Hmm... Ok. I tried switching out the film cap next to the lm3886 with a higher value, but it just raised the cutoff (no bass). Then I took it out completely, and my cutoff is around 80Hz now. I think I'll try removing and jumping the 680pf cap at the input, next. I'm really just guessing at this point.
Update: no change after removing the 680pf caps.
Update: no change after removing the 680pf caps.
Last edited:
If schematic given matches actual amp pcb, then it is obvious from amp schematic why it does't work...negative feedback loop is not closed.
Quick fix would be to connect bottom end of 1k resistor (the one between 1k24 and 31k6 resistor on the left side of servo opamp OP07CP) to inverting input of LM3886TF. Capacitor 6n8 (little above) should be like 68uF/10V (preferable 2-3 times more) non-polar electrolytic in order to have meaningful low frequency and low DC offset.
Quick fix would be to connect bottom end of 1k resistor (the one between 1k24 and 31k6 resistor on the left side of servo opamp OP07CP) to inverting input of LM3886TF. Capacitor 6n8 (little above) should be like 68uF/10V (preferable 2-3 times more) non-polar electrolytic in order to have meaningful low frequency and low DC offset.
Last edited:
As it turns out, I'm an idiot. The amp is working. The reason I can't here the bass is because of where the speakers are placed. I moved one to the corner and, wahlah, bass!. Anyway, on one of the boards I already have the 6nf cap removed so should I put in a .68uf wima, .22uf wima, 68nf wima, or just the original cap (6n8 china)? The wima's I have laying around. Also, I switched out the ne5534dd's for lme49710na's. Should I still bother with the short that you mentioned?
I guess not since the amp is working...but I would check schematic against actual pcb just to be sure.
You don't need any cap in gain setting leg from LM3886 inverting input, since LME49710 is used as buffer have much lower intrinsically Vos then NE5534 and output DC offset is handled by servo opamp (use one with J-fet input).
The DC offset is ofcourse handled by the DC servo but if you got a DC offset in the input, the signal will still be off-set, making a larger excursion to one side. Since the driver(s) act the same like with a DC offset and having a stronger spring rate and stiffness of the spider etc (pre-tension), the nonlinear distortion will rise quite noticable.
Last edited:
Well, I did some thinking and I'm sure that there was indeed something wrong before. For one, I had connected an oscilloscope to the output and saw a mess coming through around 600-800Hz and 1000Hz, at different times. So, now I'm wondering if some of the components needed to burn in.
ICG, I noticed you mentioned "larger excursion" and "stronger spring rate". When I first connected the amp I noticed the bass would cause the woofer to exhibit an echo-like effect. Basically, there would be bass (which was not really seen or heard), and then the woofer would take 2, or so, deep breaths. Not sure if it's related.
The LF411, is that a servo op amp with a j-fet input? Would that be the best pairing with the LME49710?
ICG, I noticed you mentioned "larger excursion" and "stronger spring rate". When I first connected the amp I noticed the bass would cause the woofer to exhibit an echo-like effect. Basically, there would be bass (which was not really seen or heard), and then the woofer would take 2, or so, deep breaths. Not sure if it's related.
The LF411, is that a servo op amp with a j-fet input? Would that be the best pairing with the LME49710?
What exactly is a DC offset at the input, and how can I measure it? My output is going to be a minidsp 2x4 hd. I already know that EVERYTHING comes through from my computer, so I'm going to try connecting to the same ground this time. All of the noise (except the Ethernet cable noise) is eliminated when I use fiber optic, but I'd rather use the USB cable (sounds better).
Just got done testing the boards, bi-amped with my speakers. Incredible amounts of noise coming through, even with the fiber optic cable. When I play a low frequency signal, like 20Hz, there is a lot of buzz coming through as well. I guess I'll try switching to a better power supply on the minidsp, since it came with a wall-wart.
Switching the power supply helped reduce some of the noise with the spdif cable, but the USB is still unusable. Also, the low frequency distortion sounds like oscillations. Starting at about 80Hz they are slow, but as work my way down to 20Hz they speed up. Any ideas of how to get rid of this?
It's not the burn in question, it's the measurement or amp design or a bit of both maybe, I would say.
It could be a symptom of "bass pumping"...if it is DC servo design is one to blame.
J-fet input LF411 opamp will be fine.
Grab your DVM, set it VDC (less then 1V scale) and measure voltage between +(hot) and -(ground) at source RCA connectors (left and right channel)...you want to see 0mV at your source output if you DC couple you amp to source. Otherwise, any DC offset at source output will be amplified by amp gain. Eg, if your source have +/-1mV dc offset and you amp is dc coupled and has gain of 21, you will have +/-21mV dc offset at speaker output connectors.
Stay away (for now) from USB connection, use fiber optic cable 'cos it will galvanically isolate your source ground from your amp ground and prevent ground loops in this early stage of troubleshooting.
While the dsp is not connected the voltage is 0, with and without a signal. However, once I connect it to the amps the one used for low frequency drivers shows 0.3mV, while the other is ~0.1mV. Is there any way to lower the gain on the amp. I've already tried reducing the output of the dsp, but it doesn't help.
I tend to think that one should not make assumptions that the source is DC free, but I think that is what you meant to say.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.