Good Preamp for myref rev.c lm3886

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hi all audio gurus,
First of all...i am a newbie of newbies.
Recently im constructing audio amplifier (stereo) using lm3886, from mauro's myref rev.c.

I was thinking of adding preamp with tone control for it (or should i??)
After searching sources on net, i came up with one that was posted in ESP - Hi-Fi Preamplifier , which suggest using ne5532, opa2134, TL072. Schematics seems simple and details are easy to understand (for me..😀). Compare to sooo many sources in the net.

On the other hand, I also noticed that lme49720 has got a lot of amazing reviews from many audio forums.
And some posts are saying ...that 49720 are eol, so im thinking of getting some while they are i still in the list my local onlone shop (Bali).
Soo
1. can i just replace the suggested opamps (by ESP) with LME 49720NA dip8 with the same schematic (and the components listed)?



Cheers
 
NE5532 will work absolutely fine, but the decoupling caps are missing, not something you can afford to risk with NE5532's, 100nF between + and - rails on each chip will ensure HF stability.


I'd suggest adding RF suppression capacitors on the inputs, you don't want to pick up
mobile phone signals.
 
You put the mixer in a steel box. Safety ground the steel box to the 3rd pin of the power plug. Isolate RCA jack rings from the steel box with o-rings. Connect RCA jack centers to 33 pf disk cap to analog ground.
Decouple RF coming in power lines with a 11 turn on a toroid choke. Salvaged mine from a dead PCAT power supply. Also reuse the blue MOS suppressor on the AC input to reduce pops induced by lightning.
I moved power transformer out of the mixer box to reduce hum. A second steel box inside the main mixer steel box can also be used to suppress steel wrapped transformers.
LM49720 has higher slew rate than 5532 and is more prone to oscillation. Layout of PCB becomes critidal. Start with 5532 or at lower gains (under 20) BA4560. TL072 and 4558 are hissy at higher gains. MC33078 and NJM2068 don't have the hiss, but the higher slew rates require a 33 pf cap bypassing the feedback resistor to cut gain at high frequencies. Else they oscillate.
 
Last edited:
Indianajo...thank you,
3rd pin of power ground to metal body,
33pf between RCA center jack to Ground
11 turns of power line on toroid choke, do i still need to do as Mark said ? "RF suppression capacitors on the inputs"..i thought what he meant was the audio input, or is it actually the same idea? ..blocking RF
Seperate the transformers to another box, can i just place a steel divider that reach bottom to top of the metal box, between transformers (i'll be using multi) and signal?, but still in same box.
btw...appreciate the helps..(this is my first thread in this forum).. :)
 
NE5532 will work absolutely fine, but the decoupling caps are missing, not something you can afford to risk with NE5532's, 100nF between + and - rails on each chip will ensure HF stability.


I'd suggest adding RF suppression capacitors on the inputs, you don't want to pick up
mobile phone signals.

I think all of this is covered in the ESP design, figure 5 shows the schematic for the opamp bypass components and RF suppression.
 
Yes, a bulkhead partition can turn your steel box into two boxes. The fasterners have to keep it well bonded to the safety grounded case. One side EMF dirty: power switch, transformer, RF supression inductor, main filter caps & regulator.
Other side: op amps inputs. output, amps, volume pots, local filter caps.
Signal Inputs are shorted by ceramic caps to analog ground. The center of the power supply regulators (+-15?) . Analog ground is separated from case ground, often by a 10 ohm or 1 ohm resistor, sometimes series an inductor. Sometimes both parallel a type Y cap. I salvage Y caps from PCAT supplies also. Rated 250 vac or more, that is how you know it is a Y cap.
Ron E lives in Australia, he may not have the RF rich enviroment I live in. Without careful control (and my RA-88a Herald mixer was extremely sloppy) I get the lamp dimmer howl, police radios, CB radios, dogs barking the CSA anthem Dixie, an AM radio station doing sports talk. With the added controls, that $15 mixer sounds great, dead silent, except for lightning pops. That mixer Also upgraded to 33078 op amps and the transformer/switch moved out of the case.
 
Last edited:
BY The Way

Mr. Neal..i guess it is..i'll stick on ESP schematic, but THANKS anyway mr Mark.
Mr Indijo, much much clearer...analog (power)><digital (audio path)

BY THE WAY GUYS...here is the story,
I'm was planning (opened a thread) on the preamp, so that i have ideas for nest step while i'm building the amp, so last night is done ( i thought), so when i powered up..and there is no sound came out (i touched the 3.5 mm audio jack, connected to the amp), and the led is off ( waited for about 10 seconds).
BUT..when i unplugged the main AC power (no switch), in about 6 seconds, led is on for about a second and slowly dimmed off, sound is speakers to..just a soft high tone.
then i touched all components, none of them is hot ( i was worried)
what could go wrong..Gurus...?
i have checked the input power, and it was 20V from each of the phase (before plugging to the pcb spade)
the short lighting led (with sound on speaker) is my main curiosity.

