Author Topic: DIY SCPI programmable dual channel bench PSU 0-50V/3A (now EEZ H24005)  (Read 451942 times)

0 Members and 1 Guest are viewing this topic.

Offline uChip

  • Contributor
  • Posts: 35
  • Country: us
Re: DIY programmable dual channel bench PSU 0-50V/3A
« Reply #25 on: February 06, 2015, 06:05:21 am »
Kevin,
I think you are interpreting my posting above incorrectly.  The JYE Teck kit I refer to has nothing to do with PS design.  As far as I am aware prasimix' design is his own based on previous designs from Ian Johnston and Gerry Sweeney.  Rereading my posting, I can see how you might have interpreted it otherwise.  I'll go back and edit to clarify.
Thanks,
Chip
 

Offline prasimixTopic starter

  • Supporter
  • ****
  • Posts: 2022
  • Country: hr
    • EEZ
Re: DIY programmable dual channel bench PSU 0-50V/3A
« Reply #26 on: February 06, 2015, 07:29:22 am »
Nice project prasimix.
How much did all those PCBs set you back?

Huh, it's not easy to say since I made a panel with some other PCBs that not belong to this project. A good way to estimate a cost is to use their online calculator.

Offline prasimixTopic starter

  • Supporter
  • ****
  • Posts: 2022
  • Country: hr
    • EEZ
Re: DIY programmable dual channel bench PSU 0-50V/3A
« Reply #27 on: February 06, 2015, 07:45:39 am »
Nice project. I can see you put alot of effort into it, is it your schematic or from an already  published/tested project ? The reason I ask is that I think there may be a stability/CC performance issue due to the way Q4 is driven. Do you have a spice sim available to post (just the basic main control loop components is all thats needed to show if this is so Q4,Q5,Q6, IC7A-IC7D and associated small  stuff) .

Regards

Edit .I should have read the other few posts above :) I  see someone says it's a kit from  JYETech.com .Still do you have a sim available ?

Thanks Kevin.D. As I mentioned in the first post, I don't have any simulation available. LTspice looks nice but it lacks most of components that I'm using. Any suggestion or help regarding this topic is highly welcomed. When we talks about power board everything is based on many hours spent in front of breadboard. I started with something that was unstable but over time with adding few components and fine tuning (like C49, C54, C55, C56, R39 and R48) I think that I got a stable solution over whole range (0-50V, 0-3A) at least when testing with pure resistive loads (16R8, 33R and 66R). Of course all suggestions how to possibly improve the design are also highly welcomed.
 
The following users thanked this post: uliano

Offline void_error

  • Frequent Contributor
  • **
  • Posts: 673
  • Country: ro
  • I can transistor...
Re: DIY programmable dual channel bench PSU 0-50V/3A
« Reply #28 on: February 06, 2015, 11:52:32 am »
For simulations I use SIMetrix, the free version. Its only drawback is that it can only simulates circuits up to 140 nodes. I tried LTspice myself and found it too fiddly because of it's quite primitive UI.

Other tan that, what can I say? Nice design, although a bit too complex for my taste. Keep up the good work :-+
Trust me, I'm NOT an engineer.
 

Offline prasimixTopic starter

  • Supporter
  • ****
  • Posts: 2022
  • Country: hr
    • EEZ
Re: DIY programmable dual channel bench PSU 0-50V/3A
« Reply #29 on: February 07, 2015, 07:22:32 am »
Thanks for recommendation. It seems that it cannot be installed thru Wine on Linux but I'll try it using Virtualbox.

Offline prasimixTopic starter

  • Supporter
  • ****
  • Posts: 2022
  • Country: hr
    • EEZ
MCU board and keypad assembled ...
« Reply #30 on: February 11, 2015, 01:10:31 pm »
I know that this is not a correct order of presenting what is already done under PSU project but I didn't have a chance to make a lot of progress on power board (I got yesterday required inductor for bias ps) so I decide to test how MCU board is working.
The following pictures shows MCU board with minimum components required to make connection to ISP programmer (I'm using Arduino for that) and run simple test melody on buzzer. I reused the following code:

Code: [Select]
/*
 Melody
 
 Plays a melody
 
 created 21 Jan 2010
 modified 30 Aug 2011
 by Tom Igoe

This example code is in the public domain.
 
 http://arduino.cc/en/Tutorial/Tone
 
*/

#include "pitches.h"

