Hello Tony, let me assure you, it is almost impossible to find somebody who is able to explain things as clearly as you do. Moreover it is your relaxed attitude that makes watching your videos that comfortable. Thus I'am listening to you very patiently even if the video lasts more than one hour.
I'm not one of you guys. Never have done anything like this . But my uncle work for sear back in the 50s 60s an part of the 70s as a TV an radio repair man . He fixed my radio on my 56 chevy one time . He was always in the garage working on something an had the ball game on . I loved him a lot he was a good man ! I fine this fascinating your kind of like him an a great teacher ! I wish i could start doing this but i'm to old my eyes hands cant do it . But love what you are giving to us . Thank You for Your HELP an TIME . Live Life !
My favorite "Amplifier" book is "High Power Audio Amplifier Construction Manual" by G. Randy Slone (RIP) published in 1999. This book explores all the various sub-circuits you'd find in most transistor amplifiers (including MOS-FET amps) with a minimum of theory and math. It's also a great book for folks who like to build amps using chip amps and/or the various class-D modules. Probably not the ideal book for restorations.
I've been binge watching your channel the last couple of weeks. A lot of it is over my head but I love seeing these classic pieces of audio gear. Thank you!
Recommended book: "Designing Audio Amplifiers", by Bob Cordell. It is quite a brick of paper and rather expensive, but it is one of the most complete books on amps available.
Another option, depending on how many diodes there are in series in each package, is to replace them with LEDs. So two Si diodes in series is roughly 1.2-1.6 volt drop, or close to a red LED (around 1.5-1.7 volts). Three in series is probably close to a yellow or green LED, and so on. Just make sure to replace them in pairs (one on the - rail, and one on the + rail, for example), and pick out two LEDs of the same color with as close to a match for Vf as you can. Plus, they may even give you a visual indication if there’s a biasing mismatch (i.e. one will get brighter as the other gets dimmer.) HTH! 🤠
One of my first electronics instructors in college, back in the early 70s, was some kind of aerospace/radio telescope engineer. His favorite words were: step input, transient, roll off, instability, overshoot & oscillation. I sometimes wondered where he was coming from, and other times just didn’t get it (yes, I mean it the way I just said it). Now, you’ve given me a much better understanding. BTW, op amps and such (DIP chips), were just coming out (maybe just then becoming available at RS, & parts houses) .. he personally assisted me starting a sort of development lab at the school (Nearly 50 years ago). I remember it like it was yesterday, and have recalls from it all the time .. to this day ...
Many thanks for that explanation on crosstalk Tony! In my 60s and electronics is far easier on my arthritic old self than my previous interest have been, my thanks to you for sharing your vast knowledge!
This video about Transistor matching is a keeper for me. I added it to my favs and will refer to it for some time. I had a lot of questions years ago and less now. Other videos where way above my head. I love the level you show this material at Tony. It's very appreciated. Thank you.
Hi Tony, love your channel! I’ve learned so much from your videos. You work on all the equipment that I enjoy so much. I even own a number of units you have worked on including the Pioneer SX-1250 and the Pioneer SX-1980 as well as other Marantz and Heathkit units. I don’t mind at all when you go off on tangents and explain other related subjects such as books and test circuits. In fact I look forward to it! You nailed one of my issues with your discussion of stabistors. The old Motorola MZ2360, MZ2361, MZ2362 have been an issue I have had for some time. They were used on a lot of vintage equipment and have become quite hard and expensive to come by if you can find them at all. The solution you present is pretty slick. I don’t think I have watched one of your videos that I didn’t walk away learning something I didn’t know. Keep up the good work.
The NTE app is great. Occasionally their substitute will not work. Good example is RF driver and final transistors. They rarely work as good as a original. Otherwise RCA /EGC / SK / NTE substitute parts are usually just fine. They don't make parts. They are just relabeled parts from another company / manufacturer. For over 20 years I inventoried everything using NTE part numbers. For inventory control done with pencil and paper they have the easiest numbering system. I had everything crossed to a NTE number and had a spiral notebook with literally every NTE number written in it. Every part I could cross I did and it was put in the book under the NTE number. Beside the NTE number in the book I wrote the actual part number and the parts cabinet and drawer number like "R16". A few years ago I spent a few months and finally cataloged everything on a inventory program on a computer. Honestly I still use my old book inventory. It's just faster. Grab a phone or tablet with the NTE app and cross the part number. Grab the inventory book and look up that NTE number and find the cabinet and drawer number it's in. Make it simple stupid inventory system.
I am not an electronics person (Applied Scientist) but can find my way around a circuit. Your videos and explanations are a tonic. Your reference to Horowitz/Hill was much appreciated. I bought a copy about 30 years ago and it was great to have a reason to get it off the bookshelf.
