On ‎6‎/‎20‎/‎2018 at 3:52 PM, kevin gilmore said:

It really depends on which resistors were changed. If they gave you a schematic, please post it.

Edit: found the schematic.

you can replace the 1k with 10k resistors. The power supply will come up a bit slower, but it will otherwise work.

 

Thanks for the help. 

I have started up the PSU and i have 35V DC on the output. + - 70V DC together. Should be 21 + 21 -?  All resistors are correct. 
RAW  35V DC is just over the rectifier? not on the output? If i understand correct? 

 

If you have 35VDC on the output it means the PSU is not regulating, so something is wrong. Check soldering, component positioning and also the values of all components.

If everything looks ok, measure the voltage across the voltage reference and zeners (U3/U4 and D1/D2 in the schematic)

45 minutes ago, UFN said:

If you have 35VDC on the output it means the PSU is not regulating, so something is wrong. Check soldering, component positioning and also the values of all components.

If everything looks ok, measure the voltage across the voltage reference and zeners (U3/U4 and D1/D2 in the schematic)

I have checked all the resistors. 

U3 and U4 have 3.5v between anode and chatode. And D1 and D2 have 12.9v across them.
 

Do you have the regulators in the correct position? 

1 hour ago, jose said:

Do you have the regulators in the correct position? 

Yes, and have changed the mpsa06 and mpsa56. Checked so every pin is correct. Don´t know what is wrong ?

Measure the voltage before each component . And 15024 and15025? are OK?

Edit: These kind of problems require patience

Edited July 7, 20187 yr by jose

2 hours ago, jose said:

Measure the voltage before each component . And 15024 and15025? are OK?

Edit: These kind of problems require patience

Thanks for mention the 15024 and 15025, i had swapped them around. Now i have 21.8v + and 21.8v - stable voltage ?

 

Amp is working great, and sounds good.

Is it normal for the 15024 and 15025 to run very hot?

Not "very hot" I would think. If you have 35VDC input and app. 22V output you are losing 13V*the current draw of the amp per transistor. IIRC the bias is around 350-400mA, so each transistor dissipates around 4-5W at idle. That should warm up the heat sinks a bit, but they shouldn't be to hot to touch. Have you got measurements (of current draw and/or heat sink temp.?)

On 7/7/2018 at 11:31 PM, UFN said:

Not "very hot" I would think. If you have 35VDC input and app. 22V output you are losing 13V*the current draw of the amp per transistor. IIRC the bias is around 350-400mA, so each transistor dissipates around 4-5W at idle. That should warm up the heat sinks a bit, but they shouldn't be to hot to touch. Have you got measurements (of current draw and/or heat sink temp.?)

Heatsink temps are around 50-60C, so it´s normal. 

Have question about the trimpots. 
Trimpots (500ohm) is turned all the way down (clicking) to get 105mv (hot) .
Can´t get any lower mv, are they just for if you to run higher mv? Cause no need for adjustment then. 
 

Thanks 

 

Wich opa445 should i use? i have opa445AP and they wont work in my psu. I have tested two pairs of them, and psu is dead with them. 
 

Thanks

I’d you need an OPA445 in a DIP-8 package, the AP is pretty much your only choice. Does it work with other opamps? Your post makes it sound like it does. The OPA445 uses a standard single opamp pinout, from what I recall. Where did you source the ones you have from? Are they genuine?


Sent from my iPhone using Tapatalk

21 minutes ago, Pars said:

I’d you need an OPA445 in a DIP-8 package, the AP is pretty much your only choice. Does it work with other opamps? Your post makes it sound like it does. The OPA445 uses a standard single opamp pinout, from what I recall. Where did you source the ones you have from? Are they genuine?


Sent from my iPhone using Tapatalk

Yes I am using OP27. I bought them from ebay, guess not the best source. I bought from a top rated seller. Logo on opamps are Texas Instruments 
Only difference between opa445 and op27 pins are

Don´t know if it matters. 


opa445
pin 8 NC
pin 5 offset trim

op27
pin 8 Vos trim
pin 5 5 NC 


And on my KSA5 the 500 ohms trimpots are turned all the way down (clicking) then i have 125mv cold. If i am gonna rise the voltage i need to be able to adjust the mv down even more...
Voltage are 21.7 v with op27.

 

 

Why do not you use OPA445 directly?

You will save time and money. Buy them from a Texas provider like Mouser.

Indeed, there are a lot of fake OPA445's out there as it is an expensive part. 

I will order them from mouser. 

But how to get lower mV´s on the transistors?

Trimpots turned all the way down and have 130mV cold. 
 

 

Thanks

 

Which capacitors 1000x35 are best used in this KSA5 amplifier? Elna RJH or Nichicon KA ? Or other ?

I personally like Panasonic FR or FM, but haven't built this amp. Something that fits properly would be a good start. Right lead spacing, etc.

There are not such 35 pieces of Panasonic, there are only Nichicon KA, Elna RJH, Hitano EXR, ESX, EWH, EHR. Samwha SD, WB, RD.

I used Pana FMs on at least one of mine. Mouser p/n 667-EEU-FM1V102 is the right size as I recall (1000uF/35V ø12x25) 

Mikhail rebuilt an original with black gates. Probably not possible these days. A few were the wrong voltage.

I used Nichi KW... I guess there is no fixed rule, we have used capacitors that we know.

Than to replace  diode bridge G3SBA20-E3 / 51 ?

Hi KSA-5 builders ?

Very interesting topic. I plan to build this amp and there are small questions - 
does anyone compare sound with 1oz/2oz/3oz copper PCBs?
There are any objective reason to use 3oz copper? 
Any information about original Krell copper weight?

Thanks!

Hi AlSp

I built KG's device as a pre. There is not much information in this thread about how this amp. reproduces sound.
IMHO there is no need to use a thicker layer of copper, bearing in mind that you will be using a maximum of 30 Watts.

 

I believe we could use better capacitors and resistors.

 

Cheers

2 Oz boards tend to be a little more tolerant if you need to replace components because the traces will take a bit more heat. Thick 3 Oz traces may actually require more heat than you usually need to get good joints and make the boards seem difficult to solder, so if you don't need the high current I wouldn't bother with that (they also tend to be much more expensive).