Head-Fi.org › Forums › Misc.-Category Forums › DIY (Do-It-Yourself) Discussions › stax srm717/srm727-11
New Posts  All Forums:Forum Nav:

stax srm717/srm727-11 - Page 4

post #46 of 69
Quote:
Originally Posted by spritzer View Post

Put the largest cap you like in there but it won't change how badly overdriven the tubes are. 


I guess I'll take a look at the 6S4A mod.

 

What's the best way to remove the top layer of the PCB to expose the bare copper trace?  I noticed in Justin's high res photos of the mod, that he did this several times?

 

No need to change any resistor values?

post #47 of 69
Thread Starter 

Thanks for the positive comment Argybargy I have been put off adding to this because of the general negative comment.

                       That I wasn't presenting this by the "way it is done in Head-Fi"

                          And yet looking at some posts they go on for greater lengths to mine --in the same lay-out.

                            My whole approach and attitude is bound up in the way that JLH [john linsay  hood ] .

                                approached it but very condensed- I love music but reproduction of it isn't always perfect

                                 Like John I believe in complete openness so that any fault in the electronic components is shown up    NOT to be "dummed down" to make it smooth by adding many frequency changing and stability components like small value capacitors . To me that is cheating.

                                     A certain ADE wrote in EW/WW that while on a breadboard his new design performed beautifully BUT transferred to a lovely to look at PCB he had to add comp capacitors.They affect the reproduction -yes to stop HF oscillation but also the perceived sound reproduction. 

                         Johns designs were always open -and latterly -high input impedance/high sensitivity any ADE will tell you that that is harder to achieve  low distortion than a low sensitivity/low input impedance.

                           John also used Mos-Fets at that time he was criticised for using them under the heading-"they are not as LINEAR than BJT"-quite right -BUT a lot more neg feedback can be applied as they have a higher bandwidth and as some know here have a more clear presentation.

              MY whole aim here is to achieve  a lot more open and natural sound-as you would hear live at a concert hall 

         Not smoothness for smoothness sake with rounded ends which after a while you feel inside yourself that you are missing something in the music but a feeling you are in the stalls listening to it live. I have no intention of changing my values learned from JLH. I know ADE have blazing  rows with each other on audio issues but I wont be changing mine.

post #48 of 69
Quote:
Originally Posted by Argybargy View Post


I guess I'll take a look at the 6S4A mod.

 

What's the best way to remove the top layer of the PCB to expose the bare copper trace?  I noticed in Justin's high res photos of the mod, that he did this several times?

 

No need to change any resistor values?

 

Probably just scraped it off.  I cut off a part of the trace to make sure there was no chance of arcing. 

post #49 of 69
Thanks. I'm a little hesitant to take a razor blade to an amp I paid $900 for, but, hell for better sund quality, why not?
post #50 of 69
Duncan, keep up the good work.

It's the ideas that matter, not the form of the message nor the messenger for that matter.

I don't currently have a 727 or 717 but will probably get one in the future, so I copied and saved your posts into a word file for future reference.
post #51 of 69
Quote:
Originally Posted by Argybargy View Post

Thanks. I'm a little hesitant to take a razor blade to an amp I paid $900 for, but, hell for better sund quality, why not?

 

The first amp I modified was a brand new 007II...  redface.gif

post #52 of 69
Thread Starter 

The 727-11 produces quite a bit of heat-it does run in class-A. But heat means noise and this too will have be removed from the amp by  NEG. feedback.[more work for it]

        It is supposed to work by convection but it only takes a quick look at the size of the holes on the base plate to realise that the efficiency will be near nil.

               I understand that companies are  under legal restraints by governments  as tiny fingers could reach underneath and touch a live part. causing a massive legal claim.

      For those who dont have that problem- I got my old 717 base plate   and was going to drill  large holes in it bit as I wont be selling it I just removed the whole area of holes.Try using that base in front of your PC extraction fan and watch the CPU temp rise.

           That isnt enough as the bottom clearance to allow a good stream of cold air isnt enough.

              Using the old base plate I added new feet to it these I got from an old 70s  Japanese record player being used as suspension these raised up the base high enough to allow more air to get in.

                      I realise this isnt for everybody but just as a test  IF you have re instated the feedback as I suggested

                           run the amp without the top cover for 2 days to see if it makes an improvement.

                              I am talking about a very small   improvement as it sounds very good as it stands. 

                                   And remember IMO- SS equipment should never be switched off.

                                        The company I bought all my input equipment says the same -CYRUS Hi-Fi-UK

post #53 of 69
Thread Starter 

I said in a previous post that to work on the Stax 717/727 1/2 energizers that you would need  a higher impedance than your normal MM of 10M ohm input  because of the very high  impedance of the circuits and that 10M would cause the safety relay to cut out the power to the units inside.  I said I had a HP bench MM industrial version of very high input impedance.

