electro-music.com   Dedicated to experimental electro-acoustic
and electronic music
 
    Front Page  |  Radio
 |  Media  |  Forum  |  Wiki  |  Links
Forum with support of Syndicator RSS
 FAQFAQ   CalendarCalendar   SearchSearch   MemberlistMemberlist   UsergroupsUsergroups   LinksLinks
 RegisterRegister   ProfileProfile   Log in to check your private messagesLog in to check your private messages   Log inLog in  Chat RoomChat Room 
go to the radio page Live at electro-music.com radio 1 Please visit the chat
poster
 Forum index » DIY Hardware and Software
Thomas Henry - PSU for Electronic Music
Post new topic   Reply to topic Moderators: jksuperstar, Scott Stites, Uncle Krunkus
Page 1 of 1 [21 Posts]
View unread posts
View new posts in the last week
Mark the topic unread :: View previous topic :: View next topic
Author Message
mrmrshoes



Joined: Feb 19, 2011
Posts: 73
Location: Newcastle Upon Tyne
Audio files: 4

PostPosted: Fri Jul 26, 2013 10:52 am    Post subject: Thomas Henry - PSU for Electronic Music
Subject description: Advice needed
Reply with quote  Mark this post and the followings unread

I'm in the process of building a PSU and wanted some advice/guidance before I commit.
I thought I would use a Thomas Henry design because this comes with an well written article, that breaks down the design choices and construction. I will also be making use of resources provided by our very own PSU wizard Osal, so I'm hoping he can help out here.
Links to the resources:

Linear Power Supply For Modular Synth Application
http://electro-music.com/forum/topic-51694.html

Nuts and volts article – Thomas Henry – Power Supplies for Electronic Music
http://www.nutsvolts.com/uploads/magazine_downloads/11/January%201998%20Thomas%20Henry%20-%20Power%20Supplies%20for%20Electronic%20Music.pdf


I've attached a redrawn schematic of TH design as well as a PCB layout depicting the parts placement. I pretty confident I’ve covered and understand most of the theory in this design, but I just want to double check a couple of things I’m not a 100% on. Also it would be good for someone else to review the schematic and PCB before I build it, just in-case I've missed or overlooked anything. I have made a couple of slight changes to the basic design but these are really related to component choices and PCB layout. These include a dual primary/ secondary transformer, 1N4002 diodes replaced with a bridge rectifier package (GBU4D) and added an extra set of smoothing Cap's in parallel.

So with this first post I was wanting to concentrate on the Mains wiring, Dual Primary / Secondary Transformer wiring and how the transformer secondary windings connect to the rest of the circuit (bridge rectifier - GBU4D)

If you want to do some extra reading on Mains / Transformers I recommend these links

Afroman tutorial
http://www.youtube.com/watch?v=GMePE7NZcxw&feature=fvwrel
Transformer Basics
http://www.electronics-tutorials.ws/transformer/transformer-basics.html
Multiple Winding Transformers
http://www.electronics-tutorials.ws/transformer/multiple-winding-transformers.html
Useful Book – Contains a couple of chapters on handling Mains voltages, mounting heat-sinks ect.
http://www.amazon.co.uk/Starting-Electronics-Construction-Techniques-Equipment/dp/0750667362
Mains Safety in DIY electronics
http://www.penguintutor.com/electronics/electrical-safety
Power Supply Basics
http://www.mhennessy2.f9.co.uk/articles/power_supplies.htm
Article about wiring an 1176 Power Supply
http://mnats.net/1176_reva-d_hairball_wiring_power.html
Wiring Transformers
http://www.e-dan.co.uk/electronics/wiringtrans.html
Grounding and Shielding for your DIY Audio Projects
http://diyaudioprojects.com/Technical/Grounding-Shielding/
Grounding 101
http://www.modular.fonik.de/files/grounding_101.pdf

Right down to brass tax. In the UK Main voltages are 230V + / - 10% and has a frequency of 50Hz. The colour code for Mains wiring is as follows

Post-March 2004 the UK standard was changed to:
Live = Brown 
Neutral = Blue 
Earth = Green and Yellow sleeving (no change)
(Note: This new colour coding is the same as used for appliance flexes in the UK for many years.)

Useful page for UK Plug wiring
http://web.onetel.net.uk/~uncletony/mains%20stuff.htm

I've opted for the safest Mains wiring scheme by using IEC Inlet and insulated Crimp connectors. The crimps are 6.3mm in size and can be attached to the IEC Inlet solder lugs as-well as PCB inter-connectors with the use of 6.35mm PCB mount blades.
Here are links to the parts and their respective data-sheets.
Connectors, Inlets
Insulated Female Connector 6.3mm
http://www.rapidonline.com/Cables-Connectors/Davico-Insulated-Female-Connector-Red-6-3mm-Pack-of-100-33-1062
Blade PCB Connector (Vertical) 6.35mm
http://www.rapidonline.com/Cables-Connectors/Blade-PCB-Connector-Vertical-6-35-x-0-8mm-100-33-4022
IEC inlets- switched and fused
http://www.rapidonline.com/Cables-Connectors/Polysnap-horizontal-switched-and-fused-inlets-66200

The IEC inlet has Live, Neutral & ground lugs marked clearly on the back of the device. It also has a box draw for a fuse, as well as three extra lugs that make up connections for the inbuilt Illuminated SPST switch. So the three unmarked solder lugs make up the SPST switch. Two lugs are accounted for by the SPST connections but that leaves an extra lug unaccounted for. (I’m not sure what the third one does as of yet, something to do with the LED in the switch).
By testing with a multimeter you can figure out how this IEC inlet works. Live Mains (230V) has to pass through the 20mm fuse before it passes to its solder lug on the back of the device. Neutral & ground pass straight through to their respective connections on the back of the IEC inlet. From there you have to wire the Live Mains to one end of the SPST switch.


