QUESTIONS arised - Unique sound of Pioneer Exclusive M4

Hi, ( here are some answers I found out later on post : http://www.audiokarma.org/forums/showpost.php?p=7405538&postcount=6 )

I recently obtained one of them. I gave my out of box and restored terrific sounding ta5650 to my brother being maybe a second best amp on planet. M4 rised musical flow to a degree of nuances. It had problem on left channel so I am restoring it. But many questions arised with it:

1 - Resistors on output and protection boards look to have a final blue band meaning they are 10ppm resistors. Is it correct? Fantastic, nothing special on the amp but sound and resistors. Simple classA, big transformer, 66.000uF, again nfb, simple differentials, nonspecial outputs, not regualted for output stage but drivers and pre stages, everywhere is full of sockets being far from simple path teory.. They are %5 but I see they are clasified ase %1 variances, everything is matched here and there even mostly equal values here and there. What do you think about x-x-x-gold-gold resistors on newer m4a? What are resistor kits that you use covering wide values maintaining 10ppm, and also at least %1 tolerance due to high cost of thermal unshift that all must bu used without being garbage?

2 - Do 2sc1116 and 2sa747 outputs have ranking like vfets or many others or all those sub numbers are only for production specific label? I always see another new number on them on adverts or pics.

3 - I think noone has a service manual for m4 amp. What do you think of correct bias and supply voltage settings if you worked on these? It i snot simple rail V /2 of course. 4 of 10A push pull are too many for only 50W x2.

4 - For pre stage and drivers, suplly has only 1000uF x 2 comparng to output main filter caps of 33000 x 2 . Do you prefer increasing it to somewhat 6800uF etc? Any safety issues? Output is not 33 times hungry than driver stage. rDver rails have 1A fuse for each, so?

5 - Is it better to replace all 40 years old small diods and zeners? Do you always replace rectifier diods with fast and more powerfull diods?

6 - I see pioneer replaced all ceramic caps with poliester film caps or nonpolar electrolitics on supply and output boards. Values are 10pF, 47pF, 68pF, 470pF. Is it always better to replace pF nF ceramics with metalised film or any simple poliester caps? No poliester or electrolitics thogh easily for these values. I replaced some 10nF's on power supply regulator with polypropylene available in my stock. I replaced 10nF paralel ceramics to rectifier diods with 400V ceramic ones. But I dont want to loose the sound due to changes.

7 - Pioneer also replaced 4,7uF 16V 2-3ohm esr tantalum input caps with standard nonpolar electrolitic. I know tantalum has bad reputation on audio path, but the amp gives nuances nowhere available. This means amp still has much more potential or we loose many things du eto electrolitics. I replaced them lower esr 0,77ohm 20V smd type tantalums. Any idea about Muse, black, silmic if you compared to tantalum?

8 - Do you think it is more stable and safe to remove output tr sockets put only isolation knobs and solder emitter and base legs directly by cables? Also cable to collector with circle terminal with the help of spring washer on it but under screw? Any side effects of this comparing to sockets? I did this for my ta5650, plus soldering poweramp board socket cables to boards, then fresh solders everywhere. It obtained way better bass depth and treble extension. Better control on bass, more details and involveness, simply beats taf7 all the way but dynamism that taf7 had only for first 3 hours effective on my perceive.

9 - Is covering whole roads of pcb with solder worse due to unmatching thermal resonses between solder and pcb? cracks due to classA heat?

10 - Do you know strange socket systems so that you can also simply solder pin of pcb to female socket after pushing female? It has nearly rusty 150 pcb pins, corrosive crimp points on sockets, too short cabling to try other ways. But this amp should stay with sockets due to test, analyse purposes. While female sockets maintain cables healty due to hard bends and turns around while working, also isolate free cables when pulled from pcb.

11 - What are your solution to easily get rid of flux residue after resoldering? What is suitable to paint pcb green as original also I think it provides electrical isolation on solder side? Any side effects of flux cleaner and pain? I newer use wd40 on switches monted on pcb, it gives oily resude and after months all the conductive dust are stuck on it, yet I dont know if it is perfectly nonconductive on pcb. I suspect some crostalk on repaired units are du eto this, only a suspect, I dont know.

