NE5532 Buffer & Pre-amp. Full test & analysis.

[@MichaelBeeny]

PLEASE READ THIS. The capacitor between pins 4 and 8 needs to be around 1 uF, not 1 nF, or in other words 1000nF.

[@olebjrklund8648]

Yet another great video Michael, Thank you and have a great week and coming weekend.

[@ryanchappell5962]

This video was posted 6 days ago, how did I miss that? Great video as always Michael!

[@ronschauer839]

Nice video..! We have been using one of these boards for a long time now (with a genuine LM4562 chip and an ALPS RK27 50K volume pot). It works extremely well and is part of an A/V amp that is used for hours daily. A couple of things are worth mentioning- Mine was received with both voltages adjusted uncomfortably high so I had to adjust them. An older production batch perhaps? To use an LM4562 or LME49720, the capacitor across pins 4 and 8 should be on the order of at least 100nF, if not 1uF, **not 1nF** (sorry...). And I am powering mine from the same +/- DC rails that power the power amp boards. Just connect DC common to the center power input pin, and the plus and minus DC source voltages (separately of course) to either of the outer pins. The on-board diodes will take care of routing the appropriate polarity to the right place, no problem. 😀 Very nice..!

[@sitaram.electronics]

Very nice and very useful video sir thank 🎉🎉🎉❤❤❤

[@cobar5342]

So Michael, after your glowing report on the Accuphase 3850 clone, how would you compare the two now?

[@MichaelBeeny]

When you get to a certain level in performance, the differences in both measurements and any audible differences are very small indeed. I'm using both at present, personally I would be happy with either.

[@TerryClarkAccordioncrazy]

​@MichaelBeeny These differences are small indeed, I certainly can't hear them. I wonder if we cascaded 10 or 100 stages of NE5532s whether it would be audible. noticeable.

[@MichaelBeeny]

@@TerryClarkAccordioncrazy If you do a search Terry, you will find circuits using 30+ 5532s, not to get better sound but to turn them into a power amplifier. 15 watts if I recall. In the next few weeks, I will be testing to see if I can measure and more importantly, can I hear any difference between 1,2 or 3 NE5532s in parallel suitably isolated of course. NEVER just put them on top of each other. We shall see, truth or snake oil?

[@TerryClarkAccordioncrazy]

@MichaelBeeny Those are parallel designs but I meant put many unity gain opamp stages in series purely to see at what point they become audible. I'm guessing the noise would be noticed before the distortion but who knows? Some claim to hear differences between opamps. I can't.

[@MichaelBeeny]

@@TerryClarkAccordioncrazy I guess noise would become noticeable. I think probably half the world has fake 5532 in use. If you have no way of testing them or comparison you will never know. Fakes are not another chip maker making a copy but usually simply an older, poor spec similar IC.

[@PoPanzPoPanz]

Long time ago we compared and listen to NE5534, TL071, TL081, TDA1034, LM49720(?)... No difference heared.

[@MichaelBeeny]

I have listened to many ics over the years. Most modem ICs are very good indeed. There would be no point in bringing a new IC to market if it was not that good. I simply would not sell. I have concluded that very small differences do exist and can often be detected after many hours of listening. But which of these differences are better?? Some of these discreet ICs cost silly money but contain more snake oil that fact.

[@robertjianu7434]

greetings from Romania!!! regarding an older video of yours about the NE5532 preamplifier, with that L7812 voltage regulator...I was wondering if I can change to L7815 without causing damage. I mention that my kit is identical to the one presented by you in the video

[@utenatenjou2139]

Hi, thank for the post, about LM337/317T, I heard about the ripple performance similar to what you had mention. Another, (may be hearsay) is 33T (the negative)'s performance is less that the positive one, but I don't know if it true or not. I still use the 337T anyway. Though I used the variable "T" on basic of conveniences, from where I came from the cost between the fix and variable one is not that much. Again thank, the board look nice, I'll order one.

[@MichaelBeeny]

The 5532 has a very good ripple rejection. Noise on the supply rails can be quite poor offering no performance changes at all. Out of the box we have no audible hum whatsoever, it's completely silent.

[@ronschauer839]

If you look closely at the way they applied the regulators on this board, they have capacitors in their feedback loops. That causes the regulators to be even more stable and to reject source noise and ripple to a much higher degree. Very clever actually, and an excellent reason to use LM317/337 regulators rather than 7815/7915. This really is a nice preamp board.

