So let me tell you one of my experience about a problem like yours.
A year ago i had a Klein Hummel SB280 to be fixed, its problem was there was rail voltage at the output (i dont recall positif or negative rail supply but doesnt matter) and worse still the owner of the amp burned one of his Usher speaker
Anyway i checked the all transistors ,resistors and capacitors they might be the problem and i find the two input transistors corrupt, please check the TR1 and TR2 transistors if you didnt.
I hope it will be help and you fix the problem,
good luck.
A year ago i had a Klein Hummel SB280 to be fixed, its problem was there was rail voltage at the output (i dont recall positif or negative rail supply but doesnt matter) and worse still the owner of the amp burned one of his Usher speaker
Anyway i checked the all transistors ,resistors and capacitors they might be the problem and i find the two input transistors corrupt, please check the TR1 and TR2 transistors if you didnt.
I hope it will be help and you fix the problem,
good luck.
Yes, I realise it is confusing but when the simple problem is found, you will wonder why it took so long to find.Removed TR4, TR5, TR7 and TR8 and tested unconnected: they are fine.
It's good that these parts are OK but was there some reason not to test TR6 at the same time? Anyway, before refitting TR7,8 we can test the input stage.
First, check that the output terminals are insulated from the chassis properly - there may have been a shorting problem there in the past. Then check the input capacitor (10uF) is OK by measuring any small DC across it, which would be OK. A short or blown Tantalum capacitor is not easy to measure in circuit, so a substitute electrolytic type may be used for checking its operation. These caps are a known issue with audio amplifiers generally.
Then, with power on, measure voltages at the pins of TR3 with respect to the -ve rail. i.e. connect the negative terminal to the -33V rail and read voltage at emitter and base. The emitter resistor could be 560R here - please check.
With 30V on the feedback loop, nothing will measure normally to compare with the good channel but we can still check all possible problems with the semis themselves and eliminate the major suspects. Then test voltages at TR1,2 for Vbe etc.
BTW, have you tested the output transistors for shorts to the heatsink when you replaced them?
Last edited:
Re Nait:
The Nait power amp section = NAP90 and electrically equals this NAP110 variation. All early Naim amps share the same schematic already posted and explained in few Naim threads. You can track down preamp design issues in the Nait (thats probably more the matter for noise) from this Naim legacy site:
Modifying Naim Audio power amplifiers
Then follow the link at the bottom of the page for Nait details and links to Pinkfishmedia forum.
However, the power transistors, drivers, rail voltages and few resistors are scaled according to bias currents required for each model. The Nait for example, uses a pair of little TO220 output transistors - allegedly BD743 type which is obsolete but similar to current type BD911. The drivers are just another pair of ZTX652/752. A very low cost parts model!
If you were hoping for full individual schematics for each model, I doubt you will find any, except those reverse-engineered from the real thing or those of for more recent models. Because of service agreements, I don't see those being shared free on the net any time soon.
The Nait power amp section = NAP90 and electrically equals this NAP110 variation. All early Naim amps share the same schematic already posted and explained in few Naim threads. You can track down preamp design issues in the Nait (thats probably more the matter for noise) from this Naim legacy site:
Modifying Naim Audio power amplifiers
Then follow the link at the bottom of the page for Nait details and links to Pinkfishmedia forum.
However, the power transistors, drivers, rail voltages and few resistors are scaled according to bias currents required for each model. The Nait for example, uses a pair of little TO220 output transistors - allegedly BD743 type which is obsolete but similar to current type BD911. The drivers are just another pair of ZTX652/752. A very low cost parts model!
If you were hoping for full individual schematics for each model, I doubt you will find any, except those reverse-engineered from the real thing or those of for more recent models. Because of service agreements, I don't see those being shared free on the net any time soon.
Thank you huseying, they act a bit unpredictably...
Yes, I realise it is confusing but when the simple problem is found, you will wonder why it took so long to find.
Are you saying that the fault is just in front of my eyes? Of course it is but I don't know where I have to look at
Input cap is ok. Output terminals are insulated and so are power transistors cases. TR3 emitter resistor is 560R.
I did some other measurements and I found a suspect thing.
TR1 had CV 1,3v and so TR2. I repeated the measurement and they had 31v like the other channel and in that moment I had a few mV offset. Then they returned to 1,3v and so did -31v offset.