// notes in the melody:
int melody[] = {
  NOTE_C4, NOTE_G3,NOTE_G3, NOTE_A3, NOTE_G3,0, NOTE_B3, NOTE_C4};

// note durations: 4 = quarter note, 8 = eighth note, etc.:
int noteDurations[] = {
  4, 8, 8, 4,4,4,4,4 };

void setup() {
  // iterate over the notes of the melody:
  for (int thisNote = 0; thisNote < 8; thisNote++) {

    // to calculate the note duration, take one second
    // divided by the note type.
    //e.g. quarter note = 1000 / 4, eighth note = 1000/8, etc.
    int noteDuration = 1000/noteDurations[thisNote];
    tone(A1, melody[thisNote],noteDuration); // Arduino pin A1 is PF6 (Pin 37) assigned to BUZZER

    // to distinguish the notes, set a minimum time between them.
    // the note's duration + 30% seems to work well:
    int pauseBetweenNotes = noteDuration * 1.30;
    delay(pauseBetweenNotes);
    // stop the tone playing:
    noTone(A1);
  }
}

void loop() {
  // no need to repeat the melody.
}

... and required header file:

Code: [Select]
/*************************************************
 * Public Constants
 *************************************************/

#define NOTE_B0  31
#define NOTE_C1  33
#define NOTE_CS1 35
#define NOTE_D1  37
#define NOTE_DS1 39
#define NOTE_E1  41
#define NOTE_F1  44
#define NOTE_FS1 46
#define NOTE_G1  49
#define NOTE_GS1 52
#define NOTE_A1  55
#define NOTE_AS1 58
#define NOTE_B1  62
#define NOTE_C2  65
#define NOTE_CS2 69
#define NOTE_D2  73
#define NOTE_DS2 78
#define NOTE_E2  82
#define NOTE_F2  87
#define NOTE_FS2 93
#define NOTE_G2  98
#define NOTE_GS2 104
#define NOTE_A2  110
#define NOTE_AS2 117
#define NOTE_B2  123
#define NOTE_C3  131
#define NOTE_CS3 139
#define NOTE_D3  147
#define NOTE_DS3 156
#define NOTE_E3  165
#define NOTE_F3  175
#define NOTE_FS3 185
#define NOTE_G3  196
#define NOTE_GS3 208
#define NOTE_A3  220
#define NOTE_AS3 233
#define NOTE_B3  247
#define NOTE_C4  262
#define NOTE_CS4 277
#define NOTE_D4  294
#define NOTE_DS4 311
#define NOTE_E4  330
#define NOTE_F4  349
#define NOTE_FS4 370
#define NOTE_G4  392
#define NOTE_GS4 415
#define NOTE_A4  440
#define NOTE_AS4 466
#define NOTE_B4  494
#define NOTE_C5  523
#define NOTE_CS5 554
#define NOTE_D5  587
#define NOTE_DS5 622
#define NOTE_E5  659
#define NOTE_F5  698
#define NOTE_FS5 740
#define NOTE_G5  784
#define NOTE_GS5 831
#define NOTE_A5  880
#define NOTE_AS5 932
#define NOTE_B5  988
#define NOTE_C6  1047
#define NOTE_CS6 1109
#define NOTE_D6  1175
#define NOTE_DS6 1245
#define NOTE_E6  1319
#define NOTE_F6  1397
#define NOTE_FS6 1480
#define NOTE_G6  1568
#define NOTE_GS6 1661
#define NOTE_A6  1760
#define NOTE_AS6 1865
#define NOTE_B6  1976
#define NOTE_C7  2093
#define NOTE_CS7 2217
#define NOTE_D7  2349
#define NOTE_DS7 2489
#define NOTE_E7  2637
#define NOTE_F7  2794
#define NOTE_FS7 2960
#define NOTE_G7  3136
#define NOTE_GS7 3322
#define NOTE_A7  3520
#define NOTE_AS7 3729
#define NOTE_B7  3951
#define NOTE_C8  4186
#define NOTE_CS8 4435
#define NOTE_D8  4699
#define NOTE_DS8 4978


« Last Edit: February 11, 2015, 01:17:53 pm by prasimix »
 

Offline prasimixTopic starter

  • Supporter
  • ****
  • Posts: 2022
  • Country: hr
    • EEZ
LCD display on MCU board ...
« Reply #31 on: February 11, 2015, 01:15:00 pm »
... and some pictures where a fully assembled board with connected LCD 128x64 and test message (using u8glib) is showed. Here is the code:

