"If you are making in amp that is gonna get very hot (class A) then the TO-3 transistors are definatly better at keeping cool."
__________________________________________________
Actually I disagree they they are better at keeping cool. They should be as good at keeping cool as the TO-247 casing. The TO-247 cases also have the silicon wafer mounted to a metal substrate that is in turn making contact with the heat sink. Same as in a TO-3. The main difference is the ruggedness of the TO-3 vs. plastic, and the fact that thermal failures occur more often in plastic packages. The thermal aspects of plastic encapsulation has gotten alot better though in the past decade....but failure from repeated thermal expansion and contraction would be my main concern in a plastic vs. metal debate. So, when building "hot running amplifiers" that utilize plastic devices you should over heatsink the devices where ever possible so they run alot cooler. Also leave things turned on for longer periods of time so there is less thermal cycling of the plastic devices.
Mark
__________________________________________________
Actually I disagree they they are better at keeping cool. They should be as good at keeping cool as the TO-247 casing. The TO-247 cases also have the silicon wafer mounted to a metal substrate that is in turn making contact with the heat sink. Same as in a TO-3. The main difference is the ruggedness of the TO-3 vs. plastic, and the fact that thermal failures occur more often in plastic packages. The thermal aspects of plastic encapsulation has gotten alot better though in the past decade....but failure from repeated thermal expansion and contraction would be my main concern in a plastic vs. metal debate. So, when building "hot running amplifiers" that utilize plastic devices you should over heatsink the devices where ever possible so they run alot cooler. Also leave things turned on for longer periods of time so there is less thermal cycling of the plastic devices.
Mark
Per-Anders, drilling holes in heatsinks is done by machines in the industry so mis-drilling is not common. It's 4 holes against 1 hole so time is an important factor to consider. But what does make a difference is that there is need for manual labour when using TO 3. Remember the solder lugs ? Manual labour is a curse in the western world because it is costly.
The qualification of older devices as "yesterday"and "rubbish" is strange to me and the example of valves is a very effective one to withstand this modern "newer is better" moral. There are in fact old semiconductors that are very good in quality. For the industry there is the question of continuity so when newer devices are available it's best to adapt production lines for those newer components as the supply of the old ones will dry up. The MJ series may be good for controlling car window movers but Quad 405 owners can't do without them !
You have to agree with me that the factor "quality" is becoming less important by the day. It's all about quantity. Stereo equipment must work for a few years and then it's worn out.
Please take a look at Dolby 5.1 devices. What is positive is that newer devices are made environmentally safe. It is a fact that newer semiconductors are much faster, but that's not always a pre for using them.
Wish I would still have my 23 year old Kadett 1.1. It got me to Hungary, Tchechoslovakia and back with the only problem that I had to fill it up now and then. And then there was this shiny new Astra 2.0 D. After having driven this modern newer piece of junk with stuck brakes, refusing sensors, broken turbo and gas leakage I am reluctant to say "newer is better".
A strong factor in the "newer is better" moral is that newer devices are designed for a shorter lifespan to fulfill the self-fulfilling prophecy that "newer is better". It breaks down after some years and you'll buy newer and better again
But you know what marketing can do. My daughter did learn something new at school. She asks with everything she sees if it is new or old. When I answer it's old she just looks negative because she learned at school that older things are less in value than newer things.
I am not surprised that there is a discrimination for age in the west.
The qualification of older devices as "yesterday"and "rubbish" is strange to me and the example of valves is a very effective one to withstand this modern "newer is better" moral. There are in fact old semiconductors that are very good in quality. For the industry there is the question of continuity so when newer devices are available it's best to adapt production lines for those newer components as the supply of the old ones will dry up. The MJ series may be good for controlling car window movers but Quad 405 owners can't do without them !
You have to agree with me that the factor "quality" is becoming less important by the day. It's all about quantity. Stereo equipment must work for a few years and then it's worn out.
Please take a look at Dolby 5.1 devices. What is positive is that newer devices are made environmentally safe. It is a fact that newer semiconductors are much faster, but that's not always a pre for using them.
Wish I would still have my 23 year old Kadett 1.1. It got me to Hungary, Tchechoslovakia and back with the only problem that I had to fill it up now and then. And then there was this shiny new Astra 2.0 D. After having driven this modern newer piece of junk with stuck brakes, refusing sensors, broken turbo and gas leakage I am reluctant to say "newer is better".
