I know the nytech well, I know the Ion well, but this one, I can not understand his problem and I'm tired of the disassemble / reassemble to repair it.
one channel works well, the other explodes.
I repair the exploded channel, it is the other that fails, and so on 9 times.
no preliminary symptoms, bias / temperature / stable tension,
then he explodes.
all the transistors have been changed several times, only Philips and ST semi originals, all the diodes changed, all the capacitors, almost all the resistances.
several attempts with transistors of the same family but slower, nothing works, it explodes.
I could see the last time for a split second, the consumption mounted on my meter, I did not have time to note the final value but when I looked it exceeded 1A.
the only thing I did not try is another power transformer, but then I'm fed up, I gave up.
maybe one day in the future, someone will have the same problem and find the solution and maybe he'll post it here.
one channel works well, the other explodes.
I repair the exploded channel, it is the other that fails, and so on 9 times.
no preliminary symptoms, bias / temperature / stable tension,
then he explodes.
all the transistors have been changed several times, only Philips and ST semi originals, all the diodes changed, all the capacitors, almost all the resistances.
several attempts with transistors of the same family but slower, nothing works, it explodes.
I could see the last time for a split second, the consumption mounted on my meter, I did not have time to note the final value but when I looked it exceeded 1A.
the only thing I did not try is another power transformer, but then I'm fed up, I gave up.
maybe one day in the future, someone will have the same problem and find the solution and maybe he'll post it here.
Rapidly rising current is often a sign of instability or oscillation. The duration may be too short even to see it on an oscilloscope but an old CRT type might show it before the image dissappears and the amplifier fails again. I'm not familiar with Nytech as you and other European folk are but I know that member Lohk is, since I believe he was a service agent for Nytech in Austria. Perhaps he would respond to a PM if you quote the model number.
Otherwise, there is a long thread at PFM that among other topics, discusses the potential for instability of the Ion models due to semis having excess gain. You could check that thread for clues: Nytech Audio / Ion Systems Appreciation Thread | Page 24 | pink fish media
Otherwise, there is a long thread at PFM that among other topics, discusses the potential for instability of the Ion models due to semis having excess gain. You could check that thread for clues: Nytech Audio / Ion Systems Appreciation Thread | Page 24 | pink fish media
Current limiting resistors on the rails to the offending amp are needed, they you can debug this without blowing everything up...
Frequency instability is for me the most likely cause.
I thought for a moment to put it back in its original configuration
(see the diagram of the nytech ca 202) and I'm sure it would work today, but I can not do that, it would not be honest
for the customer who bought an Ion and not a Nytech.
I have some of the original transistors of the amp, I can always see what average gain they are and compare with the ones I put in place.
for output transistors, I have only tried ST micro darlington (BD93 / 94 BD63C / 64C BDX53C / 54C).
I'll see with the client if he agrees that I do a pose with a try again because if I start today, I think the amp will go through the window.
thanks for the tip Ian, I did not know this thread on PFM
thank you Mark too
I thought for a moment to put it back in its original configuration
(see the diagram of the nytech ca 202) and I'm sure it would work today, but I can not do that, it would not be honest
for the customer who bought an Ion and not a Nytech.
I have some of the original transistors of the amp, I can always see what average gain they are and compare with the ones I put in place.
for output transistors, I have only tried ST micro darlington (BD93 / 94 BD63C / 64C BDX53C / 54C).
I'll see with the client if he agrees that I do a pose with a try again because if I start today, I think the amp will go through the window.
thanks for the tip Ian, I did not know this thread on PFM
thank you Mark too
Lets try some logical thinking. What could cause this behaviour? If the load on the power supply is one channel, all is well. It doesn't matter which channel is OK, as long it is only one, it is OK. Is that correct?
If the load on the power supply is two channels, one explode.
Question: if you repair one channel, is it the other that explodes next or is it sometimes one, sometimes the other?
Does this have a regulated (feedback) power supply?
The above suggests one possible issue: incorrect connection between the signal- and power grounds for the two channels that causes oscillations when two channels are present. Do you have a layout of the grounding scheme, or did you make changes to the grounding scheme?
Jan
I spent a lot of time mending a couple of these a few years ago. Nice amps but they live life on the edge. While the root cause of your problems may be oscillation/instability I found problems with the poor temperature tracking of the Vbe multipier TO92 transistor that is clamped to the heatsink. I ended up replacing it with a slower TO220 (TIP41C from memory) and it then kept the rather small heatsink under control better. I went through a couple of pairs of the output darlingtons before I got the temperature regulated I recall.
Other things to check: The Zobel capacitor and the DC blocking feedback capacitor are stillgood. In my limited experience the electrolytic caps used in these amps can go very high ESR when old and this can dramatically reduce the phase margin and gain of the amp.
John
Other things to check: The Zobel capacitor and the DC blocking feedback capacitor are stillgood. In my limited experience the electrolytic caps used in these amps can go very high ESR when old and this can dramatically reduce the phase margin and gain of the amp.
