Tarun this one just proves your point on spurious stuff. While in mkt I decided to visit a real old timer in field now he showed few examples of BJTs costing Rs150-200 originals fakes Rs50/ All appeared same except for very small differences fakes were even LASER marked. So it is back to square one when dealing with the local dealers If I could not distinguish I do not expect them to. Authentic source is all we can rely for now.
For now I go ahead with the 2Sxxx configration. In things get bad we can always put in BC. BD devices after a bit of pulling of legs.
Main point if we get the original stuff then OOOOhhh Lala lah
REgards
Rahul
For now I go ahead with the 2Sxxx configration. In things get bad we can always put in BC. BD devices after a bit of pulling of legs.
Main point if we get the original stuff then OOOOhhh Lala lah
REgards
Rahul
Another thing now I feel is 25 is a reasonable figure we can go ahead and expect to place any order be it samples or regular.
Regards
Raghul
Regards
Raghul
Sunil,
At this point On Semi looks best to me. Last night I checked out the prices at RS India and the MJ15003 at Rs. 350 each don't appear appealing to me. I'd rather request On Semi for 25 of these, plus any number of other stuff (ultra fast rectifiers, small signal NPN's, etc) and pay the $20 processing fee. I just have to find if my CC still has enough Forex allotment left or not. I have 10 MJ's with me and 6 I can spare.
RS doesn’t seem to be very DIY'er oriented. A shame indeed. They will only ship to an official looking address. Did you speak to their regional office? They can help. Do ask them for their free catalogue. Its got tons of info.
And did you get the Tube XO schema I sent via the forum email?
Rahul,
FWIW, even the 'original' ones your shopwalla had could be fakes. Just assume that 99% of all such devices (particularly if they claim to be form vendors such as On Semi, Hitachi, Toshiba) are rip offs.
Aur mian...Mumbai kya fillum ka hero banne jaa rahe ho?? Kya missus jaanti hai ye baat????
At this point On Semi looks best to me. Last night I checked out the prices at RS India and the MJ15003 at Rs. 350 each don't appear appealing to me. I'd rather request On Semi for 25 of these, plus any number of other stuff (ultra fast rectifiers, small signal NPN's, etc) and pay the $20 processing fee. I just have to find if my CC still has enough Forex allotment left or not. I have 10 MJ's with me and 6 I can spare.
RS doesn’t seem to be very DIY'er oriented. A shame indeed. They will only ship to an official looking address. Did you speak to their regional office? They can help. Do ask them for their free catalogue. Its got tons of info.
And did you get the Tube XO schema I sent via the forum email?
Rahul,
FWIW, even the 'original' ones your shopwalla had could be fakes. Just assume that 99% of all such devices (particularly if they claim to be form vendors such as On Semi, Hitachi, Toshiba) are rip offs.
Aur mian...Mumbai kya fillum ka hero banne jaa rahe ho?? Kya missus jaanti hai ye baat????
Guys,
You are the experts. You decide on the output device & topology of the amp. BTW, the design you have chosen is the JLH class A. Am I right. What is the power output we're looking at ? Can it drive 4 ohms. Is there some specific site I can read up on ? If it is class A, I have to figure out some silent forced cooling as I don't have place outside of my present cabinet.
Ashok,
Good idea to do all the shopping in one go. I had a word with Ajju ( Ajayan ) & he is interested as well. I'll check my mail this evening for the tube X/O. How are you planning to use the knobs, with or without indents.
So, who is going to do the ordering & co-ordination of the whole thing. Let's just go ahead.
cheers,
sunil
You are the experts. You decide on the output device & topology of the amp. BTW, the design you have chosen is the JLH class A. Am I right. What is the power output we're looking at ? Can it drive 4 ohms. Is there some specific site I can read up on ? If it is class A, I have to figure out some silent forced cooling as I don't have place outside of my present cabinet.
Ashok,
Good idea to do all the shopping in one go. I had a word with Ajju ( Ajayan ) & he is interested as well. I'll check my mail this evening for the tube X/O. How are you planning to use the knobs, with or without indents.
So, who is going to do the ordering & co-ordination of the whole thing. Let's just go ahead.
cheers,
sunil
As a general rule, let's not decide on our building strategy based on what is available or not available in a particular local market. Sort of defeats the purpose of the Internet and our discussion forum. Just like you assume that your PCB, and everyone else's PCBs, can be made in Bombay, let's also assume that components will be sourced from wherever any one member of our group can source them. Anything in Bombay, I'm willing to source for others. RS is of course available to anyone in India. Samples, ditto. Importing, ditto, if we are willing to pay courier charges (which will be quite affordable for light items ordered as one large group buy). Therefore, let's not confine ourselves to any one town's local market? Please?Rahul said:
Call them up. The Bombay numbers are 24114716 and 24114722. Follow the following algo:sunil said:
- First, look up their Website, get the precise part numbers for all the parts you are interested in. This can be dozens of part numbers sometimes. Note them down on paper. Against each part number, write in brief what that part is, for your own future reference.
