0:15 Calibrating offset & bias on power amplifier 4:39 A little bit of fun with RUclips Audio Library song 6:48 Testing the output of the amp with 1kHz sine wave with explanation of test equipment setup 14:19 Transistor temperature: How hot is too hot? 21:32 Checking THD at 1kHz & rated power. THD is tested with Keithley 2015 out to 5th harmonic. 23:58 Frequency analysis from 20Hz to 20kHz using the Instek MDO scope and FA software 29:03 Slew rate calculation and testing: How slew rate determines max frequency response. 44:53 Closing comments
Great video on the sx980 Pioneer. I had it . I wish I had it today. But I did it in with tuner cleaner. I paid 600$ for it in 1978. Loved the video. Dennis. Thanks.
Love the commentary and your videos. Please keep them coming. As to the two capacitors at the input to the amplifier that are electrolytics mounted back to back to create a nonpolarized electrolytic of a smaller value, it may have been done due to cost. At the time this receiver was designed large value film capacitors may have cost more and they were just doing it to save money. And as it is much of the time when electronics are designed, you utilize parts that are readily available at the time you're producing your device. So what I'm saying is it may just have been bean counters or supply issues that were inputs to the engineers decision to do what they did. I've worked in electronics for over 40 years and I've seen it time and again where parts chosen to produce whatever unit you're engineering were chosen due to cost and availability.
I think you are correct, in that they are wanting some ESR at the input. A good experiment you did replacing these with a 1uF film cap to see what happens. So, it may be a "sonic tuning" sort of thing they are doing. Similarly, I have found from my experimentation that using a lone electrolytic cap for the feedback loop produces much better sound than say, an electrolytic with a film bypass on the feedback cap with simple long tail pair setups. The results have been consistently better sound in doing so, and a nice audio grade electrolytic here, works very well. So it may be safe to assume that some ESR in these signal path locations may produce a better sonic result for some amps. I also believe you do not need a bipolar cap at the input, so simply adding ESR is likely all they are doing, using 2 small value caps. I think you should do a critical listening test comparing both the electrolytics and the film caps in this location.
Tony, Thanks much for sharing your wealth of knowledge! My biggest challenge is remembering which video has the relevant info. The time stamps in the comments help, Thanks! In this video you replace the heatsinks on the TO-220 transistors in the power supply. Where do you find the tall extruded TO-220 heatsinks? There's a ton of TO-220 heatsinks out there, but having trouble finding the tall small footprint that will fit in these receivers. Thanks, Todd
This has been a very instructive video, much better than watching you solder components in. The tests were not strange to me but I particularly liked how you worked out mathematically what the reading should be, Then ran a test, thus proving that the amp was working beyond pioneers specs. Thank you Prof.
Just be good to me. "Great song by the way. I love it." Your cool points just went up by a factor of 10 with me. Excellent instruction and excellent taste in music!
Really appreciate your style of explaining the electrical knowledge you have to help us dummies. Kinda like the books they like football for dummies etc. You should write your own book
Excellent as always Tony! Especially liked the slew rate test. I was lucky enough to own a Sansui G22000 monster receiver in this same era (1978) which had a super-fast slew rate & 'rise time'. Suffice to say, it was an incredible sound reproducer, and was my replacement right after my SX980. Would LOVE to see you work your magic on either the G9000, or (if you can get hold of one!)
Great video Tony. The SX series seems so well designed. I know it's not your thing, but I'd love to see you test (compare) a contemporary stereo receiver or AV receiver with one of the Pioneers.
You are great . An I like your teaching. I had that sx980 Pioneer receiver but I ruined it with tuner spray I wish I had that receiver today I enjoyed it and 80 watts per channel is all you need as you know.
I believe I know a (partial) answer to the "series capacitor" design. The problem has to do with how much DC is being applied to either end of the caps under normal use. If you have very little DC relative to the cap rating (across the cap or caps) you want to use the "series design" (ie the virtual non polarized cap) in order to minimize any "wave bending" (aka distortion) around the 3db (f0) frequency which is typically between 10 and 20 hz for a first order high-pass filter. (which is what most of these circuits are...) The problem has to do with "asymmetric charge distribution" inside the cap when the internal polarity inside the cap itself is changing and how this may affect the overall shape of the output waveform. Electrolytic caps basically "like" to be properly "polarized". They don't really play nice when just lightly or slightly biased and they have tendency to "oppose" the change in charge of slowly varying ac signal (ie low freq audio). According to design engineers tantalum caps respond much quicker to this low freq AC charge-change around the transition frequency. However the response is still "asymmetric" which is where the second capacitor comes in. ("asymmetric" as is..the bottom half of the waveform will be flatter than the top half) This design dramatically improves the "linearity" of low frequency AC signals around the 3db point. The output signal should just be attenuated..while NOT distorted (waveform bending). You can can actually set up a simple filter circuit to test this. I used about four different caps between 1 and 2uf and 10k resistor. 1uf&10k gives you an f0 (-3db) of around 16hz. The trick is having a decent analog (dual trace) scope where you can visually inspect the input and output low freq (< 20 hz) AC signals. You can visually see output waveform asymmetry! (Most digital scopes can't accurately sample below 20 hz.) So the series design is basically a distortion minimization "trick" when the DC bias is relatively low. If the DC bias high enough above maximum (AC) input signal you won't get this charge reversal affect and the distortion will be much lower even without the second cap. Sony used this design on their pre-amp output stage way back when and Randy Slone (RIP) does it on his power amp input stages with tantalum caps.
I love way you explain things. The only thing I see is that, when you use your test equipment , like oscilloscope or spectrum analyzer, show us where to hook up or connect to. Too someone who new like me will not know and that would help us.
I have also done a test and calculation of the output power of a sx-980 and came to approximately 85W into 8 ohm dummy-loads. Using an analogue scope and reading to the point of flat-topping. The transistors in the PSU of this amp is coking, i think I measured them being close the boiling point (100 C), but they seem to last. this is the original transistors on original heatsinks.
