HiOh hi, I don't know how I could miss your post, I apologize. In the meanwhile I replaced all resistors directly in the signal path with naked z vishays, the 1uf coupling caps with mundorf silver gold and the volume pot with a stereo khozmo attenuator. I still find the midrange and highs somewhat annoying, compared to my all-tube amplifiers and I think soundstage consistency could be better, which I always felt with SS amplifiers. The annoying quality of the highs and midrange is hard to describe accurately, obviously. I have gotten used to it, but it is still very much there. I appreciate the ease with which it controls the bass and its amount, again compared to my anemic flea watters.
I received the new parts today from partsconnexion and I hope to operate on the feedback circuit tomorrow, then i will let you know how it went and how I think it sounds, if i get to that stage
I replaced all rectifying diodes with FR shottkys and the 100uf caps in the tube stage with large film clarity caps. I got rid of the .1uf cap in parallel with the last 100uf cap and I believe I should also lose the .01 caps straddling the rectifying diodes, would you agree?
I am powering the output stage with an entirely separate power supply unit that I had built for a pass lab F4= improvement in smoothness and tightening of bass + visibly better dynamics.
my impression of LM 1875 is that if it is well implemented, the mid and high are very smooth , even when compared to my tube amp.
l usually add all the stability parts as described by Tom in the forum before making the final judgement. These are Thiele network, R+C across the feedback resistor and 221 cap from IN+and In- inputs .I also experiment with placement of loudspeaker and signal grounds.
Regards
Kp93300
Hi, I'll have to read up on those stability parts, I do not know what they are. I noticed there is an unpopulated position for a 220pF cap on each channel on the output board, is that the 221 cap you are referring to? I find the grit has gon from the sound after a change of intersatge coupling caps. I am using .47UF cutf vcaps in parallel with duelund cast copper, same value, to achieve 1uF. I did not expect much, yet they sound excellent. I am awaiting some V-cap ODAM 1uF caps in the mail. I reverted to the original negative feedback configuration, which to me sounds best. I have another plan for the amplifier. I am intent on replacing some of the dropping resistors in the input power supply unit with VR tubes. Here is what I read this morning:
Under load:
249.1 ac from PT
335 dc rectified
273.5 after c1
212.9 after c2 (B+)
122v total drop
61.5v first drop after 1st 33k
60.6v second drop after second 33k resistor.
The VR tube has to be in series with a resistor, 1k 2w was suggested to me by Steve Deckert. But I am not savvy enough to understand where to place the VR tube. I could replace the second 33k resistor with an 0A2, which drops 75v and that wold also lower B+ a bit. Would you know what the B+ measures under load? I wrote to howard wang from jolida and am waiting for his reply on B+ voltage too.
Obtenir Outlook pour Android
Under load:
249.1 ac from PT
335 dc rectified
273.5 after c1
212.9 after c2 (B+)
122v total drop
61.5v first drop after 1st 33k
60.6v second drop after second 33k resistor.
The VR tube has to be in series with a resistor, 1k 2w was suggested to me by Steve Deckert. But I am not savvy enough to understand where to place the VR tube. I could replace the second 33k resistor with an 0A2, which drops 75v and that wold also lower B+ a bit. Would you know what the B+ measures under load? I wrote to howard wang from jolida and am waiting for his reply on B+ voltage too.
Obtenir Outlook pour Android
Last edited:
Hello,
I have a question about my Jolida 1301. The cap C10 (100uf/16v) seems to be reversed on one of the channels. The marking on the PCB itself is reversed and on one channel the positive + side of the cap goes to ground and on the other channel the negative side does. The other side is connected to the R13 (1k) resistor in both cases. What is the correct polarity? Shouldn’t they both be identical, since everything else is perfectly symmetrical?
Thank you in advance
I have a question about my Jolida 1301. The cap C10 (100uf/16v) seems to be reversed on one of the channels. The marking on the PCB itself is reversed and on one channel the positive + side of the cap goes to ground and on the other channel the negative side does. The other side is connected to the R13 (1k) resistor in both cases. What is the correct polarity? Shouldn’t they both be identical, since everything else is perfectly symmetrical?
Thank you in advance
Thanks for the reply. I wrote to jolida and keep my fingers crossed, as they don t always answer. I have no idea what would warrant that discrepancy btw channels. I am building two mono blocks based on the 1301, with separate transformers for each stage and i am stumped by it and don't know what to do
The easy way is to measure across the cap; it should have a very slight charge on it. Is the polarity the same for both channels? If it's absolute zero, then the mystery remains. You could measure across the output of each channel as well; at DC there's no voltage divider drop in the feedback connection, so the LM1875 output voltage should be the same as the cap voltage. If there's some current flow from the non-inverting input, of course that would have its influence.