Awesome video Tony, I would love to see a video dedicated to your transistor matching device even if it's a short one. You showed it on one half of the screen but I think that it deserves a lot more credit than that... I would definitely want to build that circuit ( the TTT Tony's Transistor Tester)... I have lots of transformers lying around but it's just a simple matter of how much voltage and current is needed to perform the test... I had a quick look through the comments to see if anyone else wanted to know more about it but so far it seems that I'm the only one!... Thanks in advance 😊
Super interesting video! Thank you again for sharing your knowledge and tricks. These stabistors replacement STB400 are somewhat more precise than 1N4148 associations, but are not as easily available nor needed for a higher precision. An association of diodes thermally coupled is always better than these specific old Sony or else components, and are a mandatory first thing to do one these amplifiers. Once they fail, you mentioned that before, and your output stage is destroyed. And when it comes to special Sanken or VFETs Like Sony or Hitachi once produced, as they are obsolete, non replaceable, and you can consider your amp is dead! Your transistor comparator is quite a wonderful and good idea: it's simple, easy to use and no high money involved. Even if it's quite difficult to get results on a low sampling rate digital basic oscilloscope, there's always means to get a proper display, as the sampling and triggering devices on these machines can also use statistical modes or filtering capacities analog units don't have. The budget is anyway different, and I understand the basic needs of this application can be reached with a very low cost, basic analog oscilloscope, a 45$ sovietic 1MHz type! As for your DCA 75 pro, the curve plotting and comparison charts you can obtain are always interesting, pßarticularly on small signal transistors. Of course, this is a pain in the a.. when it comes to compare a large batch of transistors, but you can still make an easy gain or hfe classification of a batch, as it takes no more times than your ZIF support to operate. Once they have been sorted, you can try smaller numbers of similar gain transistors, record and compare their curves. I must recognise your method gives interesting other informations, thermic incidence, slew rate... And that may certainly gives you a better confirmation of a perfect or near match. I usually couple these transistors thermically using radiator mounting grease, and sort of encase them in a tiny copper part of tubing. You can also encapsulate them in a special hard plastic resin, but some can be chemically aggressive on the long term. One of my crazy friend once used dental resins to couple FETs, as it can be UV cured and is notorious for neutrality. It seems possible to mount SMD transistors on a adaptor like the SOT xx you use for stabistors, and encapsulate them with silicon, but these can be difficult to source for direct replacement.
Thanks again for your efforts and your very detailed video. You also did a pretty good job in matching modern transistors exept for two in my opinion very critical positions: The replacements for Q13 and Q15 in the gain stage. The TTA and TTC transistors in my opinion are only suitable for driver applications in this amplifier because of their high Cob of 12pF/17pF which would create a dominant pole in the gain stage with a too low transition frequency. As you are already using KSC3503 and KSA1381 as predriver replacements you should really use them in the gain stage and for the predrivers Q25 and Q27 the TTA and TTCs. There may be another not so critical replacement error in the differential amplifier. The 2SA992 has almost double the beta of the original 2SA979. You could get away with the boost in open-loop-gain, but it might be a little too much. If my calculations are correct you are about to increase the open-loop-gain of the amplifier by about 8dB at 1kHz if you are using those high beta transistors in the differential amp. With best regards, Paul
Hi Paul, I am just replacing Q13, Q15, Q25 and Q27 with KSC3503 and KSA1381. But Im wondering why the manufacturer put different transistors on Q13, Q15 vs Q25, Q27? 2SA939 different than 2SA899A, and 2SC2071 different than 2SC1904A ? Is there a specific reason on the circuit? Regards
Hi fmoyano, this is a very good question but there is no short answer, so here we go: There were in my opinion two good reasons why Pioneer used the 2SA939 and 2SC2071 in the gain stage instead of also using the 2SA899A/2SC1904 pair. The electrical reason was that they wanted to use a transistor with comparatively high beta/hfe and the 2SA899/C1904 have almost too little for a perfectly performing gain stage (same for A1381/C3503) and the second reason may have been cost optimisation. For the gain stage especially in higher power audio amplifiers with high supply voltage rails we have the problem that the gain stage transistors have to fulfil at least four important characteristics -relatively high Vceo, high beta/hfe, low beta/hfe modulation over voltage and very low parasitic capacitance all in a package that can dissipate more than 700 mW. In the past there was a whole subgenre of suitable mid power transistors - all long gone. Today we only have the A1381/C3503 for this application. If you find some older stock with the E or F hfe grade you would have the perfect replacement. As far as I know there are only the D/E variants still in production, but fear not. I may have exaggerated the performance impact a little bit. We are talking about 0,0003% THD against 0,0007% at 1 kHz in the case of the Pioneer SX-1980 (when driving the power amp directly) But if you plan to use the 1980 as a whole: The preamp has far worse THD and noise, so these figures should not be meaningful in any way.@@fmoyano
A tip on using modern digital oscilloscopes. If the waveform is extremely complex and fuzzy looking making it hard to compare two waveforms you can clean / clear it up a lot by going into "average" mode. The acquisition mode is usually found under the "Acquire" menu. Also play with memory depth, sampling rate and anti aliasing settings . It can also greatly effect waveform appearance.
