Don't Blindly Substitute Transistors - Marantz 2215B Restoration Gone Wrong And Other Mishaps

Need tools? Here's a video showing what I use (with links!): • Must-Have Tools For Wo...
In my work, I break rules and make mistakes all the time. Whenever I do, I try to show it in my videos so others can learn from my mistakes just like me. I found a list on AudioKarma of replacement transistors for the Marantz 2215B. It recommended replacing all 2SC1384 transistors with KSC2690. This video proves that this is not a good replacement.

[@branhicks]

I like that you show your mistakes. I tend to get discouraged easily when doing stuff like this.

[@robfors]

I always check the operation after I replace a transistor or pair of transistors. Much easier to diagnose! We all have been there and it is always good for a refresher. Thank you for your excellent videos!

[@rlgrlg-oh6cc]

The person who designed this amplifier put a lot more care into the details than you may realize. If you use new transistors with "better" high frequency response, it can make the circuit unstable, leading to the oscillation you were getting on the waveform peaks. Your waveform was also showing what is called "crossover notch", which is caused by inadequate biasing of the class AB output transistors. There was pretty much always an adjustment for this in hifi amps and receivers, as it depends on the characteristics of the installed transistors. To replace any transistor, ideally you should compare the data sheet for the original part with the new one to avoid problems. It's my understanding that transistors don't "wear out", so replacing ones that are working fine with different types is probably not a good idea.

[@SaturnV2000]

The correct procedure: Download the datasheets for both the original and the given substitute. Compare operating parameters and make a determination if it's suitable. Don't take anyone else's recommendation without vetting it yourself. Fortunately, I have a knack for locating sources for some of these very old transistors, so RARELY do I need to substitute. I highly discourage substituting for the purpose of "upgrading". Plenty of work was already done in the engineering and design stage. If the manufacturer made a mistake, it would have been identified along with a remedy in a TSB (Technical Service Bulletin). Also - be aware that the Chinese will counterfeit anything. They will grind off numbers and re-stamp a .05 cent transistor with the numbers from a popular high-demand original part and sell it for .99 cents as genuine . . . and .94 cents is a big profit for a lot of people in that country.

[@williammanganaro9070]

Very good video and glad you learned from your mistake. Just some words on probing circuitry. For myself, to prevent shorting things when probing around power stages or anything with higher voltages or current potential, I usually instrument the unit (place probes) with power off. Place your probes (clips) with the power off, make sure they are not shorting to anything and they are stable, then power up and make your measurement. Before removing the clips when your done, turn the power off again then remove to avoid shorting out anything. I generally use this technique and I have blown things up much less frequently. LOL !

[@Brian-yt8fu]

Signal tracing with scope some good advice you gave. One person here suggests one long ground clip to the amp to ground it. This way you don't have to use a ground clip on your probe that might slip off.

[@g.fortin3228]

Nice video.. one of your best. Good teaching and can see actually what your doing more. And good explanations too. Much appreciated. I'm just starting working on audio, worked on many different thing in past but this is a new avenue for me. Your channel is a big help.

[@JayanthKDass]

You are not only a good Engineer, you are a good teacher too! Thank you very much!

[@jeffdowning4877]

You skipped the part where you go F#@!!! That is also a common part of vintage audio repair. Haha. :)

[@veitheld167]

You seem to make steady progress and it is interesting how you drive this home! However, a couple of times I was a bit surprised about how you approached this task, I expected you to refer to the standard amp designs more often when you see an effect and then look for the most likely causes. For example you had those 28 volts at the speaker terminals. Now, in most cases the output is decoupled through some capacitors or maybe a transformer, so I would have checked the output capacitors as the natural next step. Checking all the capacitors is a very good idea in any case. Then in this video you had these very pronounced crossover distortions. In almost all of these old amps the power stage is a push pull configuration of a pnp and a npn pair of transistors that needs to be biased. This is done by diodes or a Vbe-multiplier which is normally adusted with a potentiometer. I did not see you search for this part? Also, you can very often get the circuit diagrams on the internet and it should be very interesting to analyse these schematics. But, as the saying goes - many paths lead to Rome and all that counts is to get this thing to work. Im am very curious already to see the final result!

[@blackarrow8683]

Thank you!