The Transformer.

Here is the link to the transformer I’ve obtained. http://www.rapidonline.com/pdf/82719.pdf
It is rated as a 60VA and has Dual Primary/Secondary windings. I'm based in the UK so I will need to configure the primary windings to 230V Mains input by connecting them in series. If you look at the schematic or the transformers data-sheet you will see the colour code and schematic symbol for the transformers inputs.

Red wire = Live
Blue Wire = Neutral
Grey & Violet = connected together

By making these connections I should step down 230V to 18V - 0 – 18V.
The Thomas Henry design calls for a 36V centre tap Input to the rest of the circuit. With our dual secondary we accomplish that by once again connecting the winding in series.

RED = AC+
YELLOW + BLACK = Centre tap / Ground
ORANGE = AC-

The problem I’m having here is related to the GBU4D bridge rectifier. Because the Thomas Henry design used individual diodes for his rectifier, so it's easy to see the physical connections to the transformers secondary windings. With the GBU4D you are given AC1 & AC2 connections.
In the schematic I have connected AC+ (RED wire) to AC1 and AC- (ORANGE Wire) to AC2 of the GBU4D. YELLOW & BLACK wires form the Centre tap and the PSU Ground.
I guess my main question here is do you have to make specific connections from the Transformer secondary winding's (AC+, AC-) to the bridge rectifier AC inputs?

I think that should pretty much cover the basic wiring and connections of the various parts. Please correct me or add anything you feel is important.

In the second post of this thread I'm going to list all the calculations for determining component tolerances and values within this design. This is basically running through Osal calculations on his linear power supply post and applying them to the TH design. The hope is to have a design that you can pick and choose what output rails you want +/- 15V, +/- 12V. Run through the calculations to determine what transformer and components you need to buy. I'll wait until my logic is confirmed to be good before completing the second post.

I have to say my PCB layout is pretty much based on this:
http://electro-music.com/forum/topic-57671.html





PSU-PCB-Parts.png
 Description:
 Filesize:  188.01 KB
 Viewed:  429 Time(s)
This image has been reduced to fit the page. Click on it to enlarge.

PSU-PCB-Parts.png



PSU-Schematic.png
 Description:
 Filesize:  40.9 KB
 Viewed:  771 Time(s)
This image has been reduced to fit the page. Click on it to enlarge.

PSU-Schematic.png



Last edited by mrmrshoes on Wed Jul 31, 2013 5:54 am; edited 2 times in total
Back to top
View user's profile Send private message
mrmrshoes



Joined: Feb 19, 2011
Posts: 73
Location: Newcastle Upon Tyne
Audio files: 4

PostPosted: Fri Jul 26, 2013 10:52 am    Post subject: Reply with quote  Mark this post and the followings unread

So I’m going to start off by listing the specs stated in Thomas PSU article. You should read this article first so you get an idea of what’s going on within the circuit.

Thomas Henry PSU - Component Specs and tolerances
Values in this design are calculated with US Mains 120V, 60Hz
Transformer is a 36V center-tap with a current rating of 1.5A
Fuse is 1A fast blow
The rectifier implements Full-wave Rectification using 1N4002 diodes
Rectifier diodes must have a Peak Inverse Voltage of 100V
Filter Caps must be rated at 50V or above
Regulators 78xx/79xx have an drop-out voltage of 3V
Regulators output voltage will be within 14.4V - 15.6V
Maximum Current output of this PSU is 0.75A

So if I want to build this design I will have to take account of a number variables, I'm based in the UK which means, Mains Voltages operate at 230V 50Hz, Also when shopping for a transformer I realized the naming conventions for the transformers specs are also different from the article. Current ratings are marked in VA and you just deal with the transformers secondary windings to configure a centre-tap set-up.
Below I will be listing all the calculations from Osal linear power supply thread and applying them to this linear PSU design. I’ve already been through them once and used this information to buy the required parts for my build. I'm going to list my part specs from their data-sheet's so we can input this information into the various Voltage and current calculations .
Also I recommend reading through Osal thread here:
http://electro-music.com/forum/topic-51694.html
I won't be going into as much detail here, so you will need to reference that post and go through the examples provided. I will be using the same subject headings so it shouldn't be hard to cross reference.


VXT Vigortronix – Dual Primary/Secondary Toroidal Transformer
VA = 60, 36VCT 18 - 0 – 18, Max Output Current = 1.66A, Regulation 15%, Input Voltages - 115V/230V

Bridge Rectifier – GBU4D
Max Peak Reverse Voltage = 200V
Max Forward Current = 4A(with Heat-sink) 2.4A(without Heat-sink)
Peak forward Surge Current = 150A
Voltage drop Per element 1V

Voltage Regulators – 78xx / 79xx
Output Voltage = (7815CV = 14.4V – 15.6V) (7915CV = -14.4 – 15.6V)
Drop-out Voltage = 3V
Input Voltage = -35V / 35V (Max)
Output Current = 1.5A
So from the parts above I’ve decided to implement an +/- 15V dual supply. If you wanted to implement a +/- 12V dual supply you would run through the calculations below and pick the appropriate Voltage regulators and Transformer.

Current Calculations

Fuse: Current Rating
Slow blow fuse or anti-surge type must be used, This counters inrush current Problems
Fuse current rating = Volts-Amps (VA) / Voltage Primary (VP) * 1.5
let's input the Transformers VA value and the Mains Voltage Value
(60/230) * 1.5 = 0.391A
So we need to use a 20mm slow blow that is close this value.

