a diode ring at the input, too. Am I correct tn that this is the usual 337/317 circuit, and I can beef up those input caps to make this board suitable for AC input from a transformer?
--ssdp--

User avatar
EATyourGUITAR
has no life
Posts: 4811
Joined: Tue Aug 31, 2010 12:24 am
Location: Providence, RI, USA

Re: FS: powered bus board PCB

Post by EATyourGUITAR » Sun May 03, 2020 6:31 pm

That is not a rectifier. You are correct that it takes two 15v DC power supplies as input. The diodes are there for protection. See the 317 datasheet for more about the diodes.

I have not designed this from the complete MFOS power supply. Only part of it. Take a look. I will publish my schematic no problem. It is the same as the 317 datasheet and the MFOS. If you need a transformer, a bridge rectifier, big caps etc... You need to make a PCB to mount it. I don't know anyone who sells that without the regulators. I have a custom solution coming with dual 16v supplies. The only thing you can do is use laptop bricks or something similar. Building laptop power supplies is not cost effective. The 16v supplies I'm building have custom filtering built around an off the shelf module. 84% efficiency. Small form factor. Managed heat. Shielded metal chassis with handle.

Obviously the total efficiency will factor in the heat produced in the 12v regulators + heat in the AC DC converters.
WWW.EATYOURGUITAR.COM <---- MY DIY STUFF

User avatar
EATyourGUITAR
has no life
Posts: 4811
Joined: Tue Aug 31, 2010 12:24 am
Location: Providence, RI, USA

Re: FS: powered bus board PCB

Post by EATyourGUITAR » Sun May 10, 2020 2:49 pm

https://eatyourguitar.myshopify.com/

The store is now open. everything that is in stock on the website is in stock in real life. I am open. here is a picture to show what the long 10 pin power cable from thonk can do for you in a 6U 96HP case.
IMG_20200510_135525247.jpg
You do not have the required permissions to view the files attached to this post.
WWW.EATYOURGUITAR.COM <---- MY DIY STUFF

User avatar
mrand
Common Wiggler
Posts: 164
Joined: Thu May 23, 2013 11:26 pm
Location: Whitehorse, Yukon
Contact:

Re: FS: powered bus board PCB

Post by mrand » Sun May 10, 2020 5:26 pm

Congratulations on the opening!
--ssdp--

User avatar
EATyourGUITAR
has no life
Posts: 4811
Joined: Tue Aug 31, 2010 12:24 am
Location: Providence, RI, USA

Re: FS: powered bus board PCB

Post by EATyourGUITAR » Sun May 10, 2020 8:47 pm

+12v 1A -12v 1A 5 minutes 45C case open cold day no fan. the heat sinks performed better than expected. the dummy loads hit 105C so the test was stopped. I think the way most people will use it. you will have up to 1A on the +12v and maybe less on the -12v rail. something like 1A + 600mA is a typical real world load. this can handle it. I am very happy to offer it up to wigglers.

also! the 15v power supply has a 0.4v drop across a 6ft (2 meter) cable 18awg. the heat I generated in the LM317 is with a 2.6v drop. basically what this means is you can use anything you have such as a dual 15v DC meanwell and you will probably be ok if you have 0.4v or less across the cable. I will officially say anything with a 15.3v output is compatible. anything with a 15v output is probably compatible. that 45C @ 1A is claimed with 14.6v at the bus board Vin. you will need to sort out how close you want to get to these extremes. I used 15v and it worked. probably you can too but think about how close you are before you hold me to it. the datasheet says 3v drop at 1.2A or 1.5A depending on how you read it. you are technically going off the datasheet to duplicate my results. I'm trying to help you find cheap options but not trying to go too far with suggestions. all the laptop power supplies I have tried with a 15v rating tested at 15.3v to 16.5v. ebay any 3A, 4A, 5A power supply. no worries.
WWW.EATYOURGUITAR.COM <---- MY DIY STUFF

User avatar
EATyourGUITAR
has no life
Posts: 4811
Joined: Tue Aug 31, 2010 12:24 am
Location: Providence, RI, USA

Re: FS: powered bus board PCB

Post by EATyourGUITAR » Sun May 10, 2020 8:55 pm

here is a stash of 5A 15V laptop power supplies for $11 each free shipping. scoop these up and cut the cable.

https://www.ebay.com/itm/Toshiba-PA3283 ... 4505344482
WWW.EATYOURGUITAR.COM <---- MY DIY STUFF

User avatar
EATyourGUITAR
has no life
Posts: 4811
Joined: Tue Aug 31, 2010 12:24 am
Location: Providence, RI, USA

Re: FS: powered bus board PCB

Post by EATyourGUITAR » Mon May 11, 2020 10:45 am

10 units of custom AC to DC converters ordered. these will sell out fast. these are proper low ripple 15.5v supplies. they are EU compliant international switch mode power supplies. they go to sleep at 0.5 watts with no load max output 2.7A each. the 65KHz switch noise is completely filtered out by all the stuff I added and then the bus board regulators.
WWW.EATYOURGUITAR.COM <---- MY DIY STUFF

User avatar
mrand
Common Wiggler
Posts: 164
Joined: Thu May 23, 2013 11:26 pm
Location: Whitehorse, Yukon
Contact:

Re: FS: powered bus board PCB

Post by mrand » Mon May 11, 2020 5:47 pm

Could you share a part number or source for those radiators?
--ssdp--

User avatar
EATyourGUITAR
has no life
Posts: 4811
Joined: Tue Aug 31, 2010 12:24 am
Location: Providence, RI, USA