12 - What is your 80mm fan (2x) solution to be able to use with amplifiers? I tested all famous pc fans but they are far from giving pressure to blow inside amp. Anyone succesful on using dead silence external pump system to pump inside classA amp? It has a bad input port, port must be at the back of amp, simple channel system inside. One of original fans run loud, teh other is not unhearable.

14 - I think of producing new protection board using Arduino uno controller, rather then traing to repair the original in analog domain that always shifts someday later again. Anyone exprienced on computerised protection system on amplifiers? Necessary data retrieve ways, dc ac analysis, ways of protections on amps, finding out location of defect, error coding, memory dump of defect state. There is none because it is too slow for real world?

Thank you for answers,

(Greetings to old "show us your sony" fans, the mad's of earth on 2000s)
Mesut B.

The only service manuals for Pioneer Exclusive equipment is in Japanese! These units were not sold outside Asian countries. It is rare and valuable. Do not modify it, it destroys it's resale value. The Exclusives were very coveted and produced in few numbers. It is special, keep it that way. If you want to mod something, please get another common amplifier and mod that.

Hi KentTeffeteller, you are very right on your explanations.

M4 and even much newer, optimised and modded version M4a are like ghosts, even japanese manual or any pic or explanation about any circuitry does not exists(there should be in paper format on some japanese services). Althogh vumeter board is the same as m3, main transformer, supply board , protection and poweramp boards are totally different than m3. M5 and m-z1 are another start on its own.

Replacements are necessary for 40 years old unit if it is not only a museum part to look at with full dried caps, some other that could not maintain their physical properties for a 40 years period. Even bias and supply voltages are not original on all of them currently. On the other hand, it still has a unique sound, maybe only a fraction of factory state. It is necessary to replace even defective parts after serious search, knowledge, with serious matching but under control against sonic changes. Modifications on parts may be done understanding its aim, current electrical aspects, its properties, current health of that part. For example pcb pins are not healty and modded by Pioneer on m4a quickly. Tantalum for input cap was posibly only choice at 1974 regarding its lower esr value, transformed to paper caps in M4a we dont know exactly why. But I replaced them even again much lower esr tantalum ones beating most of elctrolitics, by passing advancement state where m4a exists, still peserving original properties of original part. They were not matched and had 2ohm and 3,2ohm esr values in channels. It is neccesaary to see what changes if they are silmic II or Muse. However there is no specific reference speaker and source system to make sure mod is strictly betters previous state in all aspects. I dont have exclusive series speaker or p3a tt or scops, testers pioneer has. therefore modifications even with good knowledge are still so much missing and most probably harmfull on units as you mentioned.

When compared to amp8 site, I see it misses some of original transisors on its previously repaired protection board. Someone changed a blown resistor somewhere else but with a standard %10 250ppm one. On eof NTC s does not responde, I dont know iw other is shifted in value too. I hope thread becomes source to give exclusive amps another life, finding out missing things and solving issues on amps in a more correct way.

Replacing caps is OK, replacing transistors as needed is OK, but I would not modify the protection circuitry on this.

After educating a bit on protection systems, I again see that it is the most sensitive part of any amplifier to produce. Therefore they always suffer too. If someone send me transistor numbers on protection and value of ntc on poweramp board, it will be much easier to repair the board as original, currently nothing saves amp or speakers because it is totally discarded.


ANY INFORMATION / EXPERIENCE ABOUT EXCLUIVE M4 and M4a AMPS IS APPRECIATED


Pioneer at 1981, on m4a upgraded all seramic caps on poweramp boards to polystyrene, including tantalum input cap. Dissipation factor and dielectric absorbsion is much better than tantalum / electrolitic, on paper looks better than polyester or polypropylene. Continues to use tantalum on vumeter board. Tan7 used polyethylene(PET or PEN?). There is also PPS. I wonder about how tantalum sound on other amps. Feedback path is one important road that magazines miss, totally depend on parts on path. Answer what feedback does to sound may be hidden in what used parts here make to fed signal before giving to input with any phase factor. Bandwidth rised to 100khz from 70, cross modulation and thd are decresed to 0,01 from 0,1 at effective output much like with other amplifiers of the period. Driver transistors, differentials and feedback resistor values are the same, only the cap from 47 to 470uF but looks irrelevent to huge distortion optimisation. Reason is only cap types??