Transformer Current Rating - Ratio Isec/Iout
With our Transformer specs we can calculate the amount of current it can supply our PSU.
Current Secondary (Isec) = Volts-Amps (VA) / Voltage Secondary (Vsec)
60 / 36 = Isec 1.66A
De-rating the Current Output of the Transformer is an important safety precaution.
Maximum Current Output (Iout) = Isec / 1.8
1.66/ 1.8 = 0.925A (Iout - Max)

Bridge Rectifier: Maximum Average Forward Current Rating:
Current Secondary (Isec) / 2
1.66 / 2 = 0.833A

Bridge Rectifier current rating = Iout * 1.3
0.925 * 1.3 = 1.2025A

Filter Capacitors: Ripple Current Rating
RMS Ripple Current Rating = Iout * 2.5
0.925 * 2.5 = 2.3145

Inrush Current
I measured my Transformers primary and secondary windings in series to find their combined resistance, these values allow us to calculate the inrush current when the PSU is switched on.
Primary Resistance (Rp) = 46R
secondary resistance (Rsec) = 2.4R
primary inrush current (Inrush-P) = Voltage Primary (Vp) / Rp
if we input the values
230/46 = 4.97A
secondary inrush current (Inrush-sec) = Vsec / Rsec
once again we stick in the values
36/2.5 = 15A

Voltage Calculations

Mains Voltage Rise & Drop.
UK Mains Voltages can be anywhere between +/- 10% of the Stated value of 230V. Usually the power supply is calculated to be immune to a voltage rise/drop of +/- 10% of its nominal value, 230V in this case

Transformer's Voltage regulation
Quote:
The nominal voltage at the secondaries is specified at full load (Always consult the manufacturer specifications). Therefore we can expect a higher voltage at the secondaries when there is no load.
It must be considered when calculating the voltage ratings of the components that follow the transformer.

So our transformer is rated as 60VA, 1.66A and 36VCT. It also has a regulation value of 15%
When the transformer is supplying 1.66A worth of current we expect to see 18VAC on each of the secondary windings. When they is no load we should see the value of 18VAC increase by 15%.
(18 * 0.15) + 18 = 20.7VAC (Vsec + 15%)

Forward voltage drop of the rectifier diode
Quote:
After the bridge rectifier, the wave's peak loses one or two rectifier forward voltage drop depending on the rectifier's configuration.
In the case of a dual center tapped rectifier (like the PS1) there is one diode voltage drop from ground to Vin Peak:

Secondary Voltage Peak AC(Vsp) – Rectifier Voltage Drop(Vd) = Peak Voltage at regulators(Vinp)

Voltage ripple at regulator's input
The Filter capacitor determines the ripple voltage at the regulator's input.
Osal states:
Quote:
The ripple voltage in volts peak to peak is defined by:
Vr=(Δτ*Iout)/C

Where (Δτ) means deviation in time and refers to the time in which the capacitor is discharging. 
To estimate the ripple, usually is assumed as worst case scenario that the capacitor discharges during all the cycle. However, in this context, according to my measurements, the longest discharging time is the 70%. of the cycle.
If the worst case scenario is a mains electricity frequency of 50hz, this 70% of the cycle, equals 0.007s. We can use it as a factor and the formula simplifies with:

Ripple Voltage (Vr) = (0.007*Iout) / capacitance in Farads (C)

Just out of interest the Value of 0.007 seconds is derived from the Mains Voltage frequency after the bridge rectifier. This is then converted from Hz to seconds.
First we need to look at the rectification process. If the Mains voltage frequency is 50Hz and we are using a bridge rectifier that implements full wave rectification, this process uses both positive and negative parts of the Mains AC input. This in-turn doubles the frequency at the bridge rectifier outputs. So Mains Frequency after the bridge rectifier = 50Hz * 2 = 100Hz

We then covert this value from Hz to seconds
1/100 = 0.01 seconds

We only need 70% of this value for the voltage ripple calculation
0.01*0.70 = 0.007seconds

OK lets add our values to the voltage ripple equation
(0.007 * 0.925) / 0.0068 = 0.9522V (voltage Ripple)

By the way a quick question is they a rule of thumb for the amount of ripple that is acceptable going into the voltage regulators? I guess this would guide your choice of capacitance values on the smoothing Cap's. I have read somewhere that the ripple voltage should be below 10% of the output voltage. I'm not sure if this is right.

Regulator dropout voltage
Quote:
The regulator needs an input voltage higher than the output voltage. The dropout voltage is the minimum difference within the output and the input voltage to ensure correct regulation.

78xx/79xx regulator's drop-out voltage = 3V

The Crest Factor of the wave at the transformer secondaries.
Crest factor approximation = 1.3

Transformer Voltage Rating
Quote:
The primary's voltage rating must be chosen according to the mains voltage of the country in which the power supply will operate.
The secondary's voltage rating is determined by:


Vsec=((Vout+Vdr+Vr+Vd)/CF)*(100/(100-imvd))

(Vsec) – AC Voltage at secondarys
(Vout) - Regulator's Output = 15V
(Vdr) – Voltage dropout regulators = 3V

(Vr) – Ripple Voltage
Vr = (0.007*Iout) / C
C = 6800uF with 20% tolerance
C = 6800*0.80 = 5440uF (minimum Cap Value)
Vr = (0.007*0.925) / 0.005440 = 1.1V

(Vd) – Forward Voltage Drop(Rectifier) = 1V
(CF) = Crest Factor = 1.3
(ImVd) = Immunity to mains Voltage drop = 10%
When we Input these values we get:
Vsec = (( 15+3+1.1+1 ) / 1.3) * (100 / (100 – 10)) = 17.2V
Quote:
So according to our calculations the minimum voltage rating we need is 17.2V.
We choose the next higher standard value which is 18V. This calculation was for one secondary. Thus, for two secondaries, the voltage rating we choose is 36VCT.


