there is still a bit of noise too when i do that but less. as i said in another reply, all units throw a bit of noise, that i try to reduce at this point.
-the cross over alone has a bit of noise
-when the preamp is added, even with volume pot at 0 there is a bit more noise. id say each has 50/50 of the overall final hiss noise.
-the cross over alone has a bit of noise
-when the preamp is added, even with volume pot at 0 there is a bit more noise. id say each has 50/50 of the overall final hiss noise.
Either the LF or the HF should be at max. Otherwise, there's some sacrifice of available SNR.
One other subtle issue. The crossover seems to have decade capacitor options. Try to use the maximum workable selections and consequently minimum R20, R21 settings.
lf and hf pots are at max (lf beign not important anyway since hiss comes from HF)
you mean for the capacitor, to try using one single cap that would summarize the whole ones (C3-6, c7-10 ) ? the decade option is to use X1 or X10 xover freq, f is set at around 250hz for now because of the actual set-up. So J2 and J3 goes to switch and they are closed. so all caps are in the loop.
you mean for the capacitor, to try using one single cap that would summarize the whole ones (C3-6, c7-10 ) ? the decade option is to use X1 or X10 xover freq, f is set at around 250hz for now because of the actual set-up. So J2 and J3 goes to switch and they are closed. so all caps are in the loop.
What’s your decoupling like on the opamps?
If you don’t have a small cap across the power, directly on the pins or as close as possible, that may be a possible cause.
If you don’t have a small cap across the power, directly on the pins or as close as possible, that may be a possible cause.
its all decoupled, all ic has 0,1 right next to the +v/-v, main supply is 2200uf/7815-7915 and 470uf on both seperate boards and all 0,1uf decoupled.
The noise of this preamplifier is probably dominated by the thermal noise of its potmeters. Replacing the op-amp will not help, and it might make the circuit instable.
As a review of perspective/expectations, were you content with hiss level when using conventional speakers without active crossover? Were you using the same preamp and power amps before? Do you have any instrument that would measure noise amplitude? If a subsequent stage has 10dB lower noise re the driving source, it contributes only about 1dB to total output noise power. And of course the ear is logarithmic--- 10dB sounds about twice as loud. Point being, don't make performance goal harder than is needed.
Things that might help the crossover:
Try making R12 and R19 = 0 ohms. They aren't necessary and R19 thermal noise adds directly to output with unrestricted audio bandwidth.
All crossover opamps are LM4562? Make R4, R6, R22 = 1k rather than 10K. Similar scaling of R5, R13, R15, R16 will help.
Things that might help the crossover:
Try making R12 and R19 = 0 ohms. They aren't necessary and R19 thermal noise adds directly to output with unrestricted audio bandwidth.
All crossover opamps are LM4562? Make R4, R6, R22 = 1k rather than 10K. Similar scaling of R5, R13, R15, R16 will help.
i will focus firs on the cross over noise issue since there is a fair amount of it even when input is shorted.
i ordered few opamps with lower noise level and some ic socket and will test. rememeber this forum is DIY and do experiements and try to learn by my expensives mistakes.
i ordered few opamps with lower noise level and some ic socket and will test. rememeber this forum is DIY and do experiements and try to learn by my expensives mistakes.
noted
it was ok before i went biamped, only the nikko preamp and a BGW 7000 proline to custom made 4351 jbl alike speakers. i decided to go biamped just to mess it all apparently....
yes all opamps are LM4562
isnt R19 only affect LF output ?
can i change first only the 1k=0, and r4-6-22 down to 1k ? or i need to change at the same time 5-13-15-16 ? just because i dont have theses values on hand, well i have cheap carbon ones, i would prefer to change them with metal film
also, for now C11 is omited, it is actually a short. i felt it was not necessary. R7, R23 and R24 pots are 100k. should i reduce them down to 10k too ?
1k -> 0 can be done now, no need for simultaneous changes, but R4,6,22 should be changed is unison so that your eq. experiments remain unchanged. The actual value isn't critical, but should be 1% metal film for low noise, and small re original 10K. The R4,6,22 may have a bigger impact than the 0 ohm changes, but all should help the issue. Can your RTA be used to quantify noise from amp?