Great Vids Tony many thanks for all your work, time and sharing your knowledge, we all appreciate it, all the very best to you and your family for Christmas and the new Year, lets hope we all have a better one (Covid wise!) than 21, take care from the UK
Tony, when you were measuring the slew rate on a square wave, how much of the delay was in the signal generator? I would like to see the the signal generator connected directly to the o-scope. Would you subtract the signal generator delay from your measurement?
Good question! The signal generator I used had a frequency range up to 160MHz. As this signal generator has a maximum output amplitude of 5volts, the slew rate would need to be 5024V/uS to achieve that frequency. This makes the delay of the signal generator at audio frequency negligible. Additionally, the cables will have a velocity factor that will add delay. The cables will also add some delay, but again, it wouldn't be a huge factor when compared to the amplifier's components, such as the capacitors, miller capacitance of the semiconductors, etc. Hope that helps.
Another great video, I'm following this one closely because I have the same receiver. Question: where did you get the heat sinks, for the transistors in the power supply? Thanks
I have a bunch of those heatsinks. I got them on eBay. They also had the 1" tall ones as well. I think I got 100 of each at the time for a really good price.
The slew rate for a Crown CT-210 is "Greater than 17 volts per microsecond" It is rated at 105 WPC into 8 Ohms or 135 WPC into 4 Ohms at 20Hz-20KHz. That's more than twice as fast as necessary.
Great as always. I learned a lot. As a general comment, it would be helpful (at least to me) if you listed the major items of test equipment used by manufacturer and model.
Thanks for the comment. I actually did this in a number of previous videos. I always have to leave some things out because I don't always have the time to film everything. If you have any specific questions about a particular piece of equipment you see, just post a comment and I'll try to answer you, as long as I see the comment. Thanks again!
This is a great video sir! I've long considered a THD Analyzer for my bench to prove the health of my projects aside from verifying offset, bias, power supply voltages etc. If you could make a similar video about THD calc as you did for the slew I'd buy you lunch!
Apologies for the noob questions in advance! "We are going to apply a 1kHz sine wave tone into the auxiliary inputs of the amplifier and we're going to use about 500mV of drive signal". Why 500mV? When I look at the specifications for the SX-980, I see that the input sensitivity is 150mV for Aux/Tape, 7.5mV for Mic, and 2.5mV for Phono. Are those specifications minimums and is there a maximum? Can you elaborate on the, seeming random, choice of the 500mV? The phono overload shows to be at 200mV or roughly 100x the input sensitivity. The Tape and Pre out levels are listed at 150mV and 1V respectively. Would just like to know your thoughts on all of this. Thanks Tony!
Those are the minimum input levels required to get the rated output. Different amplifiers will have different amounts of gain, which will affect the minimum signal required for rated output. I usually apply a signal level of 500mV RMS to line level inputs as a standard. Most line level amps/preamps will have a range of anywhere from 150 mV to 1.5 or 2 v. If there is no attenuation control on the amplifier, then I will vary the output of the signal generator to get the output level I'm looking for.
40:30 No confusion at all, it's all very interesting; you're just calculating the gradient. Theoretically, the sine curve is g(f,t) = sin(2πft); so to calculate the rise time, simply take the first derivative (to find the gradient) and multiply by the peak voltage to get the rise time in V/s, i.e.: dg(f,t)/dt*Vp =2πf cos(2πft)*Vp. At t = 0 (max gradient) for 20000Hz the gradient is dg(20000,0)/dt = 2π*(20000)*cos(0)*Vp = 5.027V/μs. On the oscilloscope it's just ΔV/Δt to get the gradient.😁
The principal failure mode of capacitors is short circuits, so it makes sense to use two identical caps in series as a safety measure against failure of one cap due to accidental over-voltage. Generally, components that fail short circuit are used in series (like caps) and components that fail open circuit (like resistors) are used in parallel as a safety against accidental failure. Two identical caps in series have double the voltage rating and two identical resistors in parallel have double the current rating, adding to the safety margin .
Thank you for another fantastic video, you work so hard on them and are a a superb communicator. I don't really have much electronics knowledge but over the years have as a consumer of mid range audio gear picked up the ideas behind the measurements. I am curious to know what the significance of the phase shift curves are. They look quite alarming to to a person in the street and imply significant distortion of a kind not measured with fixed frequencies. Do these things matter or just an inevitable consequence of driving a load which of course would be far more variable in a real speaker?
Great video Tony as always, thanks! In the end you talk about the 1uF film capacitor to replace those two polarized 2.2 uF in series. I am dying to know the final answer to this question, if there is one definitive answer based on laws of physics. I would just replace those with the 1 uF film capacitor as you did. But from the schematics you see that on the right side of resistor R3 we have 0 VDC, and at point #6 we also must have 0 VDC, so the capacitor to take place of C1+C2 will have zero VDC on both of its extremities, therefore I think that the combination C1+C2 needs to be non-polar (or bipolar). (Given that point #7 is connected to ground per the schematics)
Well I dunno about "laws of physics" for parts choice all around everywhere but my likely bet here is cost. Working contract mfg. at a PCBA outfit myself, I've seen this countless times by way of using say 10k resistors in parallel to get other resistances without adding another parts feeder to the machine (SMD), series or parallel caps, etc. Likely they had a spec for a 1.0uF cap in that spot in the design. To get that easy and quick and with fewer added special parts, what they did was probably their cheapest easiest solution rather than up the BOM parts count. And as he mentioned, film caps were larger in those days so might have been mechanical planning/insurance in the event they couldn't source a certain package size.