Code: [Select]
#include "U8glib.h" // Universal 8bit Graphics Library, http://code.google.com/p/u8glib/

// Define LCD type
// devices with all constructor calls is here: http://code.google.com/p/u8glib/wiki/device
U8GLIB_ST7920_128X64_1X u8g(A0); // Arduino Pin A0 is PF7 (pin 36) assigned to LCD_SELECT

void draw(void) {
  // graphic commands to redraw the complete screen should be placed here 
  //u8g.setFont(u8g_font_unifont);
  u8g.setFont(u8g_font_tpss);
  //u8g.setFont(u8g_font_osb21);
  //u8g.setFont(u8g_font_04b_03b);
  u8g.drawStr( 0, 10, "Welcome screen");
  u8g.drawStr( 0, 22, "PSU dual 0-50V/0-3A");
  u8g.setFont(u8g_font_04b_03b);
  u8g.drawStr( 0, 30, "- SMPS pre-regulator");
  u8g.drawStr( 0, 36, "- Linear post-regulator");
  u8g.drawStr( 0, 42, "- CC/CV, OCP");
  u8g.drawStr( 0, 48, "- 16-bit DAC, 15-bit ADC"); 
  u8g.drawStr( 0, 54, "- Isolated USB and Ethernet");
  u8g.drawStr( 0, 60, "- Arduino sketches");
}

void setup(void) {
  // Set LCD brightness
  pinMode(11, OUTPUT); // Arduino Pin 11 is PB7 (pin 12) assigned to LCD_BRIGHTNESS
  analogWrite(11, 200); // (value from 0 to 255, 0=dark)

  if ( u8g.getMode() == U8G_MODE_R3G3B2 ) {
    u8g.setColorIndex(255);     // white
  }
  else if ( u8g.getMode() == U8G_MODE_GRAY2BIT ) {
    u8g.setColorIndex(3);         // max intensity
  }
  else if ( u8g.getMode() == U8G_MODE_BW ) {
    u8g.setColorIndex(1);         // pixel on
  }
  else if ( u8g.getMode() == U8G_MODE_HICOLOR ) {
    u8g.setHiColorByRGB(255,255,255);
  }
 
  u8g.firstPage(); 
  do {
    draw();
  } while( u8g.nextPage() );
}

void loop(void) {

  // Nothing to do
 
}
« Last Edit: February 11, 2015, 01:26:07 pm by prasimix »
 

Offline prasimixTopic starter

  • Supporter
  • ****
  • Posts: 2022
  • Country: hr
    • EEZ
Aux power supply
« Reply #32 on: February 12, 2015, 08:48:06 am »
Aux power supply PCB is assembled and tested. LM5574 works nicely and I choose it because it is also used for bias power supply on power board when input voltage could go up to 67V. The board is intended to provide separated +5V for MCU board and Binding posts board. Max. consumption is ~400mA (mostly consumed by binding posts LEDs and relays when all are active) which LM5574 is delivering without problem.

Here are some pictures of assembled board and basic measurements.

Offline prasimixTopic starter

  • Supporter
  • ****
  • Posts: 2022
  • Country: hr
    • EEZ
Power pre-regulator inital testing
« Reply #33 on: February 14, 2015, 02:24:18 pm »
Power pre-regulator is located on the power PCB (Reply #22) and initially presented on the Sheet 1 (Reply #1). I'd like to share some information about initial testing of this part of the system.

First I'd like to starts with some mistakes. Input filter capacitors are located optimistically too close to the rectifier diodes D2 and D3. Since that diodes has to be mounted with screw to the main heatsink there is not many spaces for maneuvering with a screwdriver. It's so difficult to mount mentioned diodes but they could be more accessible.
Another mistake that cost me one LM5118 is that I forgot to add a small resistor in serial with P1. In normal operation a whole section P1, R1 and R2 is not necessary since PREG_FB signal (Sheet 2) will be used for controlling PREG_OUT. I added this section deliberately because in that way you can assemble and test power pre-regulator section without tracking control and the rest of the post-regulator (Sheet 2). So what's happened? P1 was initially set for Vout below 10V and I started to turn P1 too fast (approaching ground) and since LM5118 is buck-boost controller it has no problem to increase Vout to 100V! I heard click and the next action was removing LM5118 with rose's metal from the PCB and mounting another one :headbash:.

