This is one of the best videos on this channel! Tony in his element. Very educative and entertaining finally of 2252b series. I have learnt a lot about testing amplifiers and, frankly it is encouraging to witness expert to struggle from time to time . However Tony prevail as always :-)
I wonder if the unterminated PCB traces from the original cap locations may contribute to the oscillation? Brilliant work Tony! This is why I love your channel.
Very nicely done Tony! No way in hell would I have been able to figure that out... but now I think I'll know what to look for in the future when replacing transistors.
Thank you for the nice video Tony! @ 32:14 about the miller effect it should be "base-collector" in your captions. Another way to deal with the oscillation caused by the faster output transistors is adding a smaller miller cap on the VAS (Q705, Q706) instead and I'm absolutely sure you are aware of this.
Its a bit stupid how modern class D amps are only now beginning to catch up with the low impulse response/intermodulation distortion of these legacy A/B receiver units from the 70-80's. Today's "music" and how its consumed mostly with low cost headphones surely played a role too. There is enough HD audio tracks available now to allow more high end class-D amps to be sold thus produced. Thank you for these videos. As an 80s kid we grew up with these systems and they sounded exactly as we remember, making even the better modern high-end speakers shine.
Interesting! I was thinking "faster transistors?" when I saw this. Glad you found the cure, and corrected the long path from the transistors to the Miller caps
Excellent troubleshooting! I’ve heard that using film capacitors in place of electrolytics in signal paths can sometimes ring and induce oscillation too.
AWESOME! Good info to know! SO, have to wonder, did they actually put in the wrong caps originally, or did they find out later that the 100pf were too large and had to make a revision that didn't get changed on the schematic? 🤔
I suspect this was an undocumented revision. Installing transistors with really low gain was probably an attempt to reduce the chance for oscillation and still get better high frequency response. When I installed a faster transistor with higher gain, it was necessary to increase the miller capacitor value.
Excellent video Tony! Thank you. I wonder are there any significant differences in parasitic oscillation using 1k vs. 20k test signal (before upping these small miller caps)? Maybe production was already half way in the making and someone has (for)seen the potential problem so Marantz tried to work around it. It would be amazing to see were there any (other) changes in the PCB design and or component usage liat of this specific model as the model production run went on. You caught some differences in SM and on board and that is great! Doesn't this makes you wonder on how many of other Marantz designs are flaws just like this one since so many of similar receivers used the same topology and almost the sane circuit architecture. That is something to think about, isn't it? Anyway, I am just thinking out loud. Keep these super interesting videos coming! Regards, Krešo
20kHz will light it up a bit more than at lower frequencies. Not sure if I showed it in the video, but it was a bit more visible at 20kHz than at 1kHz. That said, depending on the circuit, parasitic oscillation can happen at any frequency and have more or less amplitude depending on the fundamental frequency. Thanks for the comment!
Wow Tony what timing ! I'm experiencing the same exact thing ( on the scope) with a Nikko alpha 220 I just repaired. And when cranked up the bass peaks cause protection mode to kick in. No DC to speak of at idle on the speaker outputs. I see, and it was pointed out to me by Ray G, that the model 230 had caps added/it was updated. So I need to study the schematic and hopefully figure out where to add them. It looks as though the board was updated to accommodate the change, but can I just add them directly between the Base and Collector ??? that would be my hope! Your demo's are great ! The face of that unit is really great after LED pushed the color.
Super interesting troubleshooting. I thought it looked like the distortion in the sine wave always occurred when it hit a certain amplitude/voltage and then stayed on the wave when it crossed that voltage. Maybe showing why the THD moved down again as the “glitch” in the wave moves from the “bottom” to the “side”. Could this be any clue as to a cause when troubleshooting? Hope I’m stating this in a way that makes sense. Thank you for sharing your knowledge with us. I love the way your different videos cover the range from beginners to advanced.