@@wonderbars36 Hi Nonya, I agree with you about keeping costs down, in this case. I am saying "in this case" because if the equipment has to meet rigid specs, then it may not be so simple to do that. Tony's questions on why there are two 2.2 uF polarized capacitors in series in that part of the circuit included his sub-question of 'Why didn't they install a single 1.0 uF polarized capacitor in that place instead?' I have seen also many times companies like Pioneer, Toshiba, JVC, Sony doing that in their audio equipment from the 70s and 80s, I mean, installing a polarized decoupling capacitor in the audio path like that. In my restorations, I replace those with a film capacitor like he did (usually KEMET, Panasonic or WIMA) if there is space, otherwise I replace with a polarized capacitor (such as Nichicon) if it is in accordance with the manufacturer's schematics and original installation. When I mentioned "laws of physics", I really meant that because a polarized capacitor can't arbitrarily work under AC voltage, or can't work under AC voltage at all, some may under certain specific parameters or problems will occur. Since those two 2.2 uF polarized capacitors are directly in the audio path, I have to assume that the Pioneer engineers were 100% sure what they were doing. Per my explanation above, and per the schematics, both extremities of the capacitors in series have zero VDC, so Pioneer needed to install a non-polarized 1.0 uF there but they used 2 x 2.2 uF polarized in back-to-back series. I have seen single polarized capacitors (some 10 uF) in the audio path in my restoration projects -- in places that the capacitor has also 0 VDC on both sides, and will be under an overlaying AC voltage due to AC audio signals being injected.
Interesting video. I think you may have used the wrong value for the rise time when calculating the slew rate. Oscilloscopes rise time measurements are usually the delay from 10-90% of the min and max values. For slew rate, wouldn't you want the delay from 0-100%?
Hello , I am working on the SX980 power supply board and I am replacing the 2200uf 35v cap on the board. The factory cap has 3 leads and replacement has 2 . How should this be oriented on the board. I do know the stripe is negative but which hole on board should positive lead go to. Please help. Maybe you could show underside of board. Thank you. Chris.
Isn't it amazing! In 11 hours you had 1400 students attend your class! If you had any doubt of how good of a teacher you are, you shouldn't... I think that you really missed your calling in life LoL....
Thank you for posting Tony! Question; if a person were to take the time, and had access to enough parts inventory, would precision testing/matching components like caps, resistors, diodes, etc., be a worthwhile time investment when it comes to circuit balance/stability? Thanks for your time
That's really subjective. I do believe that certain components benefit more from this than others. Case in point would be the transistors used in the differential (long tail) pair at the input of the amplifier. Output transistors don't really need to be closely matched, as the output stage has little or no gain and they have the emitter resistors to "take up the slack". In addition, the PNP and NPN transistors will often have very different gain characteristics just by their nature. Coupling capacitor values are usually chosen to be high enough so they won't attenuate the signal much within the frequency range of the circuit. If they are a bit different, it won't make a big difference. Now if the components are out of spec (leaky/high ESR caps, leaky transistors, etc.) then it could affect performance. That said, there are some who say they can hear a difference, which is entirely possible. Coming around full circle, it's subjective. Not a big problem for "Tin-ear Tony" ;))
Hi Tony, thank you for all the hard work and effort you put in all your videos and shearing you knowledge onto us. God blessed you with an amazing gift , learning more every day from you. So I say God bless you and your family. Thanks
I have an SX 980 probably one of the cleanest around anyway. Started blowing fuses actually just one looking at the front it’s the far right glass 1 amp fuses is there a simple way to trouble shoot the issue I assume it’s a transistor? I have been watching your videos and tried looking for obvious issues and I cannot find anything obvious? I’m pretty upset it’s my baby! Any kind of advice would be wonderful 😳
I think also you have to make a critical listening test and also to make experiment with replacing the electrolytic caps with two in series film caps , not only one equivalent to see the sound and to measure also the distortion with the distortion meter. Why ? Because may be the sound will go more metallic with only one equivalent film cap , as stated the comments below.
Tuner Test comment / ? Tony can you bring your expertise to bear on answering a question? We can all get a good idea of evaluating the difference receivers by what stations they can pick up in our area. But using our test equipment how can we run good comparison tests between different receivers for sensitivity selectivity etc.?
As long as you have a signal generator that is capable of outputting an accurate amplitude and then make sure the test equipment is properly coupled to the tuner, you can make some pretty accurate comparisons. I did some videos on signal generators and impedance matching that might be helpful. Thanks for the comment!
Why didn't you use 50 watts plus transistors in the power supply won't get above body temperature and theoretically if you run 100 what you would need a heatsink at all I think with the capacitors using electrolytics non-polarized or whatever and film capacitors is just down to the sound of the amplifier signature, did you notice any difference between running a mono signal in both channels to your ears? the slew rate was slightly better on one channel which channel had the better slew rate with the capacitor the one you changed was it or the original capacitor? Wondering because it's sharing the same power supply would that mess up the results slightly might have been better running them as individual mono blocks through the same power supply and seeing what the results was each individually.
The power supply transistors get hot because of the power they have to dissipate due to the amount of voltage they have to drop at the required current. That power is converted to heat, which would be similar regardless of the power rating of the transistor. The more important factor would be increasing the rate at which the heat can be dissipated from the transistor. The heatsink and thermal interface have the most influence on this. Some transistors are a bit more efficient at this, but they will still get hot, especially if they are all in the same case type (such as TO220). As you saw from the charts, the transistor can only dissipate higher wattage when they are kept cooler through larger heatsinks. Although the transistor is roasting in this amp, it is operating well within its thermal specs.
@@xraytonyb I don't see how 100 watt transistor should be getting just as hot over a 25 watt transistor, 100-watt transistor can handle the current better then the 25 watt so at low power it shouldn't be getting as hot, they both can't be at the same temperature differential otherwise 100 w transistor overheat at it's rated current draw sure that's correct. You pump 50 watts into a 25 watt transistor and 50 watts 100 watt transistor see which gets hotter we both know the answer to that? Only if I'm missing something!
