Mark Levinson protection circuit - Need help

Hello everyone
I have an annoying issue with the protection circuit of a Mark Levinson 23.5 that I would like to solve. I need your help.

Basically, when the amp is well warmed up and I turn it off I cannot turn it on because the protection relay of the power switch is activated and I have to wait ~30-60 seconds. It doesn't happen all the times, about 80% of the times. The problem is not present when the amp is cold. The previous owner says the problem only recently showed up, so that's not how the amp is supposed to work.

I'm attaching the schematics of the protection circuit: basically it is located on two different boards linked with two cables. The first board hosts the sensors and sends a signal through the cable to the second board, the signal should activate the relay when something is wrong in the amp (temperature, DC offset etc.).
By disconnecting the cables that link the two boards the problem is still present, so the problem is not some malfunctioning in the amp that triggers the sensor (luckily). There must be a problem in the relay circuit in the second board that activates the relay coil only when turning off/turning on, and only when the amp is warmed up. I don't know whether the relay switches on during turn off or turn on unfortunately.

The second board is very difficult to disassemble because of the MANY fast-on connectors, so I didn't take it apart yet. But visually everything looks fine and I tested in circuit with a multimeter all the resistors and diodes and they test good.

The part of the schematic we are interested in is in the right upper corner. P340 carries the +83V unregulated line directly connected to an old 36000 uF big can capacitor. P332 and P333 are the two cables that connect the sensors and associated circuit from the two channel boards to the optocouplers; disconnecting both cables doesn't eliminate the problem.

Any clue about what can be wrong and how to diagnose/fix it?


Thanks anyone in advance!
 

Attachments

  • Scan_20200409_081559.jpg
    Scan_20200409_081559.jpg
    513.9 KB · Views: 992
There is not much there to get old. Optocouplers, zeners, scr are way down the usual suspects list, but are on it.
Can't be the optocouplers since disconnecting doesn't change result. Optocouplers usually fail by not transferring enough current, not by leaking. I'd change the 12 v zener first, 1n4742, then the SCR. Both parts <$1. You probably find you can't get that exact zener diode; get something of same voltage 12 and wattage, 1 to 1.3 watts. Nobody sells 300 v scr anymore, a 600 v one would be fine. TO220 package is commonly available for SCR, but look at what is there. SCR could have been damaged by a 1300 v spike coming in from the AC plug. If the board is that difficult to get out, I'd change both. Label all those flying wires to the capacitors etc before taking off with scotch tape & a sharpie, or something better.
You can test the zener with a >14 v supply a resistor and a voltmeter, but for $.07 I wouldn't bother. SCR tests are more difficult and a leaks when hot test is a real problem. You could in circuit spray circuit cool on the SCR when it is failing, but a can of that is $10 and a SCR <$1.
 
Last edited:
Thanks for the suggestions, I will try as you said.
Actually, the cable disconnects only the input of the optocouplers, couldn't they become leaky even with the input disconnected?
The weird thing is that whatever becomes leaky it only does so when switching the amp off, never during normal use no matter of how hot it becomes or how hard I drive the amp. Very strange..
 
You could measure voltage acrossR313 then R315 to see if optocoupler is leaking with pamaona grabbers or something. If you can reach it in circuit.
If replace optocouplers 4n33 only have ctr of ~20% whereas modrn ones are 100% or higher. This could cause false tripping of protection. You'd have to waste 80% of input current with a parallel resistor or something.
 
K300 is the relay involved in the soft start circuit, which seems to work fine as I can hear it click after 1-2 secs from switch in, bypassing the three 10ohm parallel resistors on the top left of the schematic. Even if it was faulty it would not cause the amp to go into protection.
 
Unless when switching off the amp k300 stays closed for a while (for a malfunctioning in the relay or it’s circuit) and when switching it on again the soft start doesn’t work. How would the absence of the soft start cause tripping of sw300 though?
That should be easy to verify though. If soft start is necessary for switch on I can replace the soft start relay with a jumper and the protection relay should trip all the times right?
 
Last edited:
If you can measure the resistor voltage, you can also measure the zener voltage hot vs cold (12v) and the SCR gate-source voltage, <1.5 vdc if supposed to be off.
Soft start k300 contact, upper left of diagram, limits AC current to transformer to that through 3.3 ohms if open. Good soft start value.
Dried up C306 4700 uf 5 kv could make the transformer act weird. Maybe it is 4700 pf? Part also subject to 1300 v spikes from the AC feed. (air conditioner compressor shutoff). Nowadays instead people put a 300 vac MOS surge supressor across the AC feed right after the mains fuse or breaker.
 
Last edited:
Ok will do that.
C306 is one of the four big filter caps (1900 if), maybe you meant c308 which is in parallel with the three 10ohm resistors? This is a 4700pf old paper rifa capacitor and I already have replacements for all of them (there are many in this board) as they start to look cracked and they are known to fail with age. Could they possibly cause spikes that can activate directly the relay coil?
 
I can try that, but why would they act up only when switching off?
Relays are not mechanically symmetrical (include electromechanically).

Typically force separating contacts comes from a spring, not TOO strong, and which can weaken with time, while closing force comes from an electromagnet with consistent "pull" and which may very well be quite stronger than opposing spring, so a relay may consistently switch ON yet under some circunstances (obviously old/worn/corroded/dirty) fail or show weakness while opening.