After a couple of months, working on my spare times, now, exclusive m4 is live. Answering my questions in the same order:
HERE ARE A COUPLE OF HELPFULL DATA FOR EXCLUSIVE M4 OWNERS

1-Resistors are so special, never replace them unless they burned. Amp is fantastically stable with those resistors and still every voltage adjustment can be easily done althogh they are designed in terrifically narrow limits. No abnormal shifts on values after 40 years. I measured & compared all of them. Big white 16 emitter resistors to common points on poweramp boards are all 0,7ohm! I replaced one that someone changed before. Replace any burned resistor matching to the other original one regarding push pull sides or nfb or anything else. There are some 10ppm resistors available (Vishay PTF56 or similar)

2-output transistors should be always matched as usual but my amp has unmatched ones, still sounds more than fantastic. You need to buy a total set of same batch 16 transistors, their ranking or production numbering is always something else than any one on advertiements. However I learned that there are fake sanken resistors, I see different logo font among avilable ones.

3- For service manuals, you need m-22 for poweramp, excluive m3 for vumeter board, and sa-1000, sm-3000 manuals to understand protection board well. After I drove schematics looking at boards, these 3 manuals are enough to solve issues, circuitry topologies are the same. m4 is different thing than m3, has one small brother m22, protection section evolves in time and unique on its own(m22 is still closest), only vumeter is the same as m3 but some resistor values. Power supply section is so simple , totally belongs only to m4, and works perfectly. In my amp, due to possibly replaced transistors in protection circuitry, PTCs need to reach to above 900ohms to trigger over heat detector. Trimpot+new ptc for suitable heat maybe necessary.

Regarding voltages, directly fed vumeter chips as in m3 board must use <18V typ 15V according to datashets. When you adjust b3 for 15, you get b2 rails because b3 is derived from b2. And b1 rails need nothing. Voltage pot on supply adjust + rail, and the other pot adjust - rail comparing itself to + rail, not a standalone +- regulator pots that behave standalone in time. Avesome, you get always matching rails by this. Under 97V ac, in my example, b2 is +- 38,5V matched. B1 is 31,4 and -31,5V. Matching b2 rails is more important than matching b3 if you have deviation from 15V at b3. I adjusted bias to 521mV. m22 uses 850mV but m4 possibly is not. It is only adjustable between 500 and around 750mV in my amp. Again looks not so wide, I think only for matching purposes always guaranting classA working. The other pot on poweramp is for dc offset, in my example I adjusted both channels to 0,0mV only 0,1mV change while working, fantastic. Adjustment is so easy too unlike many other amps.

4-I tried to fit 4700uF for prestage b2 filter caps, but there is no empty room for this size. It is hard to find axial good caps locally. Upsetting that my 7000h caps dont fit, too less room to work there, even impossible to measure while amp is working, very bad to service.

5- I replaced all zeners and diodes on protection board using more powerful ones, but they were all looked good. 1n4148 is suitable for standard diodes on amp, yet 40 years ago they used fast diodes!, chose even higher current fast diodes which beters 1n4148. Be very very careful on specs replacing rectifier diods, I let them original.

6-I replaced all the caps including seramic ones(with polypropylene or poliester) because pioneer also replaced them all on m4a. All electrolitic caps but big filter are replaced making an extensive value & esr matching. I used 2000-3000h lifetime at 105C and higher voltage than originals, yeilding much longer lifetime than egzotic costly brands, still all japan, mostly matsushita.

7-I replaced input caps again with tantalums, smd 0,7ohm, much lower esr than originals. Input cap is a standalone subject , so may improve but original was tantalum. I will try tantalums first on other amps to see sonic difference before replacing originals on m4.