John
@ Jan
there is no rule, either an exploding channel or the other.
the ground plane has not been modified.
what annoys me is that I lived with a PA amp 3 or 4 years ago.
it worked perfectly as long as I only ran one channel or used two different sources (1 channel on a CD player, the other channel on another CD player) but if I tried to make it work the two channels on the same source, it caught fire and destroyed the AOP cd player.
he had two EI power transformers and one of them had a glavanic leak.
@ John
on the ION, it is already a TIP 41c of origin, it is besides the only one that I could pretend like guilty on a channel, it had an amazing gain for a tip 41.
everything else is consistent with the origin
BC546 / BC556
BC547
TIP41C
BDT63 / 64c
all new electrolytic capacitor
all ceramic capacitors
all the diodes
most of the resitances
there remains only the power transformer to incriminate
and the two bias potentiometers that I controlled but not changed.
when at BIAS, I am very careful, I know nytech well, I go smooth mA by mA 1/4 of an hour per quarter of an hour.
there is no rule, either an exploding channel or the other.
the ground plane has not been modified.
what annoys me is that I lived with a PA amp 3 or 4 years ago.
it worked perfectly as long as I only ran one channel or used two different sources (1 channel on a CD player, the other channel on another CD player) but if I tried to make it work the two channels on the same source, it caught fire and destroyed the AOP cd player.
he had two EI power transformers and one of them had a glavanic leak.
@ John
on the ION, it is already a TIP 41c of origin, it is besides the only one that I could pretend like guilty on a channel, it had an amazing gain for a tip 41.
everything else is consistent with the origin
BC546 / BC556
BC547
TIP41C
BDT63 / 64c
all new electrolytic capacitor
all ceramic capacitors
all the diodes
most of the resitances
there remains only the power transformer to incriminate
and the two bias potentiometers that I controlled but not changed.
when at BIAS, I am very careful, I know nytech well, I go smooth mA by mA 1/4 of an hour per quarter of an hour.
I had something similar in late 90 - an amp that worked without a hitch at my place; but when returned to owner it blew channels randomly without any "warning signs"..
Short of consulting a ghost-hunter, the only thing I'd suggets is to wire up a 220V lamp (incadescent) current limiter - of enough power to get close to amps' nominal working voltages - and hook your amp to your measurements rig.
Then wait for the "ghosts" to show up...
Short of consulting a ghost-hunter, the only thing I'd suggets is to wire up a 220V lamp (incadescent) current limiter - of enough power to get close to amps' nominal working voltages - and hook your amp to your measurements rig.
Then wait for the "ghosts" to show up...
it is true that I did not think about the crucifix, even less about garlic and coarse salt but now that you talk about it, it happens that there are strange noises in my workshop!
the customer is ok to return to the original state of the circuit, namely, the Nytech CA202.
Ion
https://pinkfishmedia.net/forum/pro...HjQ.jpg&hash=9a197bd739757ee82f92b4f616d14b38
Nytech
Ion
https://pinkfishmedia.net/forum/pro...HjQ.jpg&hash=9a197bd739757ee82f92b4f616d14b38
Nytech
Last edited:
Useful info
I'm assuming that the subject of this thread is an Ion Obelisk 1 rather than a "2" or "3"
Fixed and serviced a fair few of these over the last few years so here are tips that may be of use.
First and formost - The output caps should be 35V or higher! 25V is too low and it could be the output cap breaking down which is causing your fault as a shorting output cap will destroy the upper output transistor quickly followed by the lower and pretty much everything else on that channel.
Output transistors.
I've found the modern BDT64/5 transistors are rubbish! we use BDX93/4 instead and change both channels as they do sound slightly different to the originals.
The metal clamps that hold the transistors too the heatsinks are useless and will bend if tightened up just a little too much with the result of one transistor being loose and overheating. If you have the "bent metal" heatsink then throw away the metal clamps and drill holes for each transistor and mount them individually with their own heatsink washers and thermal pads.
Check that the lower (PNP) transistors collector is correctly insulated from the heatsink. If it isnt and the collector is connected to chassis then the amp can go unstable! Its collector sits at 0v anyway so you will have to isolate the heatsink from the chassis temporarily to check.
Bias
Replace the bias pots! - The plastic on the originals degrades and starts to produce a white powder which degrades the carbon track.
Setting the bias on these amps is a real pig and requires a good quality 5 digit meter with a 2v range. They have terrible thermal regulation (mainly due to the use of darlingtons) but the early ones had totally inadequate heat sinking ( a thin bent plate of aluminium).
If the amp is cold then set the bias to about 6mV measured across the emitter resistor (18mA) as this will rise to about 25ma (8.5-9mV across the emitter resistor).
Beware!! - The value of 10.3mV shown in the "Ion service manual" PDF is a maximum value with the amp warm.! If you set the cold bias to this then the amp will almost certainly thermally run away and self distruct.
