1. Please take 30 seconds to register your free account to remove most ads, post topics, make friends, earn reward points at our store, and more!  

Anyone ever swap out the MOSFETs or other components in a modern amp?

Discussion in 'Amps and Cabs [BG]' started by two fingers, Nov 1, 2013.

  1. two fingers

    two fingers Opinionated blowhard. But not mad about it. Gold Supporting Member

    Feb 7, 2005
    Eastern NC USA
    Being an electronics student (and a generally curious individual) and working with a lot of ICs lately...........

    I go to thinking, has anyone ever taken a look at the components in a modern amp and though "I can find much better components than these. Let's swap 'em out."

    For reference, I was wondering specifically about a Mesa Walkabout. I don't have any problem with mine. I'm just a tinkerer and was wondering if it would make the amp sound "better" to put different MOSFETs in it (or swap out any other component for that matter).

    Full disclosure. I had this though just now and have not so much as popped mine open or even looked at the schematic.

    This is just a general curiosity being that we tend to have no problems putting "better" caps, tubes, resistors, pots, etc in older tube amps.
  2. Hi.

    Unless the original MosFET's, FET's or transistors are either really hard to find and/or very expensive, almost no-one bothers.

    There's literally a thousand ways to make a MosFet amp and unless You know exactly what type to replace 'em with, without other modifications the result will be a failure. IMLE anyway.
    Much easier to find an amplifier module that runs with the same rail voltages as the old one.

    Hell, You usually can't get even the exact replacements not to blow up.
    If the bias circuit is malfunctioning (as it usually is if the output devices need replacing), it's good bye and dying time for 'em.

    PS caps get replaced in SS, just as in tube amps.

    Back in the 80's some people modded SS amps as well, but it was always mods to the tone or drive section.
    They were considered fools ;).

  3. Actually a lot of us DON'T think we can make modern amps better (tube or SS) by massively swapping components out for supposedly "better" components.

    It's the circuit design that really counts, not some audiophool-grade capacitor or the latest, greatest op-amp. In fact, randomly changing things like op-amps can actually induce instability and send the amp into oscillation if someone sticks in an op-amp with a high gain-bandwidth product.

    Of course, there are situations where an amp's manufacturer did scrimp on the quality of things like switches, jacks, etc. (The circuit design in those amps would be questionable too though). And every now and then an amp comes along that was made very well, but maybe had one component that turned out to be a clunker.
  4. two fingers

    two fingers Opinionated blowhard. But not mad about it. Gold Supporting Member

    Feb 7, 2005
    Eastern NC USA
    I don't disagree with either of you. Thank you for chiming in as you both always seem to know what you are talking about. Here's my question (and I admit it comes from the angle of a somewhat electronics noob). We sometimes build circuits in school. We have trays of ICs to chose from. According to our teachers, some of the exact same ICs are better than others. Like every thing else (they claim) some companies make things with tighter tolerances, and simply "better" manufacturing processes that create a "better" product. So, what I was referring to was perhaps a "better" version of an op-amp or MOSFET that is made to the same specs only "better".

    By the way, I am back in school at 42 years old.... not young and fragile. So keeping shooting from the hip. I don't need sugar coating. I am honestly seeking knowledge. If these things aren't true, or simply don't make any difference, just lay it out there.

    Thank you again!
  5. beans-on-toast

    beans-on-toast Supporting Member

    Aug 7, 2008
    Manufactures sometimes come out with better (lower noise, better protection, higher current capacity, etc) versions of their products. Note the part numbers power transistors and drivers and check the manufacturer's site to see if there are upgraded versions or alternates. Sometimes they suggest alternatives on the part numbers web page. As noted there can be issues. In same cases, these parts find their way into amps as manufacturing upgrades so they aren't all bad.
  6. chaosMK


    May 26, 2005
    Albuquerque, NM
    Hi-fi into an old tube amp
    It's a Mesa... components are probably good quality.
  7. I'd call op-amps that aren't as good as the others in the same batch...defective. At least at the audio level. I buy a batch of NE5532's, they pretty much sound the same and perform the same in audio gear.

    Now for RF, microwave frequencies and above, I'd suspect that some of the discrete components would matter, considerably. The difference though is measurable quantities: capacitance, inductance, noise floor, etc.