- Then call them up, tell them you're from XYZ company, and tell them you want informal quotes of a list of parts on the phone. They'll sit in front of a terminal, take each part number from you, and give you its price. As they tell you each price, you should write it down.
I'm glad you liked them. To put a mark, do the following:
- Fix each knob in a jig, fixture, or vice with adequate padding or pieces of soft material, e.g. rubber or folded cloth, so that the vice's grip does not mark the knob surface.
- Take a hammer and nail/screw, and carefully, with just one sharp blow, mark a tiny indent on the knob's front surface with it. This little dimple will act as the guide for the drilling later.
- Take a drill and a 2mm drill bit. Set the drill to the highest speed it can go, so that you get the smoothest finish in your indent mark. Take the drill and touch the tip of the bit to the dimple marked earlier, and steady your hand, and switch it on. Carefully, make a depression perhaps a millimetre or two in depth, on the knob surface. That way, you'll get a circular depression with conical sloping surfaces, which will act as your marker.
Sunil.
You are right the amp indeed is Class A JLH. Power output is in range of 10W. This is a good figure from listening point of view as well as power transformer size required.
Tarun ,
Yes you are right in pointing abt local mkt stuff. I am also of opinion now since we all are making it one off for our self and a collective effort, we should go in for best of components.
No cutting corners anywhere it will be absolute high end stuff.
Just trying to make JPEG from my PCB layout so that it can be uploaded within permitted size.
Regards
Rahul
You are right the amp indeed is Class A JLH. Power output is in range of 10W. This is a good figure from listening point of view as well as power transformer size required.
Tarun ,
Yes you are right in pointing abt local mkt stuff. I am also of opinion now since we all are making it one off for our self and a collective effort, we should go in for best of components.
No cutting corners anywhere it will be absolute high end stuff.
Just trying to make JPEG from my PCB layout so that it can be uploaded within permitted size.
Regards
Rahul
JLH Class A: which version?
This version needs a few small mods, which have all been discussed in the text. In fact, I think some of us will have to work on the schematics and incorporate all the mods into whatever schematic we want to build, and keep posting the schematics here and making changes, before we finally freeze on the schematic. I am no longer sure we can simply pick up the schematic that Ashok had pointed us towards, and blindly build that one.
One of the reasons I like this version more is because of the parallel OPS transistors, thus allowing me more safety margin in case the cooling is not up to the mark. (JLH was a Britisher, who may have been quite clueless about Rahul's Lucknow summer. 😀 ) With a Class A, there's no point going in for forced cooling, because there's no point striving for the ultimate in sound quality and then raising the noise floor with fans. Hence, the cooling is always one area where safety margins will be useful. I have a few thermal switches; I'm certainly going to use them on the heatsinks for any Class A I build. 🙂
And if we build this higher-powered version, we can actually get upto 60W into 8Ohms, or 56W into 4Ohms (with a lower rail voltage). That would allow us to power some larger speakers for useful SPL levels. See the table I've attached below.
What say?
Is it still 10W after the 2003 mods? I remember that the 1969 version (the original) was 10W, and the later one in 1996 had a 15W rating and a symmetrical power supply. The latest updates are here. In this page, you'll find the schematic that Ashok had posted earlier, plus a higher powered version:Rahul said:
An externally hosted image should be here but it was not working when we last tested it.
This version needs a few small mods, which have all been discussed in the text. In fact, I think some of us will have to work on the schematics and incorporate all the mods into whatever schematic we want to build, and keep posting the schematics here and making changes, before we finally freeze on the schematic. I am no longer sure we can simply pick up the schematic that Ashok had pointed us towards, and blindly build that one.
One of the reasons I like this version more is because of the parallel OPS transistors, thus allowing me more safety margin in case the cooling is not up to the mark. (JLH was a Britisher, who may have been quite clueless about Rahul's Lucknow summer. 😀 ) With a Class A, there's no point going in for forced cooling, because there's no point striving for the ultimate in sound quality and then raising the noise floor with fans. Hence, the cooling is always one area where safety margins will be useful. I have a few thermal switches; I'm certainly going to use them on the heatsinks for any Class A I build. 🙂
And if we build this higher-powered version, we can actually get upto 60W into 8Ohms, or 56W into 4Ohms (with a lower rail voltage). That would allow us to power some larger speakers for useful SPL levels. See the table I've attached below.
What say?
