F5X -- the EUVL Approach - The Build Thread - Page 2 - diyAudio
Go Back   Home > Forums > Amplifiers > Pass Labs

Pass Labs This forum is dedicated to Pass Labs discussion.

Please consider donating to help us continue to serve you.

Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving
Reply
 
Thread Tools Search this Thread
Old 15th March 2012, 10:02 PM   #11
diyAudio Member
 
Join Date: Feb 2010
Location: Dallas Fort Worth area, Texas
Don’t install the output devices yet. In part two we will cover initial testing of the input stage and the adjustments necessary to get the bias close before installing the output devices.

Your amplifier boards should now look something like this. We'll do some wiring for testing and bias up the amplifiers in the next installment. Until then stay tuned for some other helpful details.

Dave
Attached Images
File Type: jpg IMG_0829.JPG (452.7 KB, 2098 views)
  Reply With Quote
Old 15th March 2012, 10:12 PM   #12
diyAudio Member
 
Join Date: Feb 2010
Location: Dallas Fort Worth area, Texas
Default Build_Notes - Soldering Suggestions

The PCBs for the F5X are made from high quality FR4 PCB material and feature heavy 2oz copper construction. Because of this and the inherent size of many of the components there is quite a bit of thermal mass to overcome when soldering. I like to have those nice solder filets on both sides of the boards that the pros get with a wave solder machine (even if mine never really look that good). I do prefer evenly wetted solder pads and pins on both sides of the PCB if possible and to this end I do what the machines do - I preheat. You can do this at home with a hair dryer, heat gun or oven. ALL OF THESE CAN BURN YOU AND/OR YOUR POSSESSIONS IF CARELESSLY USED - SO BE CAREFUL.

First, make sure to use an 800F/ 425C soldering iron for all soldering. Gently warm the boards up top and bottom with the hair dryer or heat gun to the point where they are hot to the touch BUT NOT SO HOT AS TO BURN YOU. You can instead place them in an oven or toaster but make sure the temperature is SET TO and DOES NOT EXCEED 70 degrees C or 160 degrees F.

Increasing the board temperature before soldering improves the wetting and flow of common solders. They boards will cool down pretty quickly so it pays to have as many components inserted as is reasonable for each heat cycle. DO NOT hold the heat gun close to the PCB, YOUR SKIN, ARMS, WINDOW TREATMENTS, etc. as it can and will burn whatever is in front of it. A hair dryer is more forgiving but can still burn you or start a fire

If all else fails you can always forgo heating and solder a handful of points then flip the PCB over and solder the same points from the top while they are still warm.

Happy soldering…
Dave
  Reply With Quote
Old 15th March 2012, 10:18 PM   #13
diyAudio Member
 
Join Date: Feb 2010
Location: Dallas Fort Worth area, Texas
Default Build Notes - Suggestions for Surface Mount Soldering

When soldering surface mount components I use a simple method that can be done without magnification, at least for my tired eyes. Place a small amount of solder on one pad. Hold the component in one hand with tweezers and the iron in the other. Reflow the solder previously applied to the pad and slide the component into the solder with the tweezers. Hold the part as you remove the soldering iron because surface tension can take the part along on the tip of the iron as it is pulled away. Now solder the other pad. If you apply too much solder use de-solder braid. Make sure to remove the braid and the iron together. If the solder in the braid cools while still attached to the surface mount component or pad it will likely all come off when you pull the braid away. This is painfully true for small pads and components.

This technique works with most passives or axial components with two connections. If you have a small active device like an IC or transistor I place a small amount of solder on one pad and set the part over this solder mound. While holding the part in place with tweezers, reflow the solder and allow the part to drop down into it. Make sure the part is aligned correctly on the other pads and remove the iron.

Keep in mind that holding an iron to a small component too long will eventually destroy it.
Attached Images
File Type: jpg IMG_0734.JPG (419.0 KB, 1997 views)
  Reply With Quote
Old 15th March 2012, 10:25 PM   #14
diyAudio Member
 
Join Date: Feb 2010
Location: Dallas Fort Worth area, Texas
Default Build_Notes - Device Matching

The F5X features a symmetrical input stage built with audio grade JFETs - the 2SJ74BL and 2SK170BL parts from Toshiba. These feature low noise, high transconductance and are advertised as complementary parts. It is necessary to match parts for both the input and the output stage. This is intended to pull relevant data from various posts in the main thread into a single reference. All links are published at the end of this group of posts.