Rectifier diode voltage rating. Peak Reverse Voltage.
Quote:
To calculate the voltage rating of the rectifier diode, or Peak Reverse Voltage, we must consider the maximum peak voltage across the diode that can happen under the worst case conditions. The two worst case conditions that we consider are:
1.When the mains voltage rises and,
2.When the power supply has no load because the voltage rises due to the voltage regulation of the transformer.

Transformer – 60VA, 36VCT, 15% Voltage Regulation
Mains Voltage Rise = +/- 10%
Quote:
Let's focus now on one secondary. The voltage at the secondary is 18VAC. This specification refers to the nominal voltage at the primary. So, if the voltage at the primary rises a 10%, this will also happen at the secondary. Therefore:

Mains Voltage Rise – (18 * 0.1) + 18 = 19.8VAC
Transformer Regulation – (19.8 * 0.15) + 19.8 = 22.77V
Secondary Peak Voltage – 22.77 * 1.414 = 32.2V (Vpeak)
Maximum Peak Reverse Voltage – 32.2 * 2 = 64.4V
So, the peak reverse voltage rating must be equal or higher than 64.4V

Capacitor voltage rating
Transformer – 60VA, 36VCT, 15% Voltage Regulation
Mains Voltage Rise = +/- 10%
Quote:
To calculate the voltage rating of the filter capacitors we must consider the same worst case situations as for the rectifier.


Mains Voltage Rise – (18 * 0.1) + 18 = 19.8VAC
Transformer Regulation – (19.8 * 0.15) + 19.8 = 22.77VAC
Peak Voltage – 22.77 * 1.414 = 32.196VAC
We choose the next higher standard value which is 35V.

If we condense all for these values down and compare them to our selected components data-sheet's we can be a little more confident about building a Mains AC Power Supply.

Voltage Ratings
Tranformer minimum voltage rating = 34.4V
Secondary Peak Voltage (Vpeak) = 32.2V
Maximum Peak Reverse Voltage = 64.4V
Current Rating
PSU Max Current Output = 0.925A
Inrush Current Primary = 5A
Inrush Current Secondary = 15A
Bridge Rectifier: Maximum Average Forward Current = 1.2025A
Fuse Current rating = 0.391A

Getting the right Heat-sinks for your PSU is pretty important. I know these calculations are not posted to Osal PSU thread at the minute, but I have found another thread were he discusses this topic http://electro-music.com/forum/viewtopic.php?highlight=heatsinks&t=54839

Last edited by mrmrshoes on Wed Jul 31, 2013 5:57 am; edited 2 times in total
Back to top
View user's profile Send private message
Osal



Joined: Aug 16, 2011
Posts: 147
Location: Here

PostPosted: Sat Jul 27, 2013 2:54 pm    Post subject: Re: Thomas Henry - PSU for Electronic Music
Subject description: Advice needed
Reply with quote  Mark this post and the followings unread

Hello mrmrshoes,

mrmrshoes wrote:

Also it would be good for someone else to review the schematic and PCB before I build it, just in-case I've missed or overlooked anything.

All seems OK

Quote:
So with this first post I was wanting to concentrate on the Mains wiring, Dual Primary / Secondary Transformer wiring and how the transformer secondary windings connect to the rest of the circuit (bridge rectifier - GBU4D)
I guess my main question here is do you have to make specific connections from the Transformer secondary winding's (AC+, AC-) to the bridge rectifier AC inputs?

Just connect the red and orange cables of your transformer to the AC pins of the bridge rectifier.

Quote:
I think that should pretty much cover the basic wiring and connections of the various parts. Please correct me or add anything you feel is important.

The ground of the power supply's PCB must be connected to the earth in order to reference it.
Thus, I would add another Blade Connector on the PCB to connect its ground to the earth

All conductive parts of the synthesizer's case must be connected to the earth of the mains electricity for safety.

Quote:
The hope is to have a design that you can pick and choose what output rails you want +/- 15V, +/- 12V. Run through the calculations to determine what transformer and components you need to buy. I'll wait until my logic is confirmed to be good before completing the second post.

The components' voltage ratings for a +/-15V power supply are good also for a +/-12V. There are no considerable price differences. Regarding the capacitors, you could use smaller ones for a +/-12V... not very relevant at all.

However, the voltage rating of the transformer is really important.
In this context, for a +/-15V power supply we use a 36VCT transformer.
For a +/-12V power supply we use a 30VCT transformer. They are the standard values. Run through the calculations to see it.

Regarding the PCB layout:
I would assemble the transformer on the case, close to the PCB, instead of on the same PCB.
The components of the power supply will be more cool and it is much more easy to assemble.
You could use then the free space of the PCB to place single larger capacitors.
And much more important, larger heat-sinks.

_________________
electronic-sea.net
Back to top
View user's profile Send private message
feggster



Joined: Sep 12, 2011
Posts: 52
Location: uk

PostPosted: Sun Jul 28, 2013 2:12 am    Post subject: Reply with quote  Mark this post and the followings unread

I'm in the process of building a power supply for a small amount of modules (nicholas woolaston's simple vco's etc)

my transformer : http://uk.farnell.com/pro-power/ctfc20-18/transformer-20va-2-x-18v/dp/1780851

20VA, 2 X 18V
Primary Voltages: 2 x 115V
Secondary Voltages: 2 x 18V
Current Rating: 556mA
Power Rating: 20VA
Input Voltage: 115V
Mounting Type: Chassis
Output Voltage: 18V
Secondary Current Nom: 556mA
Transformer Type: Isolation

I used osal's calculations for the fuse current and ended up with a figure of 129ma.... so I'm guessing I use a 125ma slo blow fuse.
I noticed in Osal's calc for the fuse there is a 'difference' in actual values..
Quote:
the current rating of the fuse (Ifuse) is, as we said:
Ifuse=Ip*1.5
Ifuse=0.218*1.5= 0.327A
In this case a 315mA blow slow fuse will do the job.

whats the safe margin for this?