It doesn't seems like a DMM problem. Can a failed transistor act like this or do this behaviour depends on other parts?
Did you read this thread?
http://www.diyaudio.com/forums/solid-state/69199-naim-nait-2-a.html
Last edited:
The best way to test transistor is to remove it from pcb and mesaure, one of the hunderd times a transistor show good healt out of the board but in the board.
I suggest you to remove tr1 and tr2 and measure , if you are sure they are in good health so i have nothing to suggest you by now, if something crosses my mind ofcourse i will write here.
good luck
Removed TR1, TR2 and TR3 from board and they are fine.
Replaced all 10uf tantalum, they were fine.
I've also replaced R22 resistors then I was thinking about bias: why the failed channel has 3,9 - just half of the working channel (which have 8)? While wondering this, the signal cable which was lying around has touched the board, probably the exposed ground wire has touched the input pin. I saw a spark and now bias is 50.
I'm crying at the moment.
I bet I have to start over.
Replaced all 10uf tantalum, they were fine.
I've also replaced R22 resistors then I was thinking about bias: why the failed channel has 3,9 - just half of the working channel (which have 8)? While wondering this, the signal cable which was lying around has touched the board, probably the exposed ground wire has touched the input pin. I saw a spark and now bias is 50.
I'm crying at the moment.
I bet I have to start over.
That is a a great shame and a royal PITA but the input pin should be at OV potential if a capacitor is fitted and there should not be any problem if you short that to ground. A spark means a power rail or at the transistor leads themselves, more likely.
Re bias: As the output stage in the bad channel now has a 30V offset, the 0R22resistor voltage drops and bias will not be correct or even equal. Measure voltage from either rail to output node and prove it for yourself. The amplifier is not functional in this state so you can't expect to make sensible adjustments like bias current, perhaps more so if a bulb limiter is being used as it should be now, to protect the remaining components.
Now is a good time to check that semis are correctly fitted. The most common error I see on the forum when people attempt repairs on their gear is that they confuse pinouts. Thats easily done with those E-line ZTX parts. Pinouts can also be different even for the same part from different manufacturers but sometimes, DIYs insert the part around the wrong way but just cannot see the error until it is checked several times or some days later.
As you seem to think the semis were all good, yet the offset problem indicates a major semi fault, this is the only sensible option along with open solder connections, broken traces and connectors. If the voltage supplies are what they should be, the options will be limited to these possibilities. (of course, that assumes no other damage to resistors etc. has recently occurred)
It may also be that your meter problems prevent you from correctly assessing transistors as good or bad. Find an alternative such as a meter with transistor Hfe test or buy a cheap basic tester as previously suggested. Unless you are certain that tests you are doing are valid for good/bad assessments, there is no point doing them.
Re bias: As the output stage in the bad channel now has a 30V offset, the 0R22resistor voltage drops and bias will not be correct or even equal. Measure voltage from either rail to output node and prove it for yourself. The amplifier is not functional in this state so you can't expect to make sensible adjustments like bias current, perhaps more so if a bulb limiter is being used as it should be now, to protect the remaining components.
Now is a good time to check that semis are correctly fitted. The most common error I see on the forum when people attempt repairs on their gear is that they confuse pinouts. Thats easily done with those E-line ZTX parts. Pinouts can also be different even for the same part from different manufacturers but sometimes, DIYs insert the part around the wrong way but just cannot see the error until it is checked several times or some days later.
As you seem to think the semis were all good, yet the offset problem indicates a major semi fault, this is the only sensible option along with open solder connections, broken traces and connectors. If the voltage supplies are what they should be, the options will be limited to these possibilities. (of course, that assumes no other damage to resistors etc. has recently occurred)
It may also be that your meter problems prevent you from correctly assessing transistors as good or bad. Find an alternative such as a meter with transistor Hfe test or buy a cheap basic tester as previously suggested. Unless you are certain that tests you are doing are valid for good/bad assessments, there is no point doing them.
If you like i give you another solition
There is a handful transistor in the board so it will be easy , please replace all transistor you can use the equivalent of them and so did all 1n4148 diodes i assume you replace all tantallium capacitors, i see that is the easy and efortles metod to fix it
good luck
There is a handful transistor in the board so it will be easy , please replace all transistor you can use the equivalent of them and so did all 1n4148 diodes i assume you replace all tantallium capacitors, i see that is the easy and efortles metod to fix it
good luck
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.