[@emmanuelvallee5386]

Funny to watch yr video right now, because 2 weeks ago i was restoring the power board of a 2215 and did the same substitution of the shorted 2SC1384 with KSC2690. I got exactly the same distorsion issue on both channels when loaded. This is an oscillation that i fixed by increasing the value from 50pf to 100pf of C713/C714 the capacitors between base and collector of the VAS transistors H703/H704 that are in the negative feedback loop

[@WFTL14]

Nicely done, good to see you admit mistakes we all could make and have made. Nice troubleshooting on the bad transistor substitution.

[@pauldow1648]

Thanks for working on these marantz'z. We love it.

[@Ms_Dana]

I am restoring restoring a 2215b on my repair bench using a rebuild kit that came with the same ksc2690 transistors. So I will scope it and see if I have the same issue when . I finish. my boards took a bath in Gatorade so will see how it goes. so far sounds good but one channel is louder then the other and it is not the fader style balance . detox that and everything else . So a little circuit tracing to do .I noticed you turning the unit back and forth . a large lazy susan would be helpful for that . that is what I use for in chassis work on the repair bench . I think I am going to go with the black face plate on this one when I am done . I was going to recommend that you should get a variac in addition to your dim bulb tester / very important on vintage units that have sat for a while . Anyway good videos .I'll update in the comments if the 2690 work out ok after checking the bias.

[@geomorphdog]

For 2SC1384, I've used TTC004 (which Digikey has in stock) in an SX-828. BUT that unit only has one 1384, and it's in the power supply. I've also used it to replace 2SC1124 and 2SC484, and the complement, TTA004, for 2SA484. The SX-828 sounds great now, but YMMV. I'm always crossing my fingers with transistor replacements, especially since so many of the old standards have become hard to get.

[@treadmillrepair754]

I bought many TTA004 and TTC004 in Digikey for fix many Onkyo receivers. Cheers.

[@nealnappa2444]

I make it a habit now to check everything before I install it. That is a nice piece, I love Marantz receivers

[@R3TR0R4V3]

Nice work dude. At least you were able to track it down and fix the problem. 👍

[@imqqmi]

I'm no expert on audio electronics, but I think I would try to get characteristics of the transistors first and compare them to substitutes. The thresholds of the current amplification and linearity vary a lot from transistor to transistor and need the bias adjusted, even the bias may not be in the range the substitute needs so you'd need to change that as well. I even might characterize the original transistor if one is still in good condition and try a few candidates and see which one comes closest. It's a lot of effort of course. My respect for even attempting to repait it!

[@greebo7857]

This is great. I found in my collection a SuperScope R340-B receiver. There is pretty much no info around on it, but I see it is almost a twin of this 221B. So now I don't have to do so much guessing.

[@larryshaver3568]

I agree completely when i retranned my astro sonic preamp and stereo fm board i carefully researched which transistors were BEST for those boards

[@dirtydon8661]

Well done. Sometimes mistakes help to learn.

[@doubledigitdreams]

I’ve just got my hands on a 2215b just the other day but there is no sound at all. Both channels are dead. The power supply board checks out ok, it supplies the power amp board with 50volts but when I start to check voltages on the power amp board I’m seeing nothing. When I plug in headphones and crank it up to max I do hear music, very faint and distorted. Anyways just wondering if you could suggest something to get me in the right direction. I’m leaning towards bad outputs, just seems strange that both channels are completely out.

[@renejensen7212]

Hello, i hope to hear a little bit of what kind of, Ceramic Tube X Ohm 5% X Watts High Power Fixed Wirewound Resistor. should you, and recommend to connect a oscilloscope like you did ? I'm always staring to work at my receiver just hobby. thanks a lot if you give my some guidelines of kind of: what Tube to buy. and Ohm / watt. 1 or 2 tube.

[@jimhibert]

There are two pairs of supposedly matched transistors somewhere that are not matched. You are seeing crossover distortion which happens when the handoff from push to pull is not perfrctly in sync.

[@JMNTN]

You mean the npn and pnp right? I thought that as well

[@jimhibert]

@@JMNTN Yes. Exactly.

[@kurchak]

I was doing an LED conversion on a Pioneer SX650 and my multimeter probe handling was a bit sloppy and I shorted the light bulb tips for a fraction of a second and blew a fuse lol. After successfully replacing 24 caps with no mistakes. Then I go and screw up something so simple as trying to see which wire was positive and which was negative lol

[@rajendrabareto8065]

I have 2 nos c1384 and 2 nos c1509. One of my 1509 is replaced with bc639 and it sounds identical. Can I replace 1509 with 1484?