@@xraytonyb Thanks, I been working w the Pioneer Power supply and I think Pioneer had a bad design w all of them? I was thinking I did something wrong when restoring the PS, After thinking about it and w no load they get run Hot and than hotter, so converting 60vdc to 5.4vdc or 31vdc is like a Big Hot Converting Resistor Transistor ,,,,,, and has nothing to w Amps since output draws low current. Tony have you ever thought about re -designing the PS, I have been looking into it Since I converting to LED's. I can remove the 5.4 vdc circuit (one less heater and one Hot step down 100 ohm resistor and add a bridge reciter (similar to what you did w a Sx-450) from the 1.5amp ac fuse. (sx-,,,,,,850,,,and more) The stereo Led voltage (sticking point) is actually drawing from 13.9 vdc not the 5.4vdc so I remove a bunch of components . What the rating of the board? the top of the board gets hot (140 Degrees in some places)
@@johnsweda2999 I was thinking the same as you But after thinking about It and even using even higher watt transistors, it made very little different w the heat and what Tony said above, what I found was converting 60Vdc to a lower voltage, its like the Transistor is a dropping resistor, so the transistor has to drop a lot voltage which it converts it to Heat. Than current is a factor but not in this design since it draws very little current.
In my imagination they had the idea to place the RF filter capacitor in between these two electrolitic .... the patch to ground this way instead to put directly into the first transistor base or attached to a stop resistor in series with the first transistor base..... this is my imagination and the several check ups they made ended with a traditional instalation of the capacitor wasting that nice opportunity to filter that way....well.... imagination only...i cannot say other reason why they did that stuff
About replacing input electrolytic capacitors with film capacitors. I did this to a sansui au-222 and a pioneer sx-636, and the result was awful! The sound became thin and a bit metallic. In the end, I had to replace the film capacitor with a electrolytic nichicon Fine gold and got the "original" sound back. The tests that you did, they are all valid and good, but they only show the performance of the receiver and not how it sounds. You can grab a sansui and a yamaha, and they will probably perform the same under tests, but they will sound completed different!
Hi Rui, those are good points, I don't think anyone here can disprove your experience with your equipment after you installed premium Nichicon caps in the inputs. We also know that all parts of any circuit have parasitic R and L and C components everywhere in the circuit, some of these parasitic components are perceived, noticeable and have to be eliminated, some don't. I am saying this because some things happen in a circuit as a consequence of parasitic R/L/C components that are not visible to the human eye (unless it is like a zig zag path in the circuit board that may visibly be causing some inductance/capacitance and therefore distorting the sound). We can be pragmatic and just replace those 2 x 2.2 uF polarized caps connected back-to-back to create a "virtual" non-polarized 1.0 uF capacitor using again new 2 x 2.2uF polarized caps back-to-back. Or just replace them with a 1.0 uF high-quality film capacitor like Tony did, and like I would do too and have done in my similar projects many times with complete success. I understood that Tony tested one channel with the 2 x 2.2uF input electrolytic polarized caps and the other channel with the 1.0 uF input film cap using a frequency sweeper and obtained success in both channels. I understand that the frequency sweeper is injecting only one frequency at a given time, so it is not like playing a real music and listening to it, but if Tony was going to analyze (on the oscilloscope) any music signal on one side of the 1.0 uF film capacitor and compare it with the signal on the other side of the same capacitor he would get exactly the same signal. (Consequently, if his oscilloscope can show a subtraction of one signal to the other, then it would show a flat line on the screen when you inject any audio signal in the equipment, be an audio frequency sweeping 20-20kHz or a real music playing).
@@kylesmithiii6150 JUST for info- I replaced the same way in Pioneer A9 the orange capacitors 0.47uF and 0.22uF on the potenciometer board with a film capacitors same capacity and the sound goes more metallic with a firmer and stronger high frequencies and also magicly adding more higher frequencies into the music. I am still listening it and till thinking to leave like this or to restore electrolytic back. The sound is very good and sweet and has a more and different obertons and is just good but different Will make test with a sweep generator and THD metr. This amplifier has range from 0.5 Hz to 200 kHz 0.003THD
Funny how slew rate is a "published spec" for op-amps but NOT for most power transistors. Instead you get "ft" (or the transition frequency..where the hfe goes to zero). However some transistor specs show rise time but not all.
Very good video.... all that is important.....thanks you have not made RF alignment as nobody uses FM now a days...at least i do not know people that does that...so.... a kind of waste feature that loose interest and need as time goes by
Perhaps where you live, there is no analog AM or FM and it doesn't interest you, but I can assure you that there are quite a few AM analog & FM stereo analog stations in my part of the world. As a matter of fact, I can't stand the digital "HD radio" that is used here in the states. It drops out, is lacking in fidelity and is very dependent on signal strength, much more than analog FM. I have to disable it in all of my car radios because of the poor performance, while the analog FM still works great. I can't speak for the DAB system used in other parts of the world, but HD radio sucks. As for satellite radio, I am not paying $12+ per month to listen to radio that we have been listening to for free since the 1920's. That said, you still need to understand how an RF carrier and modulation works, if you are going to service receivers of any type. Radio is radio. Therefore, it is not a waste of time. There are many viewers that are very interested in how tuners work. As a licensed ham radio operator, I can tell you that analog modes are very much still being used, along with the many different types of digital modes. I'm glad you like the amplifier videos and i hope you will understand that others like the tuner and tube videos.
0:15 Calibrating offset & bias on power amplifier
4:39 A little bit of fun with RUclips Audio Library song
6:48 Testing the output of the amp with 1kHz sine wave with explanation of test equipment setup
14:19 Transistor temperature: How hot is too hot?
21:32 Checking THD at 1kHz & rated power. THD is tested with Keithley 2015 out to 5th harmonic.
23:58 Frequency analysis from 20Hz to 20kHz using the Instek MDO scope and FA software
29:03 Slew rate calculation and testing: How slew rate determines max frequency response.
44:53 Closing comments
Thanks for introducing these time stamps recently. Very useful when re-watching.
Great video on the sx980 Pioneer. I had it . I wish I had it today. But I did it in with tuner cleaner. I paid 600$ for it in 1978. Loved the video. Dennis. Thanks.
