diyAudio

diyAudio (http://www.diyaudio.com/forums/)
-   Solid State (http://www.diyaudio.com/forums/solid-state/)
-   -   SymAsym - "The Sequel", AAK's PCB Builders Thread (http://www.diyaudio.com/forums/solid-state/158984-symasym-sequel-aaks-pcb-builders-thread.html)

AAK 19th January 2010 03:44 AM

SymAsym - "The Sequel", AAK's PCB Builders Thread
 
3 Attachment(s)
Welcome! This thread is for all those who bought my PCB based on Mike Bittner's SymAsym – "The Sequel” circuit design. Of course everybody is welcomed, and if you have something to add by all means feel free to jump in.

Attached is a pdf document with step-by-step instructions to help assemble my PCB. I'll provide transistor matching and testing instructions within the next couple days.

I've also attached a couple pictures that should help clarify number 8 and 9 of the Assembly Instruction section of the attached document.

For those unfamiliar with MikeB's SymAsym - The Sequel design, here's the link to the original thread.

http://www.diyaudio.com/forums/solid...ym-sequel.html

Here's the thread that I started describing my PCB design based on MikeB's design.

http://www.diyaudio.com/forums/solid...rev_1-3-a.html

I still have a few boards left for those interested.

Regards,

Al

lgreen 19th January 2010 04:17 AM

So I see that the parts list is on Post 39 of the 'new' thread.

Post 45 of that thread says that cap C7/C8 are 15 mm not 10mm, so post 49 says to use mouser 505-MKP4-0.1/630/5.

Many of us want to order the correct set of parts so I think the first issue is to identify the latest schematic and parts list.

Using the above, is this the correct parts set to order?

Also, some of us (me) will want to put off-board diode bridges + big caps so is it just shorting out/ bypassing the on-board diodes or is there anything more complex?

sng001 19th January 2010 06:21 AM

Quote:

Originally Posted by lgreen (Post 2053126)
So I see that the parts list is on Post 39 of the 'new' thread.

Post 45 of that thread says that cap C7/C8 are 15 mm not 10mm, so post 49 says to use mouser 505-MKP4-0.1/630/5.

Al confirmed the lead spacing for C7/8 are 15mm. I have checked mouser page: Mouser Part #: 505-MKP4-0.1/630/5 is equilvalent to Wima Part #: MKP4-0.1/630/5PCM15, so that lead spacing for this part is 15mm.

The Wima page also confirm the dimension & lead spacing: WIMA

- Stanley

AAK 19th January 2010 02:43 PM

3 Attachment(s)
Hi All,

Here the latest amplifier and power supply schematics plus BOM.

Igreen, I provided information noted below the PS schematic that should answer your question about using an External PS.

For the BOM I added Mouser part number 505-M10.1/250/10 which is a MKP10which according to the data sheets is better suited for audio equipment.
Thanks Stanley for pointing it out the different types of Wima caps.

I also added 3/8" nylon standoffs to the BOM used to mount the board to the main heat sink.

Regards,

Al

lgreen 19th January 2010 06:29 PM

Thanks Al.

sng001 20th January 2010 11:09 AM

Quote:

Originally Posted by AAK (Post 2053564)
Hi All,

Here the latest amplifier and power supply schematics plus BOM.


Regards,

Al


Hi Al,

Thanks for the updated schematics. I have not used any FETs before and I would like some advice on the JFET matching.

I only have a DMM and a single-channel handheld scope and a breadboard so I can only do basic device matching.

I found the following document from passdiy: http://www.passdiy.com/pdf/matching.pdf

The document mentioned: "Matching input MOSFETs is critical, because they must share equally the 10mA of bias current from the current source."

What is the normal bias current thru JFETs: Q1 & Q2?

To better match the following BJTs: Q5&Q6, Q9&Q12 and Q10&Q11, I would like to know their respective quiescent current.

Best regards, Stanley

AndrewT 20th January 2010 12:10 PM

Quote:

Originally Posted by sng001 (Post 2054708)
I would like some advice on the JFET matching.

I only have a DMM and a single-channel handheld scope and a breadboard so I can only do basic device matching.

To better match the following BJTs: Q5&Q6, Q9&Q12 and Q10&Q11, I would like to know their respective quiescent current.

first you must match at the operating currents that occur in the amplifier.
Second, the matching should still be reasonably good for a small range of current either side of the circuit operating current.

jFETs should be selected to have very similar Idss. But this requires three resistors and a 10V to 12V DC supply and an 8pin dip socket and consistent measuring procedure.
Then take a batch of similar Idss jFETs. Select one and make it your REF.
Select a DUT from the batch and thermally couple the two jFETs and insert them into the dip socket.
connect the gates pins together. connect the source pins together. connect the drains with identical 100r resistors to the +ve power supply ~10V for an Nchannel).
Connect the sources to the -ve power supply using a 200r resistor.
Connect the adjustable voltage, a pot, to the gate. Set Vgs to zero. Turn on the 10V supply. measure the voltage across each of the drain resistors. are the Vdrop the same. Then REF & DUT are passing the same current at Vgs=0.
Now adjust the pot to put -ve voltage on the gate. adjust until the Id is the operational current. Now measure Vdrop. are they the same? probably not. label DUT with the current difference from DUT.
Do this for each DUT.
Look for a pair of DUTs that have the SAME current difference from the REF.

Remove REF from the dip socket.
Insert the selected pair into the dip socket.
apply Vgs=0 and check these two have same Idss.
Now adjust Vgs to set Id to operating current. measure the ID of each jFET.
are they within the tolerance that you have set for matching?
you can easily find Idss within 0.1%. you can usually find Id within 5% @ operating current. If you can find a pair of jFETs that match to 0.5% over a range of 5 currents from Idss down to 20% of Idss then you have a supermatched pair. This will take 200 jFETs and 200hours to find a dozen or so supermatched pairs. If you set your tolerance to 1% you get a higher yield of pairs. You can increase your tolerance to 10% and you will find that almost all of your 200 jFETs can find a partner.

You do not need expensive equipment, you do not need absolute accuracy. You are comparing DUT to REF. This takes time and care and patience.

You can speed things up slightly by adding a CCS to the 200r resistor. This CCS should have a fixed current that matches two times the in circuit device quiescent current. This will require a higher voltage supply.

A similar procedure and similar jig can be used to compare small signal BJTs.

juma 20th January 2010 12:22 PM

1 Attachment(s)
Quote:

Originally Posted by sng001 (Post 2054708)
... I have not used any FETs before and I would like some advice on the JFET matching...

http://www.borbelyaudio.com/adobe/ae599bor.pdf (fig. 5)
http://www.borbelyaudio.com/adobe/ae699bor.pdf

AndrewT 20th January 2010 12:45 PM

that measures Idss.
selecting matched Idss is of no value in a pair of devices that operate at a substantially lower bias current.

AAK 20th January 2010 01:21 PM

2 Attachment(s)
Hi Stanley,

Attached is the circuit that I used to match Jfets Q1&Q2. I also attached a picture of my setup using a bread board. It's a simple circuit and works well.

On the matching circuit I set Idss to approx. 2ma, which is close enough to the actual current on the amplifier which is approx. 1.6ma. Use a DVM to check Vd on each Jfet and pair up the ones that match the closest. With 10 Jfets I was able to match 2 pairs pretty much right on. Of course that may vary depending on the batch sent.

For Q5&Q6, Q9&Q12, and Q10&Q11 matching hFE is good enough. The bias current across each transistor is about 3.25ma.

Best regards,

Al


All times are GMT. The time now is 02:15 PM.


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