Very good info. Learning so much about the little things that effect scope patterns. My new to me scope has FFT, will distortion similar to this show up using the fft? I have played around with it, but none of it makes sense to me.
I remember these jukebox amps, new design, and they had to quickly mod them because the oscillation was so bad, the amp blew up. I think it was an fet design. New to them as well.
@ 9:00 You may improve your THD measurements if you connect directly to the back of the amp instead of at the far end of any extra cabling close to the load like you have it set up in this video. And, at 26:00 , that is just nuts that they didn't conform to the schematic design and put in half valued components. It would have been interesting to disconnect the zobel/snubber and sweep higher to see if any oscillating outside of the audio spectrum was happening.
The poor THD was due to the parasitic oscillation of the amp. Moving the leads to the speaker terminals would have only made a very small difference, if any. I think the circuit change was an attempt at reducing the possibility of oscillation while still having good high frequency response. The very low gain of the original transistors was probably on purpose. This also most likely made it necessary to have a lower capacitor value on the base-collector. This design probably had issues from the start and that's why they tinkered with the component values.
Why don't you see the distortion of your signal generator when you measure the distortion at the output? Is your THD analyzer subtracting or measuring relative to a preset level? Always interesting videos.
The signal generator has very low distortion. The THD meter was showing the very low distortion of the signal generator plus the very low distortion of the amp at low volume plus any added distortion of the test leads.
This is one of the best videos on this channel! Tony in his element. Very educative and entertaining finally of 2252b series. I have learnt a lot about testing amplifiers and, frankly it is encouraging to witness expert to struggle from time to time . However Tony prevail as always :-)
Thanks Tony, seeing things go a bit wrong is often more valuable than the smooth ride.
If you want to see things go wrong, you've come to the right place ;)
@@xraytonyb I love this stuff, if not for problems, neither of us would be here enjoying this!! :)
I wonder if the unterminated PCB traces from the original cap locations may contribute to the oscillation? Brilliant work Tony! This is why I love your channel.
Maybe my favourite video so far Tony. Love the troubleshooting an emphasizing the need to test thoroughly.
Very good info about modifying components with unintended consequences...
Very nicely done Tony! No way in hell would I have been able to figure that out... but now I think I'll know what to look for in the future when replacing transistors.
Thank you for the nice video Tony!
@ 32:14 about the miller effect it should be "base-collector" in your captions.
Another way to deal with the oscillation caused by the faster output transistors is adding a smaller miller cap on the VAS (Q705, Q706) instead and I'm absolutely sure you are aware of this.
I see that now. Thanks for catching it. Another case of thinking one thing and typing another :)
Its a bit stupid how modern class D amps are only now beginning to catch up with the low impulse response/intermodulation distortion of these legacy A/B receiver units from the 70-80's. Today's "music" and how its consumed mostly with low cost headphones surely played a role too. There is enough HD audio tracks available now to allow more high end class-D amps to be sold thus produced. Thank you for these videos. As an 80s kid we grew up with these systems and they sounded exactly as we remember, making even the better modern high-end speakers shine.
That was some impressive troubleshooting. Great work.
Interesting! I was thinking "faster transistors?" when I saw this. Glad you found the cure, and corrected the long path from the transistors to the Miller caps
Excellent troubleshooting! I’ve heard that using film capacitors in place of electrolytics in signal paths can sometimes ring and induce oscillation too.
You shall not covet your neighbors test equipment, great troubleshooting.
I may confess to the sin but repentant I am not.
Thanks Tony - very instructive. Food for thought on other designs !
AWESOME! Good info to know! SO, have to wonder, did they actually put in the wrong caps originally, or did they find out later that the 100pf were too large and had to make a revision that didn't get changed on the schematic? 🤔
I suspect this was an undocumented revision. Installing transistors with really low gain was probably an attempt to reduce the chance for oscillation and still get better high frequency response. When I installed a faster transistor with higher gain, it was necessary to increase the miller capacitor value.
That’s a beauty!