I can understand what your talking about as long as pointing to the screen and showing the schematic. Great teacher
Love the commentary and your videos. Please keep them coming. As to the two capacitors at the input to the amplifier that are electrolytics mounted back to back to create a nonpolarized electrolytic of a smaller value, it may have been done due to cost. At the time this receiver was designed large value film capacitors may have cost more and they were just doing it to save money. And as it is much of the time when electronics are designed, you utilize parts that are readily available at the time you're producing your device. So what I'm saying is it may just have been bean counters or supply issues that were inputs to the engineers decision to do what they did. I've worked in electronics for over 40 years and I've seen it time and again where parts chosen to produce whatever unit you're engineering were chosen due to cost and availability.
I think you are correct, in that they are wanting some ESR at the input. A good experiment you did replacing these with a 1uF film cap to see what happens. So, it may be a "sonic tuning" sort of thing they are doing. Similarly, I have found from my experimentation that using a lone electrolytic cap for the feedback loop produces much better sound than say, an electrolytic with a film bypass on the feedback cap with simple long tail pair setups. The results have been consistently better sound in doing so, and a nice audio grade electrolytic here, works very well. So it may be safe to assume that some ESR in these signal path locations may produce a better sonic result for some amps. I also believe you do not need a bipolar cap at the input, so simply adding ESR is likely all they are doing, using 2 small value caps. I think you should do a critical listening test comparing both the electrolytics and the film caps in this location.
Thanks for this video. The section about the selection of the 3 transistors on the PSU board was very helpful to me in building a parts list for mine.
Tony, Thanks much for sharing your wealth of knowledge! My biggest challenge is remembering which video has the relevant info. The time stamps in the comments help, Thanks! In this video you replace the heatsinks on the TO-220 transistors in the power supply. Where do you find the tall extruded TO-220 heatsinks? There's a ton of TO-220 heatsinks out there, but having trouble finding the tall small footprint that will fit in these receivers.
Thanks, Todd
Also looking for those Todd, let me know if you find them!
Awesome insights as always ! Thank you, Sir.
This has been a very instructive video, much better than watching you solder components in. The tests were not strange to me but I particularly liked how you worked out mathematically what the reading should be, Then ran a test, thus proving that the amp was working beyond pioneers specs. Thank you Prof.
I really look forward to these videos. Thank you very much.
Just be good to me. "Great song by the way. I love it." Your cool points just went up by a factor of 10 with me. Excellent instruction and excellent taste in music!
Really appreciate your style of explaining the electrical knowledge you have to help us dummies. Kinda like the books they like football for dummies etc. You should write your own book
Excellent as always Tony! Especially liked the slew rate test. I was lucky enough to own a Sansui G22000 monster receiver in this same era (1978) which had a super-fast slew rate & 'rise time'. Suffice to say, it was an incredible sound reproducer, and was my replacement right after my SX980. Would LOVE to see you work your magic on either the G9000, or (if you can get hold of one!)
Great video Tony. The SX series seems so well designed. I know it's not your thing, but I'd love to see you test (compare) a contemporary stereo receiver or AV receiver with one of the Pioneers.
You are great . An I like your teaching. I had that sx980 Pioneer receiver but I ruined it with tuner spray I wish I had that receiver today I enjoyed it and 80 watts per channel is all you need as you know.
Always a joy to watch your videos Toney.
All my best wishes to you and the family.
I believe I know a (partial) answer to the "series capacitor" design. The problem has to do with how much DC is being applied to either end of the caps under normal use. If you have very little DC relative to the cap rating (across the cap or caps) you want to use the "series design" (ie the virtual non polarized cap) in order to minimize any "wave bending" (aka distortion) around the 3db (f0) frequency which is typically between 10 and 20 hz for a first order high-pass filter. (which is what most of these circuits are...) The problem has to do with "asymmetric charge distribution" inside the cap when the internal polarity inside the cap itself is changing and how this may affect the overall shape of the output waveform. Electrolytic caps basically "like" to be properly "polarized". They don't really play nice when just lightly or slightly biased and they have tendency to "oppose" the change in charge of slowly varying ac signal (ie low freq audio). According to design engineers tantalum caps respond much quicker to this low freq AC charge-change around the transition frequency. However the response is still "asymmetric" which is where the second capacitor comes in. ("asymmetric" as is..the bottom half of the waveform will be flatter than the top half) This design dramatically improves the "linearity" of low frequency AC signals around the 3db point. The output signal should just be attenuated..while NOT distorted (waveform bending). You can can actually set up a simple filter circuit to test this. I used about four different caps between 1 and 2uf and 10k resistor. 1uf&10k gives you an f0 (-3db) of around 16hz. The trick is having a decent analog (dual trace) scope where you can visually inspect the input and output low freq (< 20 hz) AC signals. You can visually see output waveform asymmetry! (Most digital scopes can't accurately sample below 20 hz.) So the series design is basically a distortion minimization "trick" when the DC bias is relatively low. If the DC bias high enough above maximum (AC) input signal you won't get this charge reversal affect and the distortion will be much lower even without the second cap. Sony used this design on their pre-amp output stage way back when and Randy Slone (RIP) does it on his power amp input stages with tantalum caps.
As always, nice job Tony!
I love way you explain things. The only thing I see is that, when you use your test equipment , like oscilloscope or spectrum analyzer, show us where to hook up or connect to. Too someone who new like me will not know and that would help us.
Great series of videos on the pioneer thanks Tony, keep them coming I'll keep watching
I have also done a test and calculation of the output power of a sx-980 and came to approximately 85W into 8 ohm dummy-loads. Using an analogue scope and reading to the point of flat-topping.
The transistors in the PSU of this amp is coking, i think I measured them being close the boiling point (100 C), but they seem to last. this is the original transistors on original heatsinks.
Very well explained and very informative. Always looking out to the next video, enjoy them very much, thank you from the Netherlands.
Would be fine to see the harmonics displayed in spectrum display or FFT
Your just amazing Tony! Thank you very much.
Thanks for the ride.