In attachment you can see pre-regulator section of the power PCB. Power inductor (TR1) is Wishay IHTH1125MZEB101M5A. Selected heatsink is RAD-A52317/50 (painted in black). I'll play in the future with other inductors and try heatsink Fischer SK85/50/SA.

Currently I've measured the following temperatures on some key components under max. load when LM5118 enters buck-boost mode (@Iout=3A, Vin=48VAC, Load=16R4, Tabient=23oC):
  • Rsense (R12): 72oC
  • TR1: 53oC
  • L1: 51oC
  • Cout (3x180uF): 51oC
  • Cin (1x2200uf): 44oC
  • Heatsink: 56oC
When pre-regulator is in buck mode temp. figures are much better (@Iout=3A, buck mode, Vin=48VAC, Load=4x35W||4x35W lamp, Tabient=23oC):
  • Rsense (R12): 46oC
  • TR1: 32oC
  • L1: 51oC
  • Cout (3x180uF): 31oC
  • Cin (1x2200uf): 30oC
  • Heatsink: 34oC

Offline prasimixTopic starter

  • Supporter
  • ****
  • Posts: 2022
  • Country: hr
    • EEZ
Schematic of tested power pre-regulator and measurements ...
« Reply #34 on: February 14, 2015, 03:05:23 pm »
Attached schematic reflects changes added to the initial version of the power pre-regulator presented in Sheet 1 (Reply #1). It's accompanied with my version of TI's quick calculator that is used for defining component's value.

Unfortunately I didn't succeed with deployment of the output filter (L2, L3, C17 in the first version from Reply #1) to further decrease ripple without serious oscillations when load is attached so it's for now removed from the schematic.

For testing I've used various resistive loads incl. lamps (see Reply #10, which actually works like a huge PTC resistor) and DC motor as an inductive load. Signal traces are marked with Vout, Sync (LM5118's pin 11), HO (pin 19) and LO (pin 15). I tried to checked output ripple, possible instability with various loads, gate output signal and power up/down overshoot. I hope that picture titles are self-explaining and as you can see I went with testing Iout with up to 5A (ok, still within projected 150W).
Interesting and on the first sight possibly misleading picture is one which shows power up and down with 4x35W lamp load connected. Since when lamp is cold its resistance is similar to short circuit, LM5118 gradually reach required Vout as wire filament heated.
« Last Edit: February 14, 2015, 03:10:26 pm by prasimix »
 

Offline prasimixTopic starter

  • Supporter
  • ****
  • Posts: 2022
  • Country: hr
    • EEZ
Power pre-regulator efficiency...
« Reply #35 on: February 15, 2015, 10:01:19 am »
I tried to get some idea about power dissipation and does current heatsink will be enough for extreme case (50V/3A on output). Measuring Vin and Iin as AC (rms) and Vout and Iout as DC I got the following data:



If I substract few Watts from ~16W in 50/3A that is dissipated on power inductor (TR1) and Rsense (R12), and also add a couple of Watts because I have to go up to 52V/3A plus at least 6W on Q4 from post-regulator then it's still ~20W for max. output. Due to that probably a little bit bigger heatsink will be required. A previously suggested SK85/50/SA with thermal resistance Rth od 1.35K/W should do a good job keeping temperature to the 50oC if Tambient is 23oC.

Offline prasimixTopic starter

  • Supporter
  • ****
  • Posts: 2022
  • Country: hr
    • EEZ
Doubt about installing heatsink on back panel of Alu-enclosure
« Reply #36 on: February 15, 2015, 02:37:23 pm »
I have doubts about one rather simple issue: how to install heatsink on back panel of Alu-enclosure in the most efficient way? When I say efficient mean thermally and financially. One possibility is to open a huge rectangular hole on back panel that all TO-220 devices can be mounted directly on to heatsink. Another one is to mount them directly to the back panel and mount heatsink on opposite (external) side to assist thermal conduction. Second option is mechanically simpler (and cheaper) but with possibly lower thermal performance. But how much lower? Is it worth considering such option? Back panel area will be 280x115mm or 320x115mm (1.5-2mm thick). I'm not aware of calculator which take into account such setup.

