Dx Blame MKIII-Hx - Builder's thread

[AndrewT]

Byron, post96:
don't power up just yet.
Go back and review the resistors across the PSU supply rails.

2k0 dissipates ~ 60V^2 / 2k = 1W8.
The pair of 2k bleeders will dissipate an extra 3W6 inside the case.
The 2k bleeders will draw an extra 30mAdc from the PSU. This will increase the ripple voltage on the PSU and gives no performance gain to the amplifier.

The 500r for the LED dissipates ~[2*60 - 2]^2 / 500 = 28W. Fortunately the LED blew up long before the resistor set fire to the amplifier.

[destroyer X]

Take a good look about his points...i have not tested because extremely busy i am... i am doing exercises all day long..outside home...i am gaining weight even after gastric bypass and i am in panic... i am in a struggle to loose weight and using bike almost all day long.

Please, about any doubt, post direct questions and as soon as possible i will answer you.

regards,

Carlos

[a.wayne]

Originally Posted by destroyer X

Take a good look about his points...i have not tested because extremely busy i am... i am doing exercises all day long..outside home...i am gaining weight even after gastric bypass and i am in panic... i am in a struggle to loose weight and using bike almost all day long.

Please, about any doubt, post direct questions and as soon as possible i will answer you.

regards,

Carlos

Focus on your food groups , it's not just about the exercise, you have to change your diet and amt..

[ByronInLawrence]

Originally Posted by AndrewT

Byron, post96:
don't power up just yet.
Go back and review the resistors across the PSU supply rails.

Hi Andrew,

Thank you so much for checking my work for me. So, you actually built it? I just now put in the LED and its 500ohm resistor and the 500ohm resistor started slowly smoking. So you think 1Kohm and 500ohm are too low values? Another poster thought 2K ohm was too much for a bleeder. I have a nice article by Elliott Sounds lab about linear PS design which I've been too lazy to read carefully.
Elliott Sound Products - Linear Power Supply Design I'll have to read it again and o the math this time. I will say that the LED stayed on for a looooooooong time after switching the power off.

I do have a question: I measured the VDC from positive-ground at +58V, and negative-ground at -58V. So I figure the voltage across positive to negative, skipping ground, should be 116 VDC, but it's almost nothing, ~0.01V, both with and without a 2Kohm, 3W resistor across them (I've heard that some voltmeters will not measure voltage accurately unless a little bit of current is running). If this measurement is correct, how does 0.01V cause a 500ohm resistor to smoke? Is my voltmeter working correctly?

[AndrewT]

was your voltage meter set for measuring AC or DC voltage?

Now back to that slightly smoking resistor !

the LED has a Vf~2V.
The 116Vdc across the supply is split between the resistor and the LED. The resistor sees 116-2 ~114Vdc.
The resistor dissipation is 114^2 / 500 or approximately 26W, only "slightly smoking"?
The current passing through the LED is ~=114/500 = 0.228A or about 228mA. What is the maximum current for the LED?

[Junie]

Current limiting Resistor calculator for leds

LED current:
20mA will work for most regular LED's.
Superbright LED's can go from 30mA to several Amps.

In Byron's PSU he need a 6Kohm 2W resistor in series to LED?

[sandbasser]

I guess I missed something - what LED are you guys talking about??? I don't see an LED on the latest (downloaded 9/20) schematic or bom.

[ByronInLawrence]

While populating one of my boards I made notes of mistakes I made or things I would have done differently. Here they are:

• C08, C11 and C15 are too close together for the caps I chose to fit nicely, but they're reeeeeeely close. My solution was to insert the middle one last, and move it to the left just a bit, which requires raising it off the board a few mm and bending the leads.
• For R10 and R19 I chose 1/2W. I don't remember why Maybe I had a good reason but regardless, they are too long and the leads too thick.
• R22 should be 4.7ohm, not 2.7ohm.
• The leads for C09 are further apart than the holes in the board.
• I don't like the terminal blocks I chose. The wire slips out with a gentle tug.
• The ceramic disk capacitors I chose for most on the non-polar caps have their leads a little too far apart. I like the block-shaped caps Junie showed in his photos.

These are in addition to the suggestions made by canonnica in post #122


Junie: Thanks for the link to the cool LED resistor calculator. It won't work for rail voltages >32V, but even if my rails were only 32V I would need 3X the resistance I currently have! Knowing this I'll estimate ~2500ohms.


Sandbasser: I posted a schematic of my power supply, and people were kind enough to comment on it. There is no LED on the board.

[AndrewT]

Originally Posted by ByronInLawrence

LED resistor calculator. It won't work for rail voltages >32V,

Vsupply = Vled (Vf) + Vresistor.

If you want 12mA through the LED then you turn the equation around to get:
Vresistor =Vsupply -Vled = 116 -2 = 114Vdc
Vresistor = Rresistor * Iresistor (=Iled) = 114V
rearrange again:
Rresistor = 114 / Iresistor = 114 / 0.012 = 9500. Use 10k

If you require more LED current then substitute the alternative value for 0.012 in that last equation.

If you insert a high value Zener in series with the LED, then you can manipulate the brightness of the LED as supply voltage varies.

[Junie]

I will customized a heatsink..by saturday I will put it all up together.