Attachments
Tarun,
On the mark. Although I'll be using the amp for the tweeter I want some extra power & would like to build the 28/56 @ 8/4 ohm version. If the fans is not a good idea I'm willing to accept the largest heatsinks that will still keep my mariage intact. This means anything higher than 28/56 is out of the question for me. So how do these thermal switches work. Do they sit directly on the heatsink or what ??
Hey, Ashok has an active x/o. Tarun, you're going to make one, so why do we need anything really higher. Are you guys going to use it for the mids or the lows. GULP ! heat, Bangalore this year is already cooking. I dread to think of the months ahead. A class A amp higher than 30W is going to need some serious cooling. What about you Rahul, any ideas of an active x/o ?
Rahul, can we look at the higher rated versions ? Are we going to agree on one version or are we going to have individual preferences obliged ?
cheers,
sunil
On the mark. Although I'll be using the amp for the tweeter I want some extra power & would like to build the 28/56 @ 8/4 ohm version. If the fans is not a good idea I'm willing to accept the largest heatsinks that will still keep my mariage intact. This means anything higher than 28/56 is out of the question for me. So how do these thermal switches work. Do they sit directly on the heatsink or what ??
Hey, Ashok has an active x/o. Tarun, you're going to make one, so why do we need anything really higher. Are you guys going to use it for the mids or the lows. GULP ! heat, Bangalore this year is already cooking. I dread to think of the months ahead. A class A amp higher than 30W is going to need some serious cooling. What about you Rahul, any ideas of an active x/o ?
Rahul, can we look at the higher rated versions ? Are we going to agree on one version or are we going to have individual preferences obliged ?
cheers,
sunil
corbato said:And did you get the Tube XO schema I sent via the forum email?Aur mian...Mumbai kya fillum ka hero banne jaa rahe ho?? Kya missus jaanti hai ye baat????
Hye guys just droppin in. I did not know these threads existed. must have been started while I was offline in dec.
it'll take me a long while to get up to speed. ashok S has gotten me started on a small tube project. one day I will get time to put all that together too. till them i will lurk a bit.
He tried tubes on me too. Frankly, I'm scared. All those high voltages, that p-to-p wiring, and worst of all, that constant search for sources of tubes. The entire approach seems so strange to a solid-state-weaned chap like me. They use output transformers to get the voltage down to acceptable levels and get a modicum of current drive. All these things seem so ... well, strange. 🙂 I guess Ashok and I will have to have longer chats before I can make up my mind one way or other.navin said:
What seems more challenging to me today is to aim for a sound as good as that from tubes in every respect, but by using solid-state. I believe it's possible, but I don't know how. Typical of me. 😀
well ashok just dropped in a set of EL84s, ECC81s, and output trannies on his last visit. i must thanks him for carrying 8-10kgs of stuff all the way from calucutta. I have no relatives of friends who i would ask to do that. ashok, I am eternally grateful.
I have messed with solid state for a bit now and the romantic tube sound eluded me. so i figured what is there to learn. at best i will have a bunch of tubes and OPTs that i can ashok can alwys have use for.
I have messed with solid state for a bit now and the romantic tube sound eluded me. so i figured what is there to learn. at best i will have a bunch of tubes and OPTs that i can ashok can alwys have use for.
ok guys,
i just took some time to read and digest this thread (and some of it's sub threads and links). the 28/56 version sunil suggested seems to be idal for my needs too.
this way once i get the 4 x EL84 done as well we can compare a 30W tube amp with a 30W SS.
i now i am far behind the rst of you on this project but if you need anything from bombay let me know. last time tarun actually visited the same bldg where i work (M,W,F) and did not know it.
as far as PCBs are concerned dual layer PCBs are fine in Bombay but multi layer PCBs are better made in Bangalore.
i just took some time to read and digest this thread (and some of it's sub threads and links). the 28/56 version sunil suggested seems to be idal for my needs too.
this way once i get the 4 x EL84 done as well we can compare a 30W tube amp with a 30W SS.
i now i am far behind the rst of you on this project but if you need anything from bombay let me know. last time tarun actually visited the same bldg where i work (M,W,F) and did not know it.
as far as PCBs are concerned dual layer PCBs are fine in Bombay but multi layer PCBs are better made in Bangalore.
Great going guys...sometimes u ppl go too fast...
sitting and reading...on page no 2 now..😀.
I do respect your views.
I dint understand this one. How can a relay or fuse be nonlinear..!?
I dont think a few more bulky heatsinks are going to make a huge difference to my room..😛 Btw u had mentioned that u'll find Oscilloscopes and stuff. Is that true. Can u provide some more details..
Ok if ppl do not want to add a device on the output. Why not trying to control and restrict the power supply under such situations. May be a bit more complex stuff with a foldback regulator. This is what i propose.