I used 2 toroids in series (singles to make center tap),Attached is my arrangement layout, pls kindly observe..i maybe wrong in some ...i don't know. :)




toroid transformer.jpg

toroid in series.jpg

full pcb layout.jpg

pcb AC input.jpg
 
DVM is a useful device for determining where the problem is. Like, LM3886 usually operate on about +-30 vdc, is that what you have?
As far as the high tone, the LM3886 amps I built motorboated inexorably both channels, with the wiring looking just like the schematic. But, my layout did not look like the picture on the datasheet application. I used point to point wiring. So I gave up on that and repaired a PV-4C instead. LM3886 are prone to oscillate. The boards you bought may be okay on layout, maybe not.
Buy alligator clip leads on same order as DVM save freight. Warning, Pomona and Fluke alligator clip leads are stupidly expensive, too insulated to actually take measurements on anything, and rated to 630 vac. I'm talking the kind of clip leads I used to buy at radio shack, recently expunged from the catalog by newark when they closed mcmelectronics. 12 for $2 alligator clipleads. If you use 2 hands to take measurements you can kill yourself with 60 v from one hand to the other across your heart. (any voltage over 25) Put clip lead on analog ground and measure with 1 hand. No jewelry on hands or neck, 1 v at 1 amp can burn your flesh to charcoal through metal.
Prototype amps and those under repair are usually energized with a 100 W incandescent light bulb in series with the power transformer, usually enclosed in a grounded metal box with a circuit breaker. Mine has an AC cord and an AC socket. The CB will turn it off which is handy. Since incandescent 100W bulbs disappeared from store shelves some people use 1 cup tea boilers, although those are not for sale in stores in the US.
 
Last edited:
Founder of XSA-Labs
Joined 2012
Paid Member
752468d1556334288-preamp-myref-rev-lm3886-toroid-series-jpg


It looks like you have your two trafo output windings paralleled?

These are mono blocks and I assume the built in diode bridge and caps are expecting a center tap configuration (or dual secondaries in series) for AC/GND/AC input from the trafo. Your photo shows AC/AC and no center GND.

Is the amp making dual rail +/-?

If you get a DMM - a good brand like Fluke ($70 minimum for model 101) will be rated Cat II or Cat III and so will it’s probes. Decent Cat II (600v) alligator leads are a good idea. Look for the Cat II certification if you plan on measuring mains voltages with it.

But other brands may work as long as certified for Cat II. $5 drugstore DMMs can work but probably not certified for mains voltages and will break in short order. The Flukes last forever.
 
My sears 82140 DVM is 14 years old. They are out of the tools business but a competent $35 meter came from newark/farnell, tenma 72-7725 (I test both @ home and at a volunteer job). Also parts-express has a house brand Mastech whose C meter proved competent. The $2 harbor freight DVM at work lied to me & we spent 3 hours disassembling & reassembling a wind chest for no reason but a *****y meter.
When I use 50v rated alligator clips on mains connections I don't touch them with the power on. All that insulation around the 630 v rated alligator clips makes them a PIT* to use. Actually they are useless except on AC switchgear which have huge distances between everything to prevent arcover. I use pamona grabber clip with banana plug on leaded components but they are too small to go around transformer connections or wiring over 20 ga in diameter.
 
Last edited:
Xrk971...
I wire the toroid output in series, so primary1 start, as the +, primary 1end & primary 2 start as the Gnd/0, primary 2 end as the -.
Already measure it 2 x 20V in each rails and 40V across.
Today, there is improvement, when i turned on/powered the amp, speaker makes a hum(constant), and the cone is pushed up (not full).. Played music from my phone..nothing, and when unplugged cone is down to normal, and heard the relay "klik".
What to do next..
 
Cone pushed out is a sign of DC coming from the output transistors. Do check your speaker jack to speaker return for <0.20 V.
If so report back. You may get more help on LM3886 on chipamp forum. 3886 is supposed to limit it's own DC out current.
Handle the hum by shorting the input of the amplifier section, usually at the RCA jack in. IF still hums you've got wired wrong or too much ac on DC busses. Measure AC on the +-20 points to see if less than 0.1 vac. AC scale of cheap dvm is okay to measure for 50-60 hz. Cheap DVM AC scale lie a lot on music.
 
Last edited:
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.