Thumbs up
Great Vids Tony many thanks for all your work, time and sharing your knowledge, we all appreciate it, all the very best to you and your family for Christmas and the new Year, lets hope we all have a better one (Covid wise!) than 21, take care from the UK
Tony, when you were measuring the slew rate on a square wave, how much of the delay was in the signal generator? I would like to see the the signal generator connected directly to the o-scope. Would you subtract the signal generator delay from your measurement?
Good question! The signal generator I used had a frequency range up to 160MHz. As this signal generator has a maximum output amplitude of 5volts, the slew rate would need to be 5024V/uS to achieve that frequency. This makes the delay of the signal generator at audio frequency negligible. Additionally, the cables will have a velocity factor that will add delay. The cables will also add some delay, but again, it wouldn't be a huge factor when compared to the amplifier's components, such as the capacitors, miller capacitance of the semiconductors, etc. Hope that helps.
Another great video, I'm following this one closely because I have the same receiver. Question: where did you get the heat sinks, for the transistors in the power supply? Thanks
I have a bunch of those heatsinks. I got them on eBay. They also had the 1" tall ones as well. I think I got 100 of each at the time for a really good price.
The slew rate for a Crown CT-210 is "Greater than 17 volts per microsecond" It is rated at 105 WPC into 8 Ohms or 135 WPC into 4 Ohms at 20Hz-20KHz. That's more than twice as fast as necessary.
Great as always. I learned a lot. As a general comment, it would be helpful (at least to me) if you listed the major items of test equipment used by manufacturer and model.
Thanks for the comment. I actually did this in a number of previous videos. I always have to leave some things out because I don't always have the time to film everything. If you have any specific questions about a particular piece of equipment you see, just post a comment and I'll try to answer you, as long as I see the comment. Thanks again!
This is a great video sir! I've long considered a THD Analyzer for my bench to prove the health of my projects aside from verifying offset, bias, power supply voltages etc. If you could make a similar video about THD calc as you did for the slew I'd buy you lunch!
We'll see what we can do ;)
while measuring power You should also control distortion to do not cross spec. limit form documentation
Great video Tony.....cheers.
Great "bounty" in this series and particularly this session. I think you got a handle on the magic. Thanks for the look and listen!
Awesome video, thanks for making it.
Learned a lot.
Do you have any of those dummy load boards for sale? 😁 Also.. Where did you get those heatsinks for the power supply transistors?
Apologies for the noob questions in advance!
"We are going to apply a 1kHz sine wave tone into the auxiliary inputs of the amplifier and we're going to use about 500mV of drive signal". Why 500mV?
When I look at the specifications for the SX-980, I see that the input sensitivity is 150mV for Aux/Tape, 7.5mV for Mic, and 2.5mV for Phono. Are those specifications minimums and is there a maximum? Can you elaborate on the, seeming random, choice of the 500mV? The phono overload shows to be at 200mV or roughly 100x the input sensitivity. The Tape and Pre out levels are listed at 150mV and 1V respectively. Would just like to know your thoughts on all of this.
Thanks Tony!
Those are the minimum input levels required to get the rated output. Different amplifiers will have different amounts of gain, which will affect the minimum signal required for rated output. I usually apply a signal level of 500mV RMS to line level inputs as a standard. Most line level amps/preamps will have a range of anywhere from 150 mV to 1.5 or 2 v. If there is no attenuation control on the amplifier, then I will vary the output of the signal generator to get the output level I'm looking for.
@@xraytonyb Thank you Tony!
40:30 No confusion at all, it's all very interesting; you're just calculating the gradient. Theoretically, the sine curve is g(f,t) = sin(2πft); so to calculate the rise time, simply take the first derivative (to find the gradient) and multiply by the peak voltage to get the rise time in V/s, i.e.: dg(f,t)/dt*Vp =2πf cos(2πft)*Vp. At t = 0 (max gradient) for 20000Hz the gradient is dg(20000,0)/dt = 2π*(20000)*cos(0)*Vp = 5.027V/μs. On the oscilloscope it's just ΔV/Δt to get the gradient.😁
The principal failure mode of capacitors is short circuits, so it makes sense to use two identical caps in series as a safety measure against failure of one cap due to accidental over-voltage. Generally, components that fail short circuit are used in series (like caps) and components that fail open circuit (like resistors) are used in parallel as a safety against accidental failure. Two identical caps in series have double the voltage rating and two identical resistors in parallel have double the current rating, adding to the safety margin .
The two caps are back to back. Therefore if one fails shorted the other one has its back to the current and will pass it right through.
Thank you for another fantastic video, you work so hard on them and are a a superb communicator. I don't really have much electronics knowledge but over the years have as a consumer of mid range audio gear picked up the ideas behind the measurements. I am curious to know what the significance of the phase shift curves are. They look quite alarming to to a person in the street and imply significant distortion of a kind not measured with fixed frequencies. Do these things matter or just an inevitable consequence of driving a load which of course would be far more variable in a real speaker?
Great video Tony as always, thanks! In the end you talk about the 1uF film capacitor to replace those two polarized 2.2 uF in series. I am dying to know the final answer to this question, if there is one definitive answer based on laws of physics. I would just replace those with the 1 uF film capacitor as you did. But from the schematics you see that on the right side of resistor R3 we have 0 VDC, and at point #6 we also must have 0 VDC, so the capacitor to take place of C1+C2 will have zero VDC on both of its extremities, therefore I think that the combination C1+C2 needs to be non-polar (or bipolar). (Given that point #7 is connected to ground per the schematics)
Well I dunno about "laws of physics" for parts choice all around everywhere but my likely bet here is cost. Working contract mfg. at a PCBA outfit myself, I've seen this countless times by way of using say 10k resistors in parallel to get other resistances without adding another parts feeder to the machine (SMD), series or parallel caps, etc. Likely they had a spec for a 1.0uF cap in that spot in the design. To get that easy and quick and with fewer added special parts, what they did was probably their cheapest easiest solution rather than up the BOM parts count. And as he mentioned, film caps were larger in those days so might have been mechanical planning/insurance in the event they couldn't source a certain package size.
