Improve a Rotel amp THD by 20dB!

My gear isn't calibrated to provide absolute THD values, but if I ensure that I use the same starting point and level every time, it can measure the differences between the upgrade stages. Hence the name of this thread.

And yet you are quoting THD figures???
Forgive me for being confused.


Re PS. This is why I normally replace the old E-core with a toroidal in the RA-820AX, and will definitely do so in alp1961's 820AX. Although the old Rotel tranny isn't half bad (well built with copper belly band, etc.) it is actually one of the components in the amp that runs hottest. And if you try to take the amp to its limit - as I did in one of the earliest posts - it will blow its internal thermal fuse and go to the Eternal milli-Henry Hunting Grounds.

The 840 has two E-I cores - one for each rail. The 850 has two - one for each channel.
Did you mention this rather expensive change earlier?

Without this change, are your mods worthwhile?
Does the transformer change affect the THD results?



And if you try to take the amp to its limit - as I did in one of the earliest posts - it will blow its internal thermal fuse and go to the Eternal milli-Henry Hunting Grounds.

These amps are not designed to run at 1/3rd or at full continuous power. The heatsinks can't take the former and (as you have found) the tranformer can't take the latter. This is a sensible commercial choice.
 
Last edited:
"Does the transformer change affect the THD results?"

Please see post #161.

Per


That's noise not THD.

I did some measurements on 2 amps that have been installed in my RB850 chassis, using the RB850's power transformers (post 1098 onwards):

TGM8 - an amplifier based on Rod Elliot P3a

As part of the experiments I tried physically distancing one amp board from the chassis. It made no difference at all. I concluded that magnetic field from the Rotel transformers is not an issue. Therefore it's hard to see why toroids would be any better than the EIs.
 
Member
Joined 2012
Paid Member
"I concluded that magnetic field from the Rotel transformers is not an issue. Therefore it's hard to see why toroids would be any better than the EIs."

Who can argue with that. I wish that Rotel (and every other audio amp manufacturer since the 1980'ies) would read your post, realise the error of their ways and go back to proper EI iron core transformers.:)
BTW, could I borrow your professional Audio Precision RTA analyzer - just for a few days?

Per
 
Member
Joined 2009
Paid Member
HI i have a question. Can the noise coming from the mains earth line ? if the amp's speakers negative outputs are connected to the mains earth line could that be the reason of some noise at the speakers outputs ?
if so can a mains isolation transformer cure that ?
 
Last edited:
Member
Joined 2012
Paid Member
HI i have a question. Can the noise coming from the mains earth line ? if the amp's speakers negative outputs are connected to the mains earth line could that be the reason of some noise at the speakers outputs ?
if so can a mains isolation transformer cure that ?
I wouldn't get any sleepless nights over this. There is absolutely no externally measurable or audible hum - and isolation transformers won't help.

It is as vovkZ and others have rightly pointed out that there is most probably some devious ground loop between sound card and RTA measurement issue at play.
I have had more audible hum by accidentally running a speaker cable too close to a mains plug outlet.
Per
 
Member
Joined 2012
Paid Member
RA-820AX upgrades revisited

Right, I'm back from some needed R&R and have put the finishing touches on the RA-820AX.
I have put a speaker and "on/off thump" protection circuit in - yes, it is an inexpensive Chinese pcb, but it works well and I have modified its original 12V design to be able to directly feed off the 25Vac from the transformer with a minimal relay current of 32mA, thereby not loading the amp's rectified dc voltages.

Then I put in an AngelP emerald green power LED to replace the bland Rotel red at the front, and some AngelP Signature stickers - mainly to strut my feathers, but also to give a would-be future repair technician a chance to email me for advice on what strange things happened to this Rotel amp.

But most importantly - how does it sound?
The short answer is: Heavenly.
The amp has a completely improved soundstage detail and clarity with a commanding control over the speakers.
Punchy bass and ... well ok, I think I will stop the waffling and let alp1961 (Rui) as the amp's owner be the :judge: of these things.

But I could add that I have a setup where I can switch between my trusted Hiraga class-A amp and any second amp on test.
With this 820AX I have repeatedly had to double check which amp was actually playing, this upgraded Rotel is that good.:)

So, with only a screwdriver and a soldering iron (and a lot of time and patience) - most DIY Rotel owners could do it.
I put in 40 new components:

2 10,000uF 63V caps
12 Panasonic FM caps
6 WIMA film caps
4 PS film caps
4 MF resistors
2 Audio opamps
4 Matched input transistors
4 AngelP integrated modules
1 AngelP modified speaker protection unit

and
1 Toroidal mains transformer

If you decide not to swap the mains transformer, you should be able to readily get these components for less than £100 - (based on the online prices from Farnell UK.)
But please be very wary of cheaper bargains on ebay and other places.