[@peterlarkin762]

Opamps have good ripple rejection, but a very low noise supply does make a difference in other ways. I would highly recommend looking into Elvee's denoiser circuit for 317 regulators. A simple but brilliant addon. Also you can make dual polarity supply using two 317 instead of using a 337.

[@albanana683]

One indication of high frequency oscillation in an opamp is running at a noticeably hot temperature.

[@MichaelBeeny]

@@albanana683 Very true. Also measured as increased current draw.

[@ConorHanley]

How can you have any audiophile cred if you don't replace the NE5532 with a discreate op-amp like a Sparkos 🤔

[@MichaelBeeny]

Well Conor if you feel that spending that kind of money on such an op amp makes any kind of difference, I cannot argue. I respect your opinion. Clearly, I cannot test every such op amp available. I have tested 2, however. They are not rubbish in any way at all, but I have not found a single one that either measures any better or sounds any better. If any manufacturer would like to send one, I will of course comprehensively test and listen, and most impotently, compare. I'm really not biased against the units, but should I find one that is excellent and worth the money. I will sing its praises. One of the issues I have found is high current demand. Many circuits are not made with high current draws. Many of the class D circuits cannot cope with them. As for having any CRED? I leave that for my viewers to draw their own conclusions!!

[@ConorHanley]

@@MichaelBeeny It was a 'joke' , wouldn't spend the money.

[@MichaelBeeny]

@@ConorHanley I did not think that was really you Conor. I thought you had been taken over by the dark side lol Now you all know how gullible I am.

[@andymouse]

Awesome videos, I am still a big OPA1656 Fan Boy but it seems i'm the only one ! :)

[@JBBrown-hh4rc]

Hi!..I hadn't heard about this chip. Thanks for the info. I'll download the pdf and check it out.

[@andymouse]

@@JBBrown-hh4rc let me know your thoughts ! :)

[@JBBrown-hh4rc]

​​@@andymouseI ordered 3 of the opa1656 from Digikey. One to use with the C3850 clone and 2 to test with my A47 headphone amp.

[@JBBrown-hh4rc]

Sorry for the delay in my response as I have had some medical issues. I ordered the OPA1656 from Digikey and tried it in the Accuphase C-3850 clone preamp and conducted some listening tests. The one problem I found with this op amp is that there was little separation of instruments and the vocal. Everything appeared to be coming from a center axis which hindered the separation and detail of the music. I removed the 1uf NP capacitor across pins 4 and 8 and placed a 1uf NP capacitor from pin 4 to ground and pin 8 to ground. There appeared to be better separation but the audio had minimal dynamic range. It sounded as if the audio had been routed through a compressor/limiter. As my final choice, I used the OPA1612. As with the LM4562/LME49720, I did need to place a 1uf NP capacitor across pins 4 and 8, the supply sources, to prevent oscillations (motorboating). The OPA1612 costs between 6 and 8 dollars in comparison to the lesser cost of the NE5532 or LM4562. Also an adapter board is required for the OPA1656 or the OPA1612, additional cost.

[@andymouse]

@@JBBrown-hh4rc Oh Thanks for all your efforts ! this is great information :)

[@aronw3369]

Hello sir, i want to ask next to the opamp there is a blue capasitor connected to pin 4 and 8. I believe the value 0.1uf. So for the LME49720/ LM4562 do you still need to add 1uf capasitor?

[@rolfblomvagnes4514]

How does the sound of this little preamp compare to the previous "best preamp EVER?"

[@MichaelBeeny]

It's quite close. I think most people would be happy with either. I'm using this one at present as a preamp in my workshop.

[@ford1546]

Hello. In a NAIM NAP250 video you said that you want to modify it and play with the feedback circuit. Sorry if I remember wrongly. Can you midify it and make a video? Also try to make the frequency more flat so that it does not roll off so early in the bass and treble would also have been interesting. Can i get naim nap250 better?

[@MichaelBeeny]

It is a very simple circuit. Missing some of the circuitry that can make many amplifiers better. What it does not like is a vol pot on its input. Feed it from a preamp like this or the Accuphase one and it's sounds so much better. I think I also showed how to get rid of crossover, but you will need a scope and signal generator to do it. It is what it is, an OK amplifier for very little money.