You need two matched quads, one for each amplifier. A “quad” consists of two J74s and two K170s. The J74s are Idss matched to one other, as are the K170s, but the match between J74s and K170s need to be a bit different. Patrick (EUVL) tested the transconductance curves of some of each and found that they are not perfectly complimentary. The curve shape of the J74’s transconductance often does not match the K170’s. The first picture is an example measurement from Patrick that shows what one might find for a typical complementary pair of identical Idss devices. He also found that selecting for a specific Idss current difference and degenerating the J74s brought them better in line to a true complimentary pair.
Attached Images
File Type: png 2SJ74%20Mismatch.png (2.5 KB, 1651 views)
  Reply With Quote
Old 15th March 2012, 10:27 PM   #15
diyAudio Member
 
Join Date: Feb 2010
Location: Dallas Fort Worth area, Texas
Finding quads in the same range that he uses may be challenging unless you have a large number of parts to select from. In the table below Patrick lists calculated values for the degeneration resistors at various Idss combinations. This is a good place to start with selection based on the datasheets curves for the J74 and K70. Because every transistor is slightly different these numbers will not be exact. Select your parts to configure the front end bias at 5mA or higher (the yellow line and below in chart). A lower value here will dictate changing the drain resistor combination on the amplifier board. You can pre-select your parts with a Rdegen of zero ohms to see what you have, then change the 5 ohm resistor in the matching fixture to the calculated value from the chart. My quads ended up with the best match using 6.8/7.8mA Idss selected parts with 4R degeneration for the J74s.
Attached Images
File Type: png F5%20Frontend%20Bias.png (32.3 KB, 1706 views)
  Reply With Quote
Old 15th March 2012, 10:31 PM   #16
diyAudio Member
 
Join Date: Feb 2010
Location: Dallas Fort Worth area, Texas
Patrick has already described a three point matching method and circuit. Here is one implementation of his JFET circuit. Jumpers J1, J2 and J3 allow easy changing of the degeneration resistor values in circuit. You can optionally change the single 5 ohm resistor and jumper into multiple steps with additional jumpers across 1, 2, and 3 ohm resistors to find the best matches and Rdegen value.
Attached Images
File Type: png JFET_matching.png (81.8 KB, 1615 views)
  Reply With Quote
Old 15th March 2012, 10:35 PM   #17
diyAudio Member
 
Join Date: Feb 2010
Location: Dallas Fort Worth area, Texas
The F5X can use a number of output devices. IRF, Vishay, Fairchild, and other parts will work. If you have followed the thread for long you will note that Patrick chose Toshiba parts and a different combination of supply voltage, source resistor and bias from the original F5 circuit. Simple matching at 2A with the transistor case at 60C will do for the build as described. Use a large thermal mass and heat it in an oven as described in Patrick’s note. Make sure to use silicone or Teflon insulated wires to connect the devices in the oven to the power supply and meters outside. These insulations will not melt at these temperatures.

Doing a simple 3 point curve match will offer performance improvement over single point VGS matched parts. I purchased four “matched” sets of MOSFETs and then further selected them with a multi-point curve match. It is interesting that the best matches at the 2A “static” Vgs point were not necessarily the best curve matched sets among the parts I have.
Attached Images
File Type: png MOSFET_matching.png (69.2 KB, 697 views)
  Reply With Quote
Old 15th March 2012, 10:36 PM   #18
diyAudio Member
 
Join Date: Feb 2010
Location: Dallas Fort Worth area, Texas
If you build the amplifier with regulators it is desirable to match the power MOSFETs used there as well minimizing output voltage differences. Simple, single point matching of the MOSFETs at 4A is sufficient for the regulators. Bear in mind that it is entirely possible you will not get a perfect match between P and N devices.

Happy sorting…
Dave
  Reply With Quote
Old 16th March 2012, 02:45 PM   #20
qusp is offline qusp  Australia
diyAudio Member
 
qusp's Avatar
 
Join Date: Oct 2009
Location: Brisbane, Australia
nice work! got any more of those copper standoffs? =D
  Reply With Quote

Reply


Hide this!Advertise here!
Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off


Similar Threads
Thread Thread Starter Forum Replies Last Post
F5X -- the EUVL Approach EUVL Pass Labs 862 26th August 2014 12:56 AM
Potential EUVL F5X designed case user distribution check pchw Group Buys 1004 27th February 2013 03:32 PM
GB 2SK3497/2SJ618 for EUVL F5X or other projects pchw Group Buys 266 8th April 2012 06:05 PM
[B]PRP resistor group buy for the EUVL F5X build[/B] NicMac Group Buys 70 4th February 2012 04:20 PM
New Saburo Build thread Wha' DIYa know? Full Range 94 17th October 2010 04:49 PM


New To Site? Need Help?

All times are GMT. The time now is 09:22 AM.


vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright ©1999-2014 diyAudio

Content Relevant URLs by vBSEO 3.3.2