Re: FS: powered bus board PCB

Post by EATyourGUITAR » Mon May 11, 2020 9:32 pm

Radiators = heat exchangers = heat sinks? Or that was a typo you mean regulators? Or you want the part number for the AC to DC converter modules? I have a PCB for the AC DC converters that adds filtering. The ripple specs for the module are 150mV but it's 1.5mV after the filter based on the dB attenuation of the filter. The best laptop power supply or meanwell supply is 10mV ripple. I want to publish the schematic the same day these go up for sale in the store. There will be transparency and probably some hardcore DIY people will even source everything DIY using only the free schematic. I have to balance open source with going out of business. The mouser cart is earlier in the thread. Don't use 125C caps in the mouser cart. I have 105C caps here I'm packaging with the bus boards for $14 together. The $10 option has been replaced with the PCB + caps bundle. The bus board schematic can go up tomorrow. There is nothing secret in the bus board.
WWW.EATYOURGUITAR.COM <---- MY DIY STUFF

User avatar
mrand
Common Wiggler
Posts: 164
Joined: Thu May 23, 2013 11:26 pm
Location: Whitehorse, Yukon
Contact:

Re: FS: powered bus board PCB

Post by mrand » Mon May 11, 2020 11:18 pm

Sorry, I meant heat sinks, couldn't think of the name!
I'll look for the BOM :)
Edit: found them in the BOM,thanks!
--ssdp--

tele_player
Common Wiggler
Posts: 191
Joined: Fri Feb 15, 2019 10:50 am
Location: Sacramento

Re: FS: powered bus board PCB

Post by tele_player » Tue May 12, 2020 4:35 pm

EATyourGUITAR wrote:
Sun May 10, 2020 8:47 pm
+12v 1A -12v 1A 5 minutes 45C case open cold day no fan. the heat sinks performed better than expected. the dummy loads hit 105C so the test was stopped. I think the way most people will use it. you will have up to 1A on the +12v and maybe less on the -12v rail. something like 1A + 600mA is a typical real world load. this can handle it. I am very happy to offer it up to wigglers.

also! the 15v power supply has a 0.4v drop across a 6ft (2 meter) cable 18awg. the heat I generated in the LM317 is with a 2.6v drop. basically what this means is you can use anything you have such as a dual 15v DC meanwell and you will probably be ok if you have 0.4v or less across the cable. I will officially say anything with a 15.3v output is compatible. anything with a 15v output is probably compatible. that 45C @ 1A is claimed with 14.6v at the bus board Vin. you will need to sort out how close you want to get to these extremes. I used 15v and it worked. probably you can too but think about how close you are before you hold me to it. the datasheet says 3v drop at 1.2A or 1.5A depending on how you read it. you are technically going off the datasheet to duplicate my results. I'm trying to help you find cheap options but not trying to go too far with suggestions. all the laptop power supplies I have tried with a 15v rating tested at 15.3v to 16.5v. ebay any 3A, 4A, 5A power supply. no worries.
You seem to misunderstand an important detail about 3-terminal linear regulators. The lm317 datasheet specifies Vi-Vo (input voltage minus output voltage) has a minimum value of 3V.

This is also called the dropout voltage, below which regulation isn’t guaranteed. Running LM317 with a margin of 2.65 volts doesn’t meet the requirement. Since you’ll be feeding it from a regulated supply, it might not be obvious, but I’d suggest you take output voltage measurements at various load currents up to max load, and see how well output voltage is maintained.

Bottom line, 15V is cutting it too close for that setup.

User avatar
EATyourGUITAR
has no life
Posts: 4811
Joined: Tue Aug 31, 2010 12:24 am
Location: Providence, RI, USA

Re: FS: powered bus board PCB

Post by EATyourGUITAR » Tue May 12, 2020 4:51 pm

I have modules that take one external resistor to set the output to anything I want. 22K parallel to vref = 16v. I have it all baked in the design. I can also use 15.5v. I am basically telling people up front that the 1A was obtained by going off the datasheet. I %100 agree with everything you said. I know it is called drop out voltage. I know. I did an experiment to get the most Amps out of this setup with the least amount of heat in the heat sinks. it is basically cheating and I admitted that in my explanation. so at this point people can choose to use some laptop power supply with a 15.3v output and a 0.4v drop through the cable or they can use my 16v 15.5v etc.. power supply that I provide with it that will be exactly > 3v at the screw down terminal about 0.01 Ohms away from Vin. using a laptop supply with 2.9v drop out is still doable at 750mA. I wouldn't want that in my personal system. I'm not defending it as a good design. but I think that would be the limit for someone on a budget that just needs super cheap DIY power. I would love to sell more AC to DC converters but it would also be great if people got the bus board for both low end and high end systems. that would make me really happy to service both types of users. there is also a path to upgrade by getting the AC to DC converters from me later.
WWW.EATYOURGUITAR.COM <---- MY DIY STUFF

KSS
Super Deluxe Wiggler
Posts: 2444
Joined: Mon Jan 25, 2016 7:28 am

Re: FS: powered bus board PCB

Post by KSS » Wed May 13, 2020 2:29 am

@tele-player.

The fig.4 Droput Voltage graph of the Sept 2014 Fairchild LM317 datasheet shows that 2v65 will be fine for about 1.4A. Or more depending on temp.
According to that graph, it only takes about 2v75 to reach 1.5A full spec output from 0-125c.

Edit: Fig. 12 on the TI May2004-revised Jan2014 DS shows 2V5 easily covers full spec output at al temperatures. /edit

tele_player
Common Wiggler
Posts: 191
Joined: Fri Feb 15, 2019 10:50 am
Location: Sacramento

Re: FS: powered bus board PCB

Post by tele_player » Wed May 13, 2020 8:40 am

I stand corrected, the datasheet I looked at yesterday didn't include the (more precise) graph for dropout voltage - only the minimum of 3V.

Post Reply

Return to “Music Tech DIY”