Excellent video Tony! Thank you. I wonder are there any significant differences in parasitic oscillation using 1k vs. 20k test signal (before upping these small miller caps)? Maybe production was already half way in the making and someone has (for)seen the potential problem so Marantz tried to work around it. It would be amazing to see were there any (other) changes in the PCB design and or component usage liat of this specific model as the model production run went on. You caught some differences in SM and on board and that is great! Doesn't this makes you wonder on how many of other Marantz designs are flaws just like this one since so many of similar receivers used the same topology and almost the sane circuit architecture. That is something to think about, isn't it? Anyway, I am just thinking out loud. Keep these super interesting videos coming! Regards, Krešo
20kHz will light it up a bit more than at lower frequencies. Not sure if I showed it in the video, but it was a bit more visible at 20kHz than at 1kHz. That said, depending on the circuit, parasitic oscillation can happen at any frequency and have more or less amplitude depending on the fundamental frequency. Thanks for the comment!
Great vid
Every 2252 ive ever Worked on has had harmonics on the Spectrum analyzer . This is the same fix i incorporate into every one
Wow Tony what timing ! I'm experiencing the same exact thing ( on the scope) with a Nikko alpha 220 I just repaired. And when cranked up the bass peaks cause protection mode to kick in. No DC to speak of at idle on the speaker outputs. I see, and it was pointed out to me by Ray G, that the model 230 had caps added/it was updated. So I need to study the schematic and hopefully figure out where to add them. It looks as though the board was updated to accommodate the change, but can I just add them directly between the Base and Collector ??? that would be my hope! Your demo's are great ! The face of that unit is really great after LED pushed the color.
Nice!
WOW! It has been a long time since I have seen a slow scanning oscilloscope!
That's probably what I called it in the video, but it is actually a slow-scan audio frequency spectrum analyzer. ;)
Super interesting troubleshooting. I thought it looked like the distortion in the sine wave always occurred when it hit a certain amplitude/voltage and then stayed on the wave when it crossed that voltage. Maybe showing why the THD moved down again as the “glitch” in the wave moves from the “bottom” to the “side”. Could this be any clue as to a cause when troubleshooting? Hope I’m stating this in a way that makes sense. Thank you for sharing your knowledge with us. I love the way your different videos cover the range from beginners to advanced.
Very good info. Learning so much about the little things that effect scope patterns. My new to me scope has FFT, will distortion similar to this show up using the fft? I have played around with it, but none of it makes sense to me.
I remember these jukebox amps, new design, and they had to quickly mod them because the oscillation was so bad, the amp blew up. I think it was an fet design. New to them as well.
Great fault finding. Without a THD meter I guess you really need to look at the waveform closely.
@ 9:00 You may improve your THD measurements if you connect directly to the back of the amp instead of at the far end of any extra cabling close to the load like you have it set up in this video. And, at 26:00 , that is just nuts that they didn't conform to the schematic design and put in half valued components. It would have been interesting to disconnect the zobel/snubber and sweep higher to see if any oscillating outside of the audio spectrum was happening.
The poor THD was due to the parasitic oscillation of the amp. Moving the leads to the speaker terminals would have only made a very small difference, if any. I think the circuit change was an attempt at reducing the possibility of oscillation while still having good high frequency response. The very low gain of the original transistors was probably on purpose. This also most likely made it necessary to have a lower capacitor value on the base-collector. This design probably had issues from the start and that's why they tinkered with the component values.
what pos was the volume pot at to give 52W with that 500 mv input?
Why don't you see the distortion of your signal generator when you measure the distortion at the output? Is your THD analyzer subtracting or measuring relative to a preset level? Always interesting videos.
The signal generator has very low distortion. The THD meter was showing the very low distortion of the signal generator plus the very low distortion of the amp at low volume plus any added distortion of the test leads.
I tried to listen for distortion on my 2250b at 30 watts but my speakers caught on fire and I lost my hearing. 😀
Thanks Professor!