SymAsym - "The Sequel", AAK's Rev_1.4 PCB Builders Thread

[coolnose]

Hi Max,

Describe what you mean by slight background noise? Properly implemented Rev_1.4 should have an SNR of better than -100db with the inputs shorted. With your ears dug close to the woofer it should be dead quiet. The tweeter should have the slightest hiss with your ear right up to it. The hiss should be clean with no crackles or pops.

The bias at 40mv is too high, try 25mv.

Best regards,

Al

Thanks Al for chiming in,

with my ear close to the woofer I have a slight hiss and a faint regular pulsing sound, could be AC residue I guess.

The tweeter (94dB/W) is clearly hissing. No pops or cracks though!

I'll open it up soon and turn the bias down.

What are your thoughts about lowering the gain ?

Cheers,

Max

 

[AAK]

Hi Max,

Why mess with perfection. IMO SymAsym is one of the finest DIY amps I've ever heard. If you look at post #2 you'll see the noise floor is down -120dB for most of the audio band. My tweeter is 92dB/W, the noise becomes audible at about 6 inches. You shouldn't have any hiss or pulsing sounds coming out of the woofer, it should be dead quiet even with your ear lightly touching the dust cap. Focus on cleaning up the noise.

Best regards,

Al

 

[microx]

Hi, are there any PCB's for this project still available?

 

[coolnose]

Bought an assembled board from a forum member...

Hi all,

I know I know I did what no-one should do... I bought a set of build but "untested" boards from a forum member.

As my previous build (which was actually my first build ever) was right on spot, I did not expect any promblems in buying assembledboards...

They looked good on arrival, the soldering whas pretty ok, so I fired them up after testing the supplies...

Both boards connected to the (both) working PSU boards resulted in burning 100r resistors in place of the fuses.

Then I realized that the base resistors R40 (10R) had not been installed (propably due to the R40/R46 issue on the silkscreen) and that jumber J2 was missing opn both boards.

ALso on one board Q11 was installed backwards.

Right now I have checked all passive components and I am unsure how to proceed further...

Is taking out the OT and the drivers the way to go ?

I am wondering if I should simply rebuild the whole frontend as well...

Any help is appreciated.

Max

 

[p1p4]

Hi, I've built this amp and I have a questions - transistors Q9 and Q12 (in my version Q1 and Q2) are quite hot (I measured about 85°C), it is normal? Supply voltage 2×45VDC. I am really afraid about this higher temperature. In fact, my schematic is little bit different - from input to VAS it is the same like your version, next is similar like original symasym 5.3...
I enclose my scheme.. http://i.imgur.com/TV1Qt45.png
Thank you for answer.

 

[AndrewT]

attach your schematic.

 

[p1p4]

here is my schematic

An externally hosted image should be here but it was not working when we last tested it.

post a picture

 

[AndrewT]

Measure the Vdrop of R10. That allows you to calculate the Pdiss of Q1 & Q2, if they pass the same current.

You can also measure the Vdrop of R15 & R17, as a check for current values.

R7 looks wrong.
It should be 0r1, or maybe 0r15. 10r is far too high.

 

[p1p4]

I measured yesterday, exact same resistor you write now. Current through R10 is 9,8mA, current through R15 & R17 very similar, about 4,8mA +- few tenth miliamps.
I will probably try change the value of resistor R10 from 68R to 86R - this will help I think - current should drop to ca. 7 mA and power dissipations on Q1 & Q2 (Q5) will drop too.
Voltage drops on Q1 & Q2 (between C-E) is about 39V, so few mAps will do power dissipations in the tens to hundreds.. I think each symasym has warm transistor in VAS, hasn't it?
(sorry for my english

 

[AndrewT]

4.8mA @ 39Vce is ~190mW
They will be very warm, but I would not have expected very hot (83°C)
Tj will be even higher.

I wonder if you have oscillation making them hotter than the measured current would indicate?

 

[AndrewT]

for a 625mW device the Rth s-a will be {150°C-25°C} / 0.625W = 200C/W

with a Pdis of 190mW, the predicted deltaT is 200C/W*0.19W = 38Cdegrees
That gives a Tj = Ta+38 ~30+38, i.e. <68°C
Tc must be less than Tj
Something is wrong.

Fairchild confirm Rth j-a as 200C/W and Rth j-c = 83.3C/W
That tells you that Tc will be 83.3*0.19W=16C cooler than Tj, i.e. Tc <52°C when Ta<30°C

 

[p1p4]

I measured what you said yesterday and no oscilation in all amp, only supply voltage is not good filtered, 50Hz is visible on supply voltage rail with amplitude ca. 1V (end transistors have current 200mA each). But amplifier output is quiet and doesn't contains 50Hz. I don't think voltage rail ripple cause hot transistor in VAS..

 

[AndrewT]

As you can see from the Thermal model I presented in post151, the Tc should only be ~22Cdegrees above ambient.
There is something wrong, if you are measuring Tc~83°C

 

[p1p4]

I don't know what is wrong..
I haven't transistor from Fairchild, I have from Diotec, matched exactly in any diferential pairs and mirrors. In datasheet there is a mistake in transistors pins imo, the have E and C conversely. I matched transistors by hFE in my multimeter, when I inserted transistor to multimeter preparation by Diotec datasheet, dmm show nonsenses, by the fairchild datasheet dmm show hFE about 145, this is correct values.
Any other problem I had not, all resistors and caps is correct..

 

[aparatusonitus]

Bias voltage for vas cascode transistors (Q1/Q2 in your schematic) is wrong...for a start use 10k resistor instead LED3 so the bias voltage will be at ~1/2 +Vcc.
Also, base-stopper resistors at Q1/Q2 are way to high, use something up to 68R.

 

[p1p4]

thank you a lot, I used resistor insted of LED and transistor Q1 Q2 are only little warm now, voltage drop is now around 20V!
Q5 is hot, voltage drop between C-E is 39V, the same drop like on Q1 Q2 before using 10k resistor. But this is relatively fine for me - I use small heatsink for this.
Voltage drop simply must be somewhere

 

[aparatusonitus]

For Q5 you may use TO-202 transistors with EBC pinout like an old Motorola BD518/BD520/BD528/BD530/MPS-U55/MPS-U57/MPS-U60 pnp transistors. Now, exchange 12-15V zener for 10k resistor (mind you, zener diode pcb orientation is different from LED!) and R53 shoud be 3k/2W. And again, use up to 68R for Q1/Q2 base stopers.

 

[p1p4]

Ok, I try it at the weekend I hope, I have too many other responsibilities..
I know much things from electronics, but not good in analog..

 

[aparatusonitus]

Please don't do anything...I was completely wrong about biasing, amp would not be able to swing fully towards Vcc - it will clip (what I was thinking)...schematic is good, but you will need high power dissipation devices for Q1/Q2 in TO-202 or TO-126 housing.

 

[p1p4]

I made few heatsink for hot transistor from Cu plate and fuse holder. transistors are warm, temp about 64°C, it is ok for me now..