An opinion may have it that the cap colors the sound of the amp, by being a passive element in the feedback loop. The easy way to get the discrepancy addressed - before going into a mobius-loop over it - would be to replace the cap with something both "better" and non-polarized, like a poly cap. 10 oof shouldnt be hard to find / do, except for the physical size difference.
Or, if you're really a dare devil, short the cap. The idle DC offset at the amp output will be much bigger. However if it still happens to be a small fraction of a volt, so what if your speakers have a little DC running through it. - as long as it doesnt move the cone a perceptible amount. I think I did that and my amp version was just fine. Understanding why the cap is there (infinite gain at DC), what a manufacturing trade off is versus effect on a particular instance, I felt it was OK to simply get that cap outta there, regarding signal its path influence.
An opinion may have it that the cap colors the sound of the amp, by being a passive element in the feedback loop. The easy way to get the discrepancy addressed - before going into a mobius-loop over it - would be to replace the cap with something both "better" and non-polarized, like a poly cap. 10 oof shouldnt be hard to find / do, except for the physical size difference.
Or, if you're really a dare devil, short the cap. The idle DC offset at the amp output will be much bigger. However if it still happens to be a small fraction of a volt, so what if your speakers have a little DC running through it. - as long as it doesnt move the cone a perceptible amount. I think I did that and my amp version was just fine. Understanding why the cap is there (infinite gain at DC), what a manufacturing trade off is versus effect on a particular instance, I felt it was OK to simply get that cap outta there, regarding signal its path influence.
Last edited:
Thanks. It is going to be hard to measure in such a tiny space. If the image displays properly, you can see the c10 cap with its + side to ground (on the right side). The lowest hanging cap on the left side is its counterpart, with the negative going to ground. I tend to think it s a mistake, but I hear no difference between the channels.
Last edited:
C15 is reversed because it's connected to a negative voltage rail and ground. C10 actually has AC across it, as it's connected to the amplifier output, which can swing both + and - voltages. So it really doesnt matter which way it's installed. Lousy place they put it, for servicing; had to take the whole amp apart to put it back, after replacing with a wire, before I sold mine.
I'd recommend replacing it with a better 10 uF cap. Or just shorting it with a wire; that's easy to take off if the sale day ever comes.
I'd recommend replacing it with a better 10 uF cap. Or just shorting it with a wire; that's easy to take off if the sale day ever comes.
I don t know why that made me chuckle, maybe the relief of knowing i ll be fine either way. I ll surely replace it with nothing, as i am building two monoblocks from scratch based on the circuit. I hope that 100uf cap isn t an important safety feature. I don t know enough to answer my question. Thank you all for your input.
It exists to ensure the quiescent output voltage is zero, by making the feedback network's attenuation go to zero at DC. This leaves the full open loop gain of that LM1875 available to ensure the output voltage doest deviate from ground. It's a guaranteed way to ensure all instances of this chip behave the same in that aspect, an important one to many consumers.
There's a couple of things going on;
1. The cap versus wire debate, regarding SQ.
2. Do we really want to pull the rubberband that tight, regarding open-loop gain levels at DC? Maybe it "sounds better" if it doesnt have to move quite that much in gain, between the DC and audio frequencies? Should there not be "the right" gain at DC - versus infinity? So make the DC feedback the same as it is for AC and see if the output voltage offsets are reasonable in your amplifier.
On the off chance they're not, one could adjust for that by instead adding resistance across that 100uF cap. That will "cap" the gain of the amplifier at DC to a set level, which could still be many times more than the AC gain - just enough to bring in any DC offset at the amplifier output.
Very interesting again, thank you. I purchased audio note kaisei caps for C10, but now I want to try the wire, that is the 1k resistor to ground, bypassing c10. Pragmatically speaking I just need to bypass c10 and measure Dc offset at the speaker terminals? And aim for what kind of voltage, when is it too much? I have sensitive fullrange speakers, probably 98dB, if that matters. In terms of SQ, what can I listen for? I already reduced the bias voltage a few weeks ago and did not like the sound (the low end was flabby and boomy across a certain range)
Yes. Unsure about the DC voltage necessary to make your 98db FR speaker move off center. I'd think that would require a caliper and a variable DC source. Venturing a guess, maybe 10 - 20 mV is fine, 50 - 100 maybe biases the cone position significantly.
Being open minded about what effects SQ, I see no reason why a speaker couldnt have a little "class A" current running through it. Obviously one wouldnt want to suck it in all the way to its stops as that being ridiculous, but I said a little. Perhaps just enough to move the VC one or two winding "turns" (wire diameters) in/out in displacement doesnt effect the speaker's operation much, while allowing low level AC signals to be wrapped up within whatever DC bias it takes to do that. Change in SQ? I'm sure very few could stomach the idea enough to actually try it and see...