.. apologies Embarassed copied to http://electro-music.com/forum/topic-51694.html

Last edited by feggster on Sun Jul 28, 2013 8:05 am; edited 1 time in total
Back to top
View user's profile Send private message
Osal



Joined: Aug 16, 2011
Posts: 147
Location: Here

PostPosted: Sun Jul 28, 2013 5:08 am    Post subject: Reply with quote  Mark this post and the followings unread

Hey feggster! If there is any question about my thread I would appreciate if you can ask it there. It will help me to explain it better and in the same time we allow this thread for mrmrshoes's power supply.
Edit: Could you please copy this question in my thread?

_________________
electronic-sea.net
Back to top
View user's profile Send private message
mrmrshoes



Joined: Feb 19, 2011
Posts: 73
Location: Newcastle Upon Tyne
Audio files: 4

PostPosted: Sun Jul 28, 2013 3:53 pm    Post subject: Reply with quote  Mark this post and the followings unread

Alright Osal

Cheers for getting back to me like

Quote:
All seems OK


Cool man, That’s what I was wanting to hear like. I thought everything was in order for this linear PSU design, I just feel better about it when a second pair of eyes have gone through it because of the substitute parts in this layout.

Quote:
Just connect the red and orange cables of your transformer to the AC pins of the bridge rectifier. 


That’s good to know like. I had read for linear designs that the phase relationships of the Transformers windings wasn't critical. I believe that comes into play for switching supplies.

Quote:
The ground of the power supply's PCB must be connected to the earth in order to reference it. 
Thus, I would add another Blade Connector on the PCB to connect its ground to the earth 

All conductive parts of the synthesizer's case must be connected to the earth of the mains electricity for safety. 


**EDIT**
Sorry my last comments were poorly written, I'll try again.

I can't believe I left this out. MAINS Earth should be connected to any conductive parts of the case for electrical safety.

Things get confusing when you start connecting the various GND types together within an system.
You have MAINS Earth, Signal GND and digital GND, all end up flowing back to the source of power (PSU, MAINS Earth) but need to be handled in certain ways to reduce noise and various problems. If you want to read about have a look at this link. http://www.modular.fonik.de/files/grounding_101.pdf

Osal mention I should add a connector to the PSU GND plane so I could reference it to MAINS Earth. I called this a star connection because I've read somewhere that MAINS Earth should connect to the rest of the system in one place and one place only.

It's this grounding topic I kinda wanted to side step because I find myself getting lost down the rabbit hole pretty quickly and i'm finding it pretty hard to explain.

I have found a few resources on the net that cover transformer wiring, Mains safety and grounding issues, I will edit my first post and fire up the links. It would be cool to come back to this very confusing grounding concept because it is related to PSU building and has to be considered.


Quote:
The components' voltage ratings for a +/-15V power supply are good also for a +/-12V. There are no considerable price differences. Regarding the capacitors, you could use smaller ones for a +/-12V... not very relevant at all. 


Right I see. If it work for +/- 15V its going to work on +/- 12V. Pretty good way to look at it.

Quote:
However, the voltage rating of the transformer is really important. 
In this context, for a +/-15V power supply we use a 36VCT transformer. 
For a +/-12V power supply we use a 30VCT transformer. They are the standard values. Run through the calculations to see it. 


Yeah man. I've hand written all the calculations as notes, I'll post them as soon as I’m finished typing them up.

Quote:
Regarding the PCB layout: 
I would assemble the transformer on the case, close to the PCB, instead of on the same PCB. 
The components of the power supply will be more cool and it is much more easy to assemble. 
You could use then the free space of the PCB to place single larger capacitors. 
And much more important, larger heat-sinks.


Good points. I've been looking at heat-sinks on Rapid electronics. Once I’ve finished the calculations post we can flesh things out and update the PCB layout. The heat-sinks I bought for this this project were rated at 5.6cw. I was going to limit the PSU current draw to 0.75A so I kinda thought that would cover it. Anyway we can come back to this once I've list the current and voltage calculations for the components I have bought.

Thanks a lot for your input man, it's been really helpful. I'll get the second post typed up and posted asap.

Cheers dude

shoes

p.s do you want these posts transferred over to yours?
Back to top
View user's profile Send private message
mrmrshoes



Joined: Feb 19, 2011
Posts: 73
Location: Newcastle Upon Tyne
Audio files: 4

PostPosted: Wed Jul 31, 2013 5:52 am    Post subject: Reply with quote  Mark this post and the followings unread

I've updated the first two posts. All of the Current and Voltage calculations are listed.
I just need to figure out what heat sinks I should use and I should be set.
I thought I would upload a write up of a DIY case I'm planning on using, Generally cases and PSU's go hand in hand so it seemed like the thing to do.

For the Case I will be trying too get hold of some proper Eurorack Rails, The Rails used in the PDF are to big for Eurorack format.

I'll probably end up mounting the PSU behind a front panel so hopefully I have an easier time Earthing the Metal in the case.

Anyway if you see anything I’ve missed please comment.

Cheers

shoes


DIY-Euro-Rack.pdf
 Description:

Download
 Filename:  DIY-Euro-Rack.pdf
 Filesize:  400.92 KB
 Downloaded:  1766 Time(s)

Back to top
View user's profile Send private message
mrmrshoes



Joined: Feb 19, 2011
Posts: 73
Location: Newcastle Upon Tyne
Audio files: 4

PostPosted: Fri Aug 02, 2013 5:39 pm    Post subject: Reply with quote  Mark this post and the followings unread

Alright Osal
I was wondering if you have time to check these numbers out? I've learnt a lot from you posts but I’ve not been able to figure out a couple of things. The first was how do you select capacitance values for the Filter Caps. I know your posts cover ripple Voltages on the regulators inputs but I was picking cap values at random. I not smart enough to rearrange the calculations so I’ve been stuck for awhile. I was rooting around the internet and found Jim Patchell power supply page.
http://www.noniandjim.com/Jim/powersupply/powersupply.html
As luck would have it he's describes this very problem. I've added mains voltage drop to his calculations but I was wondering if they is anything you would change?