@@wonderbars36 Hi Nonya, I agree with you about keeping costs down, in this case. I am saying "in this case" because if the equipment has to meet rigid specs, then it may not be so simple to do that. Tony's questions on why there are two 2.2 uF polarized capacitors in series in that part of the circuit included his sub-question of 'Why didn't they install a single 1.0 uF polarized capacitor in that place instead?' I have seen also many times companies like Pioneer, Toshiba, JVC, Sony doing that in their audio equipment from the 70s and 80s, I mean, installing a polarized decoupling capacitor in the audio path like that. In my restorations, I replace those with a film capacitor like he did (usually KEMET, Panasonic or WIMA) if there is space, otherwise I replace with a polarized capacitor (such as Nichicon) if it is in accordance with the manufacturer's schematics and original installation. When I mentioned "laws of physics", I really meant that because a polarized capacitor can't arbitrarily work under AC voltage, or can't work under AC voltage at all, some may under certain specific parameters or problems will occur. Since those two 2.2 uF polarized capacitors are directly in the audio path, I have to assume that the Pioneer engineers were 100% sure what they were doing. Per my explanation above, and per the schematics, both extremities of the capacitors in series have zero VDC, so Pioneer needed to install a non-polarized 1.0 uF there but they used 2 x 2.2 uF polarized in back-to-back series. I have seen single polarized capacitors (some 10 uF) in the audio path in my restoration projects -- in places that the capacitor has also 0 VDC on both sides, and will be under an overlaying AC voltage due to AC audio signals being injected.
Interesting video. I think you may have used the wrong value for the rise time when calculating the slew rate. Oscilloscopes rise time measurements are usually the delay from 10-90% of the min and max values. For slew rate, wouldn't you want the delay from 0-100%?
Hello , I am working on the SX980 power supply board and I am replacing the 2200uf 35v cap on the board. The factory cap has 3 leads and replacement has 2 . How should this be oriented on the board. I do know the stripe is negative but which hole on board should positive lead go to. Please help. Maybe you could show underside of board. Thank you. Chris.
Isn't it amazing! In 11 hours you had 1400 students attend your class!
If you had any doubt of how good of a teacher you are, you shouldn't... I think that you really missed your calling in life LoL....
Thank you for posting Tony! Question; if a person were to take the time, and had access to enough parts inventory, would precision testing/matching components like caps, resistors, diodes, etc., be a worthwhile time investment when it comes to circuit balance/stability? Thanks for your time
That's really subjective. I do believe that certain components benefit more from this than others. Case in point would be the transistors used in the differential (long tail) pair at the input of the amplifier. Output transistors don't really need to be closely matched, as the output stage has little or no gain and they have the emitter resistors to "take up the slack". In addition, the PNP and NPN transistors will often have very different gain characteristics just by their nature. Coupling capacitor values are usually chosen to be high enough so they won't attenuate the signal much within the frequency range of the circuit. If they are a bit different, it won't make a big difference. Now if the components are out of spec (leaky/high ESR caps, leaky transistors, etc.) then it could affect performance. That said, there are some who say they can hear a difference, which is entirely possible. Coming around full circle, it's subjective. Not a big problem for "Tin-ear Tony" ;))
Hi Tony, thank you for all the hard work and effort you put in all your videos and shearing you knowledge onto us. God blessed you with an amazing gift , learning more every day from you. So I say God bless you and your family. Thanks
The awfull square wave shape can be result of 120pf capacitor directly from Q1 base to ground, or a miller unit in the VAS
I have an SX 980 probably one of the cleanest around anyway. Started blowing fuses actually just one looking at the front it’s the far right glass 1 amp fuses is there a simple way to trouble shoot the issue I assume it’s a transistor? I have been watching your videos and tried looking for obvious issues and I cannot find anything obvious? I’m pretty upset it’s my baby! Any kind of advice would be wonderful 😳
thank you for making this video
I think also you have to make a critical listening test and also to make experiment with replacing the electrolytic caps with two in series film caps , not only one equivalent to see the sound and to measure also the distortion with the distortion meter. Why ? Because may be the sound will go more metallic with only one equivalent film cap , as stated the comments below.
Tuner Test comment / ?
Tony can you bring your expertise to bear on answering a question?
We can all get a good idea of evaluating the difference receivers by what stations
they can pick up in our area. But using our test equipment how can we run good comparison tests between different receivers for sensitivity selectivity etc.?
As long as you have a signal generator that is capable of outputting an accurate amplitude and then make sure the test equipment is properly coupled to the tuner, you can make some pretty accurate comparisons. I did some videos on signal generators and impedance matching that might be helpful. Thanks for the comment!
Why didn't you use 50 watts plus transistors in the power supply won't get above body temperature and theoretically if you run 100 what you would need a heatsink at all
I think with the capacitors using electrolytics non-polarized or whatever and film capacitors is just down to the sound of the amplifier signature, did you notice any difference between running a mono signal in both channels to your ears?
the slew rate was slightly better on one channel which channel had the better slew rate with the capacitor the one you changed was it or the original capacitor?
Wondering because it's sharing the same power supply would that mess up the results slightly might have been better running them as individual mono blocks through the same power supply and seeing what the results was each individually.
The power supply transistors get hot because of the power they have to dissipate due to the amount of voltage they have to drop at the required current. That power is converted to heat, which would be similar regardless of the power rating of the transistor. The more important factor would be increasing the rate at which the heat can be dissipated from the transistor. The heatsink and thermal interface have the most influence on this. Some transistors are a bit more efficient at this, but they will still get hot, especially if they are all in the same case type (such as TO220). As you saw from the charts, the transistor can only dissipate higher wattage when they are kept cooler through larger heatsinks. Although the transistor is roasting in this amp, it is operating well within its thermal specs.