A strong factor in the "newer is better" moral is that newer devices are designed for a shorter lifespan to fulfill the self-fulfilling prophecy that "newer is better". It breaks down after some years and you'll buy newer and better again
But you know what marketing can do. My daughter did learn something new at school. She asks with everything she sees if it is new or old. When I answer it's old she just looks negative because she learned at school that older things are less in value than newer things.
I am not surprised that there is a discrimination for age in the west.
The 8 leaded TO-3 from BB was most likely a power op amp. I used to use them back in the early 80's.
TO-47's are not always easier to mount. In a well designed system they should be, but in some applications where the heat sink package simply has to be located elsewhere due to space constraints, it may actually be easie, or at least as easy, to use the TO-3 package.
It is up to the designer to determine which package is "best" for the given application. For DIY'ers, the TO-247 may indeed be the best choice in general applications, but perhaps not all.
TO-47's are not always easier to mount. In a well designed system they should be, but in some applications where the heat sink package simply has to be located elsewhere due to space constraints, it may actually be easie, or at least as easy, to use the TO-3 package.
It is up to the designer to determine which package is "best" for the given application. For DIY'ers, the TO-247 may indeed be the best choice in general applications, but perhaps not all.
Jean-Paul, I think you have mixed up something. If the industry doesn't wants to buy a certain part, then the manufacturer doesn't want to manufacture it. It has nothing to do with "newer is better" or "older is worse".jean-paul said:
The "market" desides what they want to buy, wether it's good or bad.
TO3 stuff
Jean-Paul,
I don't mean to say that older is necessarily bad, but the fact is that modern, newer power transistors have more gain, which is constant with Ic, and wider bandwidth, which translates to measurable better amplifiers.
We all are familiar with the rise in distortion starting at a few 100Hz with older equipment. Newer equipment either doesn't show it, or has much less of it. This is heavily influenced by the output stage itself and the requirements for compensation in pre-stages as a result of poor bandwidth in the output stage.
That doesn't mean that listeners are not free to like whatever they like, but if you try to reproduce the audio as faithfull as possible, it is an issue.
In this sense, I still maintain that it is not very smart to use MJ15003/4 if you are going to design a new power amp, if at least good reproduction is your goal. I still use my 10W heavily modified HF309 tube amps regularly, but that is for other reasons. I understand that Quad 405 users cannot live without the MJ's either, but when was that amp designed?
Jan Didden
Jean-Paul,
I don't mean to say that older is necessarily bad, but the fact is that modern, newer power transistors have more gain, which is constant with Ic, and wider bandwidth, which translates to measurable better amplifiers.
We all are familiar with the rise in distortion starting at a few 100Hz with older equipment. Newer equipment either doesn't show it, or has much less of it. This is heavily influenced by the output stage itself and the requirements for compensation in pre-stages as a result of poor bandwidth in the output stage.
That doesn't mean that listeners are not free to like whatever they like, but if you try to reproduce the audio as faithfull as possible, it is an issue.
In this sense, I still maintain that it is not very smart to use MJ15003/4 if you are going to design a new power amp, if at least good reproduction is your goal. I still use my 10W heavily modified HF309 tube amps regularly, but that is for other reasons. I understand that Quad 405 users cannot live without the MJ's either, but when was that amp designed?
Jan Didden
"It's much easier to get a good contact between the TO-3 and the heatsink than with the plastic cases as you have 2 screws rather than 1. Giving a more even pressure across the transistor."
__________________________________________________
Not if U use a slightly large washer under the screw( as shown in the photo). This is as Nelson Pass suggests it be done and it works beautifully! Also, one could utilize a clamping bar or machined clamp to guarantee cood contact.
I am certainyl NOT argueing against using TO-3's if available. They are however becomming quite scarce for many popular devices that are commonly used.
Mark
__________________________________________________
Not if U use a slightly large washer under the screw( as shown in the photo). This is as Nelson Pass suggests it be done and it works beautifully! Also, one could utilize a clamping bar or machined clamp to guarantee cood contact.
I am certainyl NOT argueing against using TO-3's if available. They are however becomming quite scarce for many popular devices that are commonly used.