[@aldogentile245]

Thanks for this review. I would use LM7815 and LM7915 directly. An what about the potentiomenter value, which potentiometer value do you recommend for this type of preamp. By the way, It would be nice to have the schematic of this circuit.

[@MichaelBeeny]

Why change the regulators? This type does in fact have a lower noise performance. I tested this with a 20K and 50K. Either will be fine. On paper the HF is more extended but after 100khz does anyone care? lol. The supplied pot tracks very well.

@@MichaelBeeny Hello Michael, May I also include to ask whether a fake/non-genuine 7912/7812 would affect the performance and quality of an OpAmp? I really hope you do response, I'm planning on checking out some transistors I have over here that regulates voltage linearly. Thanks

[@colombowala]

Hi, is the 1nF capacitor multilayer mylar type or ceramic? Using two 1nf , one each from +V and -V rail to ground (assuming that there is a ground track running thru the middle of opamp pins underneath ) would make any high speed opamp run more stable than just single 1 nf between rail to rail?

[@MichaelBeeny]

The capacitor between pins 4 and 8 needs to be around 1 uF, not 1 nF, or in other words 1000nF. Sorry an error in my video. A ceramic is fine and connect as recommended by TI. 2 caps are not needed, this does the job just fine.

[@colombowala]

@@MichaelBeeny thanks

[@jarlesagheim9710]

Hello. Could you give a link to purchase this?

[@MichaelBeeny]

The supplier I used is given in the video, but lots of retailers on Ali and E bay sell it.

[@nikiforosf4824]

There are several amplifiers having already 1-2 of them. Can we use it also for such products?

[@MichaelBeeny]

I'm not quite sure what you are asking me, please explain.

[@nikiforosf4824]

@@MichaelBeeny Thank you Michael. Some amplifiers like the TPA3255 for example, that you have also presented here, have 1-2 of these NE5532 on their board. So for these cases does it make sense to add this module as well?

[@MichaelBeeny]

@@nikiforosf4824 It's funny you should ask this question. Today I just happen to be testing that very combination. It does improve the sound quite abit, not sure why at this stage but it does. The gain is a little high, depending on what you will connect but lowering the gain will just be a resistor value change. It also gives you a volume control.

[@nikiforosf4824]

@@MichaelBeeny I have also tried a similar combination in the past but the pre-amplifier was introducing a lot of noise in the signal so i was a bit disappointed. I guess it depends on the amplifier that is tested every time and there is no general rule on the added value of such an equipment. But if it works properly, it can be interesting to add it.

[@peterlarkin762]

Why does my power amp weigh so much? Because it got buffer!

[@ford1546]

If you want a very simple preamplifier, this is an ok choice. But if you choose to go for one of the simplest preamplifiers. (p7 mini preamplifier) is a better choice, even if it doesn't have such an advanced power supply circuit, it has volt regulators and everything you need and takes up much less space. The preamplifier circuit itself is not that different between the two.

[@MichaelBeeny]

There are lots of small 5532 preamps on Ali, most are Prity much the same. The cheaper one does often have fake chips, test before testing.

[@ford1546]

@@MichaelBeeny Do you get anything other than a fake from China? Especially ic. chip

[@pioneeryum69v-yvy5]

Hello. What if ... Parallel ne5532??? . “If he literally just connected the legs of the chips, well, he was lucky. This op-amp survived that. But don’t rush to accept a simple explanation as a recipe for all occasions. I insist that doing so is not correct. The total current consumption should increase significantly. Connecting the outputs directly is analogous to a short circuit. Most likely, in this design the through current of the output stages of both op-amps has increased significantly. But not to a critical value - operational amplifiers usually have an output current limitation. But increasing the current through the output stage usually has a beneficial effect on quality. In general, there are prerequisites for an increase in quality. But the method is barbaric." kzbin.info/www/bejne/oX-niWqJeLNpgMU

[@MichaelBeeny]

NEVER NEVER just put 2 op amps together simply on top of each other. They will get very hot because it makes the chips oscillate and go unstable. You can parallel them, but they do need isolation resistors. Getting hot should be a warning that something is very wrong, think about it!! It could also damage the amplifier by drawing more current than it was designed to do. An average op amp will draw 10mA, some of these so-called discrete chips draw over 50mA. While these all share the same pin outputs, different chips need different compensation components to keep them stable and get the best from them.