[@kkoller8952]

As previous poster commented I would check the biasing first before going thru the effort of replacing the transistor. Ive seen exactly what you found when the amp biasing current is too high! When changing transistors more times than not requires readjusting the bias!

[@Dont_Gnaw_on_the_Kitty_1]

So you replaced working transistors with different types and now the amp clips on the lower part of the waveform. Looks like you messed up a good working amp with your substitutes. Edit: Im assuming the output stage is push-pull? Do the replacement transistors in the lower half have similar gain to the originals? This amp didnt leave the factory with a fault like this.

[@jimcabezola3051]

So instructive! Excellent lesson!

[@jimpalmer1969]

I question why replace working transistors during a restoration? I was taught a semiconductor will work or not work. They don't wear out, in as much as another component in the circuit fails and takes out a transistor. I'm about to embark on my first restoration and trying to gain some knowledge from others. I've ordered full replacement transistors and capacitors as shown from my friends on Audiokarma. But now I'm not so sure I will replace transistors that are working. Any comments?

[@AHFixIt]

Some do wear out. I've seen it plenty of times prior to starting the channel, but here's two examples: kzbin.info/www/bejne/oIiylWWwrM9ln9E kzbin.info/www/bejne/jWbdf4KHhK5ge6s It's dependent on the specific transistor at the end of the day. TO-92 size are typically ones you'd want to replace. You find these in the preamp sections and the beginning of the power amp sections.

[@richgriffith5541]

Been there done that with my 2250B main amp board. I forgot to put back the thermal sensor on the heat sink and shorted out something and ended up blowing the fuse plus output transistors. Won't do that again!.

[@mumbles552]

Your going to have to get yourself some insulated mini grab clips to get into those tight spots. There's lots of them with different style ends on them on Amazon. Is this the same unit which went thru the carwash?

[@goodun2974]

E-Z Hook probes are a must, along with a series lightbulb current limiter.

[@0ctopities]

wow, i just found your channel, i love electronics and it is my work job also,.,.lol

[@gioscervelo]

I just screwed up on a transistor replacement now. Now its worse. Pioneer sx-434. Funny that this video came up immediately after I screwed this up.

[@jackgabbert3233]

Hi Aidan, Thanks for the reply and you confirmed what had thought you'd done. So your saying the ksc2690 were good in themselves - but that in this units circuit they did not play well with the other kids, lol. Ok please help me to understand why not? That answer could really mean some good learning! Best regards Jack. P.S. a bit off topic, but after watching your "tools" video I don't recall any load resistors like you had in this video. Could you also share the important spec's, appllication and source for those?

[@AHFixIt]

The "why" would be a question for another commenter or someone like xraytonyb, a very good DIY channel. I'm learning as well! Good point on the load resistors, they should be in there. Here they are: www.mouser.com/ProductDetail/Ohmite/C300K8R0E/?qs=tdviw6CI36f2RiBBjsEH%252Bw%3D%3D

[@jackgabbert3233]

​@@AHFixIt Yes, xraytonyb has a great channel. Thanks for the link on the load resistors. Have a great day. Regards, Jack

[@iceberg789]

service manual is written as per original components, which were dated back to old ages. newer transistors have lower Vbe, which may not comply with old manual.

[@jackgabbert3233]

Hello, I'm hoping to learn from watching this and I'm not clear even after reading the other comments what was thecause and fix? Were the transisitors chosen as replacements simply bad from the factory? Or was the type of transistor not a good match as a replacement? And if so,, could you simply explain why? Finaly did you re-install the original transistors - or choose a different replacement? Again, I'm trying to learn and understand, so thank you for your help. Regards, Jack

[@AHFixIt]

In restoring the power amp board, I decided to change the transistors even though the originals worked. I did this because transistors can fail. Some will argue that if they work, do not replace them. I don't always believe in that. I found a list of replacements someone recommended. I happened to have all of them on hand so I changed everything. However, one of the subs (replacing 2sc1384 with ksc2690) was not a good sub. It was not a good sub because it did not work in this circuit specifically. This sub might work in other applications, but this one, the 2215B power amplifier, is not one of those. I reinstalled the original 2sc1384's and all was well. There were two per channel, so four total that went back to original.