Worst case scenario, Mains Voltage drop of 10%
Maximum output Current = 0.925A
One Transformer Secondary Voltage 36V / 2 = 18V
Secondary Peak Voltage 18V * 1.414 = 25.45V
Rectifier Voltage drop 25.45V – 1V = 24.45
Mains 10% Voltage Drop 24.45V * 0.9 = 22V
To get 15V out of the regulator The Voltage into the reg cannot go below 18V. So take away 18V from our worst case Secondary Voltage we get the value of our Voltage Ripple.
Ripple Voltage 22V – 18V = 4V

We need the filter capacitor discharge rate as a period(seconds), The worst case value is based on Mains voltage Frequency (50Hz). We are using full wave rectification in our circuit so you double this value 50 * 2 = 100Hz
we convert 100Hz to seconds by 1/100 = 0.01 seconds or 10 milliseconds.
Now we can input these values to determine minimum capacitance values for the filter caps:

Filter Capacitance (farads) = ( Period(seconds) * Current(Amps) ) / Ripple Voltage(Volts)
(0.01 * 0.925) / 4 = 0.0023125 F (2312.5uF)

So I guess the next value up would do the job (3300uF)

One last thing is heat-sinks. Do you have any advice on PCB mounted ones. When I ran through your calculations I was getting some low numbers. For 0.75A Current output I was getting just under 3C/W. For 0.925A It was around about 1.5C/W. I'm not sure if I am crunching the numbers right but yeah man you were dead on, much bigger heat-sinks are need.
I probably going to take your advice and remove the transformer from the PCB. I'll use this extra space for heat-sinks. When I come to mounting the PSU I’ll use sheet metal to mount the PCB and transformer, either in the case (with stand offs) or behind a front panel.

Anyway man

Cheers for any help
Back to top
View user's profile Send private message
Osal



Joined: Aug 16, 2011
Posts: 147
Location: Here

PostPosted: Sat Aug 03, 2013 9:40 am    Post subject: Reply with quote  Mark this post and the followings unread

Hey 4700 to 10000uf is fine. More large means more immunity to mains voltaje fluctuations.

About the heat sink , for worst case situations = 10% high mains voltaje fluctuaciones And 40C temperature ambient, The thermal resistance should be same or smaller tan 3C/w

PCB mounted heatsinks Will do more easy the assambling which is advisable.

I have no good internet connection, not every day, nor much time, Sorry i Will go into detail end of this month.

Oscar

_________________
electronic-sea.net
Back to top
View user's profile Send private message
mrmrshoes



Joined: Feb 19, 2011
Posts: 73
Location: Newcastle Upon Tyne
Audio files: 4

PostPosted: Sat Aug 03, 2013 11:48 am    Post subject: Reply with quote  Mark this post and the followings unread

Cheers for getting back to me man.

Quote:
I have no good internet connection, not every day, nor much time, Sorry i Will go into detail end of this month.


No probs like. I'm really looking forward to reading your take on things. your posts are proper helpful. Wink

Quote:
Hey 4700 to 10000uf is fine. More large means more immunity to mains voltaje fluctuations.
About the heat sink , for worst case situations = 10% high mains voltaje fluctuaciones And 40C temperature ambient, The thermal resistance should be same or smaller tan 3C/w

PCB mounted heatsinks Will do more easy the assambling which is advisable.


good to know man.
I've been looking at heatsinks and i think my best opiton will be a heatsink designed for euro cards made by Fischer Elektronik. The product number is SK 96 84 SA and is rated at 3.8C/W.
check out the datasheet on this page.
http://www.rapidonline.com/Electronic-Components/Fischer-Elektronik-3-8-Heat-Sink-519800

We should be able to tweek the output current of the PSU to allow this part in the PCB layout. Eagle CAD has this part in its library so no messing around designing foot prints Laughing

Anyway man take it easy

I'm looking forward to your update
shoes
Back to top
View user's profile Send private message
mrmrshoes



Joined: Feb 19, 2011
Posts: 73
Location: Newcastle Upon Tyne
Audio files: 4

PostPosted: Sun Aug 11, 2013 12:13 pm    Post subject: Reply with quote  Mark this post and the followings unread

Quick update.

I've redone the layout with the new heatsink, thought i would post an image like.

I've also added a star ground connection onto the board.


PSU.png
 Description:
 Filesize:  131.33 KB
 Viewed:  345 Time(s)
This image has been reduced to fit the page. Click on it to enlarge.

PSU.png


Back to top
View user's profile Send private message
Osal



Joined: Aug 16, 2011
Posts: 147
Location: Here

PostPosted: Mon Aug 12, 2013 3:25 am    Post subject: Reply with quote  Mark this post and the followings unread

Just to say that 3cw is for one reugulator. The same heat-sink for both regulators should be under 2cw.
_________________
electronic-sea.net
Back to top
View user's profile Send private message
mrmrshoes



Joined: Feb 19, 2011
Posts: 73
Location: Newcastle Upon Tyne
Audio files: 4

PostPosted: Wed Aug 14, 2013 5:03 am    Post subject: Reply with quote  Mark this post and the followings unread

Cheers for the head up man. well spotted like.
After carrying out some more research on heat-sinks i think I've got it cracked.