    I'm emphasizing the "measurable" part because the audio world is full of charlatans and snake oil salesman. "Improve your sound with this $1000 power cable!" for example, with absolutely no scientific basis for those claims.

    Now some of us do modify our amps to add features, change things, whatever; nothing wrong with that as long as the goals are realistic and the methods employed are fundamentally sound. And of course many of us repair older amps, so for a 45 year old tube amp, you bet I'm going to be replacing the electrolytic caps and coupling caps and taking a close look at plate resistors. Even there though we have to beware of snake oil salesman and Internet tales of mojo. Some folks claim to hear extraordinary things when going from stranded to solid wire, for example...they mutter "skin effect" but don't realize that doesn't apply at guitar amp frequencies.
  8. WretchedExcess


    Jul 29, 2013
    Nashvillebill makes an important point about gain*bandwidth product in opamps. I've had to fix amps that started oscillating after the were "upgraded" by HiFi gurus who installed modern BB high precision, low noise, high bandwidth opamps without looking at the amp's NFB circuit. if you decide to go for improved opamp that's controlling an amplifier input with NFB these considerations are especially important. So it's important to temper your enthusiasm by doing some math, looking at the poles in the NFB circuit, etc. and choose the upgrade that has suitable specs. Higher BW is not what you're looking for.

    That said, you can make real improvements to a vintage SS amp by upgrading opamps. A HiFi amp that's 30-40 years old that uses an 072 on the input of an NFB amp could definitely be improved by an opamp upgrade to one of today's ultra-low noise opamps like the OPA series. But with modern designs there's not as much to gain <rimshot> as the silicon available today is much better than it was 30-40 years ago. With that said, it really does amaze me that some audio amplifiers do continue to use pretty cheap low end opamps today, when they really don't have to. But you have to do your homework. In some situations an opamp upgrade will provide a measurable improvement, and in others it will not. Notice that I said measurable difference, not audible difference.

    Now, setting opamps aside, and addressing the subject of output transistors:

    Do some of us rip out perfectly good output transistors from a properly functioning amp and replace them with brand new production components? Absolutely. Do we do it for an audible improvement? No. Do we do it for improved reliability? Yes.

    As an extreme example: Today's silicon is far better than it was 35 years ago. In a TO-3 package the die sizes are smaller, heat dissipation is better, and current handling is better. The result is that you can buy transistors today that have 2x the SOA of transistors in your amp. Instead of just assuming that a manufacturer used the best components available, check the specs. Chances are that you might find a replacement that has a greater SOA than the transistor that was priced into your amp when it was built. Having double the SOA and twice the current rating can be significant if you're engineering a circuit to improve it's behavior into low impedance loads.
  9. Woodstockz


    Sep 23, 2011
    San Diego, Ca
    Great that you are back in school...always good to know more.

    In repairs, usually it is the power supply or power amp section that goes out. Or sometimes both. If that happens, you can "remove and replace" the original part, or you can try for an upgrade to a more robust part to prevent a repeat failure. In lower end equipment, there may be room for an upgrade for better performance. In higher end equipment, probably not much room for improvement.

    Older equipment may have obsolete parts that you may have to find a substitute for. As long as you are pretty close to the original specs, it will probably work - maybe a little better, maybe a little worse. As Beans said, look to the mfg site for recommended substitutes.

    If you are really into electronics, then the fun is in trying to make it better. If the part is close in specs, but a little better, try it. Even if you don't make it better, you can learn from trying. Try a low noise IC or a MOSFET with a higher Voltage rating. If you make it worse, then put it back to original and you have learned something. If you make it better, then you have something that you can keep...and you have learned something. Eventually, you will get an idea of what works and what doesn't.

    About IC chips...Silicon manufacturing can have high failure rates/low yield. The more complicated the chip, the lower the yield. Processes are better now, but back in the old days, 10% yield was considered good --- That means 90% failures.:eek:

    So clean rooms became very strict to get yields up. Also, parts are binned. That means that a chip is made in quantities, then tested for features and performance. After the testing, the parts are labelled according to how well they do. Part that pass all features gets one dash number. Parts that pass all tests except feature x get a different dash number. Parts that pass all tests except feature y get an even different dash number.

    For example CPU chips. Make a new Pentium or whatever. Test at different speeds. Label MHz rating according to the highest speed that passes the test. Same manufacturing, different ratings.

    Good for you. I am even less young and more fragile. I have been into electronics for a long time, and I am still learning.
  10. shwashwa


    Aug 30, 2003
  11. NightCat


    Apr 4, 2013
    Pedals! Get your Tinker Fix on by hot-rodding pedals.
    I have had a tremendous amount of fun modding pedals from the parts and plans on this site: http://www.monteallums.com/
    I am not in any way associated with this site or business, just a happy customer.
    It is very easy to swap Opamps, caps, diode or whatever, in pedals and hear exactly what changes in your tone.
    I also picked up an old dual trace O-Scope off Ebay and an audio frequency tone generator. You can put a nice clean 440hz signal into your pedal and monitor it as it goes through the system. It is almost impossible to move beyond kits without being able to map your signal.
  12. Passinwind

    Passinwind I know nothing. Commercial User

    Dec 3, 2003
    Columbia River Gorge, WA.
    Owner/Designer &Toaster Tech Passinwind Electronics
    I used to think that way, and what ended up being mostly tail chasing was often fun and instructive up to a point. Now I tend to just start at the circuit level and design and build the whole danged thing myself...double the hubris, double the fun. ;)
  13. seamonkey


    Aug 6, 2004
    Some claim that changing op-amps "changes" the sound. Coincidentally, when somebody changes an op-amp it's always boasted as an "improvement" - by their ears.

    Nobody every goes to the bother of measuring before/after. Or even recording before and after.

    Today's measuring equipment is better and cheaper than ever. Same for recording gear. Using these tools is the big advantage. Measure before and after a swap. If there's a difference it's likely due to a design error, not a component choice. Still it may be easier to swap active components than change the design. Overall it's better to do a new design based on new devices.

    In the case of class-d the output devices work entirely different than they do in class-ab. They both usually have op-amp front ends though.
  14. Yep, measuring is the way to go. As a minimum, put the scope on the amp circuit afterwards. There are instances of folks claiming improved high-end response, for example, when they are actually hearing the amp starting to go into ultrasonic oscillation (which of course can lead to failure).
  15. Woodstockz


    Sep 23, 2011
    San Diego, Ca
    Actually, I think that is the right way to do it.

    Chase your tail for a while, learning what it right, and what is wrong. Then start with a clean sheet.

    Oliver Heaviside on learning Maxwell's equations:
    “I saw that it was great, greater, and greatest, with prodigious possibilities in its power. I was determined to master the book. It took me several years before I could understand everything that I possibly could. Then I set Maxwell aside and followed my own course.”

    After that, he re-formulated Maxwell's equations into the ones that we use today.
  16. I swapped out the opamp on my Ibby EDB650 with a more modern version of the same opamp. The more modern version came in three varieties - one with low noise, one with low power, and one with some other feature I don't recall. Because I power the internal preamp on my basses with external power (not batteries) I am less concerned about power consumption so I went with the low noise version. The newer opamp also worked at a higher rail voltage so I was able to power the preamp with either 9V or 18V as a result of the swap. The old opamp only worked with 9V (I tried the old one with 18V - no go).

    To be honest, none of that noticeably changed the tone of the bass or the amount of signal noise.
  17. Woodstockz


    Sep 23, 2011
    San Diego, Ca
    +1 on measuring. I wouldn't say that nobody does it.
  18. NightCat


    Apr 4, 2013
    I'm not sure how you do this but I have never been able to use a live bass or guitar signal to analyses a circuit. There is just too much going on and you never get exactly the same waveform. That is why I have an audio frequency signal generator. you can manual sweep through the frequencies and watch your signal path react.
    I suppose if you had a Looper you could get close to perfectly repeatable tests...
  19. Bob Lee (QSC)

    Bob Lee (QSC) In case you missed it, I work for QSC Audio! Commercial User

    Jul 3, 2001
    Costa Mesa, Calif.
    Technical Communications Developer, QSC Audio
    It's usually not as easy as many people think. Sometimes substituting "better" components ends up degrading performance and/or reliability, particularly with active components like op amps, transistors, MOSFETs, etc.

    If you're into electronics it's really fun to experiment, but as long as you accept the risks involved I'd suggest exploring on surplus pieces of gear that you don't have to rely on for an imminent gig.
  20. NightCat


    Apr 4, 2013
    Yeah, he must not know the right people. "If you can't measure it, you can't improve it".

Share This Page

  1. This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
    By continuing to use this site, you are consenting to our use of cookies.