Rotel RA-920AX

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
adjusting ???means exactly what ??? these are probably fixed resistors what to adjust???

you probably mena to replace with something else that will give other type of ofset ...or replace them with a trimmer ..that could be a more realistic option

best thing will be to first verify all voltage arround there to se what is going on .... there might be a chance of something else also .... for example whta is your normal voltage down there ??? or down under ??? 220 or 240 ???

it could be that the amp is setup for 220 this lilte extra voltage might or could bring the amp outof balance ...so verify voltage before everything else ...also from the schematic to the testing points on the pcb ...

then verify close to ltp resistors for drifting .... if any suspect replace them all with metal film 1% ...that is also some type of upgrade


one other thing iwould like to do and since pcb and parts location helps is to join the ltp tranisotrs for thermal junction ...that also might solve the problem


regards sakis

Sakis, it's setup for 240v which is what our outlet voltage is here in Australia.

I might verify the output voltage on the secondaries, in case the transformer is at fault?

I tested every resistor on the PCB and they all seem to be very close to spec, they are already metal film type according to the parts list.

What do you mean by 'join the ltp transistors for thermal junction'?

Thanks for your help again! :)
 
chances that trafo is at fault is bellow zero ...

as about resistors "close to spec ) that will be actually not enough since a variation of a few ohms might cause this offset ....

option will be to install a trimmer and that will take ofset to total zero ...but as is aid it might help to join the 2 ltp transistors that will be Q601+605 and 602 +606 ...

idea is very simple : while tranistor at this stage operate get warm ...if one gets warmer that the other ( for any reason or simply just a bit ) that will result to ofset may be a lilte but still some ofset

if the leads of the transitors are long enough you may bend them a bit and glue them together super glue will do ... its not the best way but it works good enough ...

other option will be to rap them with a small pice of metal something like a belt arround them that could be cooper or alouminioum and it doesnt have to be thick

in one amplifier i just glued a small piece of aloum. aprox 2mmx4mm connecting the tops of the two transitors and wokred just fine ....i had 60mv ofset and droped to 15 which is alsmot axceptable

check it out ...might work

there is another thing that i would like to know though ...do this for me please warm up the amp measure both bias and ofset ...after that change the bias to something lower and then verify the ofset readings again see if it droped or increased ...

let me know
 
Last edited:
The power stage LTP bases have R603 47k5 on one side and R635 8.25k on the other. Current is about 0.7mA in each transistor, so expect base currents of between 1 and 4uA. This is plenty to offset the output by up to about 200mV if everything is worst case.
100mv should not harm your speakers
 
Sakis: Thanks, I will check some of these things tomorrow!
As for an incorrect reading, I have checked with 2 different DMM's now, and both times in separate locations of the office, so I think they are reading correctly.

Davidsrsb: Are you saying that the high offset may be normal in this amplifier?
 
OK guys, thanks for your patience:D

Here's a first stab at sorting out the grounding. The general principle is that grounds serve two purposes - signal reference and power supply and the two functions are best not intertwined.

The colour code is as follows :

Big green blob under E1 - star point.
Other green blobs - connect them to the big green blob.
Magenta blobs - here two choices. Either lift the connection out of the board and solder together as a local group and then link to the big green blob. Or cut tracks around them to achieve the same.
Red lines - cut tracks, or remove link.

I've only played with the input grounds on the phono and CD sources. Also I've never tried this kind of regrounding on phono stages, there might be some hum from coupling to stray magnetic fields from the trafo, so no guarantees on that one. If you want to use more than these two inputs, perform the same isolation and star-pointing as for the CD input.

There's always the possibility I've missed out noticing a place that the ground goes to (following it reminds me of the mazes I did as a kid - complete with dead-ends!). If so then I'll have caused a loop which will betray itself with hum probably - so your feedback is most welcome.

I have one or two further suggestions which go beyond grounding, to be added at some date soon.

Oh, about the offset - yes, davidsrsb is quite right, the impedances are mis-matched so the offset is there by design. It can be reduced by changing R635,6 to 47k and R611,2 to 2k7.
 

Attachments

  • RA-920AX-TM5.jpg
    RA-920AX-TM5.jpg
    445.4 KB · Views: 324
  • RA-920AX-TM6.jpg
    RA-920AX-TM6.jpg
    624.5 KB · Views: 294
The other suggestion I have is bypassing the tone controls. This is quite a simple job - just a few cut links and two added wires.

For the left channel : cut J62 and J69. Link from the north end of J69 to the north end of J62.

For the right : cut J66 and J37. Link from J66 E to J37 N.

Since the tone controls are passive and have an insertion loss around 10dB, this gain is at present made up in IC501. With the controls out of circuit, the buffer's gain should be reduced by 10dB, so change both R507 and R508 to 4k7s.
 
----by passing the tone control has absolutelly nothing to do with the ofset of course ... still it could be a nice option but on the other hand best will be see if possible any op amp upgrades before that ...that also will be worth trying

---- regarding ofset for my standards i wouldnt go for anything more than 50mv and my target will be an absolute 0 as a general rule 100mv could be acceptable

---- finally re rooting the grounds might do something .... but not in machines like that ....i think that rotel peopel worked this out in the best available way ....meaning that rerooting might help bits hereand there but i think that this will be bearly audible ...

kind regards sakis ....

( forum is a wonderfull and magic thing ....This rotel doesnt actually worth such an effort ...but here we are ...!!! a bunch of nice people giving their opinions talking about issues and adding bits and frugments of knowledge and expirience for every body to see )

enjoy !!!!
 
---- finally re rooting the grounds might do something .... but not in machines like that ....i think that rotel peopel worked this out in the best available way ....meaning that rerooting might help bits hereand there but i think that this will be bearly audible ...

Proof of the pudding... - I've had good results myself from such re-routing, (but not with Rotel as I don't have one). Let's wait for the listening verdict of the modders - that's the reason I've offered the mods - to hear what they sound like:D

( forum is a wonderfull and magic thing ....This rotel doesnt actually worth such an effort ...but here we are ...!!! a bunch of nice people giving their opinions talking about issues and adding bits and frugments of knowledge and expirience for every body to see )

The Rotel is certainly worth the effort - but can't disagree with the rest of this:p
 
Finally had success with my offset issue - with the help of some excellent advise - I replaced Q601/2 & Q605/6 with some high gain matched transistors (Toshiba 2SA970BL) - Zetex ZTX795A where recommended as superior but I could not get them in NZ without paying huge shipping.

I found that R610 was also way off spec and replacing that brought both channels to within 1mV of each other. I replaced all the elect caps and also the bias trim pots with multi turn sealed versions to make adjusting easier and more accurate compared to the single tun original.

Final results come to 31 & 32 mV at the binding posts down from 98 and 70ish at the beginning.

I have a very slight hum sounding like it is coming from the transformer and I can also hear it at the speakers and so I am hoping that the ground re routing that abraxalito is suggesting might help.....any thoughts before I start hacking up the track?
 
The other suggestion I have is bypassing the tone controls. This is quite a simple job - just a few cut links and two added wires.

For the left channel : cut J62 and J69. Link from the north end of J69 to the north end of J62.

For the right : cut J66 and J37. Link from J66 E to J37 N.

Since the tone controls are passive and have an insertion loss around 10dB, this gain is at present made up in IC501. With the controls out of circuit, the buffer's gain should be reduced by 10dB, so change both R507 and R508 to 4k7s.

hello:wave: I have this amp and have REMOVED both tone controls, should I change R507 & R508? sounds Ok!
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.