A current limiting bench power supply is pretty much a "must have" when trying to fix these amps. The lack of any fusing in the low voltage circuits means any faults will cause destructive currents to flow and theres no easy way of fitting current limiting "test" resistors without cutting tracks.
Hopefully something above will help you to get the Obelisk singing again.
I'm assuming that the subject of this thread is an Ion Obelisk 1 rather than a "2" or "3"
Fixed and serviced a fair few of these over the last few years so here are tips that may be of use.
First and formost - The output caps should be 35V or higher! 25V is too low and it could be the output cap breaking down which is causing your fault as a shorting output cap will destroy the upper output transistor quickly followed by the lower and pretty much everything else on that channel.
Output transistors.
I've found the modern BDT64/5 transistors are rubbish! we use BDX93/4 instead and change both channels as they do sound slightly different to the originals.
The metal clamps that hold the transistors too the heatsinks are useless and will bend if tightened up just a little too much with the result of one transistor being loose and overheating. If you have the "bent metal" heatsink then throw away the metal clamps and drill holes for each transistor and mount them individually with their own heatsink washers and thermal pads.
Check that the lower (PNP) transistors collector is correctly insulated from the heatsink. If it isnt and the collector is connected to chassis then the amp can go unstable! Its collector sits at 0v anyway so you will have to isolate the heatsink from the chassis temporarily to check.
Bias
Replace the bias pots! - The plastic on the originals degrades and starts to produce a white powder which degrades the carbon track.
Setting the bias on these amps is a real pig and requires a good quality 5 digit meter with a 2v range. They have terrible thermal regulation (mainly due to the use of darlingtons) but the early ones had totally inadequate heat sinking ( a thin bent plate of aluminium).
If the amp is cold then set the bias to about 6mV measured across the emitter resistor (18mA) as this will rise to about 25ma (8.5-9mV across the emitter resistor).
Beware!! - The value of 10.3mV shown in the "Ion service manual" PDF is a maximum value with the amp warm.! If you set the cold bias to this then the amp will almost certainly thermally run away and self distruct.
A current limiting bench power supply is pretty much a "must have" when trying to fix these amps. The lack of any fusing in the low voltage circuits means any faults will cause destructive currents to flow and theres no easy way of fitting current limiting "test" resistors without cutting tracks.
Hopefully something above will help you to get the Obelisk singing again.
thank you very much nytecholdie .
I am busy with other amps for now and must recognize that the Ion I annoyed so much that I take a little distance with him before working on it.
It's a 3X + X-pack
I already changed the sink for a more efficient
I also made another fixture for power transitors.
I also tried BDX (53c / 54c 73c / 74c) 2n6531 with the same result.
I think it's the previous scene that poses a problem.
I did not know about the bias value of the original schema, but I know the nytech well and I regulate the quiescent current very slowly with a True RMS device.
do you know the average earnings of pairs of previous stage?
I am busy with other amps for now and must recognize that the Ion I annoyed so much that I take a little distance with him before working on it.
It's a 3X + X-pack
I already changed the sink for a more efficient
I also made another fixture for power transitors.
I also tried BDX (53c / 54c 73c / 74c) 2n6531 with the same result.
I think it's the previous scene that poses a problem.
I did not know about the bias value of the original schema, but I know the nytech well and I regulate the quiescent current very slowly with a True RMS device.
do you know the average earnings of pairs of previous stage?
Its a lot easier to set the bias on the Ob 2,3 and 3x thant the one because the PA module is a seperate PCB. The two red wires nearest to the heatsink are connected directly to the output transistors. Cut one wire and insert a meter on 300mA and adjust for 20mA (cold) or 30mA (warm) then repeat with the other wire. When you adjust the bias be careful that you are adjusting the right one!
I've never had a case of a preamp problem causing PA failure on the Ions - The preamp is very simple and has very low gain so it doesnt usually suffer from instability issues.. Its usually within the PA module itself that any fault will be found.
Not sure what you mean by your last question "do you know the average earnings of pairs of previous stage?" - I suspect the meaning is lost in translation 🙂
I've never had a case of a preamp problem causing PA failure on the Ions - The preamp is very simple and has very low gain so it doesnt usually suffer from instability issues.. Its usually within the PA module itself that any fault will be found.
Not sure what you mean by your last question "do you know the average earnings of pairs of previous stage?" - I suspect the meaning is lost in translation 🙂
@Nytecholdie: Bias setting is clear and confirms what I suspected, thanks for sharing! I just don't understand exactly how you replace the darlington BDT64/65 with a low hfe BDX93/94. Do you substitute one BDT by two BDX in 'discrete' darlington configuration? Are any other circuit adjustments necessary for this mod?
Hi all
It was a typo! - Should have been BDW93/94
The BDX53/54 should work OK - We use the BDW's because they have slightly higher ratings.
Apolpogies for the confusion!
It was a typo! - Should have been BDW93/94
The BDX53/54 should work OK - We use the BDW's because they have slightly higher ratings.
Apolpogies for the confusion!