[@pioneeryum69v-yvy5]

@@MichaelBeeny Ok, I trust you're experince & wisdom. I suspected that a simple parallelism cannot be such a simple panacea, but I wanted an informed opinion. Thank you again.

[@MichaelBeeny]

@@pioneeryum69v-yvy5 I might look into what is exactly required to put 2 in parallel in a video soon. It will double the current capability, but that will not really reflect on the sound. We shall see if it really makes any measurable or audible difference or is it just snake oil.

[@pioneeryum69v-yvy5]

@@MichaelBeeny I'm looking forward to it. I will accept any conclusion. With transistors, the background noise seems to decrease in parallel configuration, ... a similar effect with op-amps .... would still be a bonus

[@TerryClarkAccordioncrazy]

If you put 2 opamps on top of each other then they will fight. Each will have slightly different input offsets so they will drive their outputs to different levels and the only thing limiting the output current will be the output impedance. Even if they survive it doesn't mean it was a good idea.

[@stevenbliss989]

"...it shouldn't do anything else" ...unless it is a deliberate distorter like the GARBAGE valve (tube in Yankees land) op-amps all over many seller shops. :) I get why some like it, but a few caps resistors and maybe a few soft knee diodes will give you the same distortion, but WAY WAY cheaper.

[@stevenbliss989]

10V out into a power amp, ARE YOU MAD!!!!!!!!!????????? Surely something like 2V would be way more that enough.

[@MichaelBeeny]

You are missing the point. That just shows the enormous head room available. You should be far from clipping in normal listening levels. HEADROOM MATTERS. Some preamps clip before the signal even enters the power amp. The gain can easily be changed with 1 resistor. I just showed it out of the box. I do say in the video that 500mV in gives you 3 volts out. The rest is just headroom. You will NEVER clip this, exactly what it should be.

[@stevenbliss989]

@@MichaelBeeny I get your point, and agree headroom is VERY IMPORTANT, but this amount is crazy overkill. Running a +/-5V op-amp is still HUGE headroom ( +/-3V P-P, i.e. 6V range) for what should be a 1V output for normal operation. ....that is why I made my point.

[@MichaelBeeny]

@@stevenbliss989 I take your point Steven, I just like a very wide headroom in both preamps and amplifier powers. It's so easy to offer such headroom, it's not that extra parts and ICs were added so why restrict it at all? However, you could always lower it by, as you stated, lowering the operating voltage. Buy why?? Many cars will do 130 MPH why? so, driving at 70-80 is more pleasant. Hummmm, not a good comparison, but I'm sure you get my point.

[@stevenbliss989]

@@MichaelBeeny We agree on headroom, but not on the amount. The reason is (speaking for you:) you care about having no possibility of going outside limits to cause problems, I care about headroom for the sake of not stressing components (I am an electronics design engineer - RETIRED!). So for you car example, the reason they are designed for such high speed is because that means at lower speed the engine to wheels power train is DRAMATICALLY less stressed. ALL STREES IS NOT LINEAR! At twice the push, you may get six times the stress. The same is true for electronic parts. Running an op-amp at +/15V hurts it disproportionately MUCH MORE than running it at +/-5V. You may ask why? it's just electrons moving around. Sorry NO, all electronics are QUANTUM devices, so NOTHING it it behaves with absolute predictability. The trick is to design the part so the odds are in you favor so high that the part is LIKELY (no guaranteed) a certain amount of time. The chances of an electron moving in such a way that it break some molecular bond is NOT ZERO! The hardest thing to understand about how the universe really works is that EVERYTHING is STATSTICAL CHANCES, there are no ABSOLUTES! As an old common example quoted: If you push against a wall, there is a NON-ZERO chance that you will pass through the wall. The number is INSANELY LOW, but it is NEVER EVER ZERO! I know I.ve said a lot here, and it is a lot to take in, ...and yes Physics is a hobby of mine. So in conclusion, if you take a thousand NE5532 and run half of the at +/-15V and the other half at +/-5V, then run them for ten years, WAY more of the +/-15V will fail compared to the +/-5V. The is hundreds of time more true with your car example. Anyway, this is why I am anal about unnecessarily stressing components. This is why I see the NE5532 voltage use here as needless stress, all be it a TINY one.