I enclose pictures of a proper NJW1302G power transistor next to one of my "bargain purchase" mistakes. You will see that the counterfeit appearance and markings look next to identical to the real tranny - until you crack them open and find that the chip inside the original ON component is 40 mm2 whereas the fake is a tiny 6.2 mm2 - a fake that will never even get close to the specs.

Same things apply for the ELNA LAO caps and OPA2134's. Go cheap and you will get fake crap - not worth it given the amount of hours you by then will have put into the upgrades.
Just my :2c:

Per
 

Attachments

  • DSCF4481.JPG
    DSCF4481.JPG
    891 KB · Views: 399
  • DSCF4483.JPG
    DSCF4483.JPG
    919 KB · Views: 373
  • DSCF4487.JPG
    DSCF4487.JPG
    962.1 KB · Views: 355
Member
Joined 2012
Paid Member
I don't have an RB-991 in the workshop at the moment, Paulus did the upgrades in Holland on his own pair of 991's - with only a little social distance help from me.
And the RA-820AX will soon be on its way to Portugal, I bet Rui is getting pretty impatient by now to get it back.

But the basic upgrade principle and technique is quite similar, so I would expect the improvements to be of the same magnitude for both types of amps.
The 991 is of course a beast of a power amp while the 820AX is more modest integrated domestic amplifier which you shouldn't ask to fill up the Albert Hall. - Or maybe it could? Worth a try.
Per
 
I’m sorry: this may be the wrong place to ask: would the RB-991 be a worthwhile upgrade from a RB-985 to drive my HT center channel (CC)?

Right now I use a RB-985 (5 channel) to tri-amp the same providing about 325+RMS. Due to a reconfiguration to bi-amp, power would take a hit to about 250. But I’ve read the 991 is an upgrade from the 991’s little brother, the 981. The 985 is essentially a five channel 980BX though a bit less power per channel.

Side note: 981s drive the mids & tweets of the other channels

I realize this “sounds” like a no-brained to some. Being retired I’d need a 985 buyer pretty quick to avoid the wrath of my better half. I’d like to know this change would be worth the “process.”

Thanks for your thoughts (and help)
 
Member
Joined 2012
Paid Member
Thanks Per,
Still no progress here on any audio project as I'm way too busy with the cars. Welding the Jag XJC, maintaining the Saab c900 and I'm parting out the Transam Firebird :(
What's your next project?

I recently got a decent offer to buy a souped up Mazda MX-5 (which 'just' needed a little bit of TLC) - when I got "one of those looks" from the Supreme Command :sarge:...... and of course quickly declined the opportunity:no:

Next project is an old NAD 3030 - which could be more work than the MX-5, but is easier to quickly hide away under the workbench:D
Per
 
Member
Joined 2012
Paid Member
I’m sorry: this may be the wrong place to ask: would the RB-991 be a worthwhile upgrade from a RB-985 to drive my HT center channel (CC)?

Side note: 981s drive the mids & tweets of the other channels

I realize this “sounds” like a no-brained to some. Being retired I’d need a 985 buyer pretty quick to avoid the wrath of my better half. I’d like to know this change would be worth the “process.”

You have definitely come to the right place to get a silly no-brainer answer - and I for one won't let you down!;)

Firstly, why would you need all those Watts for your center channel? Or do you plan to run a drive-in movie center?

Next, the mids and tweets will only take a few Watts of power, so it seems to me that you are already plenty overpowered with the RB-981 and 985's. The 991 is basically the same topology as the 981/5's and I doubt that there would be much sonic difference (unless you upgrade them as Paulus did).

I'd say: Obey your better half and she may continue to need and feed you "when you're 64 - Huh".:sing:
 
I’m sorry: this may be the wrong place to ask: would the RB-991 be a worthwhile upgrade from a RB-985 to drive my HT center channel (CC)?

Right now I use a RB-985 (5 channel) to tri-amp the same providing about 325+RMS. Due to a reconfiguration to bi-amp, power would take a hit to about 250. But I’ve read the 991 is an upgrade from the 991’s little brother, the 981. The 985 is essentially a five channel 980BX though a bit less power per channel.