Offline Paul Price

  • Super Contributor
  • ***
  • Posts: 1419
Re: DIY programmable dual channel bench PSU 0-50V/3A
« Reply #37 on: February 15, 2015, 03:13:41 pm »
Making a large hole in the back panel may decrease the structural rigidity of the whole cabinet. Adding a large heatsink to the rear adds to the structural robustness and heatsink compound would eliminate thermal conductivity problems.
 

Offline SeanB

  • Super Contributor
  • ***
  • Posts: 16272
  • Country: za
Re: DIY programmable dual channel bench PSU 0-50V/3A
« Reply #38 on: February 15, 2015, 03:38:16 pm »
Mount on the double finned heatsink section, and cut a smallish hole and use some flexible grommeting to cover the edges, then lead the wiring through it to the inside. Using the back panel as a heatsink works, but the holes must be deburred, and the 2 mounting faces ( between device and back, and back and heatsink) must have heatsink compound applied to them, and the back must be flat and unwarped.
 

Offline prasimixTopic starter

  • Supporter
  • ****
  • Posts: 2022
  • Country: hr
    • EEZ
Re: DIY programmable dual channel bench PSU 0-50V/3A
« Reply #39 on: February 16, 2015, 08:52:11 am »
My idea is to leave TO220 device soldered on the PCB, drill i.e 4mm "thru-hole" on back panel and mounting them with M3 screw to heatsink which is mounted on external side of the back panel. In that case no extra wiring is required (other possibility is to make a M3 hole to both back panel and heatsink). Of course, thermal transfer paste will be applied on the both side of the back panel.

Offline prasimixTopic starter

  • Supporter
  • ****
  • Posts: 2022
  • Country: hr
    • EEZ
Power board update...
« Reply #40 on: February 26, 2015, 11:02:35 am »
Power board is completely assembled and main function was tested. I have some good, not so good and bad news. Let's start with good one. Using bias power supply presented on the Sheet1 works much better and power up and down issues with overshooting caused by improper power sequencing is not exists any more. While "breadboarding" I used external transformer and set of standard linear regulators (7815, 7805, LM337) that provides required +15, +5 and -15V too late during power-up and vanish too early when power goes down. Power up and down sequencing can be found in the attachment. As a result a whole "mechanic" about controlling post-regulator power mosfet bias with TLC555 and relay is not necessary anymore freeing some PCB real estate.
Not so good news is that I found a mistake in connecting bias power supply coupled inductor. Fortunately it's not so big and with a couple of track cut and rewiring it works fine with existing PCB layout. A new Sheet1 with corrected coupled inductor wiring is also attached. LM5574 works very nicely alone and selected LDO's is probably overkill for a DIY project. I'll probably in next PCB layout make a space for lower-cost alternative like LM317, LM1117 and LM337.
Finally a bad news. Currently the only "small detail" which stopping me to announce this board finished (and functional without MCU control) is that planned tracking control does works with LM5118. It even ends up with its blowout and wiping out my LM5118 stock like a storm. So, LM5118 works just perfectly when it is controlled with "standard" feedback derived from resistor divider network (R1/P1, R2 from the latest Sheet1). When JP2 is shorted on position 1-2 I have a real mess. It even cause that ZD1 used to connect Vccx to the aux bias (as a mean of lowering LM5118 dissipation) is destroyed! I hope that TI's people will came up with some ideas or completely different approach will be required (based on simple PNP or optocoupler trackers published in some LTC's application notes).

Offline prasimixTopic starter

  • Supporter
  • ****
  • Posts: 2022
  • Country: hr
    • EEZ
Pre-regulator tracking control progress...
« Reply #41 on: March 06, 2015, 09:40:45 am »
I made some progress playing with tracking control for pre-regulator. To be sure and avoid destruction of new arrived LM5118s first I tested that everything works replacing the lower resistor with a LDR (Farnell: 2293503) in resistor divider based feedback to set pre-regulator Vout (or PREG_OUT in schematic). After that step I built a tracking control coupling LDR with 3mm red LED which could be replaced in the final stage with some ready made solution (i.e. NSL-32SR3, Farnell: 3692218).



This solution works providing difference of ~2.1V between PREG_OUT and OUT+ (after post-regulator), but not completely. It failed to set a desired voltage if during power up load is connected (i.e. 100R resistor) :palm:
If load is connected after power up everything continue to work fine: pre-regulator is "locked" to post-regulator output. Currently don't have idea why is that and yours suggestions are welcomed. Please also find in attachment some measurements during power up: without load (set Vout=10V), without load (set Vout=18V) and with 100R load (set Vout=10V or any other higher then 1.2V).