Thanks, Mike for all the great tips! I'ts really good to hear from you. I hope all is well. I'm always looking forward to another video on your channel (hint-hint). Thanks again!
The BC500 series of transistors do this "gain" suffix also. They use a trailing letter designation. I wonder how close those two transitors were in DC beta..which is easy to test for. Even within the same suffix you will see differences in DC beta (typically +-50, deviation from the stated)
Watched this thinking you'd be covering the outputs on the sx-1980. Early in the vid you said you'd re-post clips from the big sansui to cover it....... I was one of the people that commented on the sx-1980 outputs way back and it doesn't seem like this really "fits". Now, I could be wrong, but I think what some people (like myself) might have been driving at is what can you use as an alternative output transistor in the sx1980? They are a 4 pin dual/common collector that are difficult at best to come by and when you do, cost a LOT of money to replace. Back in the 1990's I had not one but TWO 1980's that needed service and the outputs were still readily available then. My tech at that time said they're getting scarce but there are ways around them-if you get my drift. Not sure if pioneer had a special reason for using them or not, but it sure seems like bgw did in their 750e versions-as well as it's brother (750d?) w/o meters. I got a really good look at the circuit board in my 750e and it LOOKS like the main reason for choosing these same outputs was simply to daisy chain the collector voltage easily from one output to the next around the heatink. Given how hard original sx-1980 outputs are to find-not to mention the expense, IF there's a "sound quality" issue I think most people would make that sacrifice just to get the thing working again. The 1980 had built in current limiting as one of it's protection methods, and people with a fine ear can hear it kicking in. I'm kinda doubting using robust 3 legged output devices would make much a difference. Look into this if you can and make another vid-not for me (both my 1980's are LONG gone) but for others in need of a cheaper alternative..............
"the main reason for choosing these same outputs was simply to daisy chain the collector voltage easily from one output to the next around the heatink" IS the reason. You can use MJL4302AG and MJL4281AG and it works perfect. I made the comparison side to side and the sound is equal on one channel with 2SB706A, 2SD746A and MJL4302AG, MJL4281AG
Wow thank you for the transistor matching lesson. What would happen if you clamp, wrap the 2 bodies with a rubber band or something, together? Would that keep them in uniform better? I guess that application would only be good for the packaged type. I don’t know, thoughts?
I just wonder why power output stages cannot be implemented using a stereo ADC a Texas Instruments Blackfin DSP processor and power DAC and program the simulation software in C++ and compile it into the object code for the DSP chip would that not be possible as something like what I am talking about is in the output of the YAMAHA Digital HIFI systems.
O Carlos Mergulhão que me mandou aqui. Não posso agradecer lá no canal do Carlos porque ele não quer que comentem sem foto no perfil, então coloco aqui. Obrigado Carlos.
Due to picture not being very clear i cannot read of screen what diodes are used in transistor matching jiig. Can someone help ? I can see is IN4148 ??? not sure tough
The appeal of stabistors, is that they have a very steep I-V curve, and so are better regulators than ordinary diodes like the 1N4148. A stack of 1N4148 is not really an appropriate substitution. Great idea to use the SOT-26 breakout boards! Here's a thread on DIYAUDIO where we discuss these stabistors. www.diyaudio.com/forums/solid-state/193699-adcom-gfa-565-diodes.html I pronounce it "stay-bis-tore", because I think the name riffs off of the word "stable"-as in-A voltage regulator using a stabistor, with it's steep I-V curve, will be more stable than one with an ordinary diode.
Who else has an AV amp that you could cook egg on it . I only use 1 pair of speakers on it and it roasts , I would not use 9 speakers on this ONKYO . The fan will not kick in . Summer as past WOW now I can use it .
Wait a sec, you said you could trace signal and voltages to individual components..but to what end?...How about the end of getting it out the door!? i have worked on a variety of old school amps like this, you'd be amazed what a heat gun and freeze spray will show you while monitoring your DC offset. Im willing to bet you have a resistor that's gone thermal(thermalaly sesnitive) or a maybe a transistor. Replacing all the components on this is a waste of time and money.
If this were just a normal repair, I would agree with you. Most units that are sent into a shop for repair are units that failed under normal use. This unit, however, was subjected to water, rust, mouse droppings and who knows what else. If one of the transistors fails, chances are that others on that board will as well. This is why I have subjected this unit to so much testing. I don't do this for a living and profit is not a concern in this case. My goal is to get this unit to be as reliable as possible and to share some techniques with others along the way. It is only a waste of time and money if this was a job for profit or if the owner didn't want it done. Ultimately, the owner decides how far to go with the restoration. Thank you for the comment and thanks for watching!