Side note: 981s drive the mids & tweets of the other channels

I realize this “sounds” like a no-brained to some. Being retired I’d need a 985 buyer pretty quick to avoid the wrath of my better half. I’d like to know this change would be worth the “process.”

Thanks for your thoughts (and help)

Wow, that is some serious power you have there to drive your loudspeakers but there are a few small catches i learned years ago when i worked in a fine audio shop in the Netherlands and that is topology and how to wire amps to loudspeakers
Perhaps i'm completely wrong because i do not have the schematics of your amps but the RB-991 is in fact a dual mono setup, up to the secondary windings of the trafo, there's even more....
Most amps have the ground from the input wired to the chassis, not on the 991...
This gives, if i'm not mistaken, a better channel separation, no crosstalk at all.....

I'm convinced that the damping factor of the RB-991 is much better then 981 and 985 and yes i know that damping factor is overrated by many people but believe me that you hear and feel the difference when an amp has a solid grip on a loudspeaker, try "Echoes" from David Gilmour's album "Remember that Night"

If you had LINN amps (LK 85, LK100, etc...) and you would like to bi-amp your loudspeakers they (LINN) recommended back then (end 90's) that you wire your highs from one amp and the low's from the other :eek:
That sounds really bad because of left and right channel crosstalk so i never followed that recommendation even when my boss said i should follow instructions, so i did a setup in the shop as recommended by LINN and then i switched to the setup with a dedicated amp on one loudspeaker.....:D

It all comes down to this, a mono amp is the best, period! Multichannel amps are less because the most of them share a single PSU and the most of them have a lot more crosstalk then i could bare in a stereo setup.
But, you play surround if i'm not mistaken.

It is not all about power, it is about control. Does your amp have a solid grip on your loudspeaker and let it do exactly what it is supposed to do, to reproduce music.

Having 2 of these RB-991's for about 20 years now i am certain that they will outperform your tri-amp setup anytime, but they are old so when you get your hands on a few of them and before you put them to the test please recap them and if you are handy do the upgrades Per wrote down here, you won't be disappointed.

Cheers Paulus
 
1000 thanks for your input, Per & Paulus!

I’m taking a short break from something unrelated.

Paulus: recap? As you did w/Per coaching or just the PS filtering? I plan to follow regardless, once I trim down my honey doo list and settle on which of these fine amps to do what.

I’ll briefly touch on omitted system info. Using Polk RTi A7 towers as LRs. Tri-amped. Each has a dedicated Rotel stack: 981 for the 6.5” mid & 1” tweet; a 980* for the pair of 7” woofers. Yes - two 981s & two 980s.
* for its 1000 DF

The two Polk surrounds each w/ 6.5” mid & 1” tweet driven full range by a single 981 - one day I’ll bi-amped them, each driven by a dedicated 981.

I realize this “sounds” extreme to some - “sounds**” effortless to me.
** the SQ! labor is not effortless, but worth it

About the time of my previous post I added the 980BX schematic to my 991,990, 981, & 985 collection. I spent time studying their respective topologies. The 990/1 and 981 are basically the same; 990/1 identical though I’d be quick to listen to Per* before myself on that. 980... I’ll leave that to Per*.
* or similar

I have a LOT more background but back to the project. Hope this helps y’all get inside my head w/40 years of this fun hobby behind me

1000 thanks Tony
 
@paulus:

I forgot to mention: I take amp damping very factor seriously. Yes I‘be read on other sites, “...above matters little...” “...can’t hear the diff above 100.” I beg to “differ.” Then again some say upgrading internal wiring* in speakers, and in some cases, amps is a waste. These are the same guys that spend 100s, and in some cases, 1000s connecting their speakers to their amps... ...go figure.
*inexpensive heavier wire. In time I will go at least a little heavier in my amps if at all possible. But that is for another session.

Rotel rates the 991 & 981 DF @ 500. I believe the 991/981 are more transparent.
Rotel rates the 990BX & 980BX DF @ 1000*
Rotel rates the 985 DF @ 180
* why 980BXs drive my towers’ woofers. In an extensive blind listening test of those woofers I determined the 980 sounded a smidge of a smidge better than the 981. The 981 did a great job, make no mistake. When we were finished, we consistently pick the 980.

Again thanks guys.
 
Last edited: