My NPM plays ok but suddenly sends dc on output and blows speakers. But meassuring it afterwords shows everything is ok and I can play music again till next time it sends dc on output. Any suggestions to where to start?
Is this essentially a solid state amp with valves added to the audio chain upstream of a hefty solid state output stage? If yes then I can think of a few generic things to check but beyond that...
We really need circuit details for anything like this.
We really need circuit details for anything like this.
First, do not connect speakers to the unit until the problem is fixed.
Second, does this happen on one channel, or both?
If in both channels, it is likely power supply related.
Can we assume you have no test equipment other than a DVM?
Connect a high power 8 or 10 ohm load to both channels, and a DVM to monitor the load voltage.
Short both inputs to ground with shorted RCA plugs.
With a long plastic or wood stick, poke around gently on the board, fuses, and connectors, trying to trigger the fault.
Especially test the fuses and holders for solid connections.
Keep one hand in your pocket while doing this.
A design flaw in the old Counterpoint hybrid amps (one of them anyway) was this:
The coupling caps from the tube/valve section to the output section would be attached to the plate of the last tube in the signal chain. Typically, these would be a WIMA 1.0uF cap rated at 250V. This was fine, while the amp was running, but B+ was often far over 250V, into the 300+V range. The amps use solid state rectifiers throughout. So, every time you start the amp up, the B+ voltage punches a hole in the dielectric of the coupling cap, until the cap begins to fail and leak voltage/current. This often goes right into the gate/base of the output transistor, and causes a huge DC offset, or in the case of the older amps, an exploding (literally) MOSFET.
You see, when you first turn the amp on, the full B+ voltage hits the coupling cap until the tube begins to conduct, then it drops to less than the rated voltage of the cap.
[EDIT] I wouldn't do anything else with the amp until those coupling caps are replaced with something rated at a much higher voltage than B+
The coupling caps from the tube/valve section to the output section would be attached to the plate of the last tube in the signal chain. Typically, these would be a WIMA 1.0uF cap rated at 250V. This was fine, while the amp was running, but B+ was often far over 250V, into the 300+V range. The amps use solid state rectifiers throughout. So, every time you start the amp up, the B+ voltage punches a hole in the dielectric of the coupling cap, until the cap begins to fail and leak voltage/current. This often goes right into the gate/base of the output transistor, and causes a huge DC offset, or in the case of the older amps, an exploding (literally) MOSFET.
You see, when you first turn the amp on, the full B+ voltage hits the coupling cap until the tube begins to conduct, then it drops to less than the rated voltage of the cap.
[EDIT] I wouldn't do anything else with the amp until those coupling caps are replaced with something rated at a much higher voltage than B+
Hello Huldra,
here is the circuit diagram of a NPM - when I understand correctly Brassteacher mean the caps C11 and C12; when I look at the circuit diagram the supply voltage for the valve stage is 240V even at the moment of switching on, so the caps should be sufficient from their voltage rating, therefore I don't believe this is the reason for your problem. The voltage at point "D" is 155V in standard mode and 125V in current mirror mode after warm up according the manual.
Nevertheless it is not a big issue to change them with higher voltage types.
I would proceed with the hints from rayma; out of my remembrance the IGBT´s made sometimes issues so maybe these are the cause of the fault.
here is the circuit diagram of a NPM - when I understand correctly Brassteacher mean the caps C11 and C12; when I look at the circuit diagram the supply voltage for the valve stage is 240V even at the moment of switching on, so the caps should be sufficient from their voltage rating, therefore I don't believe this is the reason for your problem. The voltage at point "D" is 155V in standard mode and 125V in current mirror mode after warm up according the manual.
Nevertheless it is not a big issue to change them with higher voltage types.
I would proceed with the hints from rayma; out of my remembrance the IGBT´s made sometimes issues so maybe these are the cause of the fault.
Ok, the NPM has a totally different arrangement connecting the voltage amp stage to the output stage than the Counterpoint SA series. C11 and C12 should not be an issue. I knew Mr. Elliot had converted dead or dying amps of the SA series to the NPM, but I wasn't aware the change was so radical.
I see Mr. Elliot finally learned his lesson, and started using emitter resistors on the output transistors to distribute the load evenly. The SA series used MOSFETs in parallel in the same topology as above, but he refused to use source resistors. Many SA series had catastrophic failures on power up due to this. (This was one of the other design flaws alluded to).
Why would the absence of source resistors cause a failure specifically on power on?
It doesn't. No source resistors means that current sharing among the output transistors relies on matching. It is best described as an "optimistic design philosophy". The result is a tendency to thermal runaway because the quiescent current was much lower than the point where the negative temp coefficient of the gate threshold voltage changed to a positive temp coefficient, The original transistors were designed to be used a switches in welding equipment (amongst other things) rather than as variable resistors.
The failures at turn-on were a completely different mechanism. And easily resolved.