     But I bought an Avo MK 4 VCM and noticed on the sites on that web page that Valve/Tube Voltmeters  were for sale . While these are old technology they still beat  most SS ones as far as input impedance is concerned and unlike SS measure a high voltage with the meter set at a low one only the needle would be damaged as the valve are nearly "bullet proof". At the moment they sell cheaply on EBay and others as nobody wants them but those unable to work on energizers because they don't have the right MM this is the answer. REMEMBER --there are MANY circuits for them ONLY buy ones that have an input impedance of say around 100M ohm . Many old Heath-kit  only have a 11 M ohm  input -not high enough.

     The only hard part is you might have to adjust a couple of controls  to set them up.And you have usually bought[depending on the circuit] a TT or two.  

post #54 of 69
Thread Starter 

Further to my above post I contacted a ex-government equipment seller in the UK as I saw  a Bradley CT471 very high impedance meter for sale[ 100M  ohms above 12V ]  No price was on it as it was on sale . Phoning them  I found it was selling very cheaply as it did not have a probe and no front cover [don't need the front cover and unless you are testing RF equipment   don't need the probe]. It was a total of including shipping of £37  probably $60 or so I ordered one. I have dealt with this company before all their equipment is guaranteed to work. That is a bargain! as one sold on eBay in ropey condition for £150.

                I don't want to mention the company as some business people might want to buy them as they have more and resell at a profit.

                     But if any Head -Fiers are interested I could  post it.

post #55 of 69
Thread Starter 

Further to the above post . My Bradley Electronic MM type CT471 arrived today cost £25.[approx $38]. Works just fine . Was able to test high impedance  in the 717/727 [  12V and upwards at 110 MEG input impedance]  As nobody seems bothered then the company is Stewart of Reading.co.uk  ex government electronic equipment seller for past 30yrs all equipment guaranteed working.

post #56 of 69
Thread Starter 

I realize that the feedback resistor I fitted isn't a complete answer. So I made changes to each channel PCB. It wasn't logical that both feedback's should work together on the same connections So I  decided to "go the whole hog" and rearranged the feedback to something I am sure will meet with universal approval. I cut away the original feedback connection to the inner BJTs and  connected the ends of the 2 -150K+ 150K resistors directly to spade terminal 5[from the end of each side count the spaces as well] So that leaves only them and not the inner circuit connected directly to the output on the mother board.Doing this removes any conflict with BOTH sets of feedback still being attached so now there is only  TWO feedback's from each end of each card-5 in either side. Which is like the original 717 design  --as far as feedback goes- NOT IN OTHER RESPECTS. There is also a circuit diagram on this website supposedly showing a 727 diagram- This is NOT an ORIGINAL STAX 727 diagram - There are various changes to component  numbers  even if you eliminate  the added switch section . So beware if you are using it the input side also has some differences. Tested on my scope with a 1KHz square wave it looks pretty good as there now is no longer any conflict with 2 feedback's an "improvement" should be heard . YOU decide just how much once you do it as I have been criticized for sounding SUBJECTIVE- but maybe that's because I am not a robot who hears zeros and ones . Most people on this website talk subjectively try talking another way and still describe the sound quality . I mean describe by  human ears - not my or others THD meters.I would like to add that there are 2  small value caps connected to both plus and negative INPUTS. These are bandwidth limiting caps  or called "low pass" filters They are usually there to help stability BUT they do impact on  SUBJECTIVE sound response checking out under my scope  the amps will work without them and the square wave now has sharper edges.--meaning extended response and quicker delivery.- but its up to you as to the amount of minute detail you want.so you you could try it if it doesnt suite then its easy  to replace them.--- I should also have added that the verticals can hardly be seen when caps removed signifying a much faster  "rise time" My scope is a  Tex -150 Meg with 1 nanosec rise time so it is not limited in its speed of display.Those with "virtual" scopes on PCs might have a different rise time but I dont think it equals that speed although will anybody specify the type and make of the virtual program so I can check it out.


Edited by duncan1 - 7/10/13 at 2:17pm
post #57 of 69
Thread Starter 

Well that's about 2 weeks of listening trials since I reverted the 2 -150K resistors back to where they originally were in the -717-BUT  remember the rest of the feedback between both the 717/727 is different. Actually there are now a lot less local feedback caps on the 727 than on the 717[check it out-small brown caps] and that's the way I like it as it proves it is more STABLE than the 717[although I not saying the 717 isn't stable] its just that the 717 NEEDS those caps when the 727 doesn't. Which is a good thing for openness and clarity. Every small comp caps implies a COMPROMISE in the design . If you are thinking of buying an amp check out how many-LOCAL feedback  there are  to see how good the design is. The best have very little signifying good design in circuit layout and parts used. There is now more full negative feedback [from output to input] and it shows in even more clarity without sounding--edgy/sharp/etc as more TOTAL feedback doesn't just lower the THD but also the noise floor.The music flows beautifully and realistically and does full justice to my Stax 007-MK1s . While the 009s can sound good on tube technology the same -IMOP-does not apply to solid-state.unless it has been "tuned" to sound like tubes.by that I mean some "dumbing down" of the signal. I am even happier now with the results. .I might add that the +/- 15V supplies that I separated and used 2 regulators  in-EACH  channel with smoothing caps both at the regulators and attached to the channel cards socket pins means that even on my lowest setting of my scope  I can hardly see any AC ripple and remember that's PEAK/PEAK.[RMS value much lower.] ---In case some have not read my previous posts on this . You MUST CUT-OFF[use a sharp knife] to cut the printed circuit copper that leads to the original feedback BEFORE connecting the feedback resistors to the original position of the feedback  as on the -717.


