Yes it is correct. With no source connected ther is not gnd continuity between the two amps. The noise is present also if a single channel is connected to source. If i connect source only to L amp the noise is on the L channel and the same for the R channel alone. Also with both channels xonnected to source.
Sorry. I have to change my version.
After a night of power off the noise is there also without source connected. I discovered that if I power on quickly after 20-30 seconds, noise disappears but is still there if I wait 1 minute or more.
After a night of power off the noise is there also without source connected. I discovered that if I power on quickly after 20-30 seconds, noise disappears but is still there if I wait 1 minute or more.
No problem 🙂
All this points to some weird kind of oscillation of some sort. Not having a scope makes this more difficult and so I think you need to begin by doing as I outlined earlier... get one channel right and treat that channel as separate mono unit.
There are so many variables in how you could connect it all up and also what the PCB is like with regard to layout.
All this points to some weird kind of oscillation of some sort. Not having a scope makes this more difficult and so I think you need to begin by doing as I outlined earlier... get one channel right and treat that channel as separate mono unit.
There are so many variables in how you could connect it all up and also what the PCB is like with regard to layout.
I suspect the noise ( woofer moves! Yes it does) Is a sort of bump as it disapperars if I switch on fast ( charging of the ripple caps?!?!?). The PS i am using is a capacitance multiplier stabilised by a zener in parallel to the cap (the one between base and gnd). I also put a 1n4148 diode between emitter and base as a protection.
Will I solve with a speaker delay relais?
Will I solve with a speaker delay relais?
A switch on thump is normal on an AC coupled amp but it just that, a thump and not a 'noise' as you describe.
A delay should stop you hearing the noise but it doesn't solve the problem. If you use a delay then you also need to add a 100 ohm 2 watt resistor across the amplifier output to ensure the speaker coupling cap charges during the time the relay is open.
A delay should stop you hearing the noise but it doesn't solve the problem. If you use a delay then you also need to add a 100 ohm 2 watt resistor across the amplifier output to ensure the speaker coupling cap charges during the time the relay is open.
I used that schematic in post #28.
Vin at J301 is 35VDC (25VAC transformer). R301 is 1200ohm in order to have 5mA on the zener diodes (35V-29.1V/1200ohm=5mA).
D301 should have a protection function to avoid reverse current on the transistors due to the caps.
i suspect the noise issue is there because with a CRC power supply I also have a bump but i don't have that "oscillation-like" noise.
Maybe some issue on the time constant t=R*C; in this case it should be 1200ohm*470uf (0.00047Farad)=0.564 sec. Too fast? is my calculation correct?
regards
Vin at J301 is 35VDC (25VAC transformer). R301 is 1200ohm in order to have 5mA on the zener diodes (35V-29.1V/1200ohm=5mA).
D301 should have a protection function to avoid reverse current on the transistors due to the caps.
i suspect the noise issue is there because with a CRC power supply I also have a bump but i don't have that "oscillation-like" noise.
Maybe some issue on the time constant t=R*C; in this case it should be 1200ohm*470uf (0.00047Farad)=0.564 sec. Too fast? is my calculation correct?
regards
The calculation is fine... what happens is the output voltage rises pretty much exponentially to the zener voltage.
You could try different time constants, even down to a few uF and see if it makes any difference. Its very hard to say just what is going on as the basic JLH is pretty much immune to stability issues.
The big question is whether you are masking the fault or issue, or whether there is an issue in its own right with the build somewhere.
You could try different time constants, even down to a few uF and see if it makes any difference. Its very hard to say just what is going on as the basic JLH is pretty much immune to stability issues.
The big question is whether you are masking the fault or issue, or whether there is an issue in its own right with the build somewhere.
The fact the amp is noiseless and all voltages and currenys are fine could tell me that there is no fault in whole amp but there is an issue in the PS of post #28.
Maybe is simpler than appears...what about the 1n4148? Is it located correctly? I saw Elliot multiplier with that protection diode just between emitter and collector of the upper transistor....
Maybe is simpler than appears...what about the 1n4148? Is it located correctly? I saw Elliot multiplier with that protection diode just between emitter and collector of the upper transistor....
Under steady state conditions the 4148 is reverse biased but I'm not so sure its good practice to have it in such a potentially low impedance part of the circuit. A small glitch could allow a massive current spike to flow in the diode and zeners. That's not good.
Try it without as it should really have no real effect. I wouldn't worry over reverse biasing the transistors, we never used to worry over such things years ago in circuits like this and it is after all just a classic series pass regulator.
Try it without as it should really have no real effect. I wouldn't worry over reverse biasing the transistors, we never used to worry over such things years ago in circuits like this and it is after all just a classic series pass regulator.
It ensures the power transistor can turn off when needed and in the appropriate applications (not this one) it significantly decreases the turn off time (a good thing in a power amplifier output stage for example).
It also prevents the combined pair being 'leaky' due to any slight leakage currents from the driver and it also stops interference from entering what could be classed as a floating or high impedance node.
None of these matter here, but they might if both transistors were super high gain small signal types and where the output currents were in the mA or even uA region. In those cases even high humidity or slight condensation could begin to turn the second transistor on. The resistor stops that entirely.
It also prevents the combined pair being 'leaky' due to any slight leakage currents from the driver and it also stops interference from entering what could be classed as a floating or high impedance node.
None of these matter here, but they might if both transistors were super high gain small signal types and where the output currents were in the mA or even uA region. In those cases even high humidity or slight condensation could begin to turn the second transistor on. The resistor stops that entirely.
Try putting a 100 ohm resistor in series with the base of TR302 . A basic emitter follower can oscillate a darllngton pair is more likely to . This is called a base stopper resistor .
It seems a pretty good solution to try. I am gonna try this in the afternoon.
I have a doubt: if i turn the amp on it makes the noise. If I power it off and power on again within a few seconds, the noise is absent. So is this should be compatible with a voltage-associated oscillation? In the sense that when the caps are totally discharged, we have a rising of the voltage and the amp passes through a voltage value at which it is unstable but the voltage rises quickly and instability point is overpassed. Is this possible? It means that supplying the amp with low voltage it shoul be instable. Maybe at this point time costant is too slow...
Have you considered trying without the regulator for a few moments ?
28 to 40 volts isn't that drastic on an amp like this and would see the bias current increase by maybe 80%. You could increase R1 and R2 to the values recommended for 36 volt operation first.
That would 100% eliminate the regulator as a possible cause.
28 to 40 volts isn't that drastic on an amp like this and would see the bias current increase by maybe 80%. You could increase R1 and R2 to the values recommended for 36 volt operation first.
That would 100% eliminate the regulator as a possible cause.
Great news from Italy!!
I put a 22ohm resistor in series with the base of T302 (post #28) and the noise has desappeared!
So it was an oscillation?
I put a 22ohm resistor in series with the base of T302 (post #28) and the noise has desappeared!
So it was an oscillation?
- Status
- Not open for further replies.