[@v12alpine]

@@AHFixIt I looked at the specs and it might have been differences in the VCE and VBE saturation voltages. The KSC2690 was probably not "turned on" enough as seen by the crossover distortion and jumps in gain as you turned it up higher. The rest of the specs seemed fine. Modifications to the biasing circuitry was probably needed to work right.

[@julianround4893]

I also shorted the outputs and one of the 2sc1384's by slipping the probe.

[@gioscervelo]

I changed just 1 transistor and blamo that channel is dead. It went from 6 mv offset on the right channel to 20 volts! Tried putting the old transistor back in and it is still screwed at 20 volts so this replacement blew something else down the line. Even the headphones out is gone on the right channel now. Looks like I've got a Pioneer Sx 434 for parts now.

[@rabbibacongrease88]

In the interests of high fidelity, it would be interesting to sweep it from 50hz to 10k just to make sure all is good. Sometimes it might look fine at 1K, but terrible at another frequency.,

[@marcdich9066]

well im sure your not the only one thats done stuff like that, anyway im 52 but when i was your age i was working on smaller electronics in high school back when you could take stuff to school sometime in the early 80,s anyway have a good one im working on hk 430 and replacing a 10 pin output transistor on a optonica

[@JazzCrazy]

Weird - I've used Fairchild/ON KSC2690 to replace 2SC1384 in three different receiver rebuilds over the past few years (Pioneer and Sansui units). I have not experienced the type of signal distortion that you are seeing at all.

[@wonderbars36]

Might be the hFE gain selection or V_sat can get pretty critical if things get touchy with feedback/oscillation stabilization, rail limits, etc. Depending on circuit. If this thing is a Sziklai I could see this happening even more so if the gain gets a little to wiggy (just typically the "PNP" side is pretty critically tuned). Anyway 2690? Yeah I've also encountered a not-so-great sub with it on an old Fisher. Just needed a lower gain one and it was cool after that. It is available in...I think 3 different hFE groups. Y-Gain, O-Gain and is it G?? Can't remember.

[@v12alpine]

@@wonderbars36 HFE looks good, pretty sure the V_sat differences was the issue. The rest of the specs were "close enough". Biasing mods would be required in this scenario.

[@Silent-Lucidity]

I'm not familiar with that receiver, but the reason for that output signwave distortion is due to biasing on the negative voltage. Adjusting the bias, if possible on that amp, would resolve the issue with the new transistors.

[@habsrepariert]

I agree ... this behavior looks exactly like wrong biasing

[@sirsuse]

Thanks Tyler. Good lesson to share.

[@AHFixIt]

My name is Aidan :)

[@sirsuse]

@@AHFixIt you sure? Lol

[@gordthor5351]

Look on the "bright" side. Now you will have good lighting to see when you replace the output trans.

[@janicehopkins4432]

A "dim" concept.😉

[@PioneerGuy581]

i blew out a to3 output on a 828 as i had the board loose it touched the dam metal as therei s only 1/8" clearance from the output pin wrap to the chassis. shit happens.

[@v12alpine]

Looks like the bias might be WAY off.

[@geomorphdog]

My guess: mismatch between Vbe and diodes in driver and/or output stage.

[@v12alpine]

bingo

[@jmormino5465]

Another nice one , thanks

[@petenamlook18]

Looks like you need some hook leads for your meter.

[@test40323]

How's that humble pie?:-D

[@scrook1027]

What is the rational for replacing working transistors ? Did anyone actually look at the specs and see if they are compatible? I can understand replacing the electrolytic caps as they do dry up and go bad. The old saying still holds true - "If it ain't broke don't fix it."

[@josephchavez4052]

agree!!!

[@AHFixIt]

Transistors do fail, as I've demonstrated in some Pioneer videos. I found a list of replacement transistors someone on audiokarma had specifically put together for the 2215B. I had them all on hand so I tried it out, assuming that someone had already done the looking to see if they were compatible. Clearly they didn't, or never noticed how their amp sounds like total trash after rebuild. That's actually how I first noticed the issue: music sounded very bad at higher volume. So I put a sine wave through it and read it on the scope.