Mark
Attachments
firstly i don't know yet how the mj15003 devices sound, but i am planning an amp design that uses them. if they are no longer useful, what would be a good replacement?
second, i have found it all but impossible to find good transistors in cases other than to220 and to3, however i mostly looked at local electronics stores. so i imagine ordering online allows for to-247 packages up the yinyang.
and as far as the old vs new debate goes, all i can say is i threw away or let deteriorate all of my old cd players, and bought new turntables and now exclusively purchase music on vinyl. most of the 3 or 4 crates of records i have were produced and manufactured in the last 2 or 3 years. and i know i am not the only one with such apparent backwards interests. hook up a scope to a turntable source, and one to a cd source and compare them some time. you may never listen to another cd again.
second, i have found it all but impossible to find good transistors in cases other than to220 and to3, however i mostly looked at local electronics stores. so i imagine ordering online allows for to-247 packages up the yinyang.
and as far as the old vs new debate goes, all i can say is i threw away or let deteriorate all of my old cd players, and bought new turntables and now exclusively purchase music on vinyl. most of the 3 or 4 crates of records i have were produced and manufactured in the last 2 or 3 years. and i know i am not the only one with such apparent backwards interests. hook up a scope to a turntable source, and one to a cd source and compare them some time. you may never listen to another cd again.
sam9
For production use, the mounting issue with the TO-3 isn't much concern. For hobbiests it may be more of a problem - I myself have a terrible time with anything that needs precisely arranged holes while I have seen photos posted here by a couple of people who must have been skilled machinists at some time (how I envy them). Done right the TO-3 has some advantage with heat transfer; also looking inside commercial products, I think it shows that TO-3 are probably easier to change out.
A advantage of TO-247 I did see mentioned is that in designs that need it for thermal tracking, it is much easier to mount a bias transistor directly on the output transistor. The same goes for a thermal protection switch.
For production use, the mounting issue with the TO-3 isn't much concern. For hobbiests it may be more of a problem - I myself have a terrible time with anything that needs precisely arranged holes while I have seen photos posted here by a couple of people who must have been skilled machinists at some time (how I envy them). Done right the TO-3 has some advantage with heat transfer; also looking inside commercial products, I think it shows that TO-3 are probably easier to change out.
A advantage of TO-247 I did see mentioned is that in designs that need it for thermal tracking, it is much easier to mount a bias transistor directly on the output transistor. The same goes for a thermal protection switch.
Scope picture
Hi Seanick,
MJ21193 and MJ21194 are good TO-3 transistors. Better than MJ15003.
What did you see on the scope comparing CD & LP?
The listening experience is even more interesting to me. See also:
http://www.diyaudio.com/forums/showthread.php?postid=144627#post144627
seanick said:
Hi Seanick,
MJ21193 and MJ21194 are good TO-3 transistors. Better than MJ15003.
What did you see on the scope comparing CD & LP?
The listening experience is even more interesting to me. See also:
http://www.diyaudio.com/forums/showthread.php?postid=144627#post144627
I may agree that “older” design is worse than “newer” transistor, depending on the case. Market is changing due to economic reasons: market demand, production cost, etc. Tube is not as efficient as bipolar, TO-3 is not as efficient as TO-247. But new technology and ideas are invented in amplifier design, thus making some older transistors obsolete. I agree with Peranders that MJ15003 (also 2N3055) is out of date.
I don’t know about the Quad using MJ15003, but swapping with a new transistor without changing the whole circuit design cannot be used to judge that MJ15003 is better than the new transistor. Still, I’m not surprised if there is a better “direct” replacement.
10 years ago I built my amp using 2SK135/J50 MOSFET. Now the transistors are the most expensive in Indonesian market. I have built amplifiers with modern TO-247 MOSFET, but not happy with the result. Nelson Pass mentioned that he prefers new MOSFET package than the old ones (due to transconductance I guess), but his opinion didn’t change mine, that I prefer the old Hitachi than International Rectifier’s TO-247 I have used (so I will buy the MOSFET again!).
The problem with newer MOSFET is that the input capacitance is very high (the MOSFET is mainly designed for switching anyway). With 2SK135 (TO-3), we get 150W/600pF. What newer TO-247 MOSFET can give comparable characteristics?
You are absolutely right there. My text was out of context.
I agree again for 100 %. I love 2SK135 and 2SJ50 and made many amps with them. Very rugged and good sounding, I never destroyed a single one. Shorting outputs didn't give any problems and I did not use any series resistors !! Before someone is reacting: I did use gate stoppers. Because one FET is 600 pF and the other is 900 pF I used different value resistors per FET. They were driven by a high quality opamp with feedback taken from the output of the opamp. Gainclone avant la lettre ?