If you wish, call opening "passive" and closing "active", with "active" being the mechanically stronger one.

I am not talking about this particular circuit but in general.
 
Thanks for the explanation. I understand that, but when I wrote switching off I meant the amplifier, not the relay.
The power switch relay is not supposed to close under normal functioning, so there cannot be a fail in its opening if it is normally already open.

This could be the case of K300, however, where it fails to open after switching off the amp, thereby disabling the soft start on subsequent amp turn on, but why this would make the amp go into protection I don't get.
 
Hello everyone
I have an annoying issue with the protection circuit of a Mark Levinson 23.5 that I would like to solve. I need your help.

Basically, when the amp is well warmed up and I turn it off I cannot turn it on because the protection relay of the power switch is activated and I have to wait ~30-60 seconds. It doesn't happen all the times, about 80% of the times. The problem is not present when the amp is cold. The previous owner says the problem only recently showed up, so that's not how the amp is supposed to work.

Any clue about what can be wrong and how to diagnose/fix it?
Thanks anyone in advance!
The power switch in the ML amps is a well kept secret, I never took one apart and it can’t be repaired.
It can be turned off from the outside by heatsink temp or DC at the LS output but turn on is controlled by factors like how much energy is still in the large supply caps and also by an internal ntc based circuit that has to cool down before it can activate its internal turn-on relay
So when the bias currents in the power transistors, discharging the power supply caps, are still o.k. I wouldn’t bother about the switch-on delay.
I’ve seen this in several 23.5 amps and it does not affect a proper functioning in any way when it’s finally on.

Hans
 
Interesting explanation thank you.I will check the bias currents just to be sure.
It’s interesting the switch has an internal ntc circuit controlling it, however I noticed that when the problem happen the large can capacitor of the unregulated rail supplying the relay circuit discharges pretty quickly ( it doesn’t do so when the protection doesn’t activate, nor if I disconnect the cable joining the capacitor and the relay circuit). Therefore the cap must discharge through the thyristor necessarily. This makes me think the problem must be somewhere between the optocouplers, the 2n5401 and the thyristor itself and not in the power switch right?
I agree I could just live with the issue but I would be happier if I could figure t out.
 
I will check the bias currents just to be sure.
It’s interesting the switch has an internal ntc circuit controlling it, however I noticed that when the problem happen the large can capacitor of the unregulated rail supplying the relay circuit discharges pretty quickly ( it doesn’t do so when the protection doesn’t activate, nor if I disconnect the cable joining the capacitor and the relay circuit). Therefore the cap must discharge through the thyristor necessarily. This makes me think the problem must be somewhere between the optocouplers, the 2n5401 and the thyristor itself and not in the power switch right?
I agree I could just live with the issue but I would be happier if I could figure t out.
Report Post
Question: How did you activate the protection and why ?
Notice that the thyristor switch is only connected to one channel and only to it's positive side where the discharge according to your observation goes very quick after activating the protection, which makes sense because of the thyristor unloading the cap through a 400R resistor to activate SW300 and switch off the amp.
So as long as enough current is flowing through the SW300 coil, you can't switch on but the discharge process shouldn't take very long.
But as mentioned, this is not the only mechanism that prevents the amp to switch on, because there is also some circuitry in the switch itself.

I see no logical way how an optocoupler, the 2N5401 or the thyristor can make the switch-on time any longer as it was in the past because to achieve that, a minimum of current has to flow through coil SW300 to prevent, for that needing a voltage on the positive Cap that has just been discharged by the same thyristor.

Hans
 
How did you activate the protection and why ?
Sorry I think I wasn’t clear in my previous message. I meant that when the protection goes in after switching off the amp and trying to switch it back on then the (left channel) positive filter cap discharges very quickly, supposedly through the thyristor. When the protection doesn’t kick in the capacitor discharges less quickly after switch off ( confirming that it discharges through the thyristor when the protection goes in). I’m not sure about the timing as I didn’t measure it but the feeling is that it is quite variable between 20 and 60 seconds. My explanation is that if the relay activates on switch on rather than on switch off, then the voltage on the filter cap would be different depending on how long it’s passed since switch off, because the capacitor slowly discharges anyway through other circuits in the meanwhile.

One more thing I suspect, but I cannot verify at the moment because the amp is disassembled, is that disconnecting the cap from the protection circuit seems to slow down even more its discharge (in this case that would indicate some leakage in the protection circuit even when the relay isn’t activated?). So to recap, with the cap disconnected it discharges slowly (same as the other channel positive rail filter cap); with the cap connected with the protection circuit but relay not activate it seems to discharge somewhat more quickly (but need to verify); when the protection is active the cap discharges very quickly.
 
Last edited:
This confirms exactly what I said before.
With the wire to the relay disconnected from the cap, the thyristor becomes isolated from Vcc and with zero volt on it, is 100% unable to do anything at all.
Nevertheless switching on still takes a long time.
That is because of some mysterious circuitry in the switch.
I can only repeat, don’t bother it does not do any harm.

Hans
 
This confirms exactly what I said before.
With the wire to the relay disconnected from the cap, the thyristor becomes isolated from Vcc and with zero volt on it, is 100% unable to do anything at all.
Nevertheless switching on still takes a long time.
That is because of some mysterious circuitry in the switch.
I can only repeat, don’t bother it does not do any harm.

Hans

no with Vcc disconnected from the protection circuit the amp switches on all the times with no problems. Sorry for my bad english.