Edited by duncan1 - 7/24/13 at 1:39pm
post #58 of 69
Thread Starter 

I contacted the company I bought the 727 from to tell him that I realize that I have lost the guarantee  as I had modified it and spoke to the owner. He told me that an Italian had bought a 007 MK1 from him as they are very popular and that the Italian had modified his 727 and was very pleased with it. So I am not the only person to make some major changes to the design. Its funny when you start doing something the urge to carry on and do more is strong. So in the future I intend to  convert the power supply to a stabilized version. This isn't as hard as you think its actually pretty simple as I wont be removing any parts and will just add it in series with the high voltage fuses on a small PCB . The only amount of bulk would be the heat sink and as the current drawn is only approx-60 MA it wont have to be large.

post #59 of 69
Thread Starter 

One of the faults of all amps [power] is  correcting the offset . Many things have to be taken into account= the identical spec. of the input pair/ the amount   of current through them/the stability of the CCs/ variations of the power supply and so on . Careful design in the choice of components and actual circuit design itself. There are ways of combating that by a DC servo and other means  but in this day and age as things get smaller and smaller.this means that all components must be of a high caliber. If you take the actual adjusters of a standard offset adjuster the small preset resistor. physics cant be overcome by this means of adjustment. Small track means small selective area of adjustment  That doesn't help to keep the set adjustment steady[within a small range] .I cured the same type of thing in my old cheaper of the two tube tester-Taylor   45C when testing large pentodes with large gain -needle couldn't be adjusted by "piggy backing a small value variable resistor to the main one --it worked.So in old/low tech. you use either a small value resistor to help steady the adjustment OR you remove the small variable resistor and fit one 3 times as big --much longer amount of adjustment due to the longer TRACK OR you fit a HIGH quality -10 turn pot. These will all give you finer adjustment and less chance of the offset straying  into many volts Old fashioned yes/ old technology yes/ too simple yes but not for me.Because it works for a small amount of soldering.

post #60 of 69
Thread Starter 

This might be a big money saver for some technically proficient Head-Fiers . Having had a close and extended look at the two circuit boards on each model of the 717 and the 727-11 I have to say that's its perfectly possible for a 717 board to be converted to a 727-11 board with someone who has done many changes  to electronic equipment over the years  and has various test equipment Both motherboards have differences but they are in positioning of the volume control and the cut-out relay . Why do you think it can easily be converted back  from the 727 negative feedback to the 717 version?. Many companies save money by keeping most of the layout the same on PCBs but make smaller changes to them look at the two motherboards . The essential ,components to make it operate are the same it hasn't changed . Think of saving about $2000 by DIY -BUT ONLY to be undertaken by those with years of training or knowledge of electronics and have the test equipment.----Also you will need the original circuit diagrams for both Remember ORIGINAL. Unless you like me can "read" a circuit  just by looking at it.----I will also add that only 20-25% of the layout needs to be changed . The biggest is the removal  of the last two-C5466s and replacing them with the new type . You notice heat-sinks are now fitted to them in an earlier post I fitted heat-sinks to the output BJTs as Stax didn't fit them-[717]. The more heat produced the more low level noise  you might not hear it as noise but you will hear the sibilance  rougher. And you wont need any comp.caps[small brown objects] which is a good thing because it implies a more stable circuit and comp caps effect the sound quality . That's why when they are used in the neg.feedback they must be top film cap types and small value local feedback caps should be polystyrene. They seem to have shot up in price used to be cheap not now.------I had a look at the costs it would be no more than $50 for parts and you could get better quality  heat-sinks with wings than the right angled aluminum plates fitted[better dispersion of heat]  The only real cost is time taken and precision build no cutting corners or trying to do it fast. Take your time slower is better here--less  mistakes.------Its occurred to me that there must be many enterprising people in the US or Australia and that if you charged say $450 per conversion which would still be a bargain for someone upgrading to a 727-11 as compared to buying a new one.I am sure somebody will see the financial benefits and if they buy in bulk even more of a saving in initial outlay.  ---Or they could include the mods I have mentioned above  as a "stage 2" with an additional cost of say $200.


Edited by duncan1 - 8/7/13 at 4:57am
New Posts  All Forums:Forum Nav:
  Return Home
Head-Fi.org › Forums › Misc.-Category Forums › DIY (Do-It-Yourself) Discussions › stax srm717/srm727-11