C11 = 0 ohm is fine. Reducing 100k pots to 10k or even 5k may help. Source impedance is helped helped significantly: a 10k pot at center sees 5k from wiper to ground and 5k wiper to opamp, so 2.5k in parallel.
C11 = 0 ohm is fine. Reducing 100k pots to 10k or even 5k may help. Source impedance is helped helped significantly: a 10k pot at center sees 5k from wiper to ground and 5k wiper to opamp, so 2.5k in parallel.
perfect
i just want to go pregressively so i will be able to tell which component/area was causing the utmost amount of noise
thanks very much for your help BSST
i just want to go pregressively so i will be able to tell which component/area was causing the utmost amount of noise
thanks very much for your help BSST
as i am doing tests, i have my scope always connected to see whats going on...
in normal operation, ie, shorted to test noise or with audio its all ok, but when i touch the input i would expect some 60hz beign induced...i can see some HF attached to the 60hz....there isnt any components that limit bandwidht in there....would it be adviseable to put some 10-30pf caps in the negative feedback at R22 and R26 ? its been noted on other threads about the LM4562 beign sensitive to HF instalbility....
to be continued....
in normal operation, ie, shorted to test noise or with audio its all ok, but when i touch the input i would expect some 60hz beign induced...i can see some HF attached to the 60hz....there isnt any components that limit bandwidht in there....would it be adviseable to put some 10-30pf caps in the negative feedback at R22 and R26 ? its been noted on other threads about the LM4562 beign sensitive to HF instalbility....
to be continued....
Make sure your opamp is not oscillating at HF. Oscillation well above the audio band manifests audibly as a 'shhh' sound - ie like white noise. Check also that your scope is not loading the measurement node and killing the oscillation - you can stick a 100-220 Ohm resistor in series with the probel to isolate any probe capacitance.
actually i dont think it self oscillate. i do most of the testing on my work bench with a dedicated power amp and headphones so i can hear and see/measure any changes in the level of hiss. when i probe with my scope here and there the hiss doesnt change at all it remains "constant" in the way i hear it in my real setup. my GF says she hear it too but it is accetable, an unwarned person wouldnt notice this much i think but its there for sure, just like when you use the PHONO stage of any comercial amp and you turn up the volume with no disc, but probably less.
I did all what was suggested, ie, reduce resistor values. it didnt really changed anything. tonight i swaped many many opamps, ne5532, tl062, rc4580, opa2134 NJM2043D and lm4562. i found out that either digikey sold me knock offs or i have a bad batch of LM4562. 4 out of the 8 i ordered gives signs of random pops and strange "out of universe" noise into my headphones once put the LM's to the task at U3.
The final stage follower U3 of my circuit seems to suffer the most. once swaped with cheap NE5532 the noise level reduced to an acceptable level. one thing i dont understand is when the hf level output R24 (100k pot) is at its mid position, the noise is at its upmost level (with the LM4562 only). once swaped with 5532 or opa2134 the noise is much less overall and follow R24 level/setting. since the hf level on my sytem needs to be lower than LF the noise with the 5532 is much less then....go figure. later this week i will put sockets too at U1 and U2 and swap/test opamps combinations to see if the LM's are the problems.
It gets better, but i am worried about what i have in hands.
I did all what was suggested, ie, reduce resistor values. it didnt really changed anything. tonight i swaped many many opamps, ne5532, tl062, rc4580, opa2134 NJM2043D and lm4562. i found out that either digikey sold me knock offs or i have a bad batch of LM4562. 4 out of the 8 i ordered gives signs of random pops and strange "out of universe" noise into my headphones once put the LM's to the task at U3.
The final stage follower U3 of my circuit seems to suffer the most. once swaped with cheap NE5532 the noise level reduced to an acceptable level. one thing i dont understand is when the hf level output R24 (100k pot) is at its mid position, the noise is at its upmost level (with the LM4562 only). once swaped with 5532 or opa2134 the noise is much less overall and follow R24 level/setting. since the hf level on my sytem needs to be lower than LF the noise with the 5532 is much less then....go figure. later this week i will put sockets too at U1 and U2 and swap/test opamps combinations to see if the LM's are the problems.
It gets better, but i am worried about what i have in hands.