8-Poweramp cabling sockets are removed, directly soldered, numbering them again and also gluing to board, no danger when there is breaks, no bends to deform them now. But currenly I let output transistor sockets as is, later I will solder them directly as most brands did this way on theier later designs. I saw serious benefits on my 5650 vfet amp, it got serious extension on both side.

9-To me, never cover all long road with solder, pcb is under huge heat, and road is perfectly flexible as is, but it is good to cover roads close to holes, 1-2 cm. etc. especially on socket sections which are under high pressure to crack teh road. ther may be hidden cracks, so I did at all socket points.

10-Never cut and directly solder cables to board unless you gule them. Sockets has 360C turn ability so cables never bends dangerously when you move and bend the boards. Also sockets are safe against internal breaks of cables near soldered section. The last benefit is for easy measurements and test on voltages. Thsi amp is problematic on these sockets. they measure 4-12ohm!!. unsoldering and serious cleaning of socket pins never heelps, sockets becomes even more loose thus!, tightening female sockets never helps again, they get even worse. It is possible to buy suitable and gold plated pin and socket for 4$ each, more than 100 so 400$. Your only solution is solder and glue them. But make sure you inserted pin so that there no space between base of pin and pcb, if you, this causes crack on pcb road when pressure is used on socket. Very long and careful operations that finally resulted with direct soldering for me althogh I finished soldering this way.

11- After vacuming and resoldiring every point, I removed flus residue again by hand without any chemicals against sonic elecrical anomalies, very tiring and careful job, I hate this. Be careful against bent legs of parts if yo need to remove it, very time consuming job, I dont know why pioneer choosed bending them. easy on first soldering, terrible on later servicing. Especialy pots needs huge care againts cracks on roads. First melt then correct leg a bit by anything, then use vacume iron to make them right angle.

12- I still need silent cooling method. This is a must, I still could not find better fans than original ones.

14-My protection board now delays on start, detects overloads, but thermal protection needs new PTC s so that should increase in value to trigger detetor by reaching over 900ohm in value, when abnormal heat condition above 55 C occures. C value is when amp is pen on desk so i snot perfectly correct. Maybe some modification of adding a pot will be necessary after getting suitable ptc. Dc detector of protection should work I did not test, no dc.

*Dont change alps pots, no pots I looked to be better than those. They have soldered carbon layers's contact metal to legs, next carbon is covered on top of this metal, no loosy part. Wonderful spring effect after 40 years too, and if you turn left and right several times, they scracth the dirt and works flawlessly, wd40 does not help on dirt.

*I bypassed input selector board and soldered cabling from input->board->realy->spk termnal. I dont need 2 inputs, low filter on a poweramp and sensitivity so that includies more than 20 bad contact points, thoose swiches were so bad and must be replaced but all is unnecssary, thus I got perfect immediacy doing this. I bypassed cabling terminal above relay too. It is impossible to clean it, wrapped cable connection is also rusty never resolderable.

*The issue on left channel was that, if you are not careful, you then screw white cable of output transistors to heatsink. If you lucky no short occures, if you lucky exact short occures and b1 rails's return resistor to transformer zero point simply burns so you see ther is something bad around. If you unlucky you only semi short that cable so that b1 rail decrases on - or + , bias on that side gets abnormal, so that channels gets more hot, also there is dc of 1V but sometimes while working. You cannot measure anything when heatsink is dismounted, impossible to find an error but you unnecessarily replace parts. If you mount, error pops up or not at first, but in time it may pop up sometimes. You may see no short when you measure etwen heatsink and any transistor in dismounted state or even mounted state. And you never think that a cable can be occasionally semi shorts throgh its isolation. You have 4 screws that may cause it, screwing only one doesnot pop any issue sometimes. Issue of risky cabling system.

Happy 2014
mesut b.

Dear Mesut bey;
i am an owner of M4.. are you in istanbul? I have a problem need your support. Thanks;

Well i couldnt find my problem's answer above

Yesterday when i was using my M4 the protection light suddenly lids and even after power off/power on or wait until cool down its didnt lid off. Now when i turn my amp its always in protection mode. is there any comment what can be the problem?

Thanks for the comments