Please note that "glitch" when Vout is set to 10V. It's present for Vout range of approx. 2v to 18V. Maybe that's a clue to workable solution :-//. Is it in any way connected with the fact that power mosfet driver is not connected to ground but to -2V currently derived from -15V using 13V Zener diode?

Offline Neganur

  • Supporter
  • ****
  • Posts: 1138
  • Country: fi
Re: Power pre-regulator efficiency...
« Reply #42 on: March 06, 2015, 11:15:55 am »
I tried to get some idea about power dissipation and does current heatsink will be enough for extreme case (50V/3A on output). Measuring Vin and Iin as AC (rms) and Vout and Iout as DC I got the following data:



This graph reminds me of cable losses. Were your cables thick enough to not drop any voltage at 3 A ?
 

Offline dom0

  • Super Contributor
  • ***
  • Posts: 1483
  • Country: 00
Re: DIY programmable dual channel bench PSU 0-50V/3A
« Reply #43 on: March 06, 2015, 11:52:09 am »
The highest current is to the left, the lowest to the right. So the efficiency goes down as load decreases ; not surprising but typical.
,
 

Offline prasimixTopic starter

  • Supporter
  • ****
  • Posts: 2022
  • Country: hr
    • EEZ
Re: DIY programmable dual channel bench PSU 0-50V/3A
« Reply #44 on: March 06, 2015, 12:08:07 pm »
The highest current is to the left, the lowest to the right. So the efficiency goes down as load decreases ; not surprising but typical.

Exactly.

Offline prasimixTopic starter

  • Supporter
  • ****
  • Posts: 2022
  • Country: hr
    • EEZ
Re: Pre-regulator tracking control progress...
« Reply #45 on: March 06, 2015, 12:18:11 pm »
Here is the latest circuit that appears to works even with load connected :). Still, all yours suggestions are welcomed regarding this solution or any other.



Added capacitor C1 is charging during power up via D1. D2 is added to discharge it quickly enough after power down that it is ready for next power up that can happen shortly after. Attached is also the measurement of two consecutive power ups.
« Last Edit: March 06, 2015, 12:22:17 pm by prasimix »
 

Offline prasimixTopic starter

  • Supporter
  • ****
  • Posts: 2022
  • Country: hr
    • EEZ
Re: Pre-regulator tracking control progress...
« Reply #46 on: March 06, 2015, 01:58:07 pm »
Phew almost there |O. First R8 (36K) should be 47K to reach full scale (Vout up to 50V). R18 (2K2) also need to be higher (4K7) otherwise it will be very hot with higher Vout. The problem with this "capacitor trap" is that when Vout is once set below 1.2V (LM5118 reference voltage) then tracker stop functioning locked on 1.2V regardless of Vout that is set after above that value. If I connect C1 to -2V (that is also used by power mosfet driver) that issue vanish, but if Vout is set to 0V on then next power up it will jump to 1.2V! :-BROKE

Offline mij59

  • Frequent Contributor
  • **
  • Posts: 693
  • Country: nl
Re: DIY programmable dual channel bench PSU 0-50V/3A
« Reply #47 on: March 06, 2015, 05:01:01 pm »
Had the similar problem with my power supply.
Solved it by using  the output set voltage to control the pre-regulator when the power supply output is switched off.
 

Offline teddy529

  • Contributor
  • Posts: 10
Re: DIY programmable dual channel bench PSU 0-50V/3A
« Reply #48 on: March 06, 2015, 05:14:11 pm »
Mark+ this is awesome!
 

Offline Liv

  • Regular Contributor
  • *
  • Posts: 134
  • Country: by
Re: DIY programmable dual channel bench PSU 0-50V/3A
« Reply #49 on: March 07, 2015, 09:17:32 am »
My idea is to leave TO220 device soldered on the PCB, drill i.e 4mm "thru-hole" on back panel and mounting them with M3 screw to heatsink

I'm using thick (6 mm) aluminium plate to mount transistors. The plate mounting to external heatsink through rear panel window. It's easy to disassemble - no need to unscrew every TO-220 case. And I can briefly turn on device even without external heatsink in debug state - plate is sufficient heatsink.
PSU PSL-3604 Pulse gen. PG-872 Freq. cnt. FC-510
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf