Ok I decided to mod my Yamaha HS80m since I'm pretty happy with it but like anything it can always be better

I've opened it up and it has whole bunch of Samxon electrolytic caps. (10uF/50V x12, 47uF/50V x2, 47uF/35V x2, 47uF/16V x1, 220uF/10V x1, 4700uF/50V x2, 1uF/50V x4)

First of all are Samxon eletrolytics any good ?

Anyways so I was wondering if swapping these caps with better quality such as Nichicon, Panasonics, Elnas, etc will improve the quality?

Also it contains twelve ST4558 OpAmps, would it be a good idea to swap these with something better ? What are some good alternatives ? LME49720 ?

I'm attaching the pdf file with schematics, can any of you experts give me some advice?

Thank you!!!

Great thread .

I like this monitors and I was thinking of upgrading the twitters maybe ....

https://gearspace.com/board/new-prod...d-edition.html

I'd like to do this to my 50's.

Seems they are using BB OPA2134's, an out of date opamp. LME49720MA's would be a much better choice, after all, these are monitors.

Those are crappo ******** el caps, pure crap. Use Panasonic or Nichicon.

Seems the EE's at Yamaha are stuck at 1973 technology. 4558 opamps in a studio monitor? Almost funny. Typical Yamaha sonics. They might as well offer cotton balls to stuff into your ears.

Jim, do you think LME49723 would be a good choice as well ? or do you think LME49720 is a better choice here ?

I've already replaced all the electrolytics to Elna Silmic II where possible and rest to Nichicon.

Also swapped all 1uF electrolytics to either Wima polyester or Vishay/BC MKP polypropylene depending on space availability.

Hi Jim, its the 1641.

Be careful with replacing the 4558s. The 4558 is slow but VERY stable, so higher speed replacements are likely to oscillate.

Do you own, or are able to borrow, an oscilloscope? Nothing too fancy, a basic 10MHz analog scope will do. Do not attempt replacing the opamps without an oscilloscope otherwise you may blow the tweeters if the amps oscillate.

Would better caps help against oscillation?

I am thinking about putting in some OPA1604's.

Disconnected the limiter in my 50's today, one at a time, so I had chance to compare. Definitely removed some of that processed sound. Kick drums are less klicky and the low end extends a bit. "Screams" less at you, but the high-end is still a bit messy.

You probably mean the dual package OPA1642? Don't put single opamps in there or you will let the smoke out.

That BB part is probably the safest sub for that design. It draws little current and has a huge phase margin so it's hard to make it un-stable. If you examine the layout closely and see local chip caps from the power rails (pin's 4 and 8) to ground and see a small chip cap across the feedack loops of the opamp's (pin's 1+2, 6+7) you can try out higher test opamps in there. "Trust, but verify" with a scope.

Do check out the power supply regulators, if they are well heat sinked T0-220 sized parts you will be ok, if they are smaller SOIC versions soldered to the pcb for heatsinking, watch out for extra heat from higher draw opamps.

I mention that as I've not seen one of these before and don't want any of you to let any smoke out.

Yes, its the dual...! It really does make this a whole new animal. The caps and limiter snip do a lot too.

Ok, don't use OPA2604. They cause some minor oscillation. Nothing blown though.

Will try with 1642's next.

Also, be carefull when desoldering the stock op-amps. Those lines are easy to rip up.

Don't overlook the 2134 (I will elaborate on this soon, but we did some listening tests today and inside this design as a replacement...wow). Stay tuned...but yeah give them a shot too.

*Edit: OK to elaborate...

The short story is that Bob brought in some 2134's to do a side by side mod test with before we ever shipped the first modified speaker, to compare with the 1642. For 2 days now we've tested and confirmed with our ears (not spec sheets): in this speaker and mod application, the 2134 actually delivers a more full and natural response. Therefore, without having actually shipped a unit yet (we launched this mod in the middle of a backorder with Yamaha, and none of the mail in units have even arrived to start on) our Burr Brown will be the OPA2134 for every speaker mod done here.

voidar: if you get the chance to try both, I'd love to hear your take on it. The presentation of music playback is better served, and is the best fit for this. I'm very excited about this actually, it makes better even more...better (either of these opamps will blow away the ones Yamaha uses stock).

War

Ok. The 1642's are in.

Working fine this way with no oscillation.

However I measured the rails and they've droped to about +/- 3V. I didn't check before mod'ing but the schems say +/- 15V.

The speakers don't sound under-amped or anything, but should I think I can improve on this? Will better caps help?

If it's in a through hole package , don't desolder them .
Cut each leg close to he package with a new scalpel , then heat each leg and knock it out . Knock out the legs one at a time .

They're surface mounted.

Oooo , not so good then .
Heatshield and small hotair blower if possible , or flood each side with solder and "walk" them out .
Try at your own risk , IMHO takes some experience to be doing this with an iron .
Never try without a temp controlled iron either .

I've done plenty of surface mount chips with a non temp controlled iron before. You just need to go at it and don't hesitate. in fact, I've never owned a temp controlled iron.

Respectfully I completely disagree with this post , and as such , I advise any readers of this thread NOT TO FOLLOW THE QUOTED ADVICE !

Yeah be very careful, it's very easy to peel the pcb traces on the HS80m. FYI.. I've worked with other pcb's and they were much more robust than the Yamaha HS80m's pcb.

I've messed up the 1st op-amp pcb trace so I had to order a new input pcb for about $160 from Yamaha Japan :(

It's very hard to swap the smd/smt type op-amps yourself so be very careful.

Understood, I was only speaking for myself.

Cool Talon , I wasn't trying to insult at all .
There's no prizes for winning the internets , after all .

The soldering game is giant, especially I can say it having seen the PCBs in HS80m.)))
I would not f* around without prior experience and whatnot with modifying them

Hi, i'm doing the mod with the chips i have on my stash. I the first to chips in each monitor with lme49722 and a opa1642 on mid eq and room control section.
I did the limiter snip and signal path caps with Pana FM's.
So much better sounding now, the bass is more extensive and natural and highs begin to became liquid instead of rough.

Loving so mucha that i'm thinking of swapping all opamps. In order to finish the mod in both speakers i only have 8 of the opa1612 type or ne5532 type.

Are 1612's less stable than 1642's? Worse sounding?
Same question for the NE5532.

thanks.

Should i order some opa2134's then?

I'd use the ne5532s cause I like the way they sound and you won't hear the improvment of other chips costing 10 times as much. The 5532 should be in all audio equipment 4558 and such are for stompboxes I'd never run serious audio through them.

Would swapping the active filter caps for lower tolerance ones like styroflex 2,5%, o some 1% ones, improve imaging? Since were reducing difference in a pair of speakers response i think it could make any sense.

Are Panasonic FC's ok to use? Or would FM caps be better?

Unless you're into modding for the sake of it, i can't understand why anyone would want to just swap out components in the hope it will improve the sound. There's surely more to it than that.

Also I'm wondering if removing the grille for the tweeter would do anything for the high frequencies, I've heard that grilles can create comb filtering.

That's why i tried to formulate a question based on some electronics theory insight, wandering about the effect of tolerances on possible frequency response matching.