I had been basing my layout on parts from rapid electronics, but they don't have a wide enough range. Once i started rooting through the PCB mount heat sinks at Farnell I've found a number of suitable candidates.
I've went for the original PCB Mount heat-sink foot print (25mm solder pins), So i'll have to up date the PCB layout again. I should be able to get two PSU's out of one 100mm * 160mm board

anyway here are links to 3 C/W heat-sinks.

http://uk.farnell.com/ohmite/fa-t220-64e/heatsink-to-220-black/dp/2097688?Ntt=FA-T220-64E

http://uk.farnell.com/aavid-thermalloy/6399bg/heat-sink-to-220-218-3-3-c-w/dp/1213474

http://uk.farnell.com/aavid-thermalloy/6400bg/heat-sink-to-220-218-2-7-c-w/dp/1213475

http://uk.farnell.com/multicomp/mc33267/heatsink-to220-218-3-1-c-w-notched/dp/1710610?Ntt=MC33267

http://uk.farnell.com/multicomp/mc33271/heatsink-to218-220-247-2-7-c-w/dp/1710614?Ntt=MC33271

http://uk.farnell.com/aavid-thermalloy/530002b02500g/extruded-heat-sink/dp/1578090
Back to top
View user's profile Send private message
Osal



Joined: Aug 16, 2011
Posts: 147
Location: Here

PostPosted: Thu Aug 29, 2013 11:15 am    Post subject: Reply with quote  Mark this post and the followings unread

Hey mrmrshoes,

mrmrshoes wrote:
I was wondering if you have time to check these numbers out?

The numbers in post 2 are OK.

However notice that you have calculated the maximum current rating of the components in relation to the maximum current rating of the transformer.
Due all the components determine the current rating of the power supply, we have to look to the bottle neck of the design.
In the case of the linear power supplies using standard regulators and passive dissipation for the regulators, the bottle neck is the dissipation.
I think, and according several measurements and estimations I did, that 0.8A would be the maximum current output for these power supplies.
In any case the result of the calculations are in the same way useful. I just wanted to point to this.
Quote:

Bridge Rectifier – GBU4D
Max Peak Reverse Voltage = 200V
Max Forward Current = 4A(with Heat-sink) 2.4A(without Heat-sink)
Peak forward Surge Current = 150A
Voltage drop Per element 1V

Do you have its data-sheet?
Quote:

Fuse: Current Rating
Slow blow fuse or anti-surge type must be used, This counters inrush current Problems
Fuse current rating = Volts-Amps (VA) / Voltage Primary (VP) * 1.5
let's input the Transformers VA value and the Mains Voltage Value
(60/230) * 1.5 = 0.391A
So we need to use a 20mm slow blow that is close this value.

So, the next standard value is 400mA

Quote:
Bridge Rectifier: Maximum Average Forward Current Rating:
Current Secondary (Isec) / 2
1.66 / 2 = 0.833A
Bridge Rectifier current rating = Iout * 1.3
0.925 * 1.3 = 1.2025A

These are the same parameter. The first is calculated from the current in the secondaries and the second from the current output and adds a derating factor. However your bridge is rated 2.4 to 4A so its OK.

_________________
electronic-sea.net
Back to top
View user's profile Send private message
Osal



Joined: Aug 16, 2011
Posts: 147
Location: Here

PostPosted: Thu Aug 29, 2013 12:36 pm    Post subject: Reply with quote  Mark this post and the followings unread

Quote:
By the way a quick question is they a rule of thumb for the amount of ripple that is acceptable going into the voltage regulators?


A small ripple at the input of the regulators means a smaller ripple at its output.
A small ripple at the input of the regulators means more immunity to the mains voltage drops.
These are desirable.
A small ripple at the input of the regulators means a higher voltage at the regulators input, thus more power to dissipate.
This is not desirable.
However, a small ripple has more effect over the immunity to the mains voltage drops than over the power to dissipate.

Then... what ripple is the correct? To me, the smaller the best.
But what capacitor's size is enough?
Each time that we double the capacitance we halve the ripple voltage.
It has sense to halve a ripple from 2 volts to one volt. You will improve the immunity to mains voltage drops.
But halving from one volt to half volt doesn't has a considerable effect.
So, as a rule of thumb we could calculate it to achieve a ripple voltage of one volt peak to peak.

For a 0.8A power supply it is:
C=0.007s*0.8A/1Vp-p
C=0.0056F
C=5600μF
The worst case value for a 6800μF capacitor with a 20% of tolerance is 6800μF-(68*20)=5440μF
So a 6800μF or larger capacitor would be appropriate.

I'll explain this a little bit more in a post in my power supply thread. Thanks.

_________________
electronic-sea.net
Back to top
View user's profile Send private message
Osal



Joined: Aug 16, 2011
Posts: 147
Location: Here

PostPosted: Fri Aug 30, 2013 9:16 am    Post subject: Re: Thomas Henry - PSU for Electronic Music
Subject description: Advice needed
Reply with quote  Mark this post and the followings unread

Osal wrote:

Regarding the PCB layout:
I would assemble the transformer on the case, close to the PCB, instead of on the same PCB.


I have seen that your case is wood made, a very nice suitcase I have to say.
In this case I wouldn't assemble the transformer on the case.

To house the power supply in a wood case could be dangerous. So, take special attention to the correct size of the fuse and to the sector of the PCB from the secondaries to the regulators. Protect it to avoid any accidentally short.

Avoid to assemble the transformer directly on the wood. You could assemble it on an aluminum sheet, and it on the wood, separated 1cm or so.
Another possibility could be a an external power supply, but this, I guess, is not so useful for a mobile device as a suitcase-synthesizer.

A draw·back of the suitcase could be a poor ventilation. This makes more difficult the heat dissipation of the power supply.
However this could be measured after you finish the synthesizer and if it is needed you could consider to open a few holes to add ventilation to the device.

