advertisement


QUDOS - the brilliant new amplifier boards from Avondale

Thanks Alan. I did start with the trimmer in the middle but with the bulb tester glowing I had turned it clockwise to check if the glow lowered; which it did.
I unfortunately damaged the original TR1/2 removing them to check with the Atlas. I have fitted sockets at TR1/2 positions to try different hfe transistor to try and zero the offset. The ones I tried yesterday have a 20% difference. I made the mistake as it turned out of trying a different DMM for setting the bias and had the correct voltages at the board but no reading on the screen plus I now have 25vdc at the output. Would the 20% differance account for this? I did check at tp3/4 and it was adjustable voltage reading. Struggling to get a 10% difference with the input transistors; the all seem to be in the 240's hfe wise and others 160/170.
 
if you have sockets in there experiment with different combinations and see if that changes the offset, even try 2 with a similar value just for comparison
If different combinations don't change the offset then its not TR1 & 2 thats at fault

Alan
 
On my NCC220 build I struggled to to get 10% difference between TR1/2. Both boards have offsets around 30mV.
 
Finding a fault like this becomes much easier if you remove the output transistors and get the board working properly again without them. There are even test points on the board to help you. With the output transistors removed you should get approximately 1.65 Vdc between TP3 and TP4 and you should be able to alter this with the trimmer. Getting this right before refitting the output transistors should prevent further damage.

The likely causes of your problem are the driver transistors, the feedback capacitor or the bias chain capacitor across TR5. I can't see an imbalance between TR1 and Tr2 causing that large an offset.
 
I found with my builds HFE +30% differential between TR1 & 2 will give a better than 10mv offset.

26738447518_45ef2f8c74_h_d.jpg


Or put another way:- TR1 320 HFE TR2 250 HFE but you have to buy a shed load of BC546's and spend the time testing each one but its worth it. Its easy to find 250-270 HFE or there abouts but +300 are harder to find....
 
Just swapped out TR1/2 for two matched ones and the offset has dropped to -1.2V. With the light bulb in situ the ma meter goes straight to overload and the p/s voltages sit at -15.4v and +12.2v. Regarding testing without the output trannies in place this was done and the voltage set to 1.65v. First fire up is what caused the initial problem with my not putting the dmm to the ma setting so putting negative voltage to the board before positive. The output and drivers were replaced with my first resetting the voltage at tp3/4 to 1.65v. I am confused. I guess I need to remove the output transistors again and retest. Am I right in thinking that the outputs are the things most likely to be drawing more tham 300ma of current. Obviously with the light bulb in situ there has been no magic smoke released.
 
With the best part of 300mA and ~13V something has to be dissipating ~4W. Even an o/p tr will get warm at 4W (hot if the U-bracket is not connected to a heatsink).

If the amp is anything like an NCC200 it will have 220R in each rail of the front end supply. If either of these has more than 3V across it then the front end is drawing too much current. Assuming it's not - measure every Vbe, like I said earlier.
 
Thanks for the advice re measuring across the 220R. Voltage drop is a high +- 8.3V so front end is drawing way to much current.
 
Thanks Malcolm and Alan I will check thing out tomorrow.
Regarding the voltage drop across the 220R resistors is this with or without the output transistor fitted.
 
I have removed the output transistors and have the following vbe readings;

tr1 1.0v tr2 0.641v tr3 0.644 tr4 2.529v tr6 0.216v tr7 0.588v tr8 0.564v tr5 0.667v.

The 220r resistors feeding the front end run extremely warm. The cap across TR5 and the feedback cap have been changed.
 
Tr1 and Tr4 are where the problem lies.
What is the voltage across TR6's emitter resistor (68R in NCC200, 'cos that's the only schematic I have)? It should be about 0.6V which sets the current in the VAS at about 10mA.

The cap across Tr5 cannot cause dc offset. If it shorted (or low impedance) it means you will not be able to set the o/p stage quiescent current and if it's open circuit it just means that the Vbe multiplier will be slightly less effective at hf.

If you are worried about the feedback cap, remove it altogether. The amp will work without it (in terms of dc conditions) but it will be very quiet and tinny if you try to amplify audio.
 
Further thought...
What's the voltage across the 620R emitter resistor of TR3? It should be about .6V. The sets the LTP current at about 1mA. Since you appear to have 2.6mA flowing in Tr1 collector resistor (2.6V/1K) it makes me think there is something wrong around Tr3 and Tr6. Please check you have ~1.3V from the negative rail to TR3 base. Assuming you do then TR4 is drawing too much base current and it likely dead.

Further further thought... it is possible to get E-lines with different pin-outs. Check you have TR4 and 6 fitted the right way!
 
Last edited:
Strategy for sorting gaffes:

If you suspect a semiconductor of any description has been inserted the wrong way.....forget testing - just replace it
ensuring this time, it goes in the right way round. Whilst all the semis are out, apply voltage and check for current
draw....it'll sort any reverse polarised/bad Elcos (read later)

While the semis are out of the PCB, test each resistor preferably asking someone the read out the values and checking
against the script. Tick off each as you test - it'll save you going round in circles.

Next, ensure all Elcos are inserted with correct polarity - it's easy to miss one and it causes havoc. (My money's on this)

If all else fails, send me the board and I'll have it sussed in a couple of shakes.
 
Thank you kindly for the offer Les; very much appreciated.

I will check things out with the transistors removed.
 
You must log in or register to reply here.


advertisement


Back
Top