[@scrook1027]

@@AHFixIt I agree, transistors do fail. I may be wrong but in my 45years of fixin' lectronics , the only time I've seen transistors fail is when they have been abused. I would never ever replace a transistor in any part of an audio circuit unless it failed, and then I'd do my own research to make sure the new part was the same as the old part if I couldn't get an exact replacement. I would however replace every electrolytic cap in any thing over 15-20yrs old.

[@goodun2974]

@@AHFixIt The only instance in which I would make it a point to replace "working" transistors in a more or less functioning unit would be if they are beveled edge paintbrush style transistors, and especially if the leads on the transistors are oxidized jet black. Those things often fail causing popping noises and DC offset.

[@linandy1]

I never blindly change components if everything is working. I just replace what is bad, in matched pairs if possible.

[@leguzin]

The good transistor is BC637 and you have to make shure that you have the good Hfe min and max type (Q,R or S)

[@HunterSkowronPDX]

did you go to school to learn all this?

[@vcp93]

Hey boys and girls, today's Magic Word is "Checklist". 😉✌️

[@Poppinwheeeeellllllieeeeez]

Live and learn. Afterall, its just an amplifier.

[@envisionelectronics]

So the KSC part was likely a counterfeit so your video does not prove what you think it does. The datasheet shows it’s a compatible transistor and from an engineering perspective (I’m an amplifier designer) it’s a good substitute.

[@흑석김용호]

I think that 1 bias voltage failure output transistor failure 2 emitter resistor open 3 Bad output transistor 4Fuse lamp failure

[@mat.b.]

AH break-it

[@shidongxu3410]

Why do people use 20mhz bandwidth oscilloscope to measure audio work ?

[@sanolivo6867]

I know man. They say transistors is the great invention. Not so. If is so great how come no manufacturing site can make two transistors the same. I have replaced some transistors with the same ones. And still have issues. What a pain

[@tobitechboy1461]

0:48 Click here at full volume to have your ears bleed

[@R3TR0R4V3]

Sweet! Thanks! 😎 And don't forget 4:03 too 😜

[@tobitechboy1461]

@@R3TR0R4V3 oh yup!

[@justovision]

Normalize going right to the oscilloscope.

[@justincase3880]

Hmm .. “here’s your sine” …

[@janicehopkins4432]

?should have been "internet comment of the day".

[@davewallace8219]

I think the soviets make be involved...

[@pauldow1648]

Tribulations......

[@manchinga1005]

oscillations

[@janicehopkins4432]

Fabulous song by the Polecats called "make a circuit with me" from around 1980. Fun and cute. A local alt radio DJ used to play it a long time ago.

[@richardadelberg8961]

Is this a channel to teach by mistakes? maybe I missed your objective. Why would one not confirm the proper transistors before replacing. Shotgun approach is simply dangerous practice which could lead to blowing output transistors which on some sx ….80 series are darn difficult to find. Try finding original 1280 ones

[@Skips1964]

Having watched Aidan's channel, I would have NO problem sending him work! In fact I have a Realistic STA-2000 that needs work. But unfortunately he does this as a hobby :(

[@Skips1964]

@@Magnusaudiofighter I didn't miss the objective or the point. The original comment was edited, removing a negative comment directed toward Aidan - which is why I commented the way I did.

[@goodun2974]

@@Magnusaudiofighter , my favorite definition of an "expert" is "a person who has made all the mistakes there are to make in a particular field". It's likely that there are no true experts because no matter how much experience you have there's always the possibility of making a mistake next time. I personally find that making mistakes is part of the learning curve, and if someone doesn't make any mistakes then they probably aren't trying very hard! On the other hand, it has been said that the definition of insanity is doing the same thing ---- or in my own case, making the same mistake ---- repeatedly, and expecting different results! 🤪 Keeping in mind that making mistakes is part and parcel of learning about electronics, use of EZ-hook testb probes and a series-wired incandescent bulb as as a current limiter will make it less likely for your mistakes to do major damage to the components and circuitry.

[@janicehopkins4432]

I am glad he shows mistakes. It helps viewers figure stuff out. And the school of hard knocks is the most effective and lasting form of education. Maybe AH's goal is to share his knowledge and maybe get some other young folks interested, rather than just show off. When I am too old, I might need someone younger to fix my vintage stuff!

[@thedistortedguy]

I got some , can i email you?