2 weeks ago I got the request to change caps in one of my old K135/J50 class A amps. After having done this I compared it to a Teac I-500 and guess which one was better ?
BTW You give the best example of an old semiconductor that shouldn't be obsolete.
Still have one board of the amp I made that time and was searching for K135/J50 and was disappointed to learn that they are obsolete.
To my surprise they are manufactured again by a brand that is called DSI. Please do a search on this forum as this topic has been discussed before. I paid 15 Euro's a piece for them. The production date that is stamped on them is 1997. they look exactly like the originals. Till now I have not heard them.
2N3055/2N2955
Well, go ahead and laugh everyone. But in my neck of the woods, the old 3055/2955's are dirt cheap. I am retired and on a very low fixed income, so I use a lot of them. I have built a few amps with these that actually sound great. Also, being retired, if it takes me a week to drill a few holes, it doesn't really matter. I guess each has his own circumstances. Most of my parts come from the trash bin at the local repair shops.
However, being a hard core DIY'er, I'm really looking forward to trying some of those newer more expensive transistors.
Well, go ahead and laugh everyone. But in my neck of the woods, the old 3055/2955's are dirt cheap. I am retired and on a very low fixed income, so I use a lot of them. I have built a few amps with these that actually sound great. Also, being retired, if it takes me a week to drill a few holes, it doesn't really matter. I guess each has his own circumstances. Most of my parts come from the trash bin at the local repair shops.
However, being a hard core DIY'er, I'm really looking forward to trying some of those newer more expensive transistors.
willy, I think you have a very valid point there.Good results can indeed be obtained by using not-so-expensive components.
Last year I built a few of Mr. Self's designs.Of the same design,
the worst sounding had the most expensive output transistors.
But, as Mr. Elliot points out, they might have been counterfeit
ones.And, to my ears at least, the best sounding one used
TIP3055 and TIP2955.
Last year I built a few of Mr. Self's designs.Of the same design,
the worst sounding had the most expensive output transistors.
But, as Mr. Elliot points out, they might have been counterfeit
ones.And, to my ears at least, the best sounding one used
TIP3055 and TIP2955.
A TO-3 / TO-204 has a feature over most plastic transisters, higher "Tj".
Most TO-3's have a Tj of 200+ C. Plastic tends to have Tj of 150+ C
Thus, when designing the heatsink, the higher Tj allows for either a smaller heatsink or for the same transister parameters, less of a chance of exceeding Tj under abnormal conditions.
I must admit I am not sure why the Tj is lower on plasic, is it just because of the plastic case?
Most TO-3's have a Tj of 200+ C. Plastic tends to have Tj of 150+ C
Thus, when designing the heatsink, the higher Tj allows for either a smaller heatsink or for the same transister parameters, less of a chance of exceeding Tj under abnormal conditions.
I must admit I am not sure why the Tj is lower on plasic, is it just because of the plastic case?
I have come across this thread by accident again.
Morph and others: If you want to design something, only _you_ deside which parts you are going to use regardless of what others think. If I felt like doing an amp with 2N3055 I would do it without asking for permission but it never hurts to ask around to get point of views.
It may be smart to design with new parts, especially semiconductors. Much easier if somebody else want to build or you want to repair the thing.
TO-3 parts have rather low maximun power so the argument about firmer attachment is not a very important factor. Not many amps have so heavy power dissipation from each transistor.
I could only think of one thing which is probably better: Moister absorbtion! but when is this a problem really
Morph and others: If you want to design something, only _you_ deside which parts you are going to use regardless of what others think. If I felt like doing an amp with 2N3055 I would do it without asking for permission but it never hurts to ask around to get point of views.
It may be smart to design with new parts, especially semiconductors. Much easier if somebody else want to build or you want to repair the thing.
TO-3 parts have rather low maximun power so the argument about firmer attachment is not a very important factor. Not many amps have so heavy power dissipation from each transistor.
I could only think of one thing which is probably better: Moister absorbtion! but when is this a problem really
in 1982, i built a leach super amp using 2n3055/mj2955 as outputs, tip41/42c as drivers, 92pu06/56 in lieu of the 2n3440/2n5415 transistors, bc556/bc546 small signal transistors. i scaled down to supply to +/-55 volts so i can use a pair or 56kufd/63volts caps, changed the zeners to 27volts from 40, and i would say the sound is just as good as with the mj15004/03...
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.