_________________
electronic-sea.net
Back to top
View user's profile Send private message
Osal



Joined: Aug 16, 2011
Posts: 147
Location: Here

PostPosted: Sat Aug 31, 2013 8:27 am    Post subject: Reply with quote  Mark this post and the followings unread

mrmrshoes wrote:
anyway here are links to 3 C/W heat-sinks.


Regarding the manufacturer thermal specifications:
In a set of measurements and estimations I did, I saw that the manufacturer specifications could differ greatly from the actual specifications in our application.
I guess that this is because the specifications of thermal resistance could depend on the test conditions.

Look for example the AAVID 530002B0250G, the Omhite FA-TO220-64E and the Fisherelektronik SK 129 63.5.
They are very similar in size and shape, but the manufacturers specify them with 2.6C/W, 3C/W and 4.5C/W respectively.
I made several temperature measurements using an Omhite FA-TO220-64E with a LM317 and estimated a thermal resistance of more than 6C/W.

We need to calculate and estimate the heat-sink in order to start from somewhere, but then, when the power supply, or the prototype, is finished, we should run temperature measurements to estimate more accurately the specifications of the power supply and then, to modify the power supply, if it is necessary.

I just calculated the heat sink's thermal resistance for the following situations: 0.8A, 10% mains voltage rise, max ambient temperature=40C and filter capacitors=6800μF.
The thermal resistance must be equal or smaller than 5.3C/W
However, due what I explained before, I would try a heat sink with smaller thermal resistance.

Now, that you have already build the suitcase and that you did considerable work on the power supply, in order to do it easy, I would try one of these heat sinks I have just mentioned, which are very easy to assemble and then, if you are interested you could run several temperature measurements (I have to say that they are very time consuming) and consider if they are enough, or if you want to change or modify something.

_________________
electronic-sea.net
Back to top
View user's profile Send private message
mrmrshoes



Joined: Feb 19, 2011
Posts: 73
Location: Newcastle Upon Tyne
Audio files: 4

PostPosted: Wed Sep 11, 2013 10:14 am    Post subject: Reply with quote  Mark this post and the followings unread

Alright man

Sorry about the delay, I've been on holiday for two weeks and didn't have access to the internet.
Thanks for getting back man. I really appreciate it. Your last three posts have cleared everything up and have been a great help with this project.

Here is the GBU4D data sheet
http://www.rapidonline.com/pdf/47-2907e.pdf

I need to buy the few remaining parts so I can start building.

I'll get this heat sink
http://uk.farnell.com/aavid-thermalloy/530002b02500g/extruded-heat-sink/dp/1578090

It's one of the possible options I listed above rated at 2.6°C/W and fits the 25mm footprint.

I need a couple of other bits and pieces (fuses, metal sheet, stand offs for case mounting)

Hopefully I’ll get hold of these things sooner rather than later and when I get a working unit I’ll post some pictures and what not.

Once again dude, thanks for your help.
Back to top
View user's profile Send private message
Osal



Joined: Aug 16, 2011
Posts: 147
Location: Here

PostPosted: Wed Sep 11, 2013 2:20 pm    Post subject: Reply with quote  Mark this post and the followings unread

I'll follow this thread to see the power supply progress. Very Happy

A few comments more:

To assemble the regulator on the heat sink is convenient to use a spring washer and then you tighten the nut until the end of the spring washer. Not more.
In this way you are sure that it is well hold without the risk of damage it or to bend the regulator's case loosing contact surface.

Using an insulating pad and an insulating washer the assembling is more neat and the heat sinks are insulated.
So, it could be like this:
Machine screw 3mm/metal washer/insulating washer/regulator/insulating pad/heat-sink/metal washer/spring washer/ nut 3mm

First attach the regulator to the heat sink and then solder it to the PCB, to avoid mechanical stress to the solder joints.

_________________
electronic-sea.net
Back to top
View user's profile Send private message
Colonel Steel Lungs



Joined: Oct 23, 2013
Posts: 2
Location: Blackhill

PostPosted: Sun Nov 30, 2014 7:04 am    Post subject: Reply with quote  Mark this post and the followings unread

Howdy.

So for calculating the thermal resistance of the heat sinks, what values are you substituting into the equations? From what I understand it's basically Ohm's Law (R = thermal resistance.) In that case are we just getting the VI from the regulator Vmax and Imax?
Back to top
View user's profile Send private message
Osal



Joined: Aug 16, 2011
Posts: 147
Location: Here

PostPosted: Tue Dec 09, 2014 4:13 am    Post subject: Reply with quote  Mark this post and the followings unread

Colonel Steel Lungs wrote:
Howdy.
From what I understand it's basically Ohm's Law (R = thermal resistance.)


The best way to understand thermal dissipation is making a parallelism with electricity.
Thermal resistance Θ equals Resistance R
Power dissipation P equals Current Ι
Temperature Τ equals Voltage V

The purpose of the dissipation is to avoid the destruction or damage of the regulator due high temperature.
The power to dissipate from the regulator's junction behaves like a current source.
So you need enough lower thermal resistance from the regulator's junction to the ambient (Following the parallelism equals Ground) to avoid the temperature rise over the operating junction temperature range which usually is 125C.

Ill extend the post later.

_________________
electronic-sea.net
Back to top
View user's profile Send private message
Display posts from previous:   
Post new topic   Reply to topic Moderators: jksuperstar, Scott Stites, Uncle Krunkus
Page 1 of 1 [21 Posts]
View unread posts
View new posts in the last week
Mark the topic unread :: View previous topic :: View next topic
 Forum index » DIY Hardware and Software
Jump to:  

You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum
You cannot attach files in this forum
You can download files in this forum


Forum with support of Syndicator RSS
Powered by phpBB © 2001, 2005 phpBB Group
Copyright © 2003 through 2009 by electro-music.com - Conditions Of Use