@@xraytonyb I don't see how 100 watt transistor should be getting just as hot over a 25 watt transistor, 100-watt transistor can handle the current better then the 25 watt so at low power it shouldn't be getting as hot, they both can't be at the same temperature differential otherwise 100 w transistor overheat at it's rated current draw sure that's correct.
You pump 50 watts into a 25 watt transistor and 50 watts 100 watt transistor see which gets hotter we both know the answer to that? Only if I'm missing something!
@@xraytonyb Thanks, I been working w the Pioneer Power supply and I think Pioneer had a bad design w all of them? I was thinking I did something wrong when restoring the PS, After thinking about it and w no load they get run Hot and than hotter, so converting 60vdc to 5.4vdc or 31vdc is like a Big Hot Converting Resistor Transistor ,,,,,, and has nothing to w Amps since output draws low current. Tony have you ever thought about re -designing the PS, I have been looking into it Since I converting to LED's. I can remove the 5.4 vdc circuit (one less heater and one Hot step down 100 ohm resistor and add a bridge reciter (similar to what you did w a Sx-450) from the 1.5amp ac fuse. (sx-,,,,,,850,,,and more) The stereo Led voltage (sticking point) is actually drawing from 13.9 vdc not the 5.4vdc so I remove a bunch of components . What the rating of the board? the top of the board gets hot (140 Degrees in some places)
@@johnsweda2999 I was thinking the same as you But after thinking about It and even using even higher watt transistors, it made very little different w the heat and what Tony said above, what I found was converting 60Vdc to a lower voltage, its like the Transistor is a dropping resistor, so the transistor has to drop a lot voltage which it converts it to Heat. Than current is a factor but not in this design since it draws very little current.
Rod Stewart - Passion
Just be good to me,
In the morning, just be good to me,
I Forgot X10 again!!!!
In my imagination they had the idea to place the RF filter capacitor in between these two electrolitic .... the patch to ground this way instead to put directly into the first transistor base or attached to a stop resistor in series with the first transistor base..... this is my imagination and the several check ups they made ended with a traditional instalation of the capacitor wasting that nice opportunity to filter that way....well.... imagination only...i cannot say other reason why they did that stuff
7.52V!!!!!
About replacing input electrolytic capacitors with film capacitors. I did this to a sansui au-222 and a pioneer sx-636, and the result was awful!
The sound became thin and a bit metallic. In the end, I had to replace the film capacitor with a electrolytic nichicon Fine gold and got the "original" sound back.
The tests that you did, they are all valid and good, but they only show the performance of the receiver and not how it sounds.
You can grab a sansui and a yamaha, and they will probably perform the same under tests, but they will sound completed different!
Hi Rui, those are good points, I don't think anyone here can disprove your experience with your equipment after you installed premium Nichicon caps in the inputs. We also know that all parts of any circuit have parasitic R and L and C components everywhere in the circuit, some of these parasitic components are perceived, noticeable and have to be eliminated, some don't. I am saying this because some things happen in a circuit as a consequence of parasitic R/L/C components that are not visible to the human eye (unless it is like a zig zag path in the circuit board that may visibly be causing some inductance/capacitance and therefore distorting the sound). We can be pragmatic and just replace those 2 x 2.2 uF polarized caps connected back-to-back to create a "virtual" non-polarized 1.0 uF capacitor using again new 2 x 2.2uF polarized caps back-to-back. Or just replace them with a 1.0 uF high-quality film capacitor like Tony did, and like I would do too and have done in my similar projects many times with complete success. I understood that Tony tested one channel with the 2 x 2.2uF input electrolytic polarized caps and the other channel with the 1.0 uF input film cap using a frequency sweeper and obtained success in both channels. I understand that the frequency sweeper is injecting only one frequency at a given time, so it is not like playing a real music and listening to it, but if Tony was going to analyze (on the oscilloscope) any music signal on one side of the 1.0 uF film capacitor and compare it with the signal on the other side of the same capacitor he would get exactly the same signal. (Consequently, if his oscilloscope can show a subtraction of one signal to the other, then it would show a flat line on the screen when you inject any audio signal in the equipment, be an audio frequency sweeping 20-20kHz or a real music playing).
@@kylesmithiii6150 JUST for info- I replaced the same way in Pioneer A9 the orange capacitors 0.47uF and 0.22uF on the potenciometer board with a film capacitors same capacity and the sound goes more metallic with a firmer and stronger high frequencies and also magicly adding more higher frequencies into the music. I am still listening it and till thinking to leave like this or to restore electrolytic back. The sound is very good and sweet and has a more and different obertons and is just good but different
Will make test with a sweep generator and THD metr. This amplifier has range from 0.5 Hz to 200 kHz 0.003THD
A G22000....
Funny how slew rate is a "published spec" for op-amps but NOT for most power transistors. Instead you get "ft" (or the transition frequency..where the hfe goes to zero). However some transistor specs show rise time but not all.
Very good video.... all that is important.....thanks you have not made RF alignment as nobody uses FM now a days...at least i do not know people that does that...so.... a kind of waste feature that loose interest and need as time goes by
Perhaps where you live, there is no analog AM or FM and it doesn't interest you, but I can assure you that there are quite a few AM analog & FM stereo analog stations in my part of the world. As a matter of fact, I can't stand the digital "HD radio" that is used here in the states. It drops out, is lacking in fidelity and is very dependent on signal strength, much more than analog FM. I have to disable it in all of my car radios because of the poor performance, while the analog FM still works great. I can't speak for the DAB system used in other parts of the world, but HD radio sucks. As for satellite radio, I am not paying $12+ per month to listen to radio that we have been listening to for free since the 1920's. That said, you still need to understand how an RF carrier and modulation works, if you are going to service receivers of any type. Radio is radio. Therefore, it is not a waste of time. There are many viewers that are very interested in how tuners work. As a licensed ham radio operator, I can tell you that analog modes are very much still being used, along with the many different types of digital modes. I'm glad you like the amplifier videos and i hope you will understand that others like the tuner and tube videos.