Basically, the Twisted Pear kit, the 1uF and the 220uF are bypassed using 0.01uF. The PS I mentioned are per channel. The filter caps are only 1000uF, but I plan to remove them and see what happens.
There are certain sonic signatures that I have not determined what is causing it because a more revealing amp can sometimes show problems in other areas.
I will get some pictures when I open it up the next time.
Hmm, the relays do not seem to delay with power up.
There are certain sonic signatures that I have not determined what is causing it because a more revealing amp can sometimes show problems in other areas.
I will get some pictures when I open it up the next time.
Hmm, the relays do not seem to delay with power up.
An externally hosted image should be here but it was not working when we last tested it.
Interesting, the measured DC at the input is more than the output, about 5 times as I recall. Generally not a good situation. The sound seems to change a bit as I let the unit run in.
Last edited:
Just replaced the tweeter. The sound quality seems to have an echo effect probably in the 400Hz~800Hz region, like a karaoke. First time listening will attract attention, but the sound stage is not clearly focused, and the depth is not there. Since I also have the speaker on Ultimate Equalizer, it could be the phase change due to the amp characteristics compared to a voltage source amp.
Hmm, without using the Ultimate Equalizer, it does not sound right either. This is not a simple task. Need to go through a whole suite of test and listing to figure out what is going on.
Hmm, without using the Ultimate Equalizer, it does not sound right either. This is not a simple task. Need to go through a whole suite of test and listing to figure out what is going on.
Last edited:
TP's and first GB BOM don't make justice of the My_Ref and some components, like Panasonic FC and AVX BQ are detrimental.
If you can make a picture of the My_Ref PCB (or indicate clearly which BOM you 're using) I can point for you the most important components to swap.
As I said, aside from the main power to the amplifier chip, all components came from TP. From the sound, it seems like damping related problem in the 400~800Hz region. Also with measured DC on the input (no input connected), the problem is not as simple as changing components unless those actually change the damping characteristics of the amp. Which components do you suggest?
Panasonic FC and AVX BQ? Are those in the power supply?
Panasonic FC and AVX BQ? Are those in the power supply?
So if it's an original TP kit you should have FCs and BQs
The first thing to do is to remove C17, C18, C19, C20 (BQs or MKS2), they're useless and in fact detrimental.
I'll publish later an updated BOM.
Yes, also C9 bypass (C21) is an AVX BQ
Dario, I dont know whats wrong with it but he has more problems than simple tweaking. Ever seen a huge inductor on a MyREF? Spose that could be part of his switching supply which could also be part of the problem. Who knows? But with an echo at 400-800hz I dont think thats a problem fixed with tweaks.
The inductor is a common mode choke, part of the power line filter so that the PS does not pass conductive noise to the mains. This design had been used on other chip amps with relatively good results EMC and sonic qualities. My initial guess is the 1000uF power filter caps, which I plan to remove first since this is an uncontrolled factor in this kind of floating circuit. udailey is right on, this needs to be solved first before getting into the tweeking.
Last edited:
Yes Uriah you're probably right, I would first build the 'stock' MyRef, and only when I'm sure it works I would replace the linear supply with the switching one...
But I would remove the bypasses anyway, since they disrupt the floating ground of the MyRef, maybe the two things are related...
This is not a simple chipamp...
I'm amazed I did't burn anything yet. But I did find out that the 24V to the protection circuit was not connected because it picks off of the transformer. Plus the fact that I got the grounds of both channels connected together, the mess let the sound pass though somehow. I'm guessing that applying 27V the through D1//D2 is not going to damage anything is it?
BTW, during power on, the input can have a DC of up to 1.5V across it, I think this is the main reason the decoupling cap needs to be there. It sort of fades out in about a minute.
BTW, during power on, the input can have a DC of up to 1.5V across it, I think this is the main reason the decoupling cap needs to be there. It sort of fades out in about a minute.
Last edited:
Okay, seems that the protection relays are operating correctly, input DC seems to float around 100mV, output DC seems to float around 1mV. Connected to my notebook source, the first thing I noticed was the hum in the background which has never been a problem with other amplifiers using the same speakers.
Sound:
The transients and detail is good, but the sound stage image is mixed up, the decay of sound is not so natural which seems to be partly the cause of sound stage, but I guess the phase also plays a big role.
The way the transients and dynamics are presented, I can imagine how it can be liked by many.
I still think there is some problem with damping. But I will investigate it a bit more and see if anything can be done.
Sound:
The transients and detail is good, but the sound stage image is mixed up, the decay of sound is not so natural which seems to be partly the cause of sound stage, but I guess the phase also plays a big role.
The way the transients and dynamics are presented, I can imagine how it can be liked by many.
I still think there is some problem with damping. But I will investigate it a bit more and see if anything can be done.
Fantastic!
Try some of the mods I've suggested... You'll be amazed
PS
The new LM318's PS I'm testing sounds fantastic...
I had used snubbers on other amps, but not this one. So for the sake of power supply configuration, I'm going to add them in first. Then I'm going to check the input DC again to see how it measures. Still would like to be able to take the input cap out of the signal path if at all possible.
Let us know your results
It's possible, just jumper the input cap pads.
Anyway you'll discover that bass performance suffers a bit.
DC on the input side is around 50mv and 100mv for each channel. So bypassing the input caps is out of the question.
Sound is detailed as mentioned before, there are some frequency range where there are changes image focus. Image depth is somewhat improved, but not yet what I'm used to. The decay of instrument tones is not natural seeming too slow resulting in feeling that the trailing sound is not totally from the instruments. This characteristic seems to be the same that is effecting image depth perception because the timbre and delayed reflections in the recording environment is mixed with this.
Hmm. I'm beginning to think that the snubber is not going to do much because the circuit is floating from the power ground. Need to think more about this.
Just going to let it run this week and come back on it next week.
Sound is detailed as mentioned before, there are some frequency range where there are changes image focus. Image depth is somewhat improved, but not yet what I'm used to. The decay of instrument tones is not natural seeming too slow resulting in feeling that the trailing sound is not totally from the instruments. This characteristic seems to be the same that is effecting image depth perception because the timbre and delayed reflections in the recording environment is mixed with this.
Hmm. I'm beginning to think that the snubber is not going to do much because the circuit is floating from the power ground. Need to think more about this.
Just going to let it run this week and come back on it next week.
Grounding Q
Hi,
On the versions of this circuit that were posted by Mauro Penasa, the ground for the LM3886 and the speakers is returned to the power ground, which seems OK to me.
On latter versions of the Rev C variant, I noticed that the LM3886 and speaker grounds are returned to the signal ground ("IGND").
Which is "more correct"?
Thanks,
Rob
Hi,
On the versions of this circuit that were posted by Mauro Penasa, the ground for the LM3886 and the speakers is returned to the power ground, which seems OK to me.
On latter versions of the Rev C variant, I noticed that the LM3886 and speaker grounds are returned to the signal ground ("IGND").
Which is "more correct"?
Thanks,
Rob