Did you measure the current through R1?
Really strange behavior 🤔, my JLH works stably with 60MHz output transistors.
I think i gave a sensible answer. I also had an experience of defective but new genuine transistors.
Mooly, thank you very much for your active help in finding the problem.
I checked the resistor values several times - everything is correct. Everything is fine with the supply voltage as well. Installing fake Chinese transistors on the same board while maintaining the ratings of all elements removes the problem - the oscillograms become normal. This also speaks in favor of the fact that the element values are correct.
"The weird thing though is that zillions of JLH's have been built using every type of transistor you can think of and this issue has never been seen (as far as we know) before. It is odd".
I completely agree with you - I have assembled several Hood amplifiers myself (mostly on 2N3055G and TIP41C transistors) and have never encountered this problem. Unfortunately, speaking about MJ15003/01G transistors, I have never seen oscillograms with a frequency sweep - only in words. I guess I didn't search well. If someone gives me a link to oscillograms of MJ15003/01G transistors at different frequencies with sine and square waves, I would be very grateful. Many consider the MJ15003 transistors to be the best choice for a Hood amplifier in this type of package. I have never dealt with MJ15003, so it was interesting to install them. Unfortunately, everything went badly - I have never encountered this type of problem and do not know how to solve it...
So far I am inclined to conclude that the MJ15001G transistors from Onsemi may be defective. The best way to check this version is to buy other original power transistors from another place and see what happens. If the problem disappears, then the current original transistors are defective.
You're welcome 🙂 Its an interesting one for sure.
That would be the most conclusive.
This is a very good remark and suggestion! Unfortunately, I have no "clean" Onsemi transistors left. It is difficult to say anything, but it seems to me that I could not damage the transitions on the measuring stand - despite the fact that the collector current was quite high (2A), the total voltage was 24 Volts, 12 of which were quenched by the power resistor. The transistor base current was 0.01 A on average. This is still not much for a 200 watt transistor. Everything was fine with heat dissipation - there was no overheating of the transistors. In addition, the Chinese counterfeit transistors were subjected to the same procedure - they could certainly be damaged, since their crystal is small, and the actual power is about four times less than the declared one. But nothing happened and the Chinese counterfeit transistors showed good oscillograms, unlike the original Onsemi. Of course, these are just my assumptions... By the way, the gain of the Chinese counterfeit transistors was only about 50-55.
Last edited:
Random thoughts... not sure where this is going 🙂
If you set the output voltage to say -/+10 volts peak/peak with no load, what sort of change do you see in that voltage as you add the 8 ohm load. Is there any change? Does the output voltage decrease or does it increase? If it increases under load there could be an issue with the ground and wiring configurations. Would higher and lower gain devices make a difference to that... I dunno.
Like I say, just random thoughts.
If you set the output voltage to say -/+10 volts peak/peak with no load, what sort of change do you see in that voltage as you add the 8 ohm load. Is there any change? Does the output voltage decrease or does it increase? If it increases under load there could be an issue with the ground and wiring configurations. Would higher and lower gain devices make a difference to that... I dunno.
Like I say, just random thoughts.
Another random thought. What does the output from the driver look like with power transistors connected and without?
I don't think that would work tbh. The driver transistor would see a low value collector resistor and a high value on the emitter. The feedback loop would be open as well.
I have the Chinese jlh1969 with 2 2n3055 transistors.
As power in using the sound-au.com capacitor multiplier.
It delivers around 25v DC.
Each channel has its own transformer winding and rectifier
board.
I have around 9v DC on the speaker terminals when I power up the amplifier. The DC protection is activated. Slowly the DC voltage drops. Safer around 1 minute it reaches 0.5v DC and the protection goes off and drops the DC instantly to 0.00v.
When I power off the amp, there is a negative DC voltage of around 3v DC on the speaker terminals. Slowly going back to 0v.
Any idea what may cause the DC voltage at power up/down?
It's very annoying.
As power in using the sound-au.com capacitor multiplier.
It delivers around 25v DC.
Each channel has its own transformer winding and rectifier
board.
I have around 9v DC on the speaker terminals when I power up the amplifier. The DC protection is activated. Slowly the DC voltage drops. Safer around 1 minute it reaches 0.5v DC and the protection goes off and drops the DC instantly to 0.00v.
When I power off the amp, there is a negative DC voltage of around 3v DC on the speaker terminals. Slowly going back to 0v.
Any idea what may cause the DC voltage at power up/down?
It's very annoying.
This output voltage is normal if you have added a protection device or a time delay because your output capacitor is not charged while the relay is open.
You don't need all that, the output capacitor "protects" your speakers.
You don't need all that, the output capacitor "protects" your speakers.
Or you can add a 10R 10W resistor connected from NC contact of the relay to ground. It'll eliminate turn on thump & you can keep that delay circuit too.
I found fake 2N3055G transistors and installed them. The oscillograms are normal - the signal behaves normally up to 50 kHz. From this I conclude that the problem is most likely in my original Onsemi transistors, the oscillograms of which are bad. This means that everything is fine with the board element ratings. If installing MJ15003/1G transistors required changing the ratings of the strapping elements, this would have been known for a long time. Therefore, I will order other original transistors, install them and see what happens. Thanks to everyone who gave advice! I apologize for not answering everyone - I read everything. But first, I need to exclude the possibility that my original transistors are not defective.
Will this not place the resistor in parallel with the speaker when the relay is closed and so reduce the impedance of the load?
It is a super strange issue. So you are going to try some more of those same MJ's from elsewhere? It would be fascinating to see the result.
As you say, this kind of issue or even hint of an issue should/would have surfaced long ago.
What @NanoFarad says is the best way to do this. No extra load is added... think it through 🙂
Hmmm... unfortunately it seems yours doesn't have that contact available.
http://www.songlerelay.com/Public/Uploads/20161104/581c81ac16e36.pdf
Does your board give a proper delay before activating the relay or is it just DC offset protection? If it does give a delay then how long is that delay?
http://www.songlerelay.com/Public/Uploads/20161104/581c81ac16e36.pdf
Does your board give a proper delay before activating the relay or is it just DC offset protection? If it does give a delay then how long is that delay?
What I'm thinking is that if there is a delay then you can still add a resistor across the amp output before the relay. Everything depends on the delay time as to what value resistor you use. The longer the delay and the higher the value of resistor... and higher means less load on the amp.
@Mooly
Delay depends on how long after the last time you turned the amp on.
From cold it takes about 1 minute before the relay will close. It has a DC voltage of 10-9v on the speaker terminals.
When it drops to below 0.5Vdc the relay close.
If I turn on the amp, just after I disconnected it, the relay close almost immediately .
I don't understand why they put this relay but still foresee a hole on the pcb for the NC contact. The one with NC contact cost 1-2 cent more.
Delay depends on how long after the last time you turned the amp on.
From cold it takes about 1 minute before the relay will close. It has a DC voltage of 10-9v on the speaker terminals.
When it drops to below 0.5Vdc the relay close.
If I turn on the amp, just after I disconnected it, the relay close almost immediately .
I don't understand why they put this relay but still foresee a hole on the pcb for the NC contact. The one with NC contact cost 1-2 cent more.