Ibn addition to ingerent foldback with a simple opamp, or even a single transistor or something to monitor the output going into
1) clipping,
2) short circuit. (drawing a heavy current)
detect this and feed it back to the power supply through a opto isolator. At the powersupply end the opto coupler output will trigger a bistable latch which will be tied parallel to the foldback control. The fold back nature of the power supply will ensure that
no heavy current is drawn from the source but at the same time wouldnt trip with the power drawn at a peak passsage.. but the moment the current drawn crosses the limit it triggers and the current will settle to the fold back current which will be a few milli amps thus enforsing the protection. In such a case we can even do away with the output monitoring circuit i referred to above. The power supply will be self protecting...
additional monitors can be used for temperature ( to prevent a possible thermal runaway).
Also we might be requiring two regulators for + and - lines. more importantly the two sections should talk... we would be requiring the above optocouplers and latch may be, to make sure that once the thing triggers u need a manual reset to start the thing again... So this will get rid of any relay contacts or fuse at output.
A soft start can also be incorporated easily into the design...!!
well need to figure out how this can be made to work relibly.!
ajju
sitting and reading...on page no 2 now..😀.
I do respect your views.
I dint understand this one. How can a relay or fuse be nonlinear..!?
I dont think a few more bulky heatsinks are going to make a huge difference to my room..😛 Btw u had mentioned that u'll find Oscilloscopes and stuff. Is that true. Can u provide some more details..
Ok if ppl do not want to add a device on the output. Why not trying to control and restrict the power supply under such situations. May be a bit more complex stuff with a foldback regulator. This is what i propose.
Ibn addition to ingerent foldback with a simple opamp, or even a single transistor or something to monitor the output going into
1) clipping,
2) short circuit. (drawing a heavy current)
detect this and feed it back to the power supply through a opto isolator. At the powersupply end the opto coupler output will trigger a bistable latch which will be tied parallel to the foldback control. The fold back nature of the power supply will ensure that
no heavy current is drawn from the source but at the same time wouldnt trip with the power drawn at a peak passsage.. but the moment the current drawn crosses the limit it triggers and the current will settle to the fold back current which will be a few milli amps thus enforsing the protection. In such a case we can even do away with the output monitoring circuit i referred to above. The power supply will be self protecting...
additional monitors can be used for temperature ( to prevent a possible thermal runaway).
Also we might be requiring two regulators for + and - lines. more importantly the two sections should talk... we would be requiring the above optocouplers and latch may be, to make sure that once the thing triggers u need a manual reset to start the thing again... So this will get rid of any relay contacts or fuse at output.
A soft start can also be incorporated easily into the design...!!
well need to figure out how this can be made to work relibly.!
ajju
Hi Ajju and friends,
Nice to see some active participation from all corners.
Now for the non linear devices Ajju a fuse still depends upon its action as a result of current flow, obviously it would depend upon its series resistance, there are many purist who would not tolerate this in similar manner as a no cap in signal path logic. I do not think in those terms but again another logic is justified. Here we are building the ultimate biasing topology amplifier at par with world best and if there is any slight scope of some sort of compromise it should be avoided.
Now on topic of relays it should be taken in account that all relay have a specific contact resistance, again it is a purely mechanical device and lends itself to shortcomings like degradation with time. I am sure many of us have encountered problem of loose contact in switches and rumbblings and erratic operation in case of dirty or loose contact, after all this is what keep companies like WD40 alive.
In this design I feel it is practical if we provide some sort of DC protection , being a series pass regulator supply it easily lends itself to modification. While designing this we can easily incorporate power on delay also without introducing something in series with signal path.
As for the stuff available all is true, just think in terms anything used by the various services, Its all there. Possibly as condemmed, with a slight chance of restoration but its there.
It is a gamble its all your luck. He who plays stand to win.
Now Ajju few heatsink I too gladly procure for my friends but think of what I have picked up over years just to satisfy my own whims. Tough job convincing my wife that all the disposed stuff is realy worth many times over. If you have anything specific in mind drop a line direct I will make all efforts to arrange it for you.
Regards
Rahul
Nice to see some active participation from all corners.
Now for the non linear devices Ajju a fuse still depends upon its action as a result of current flow, obviously it would depend upon its series resistance, there are many purist who would not tolerate this in similar manner as a no cap in signal path logic. I do not think in those terms but again another logic is justified. Here we are building the ultimate biasing topology amplifier at par with world best and if there is any slight scope of some sort of compromise it should be avoided.
Now on topic of relays it should be taken in account that all relay have a specific contact resistance, again it is a purely mechanical device and lends itself to shortcomings like degradation with time. I am sure many of us have encountered problem of loose contact in switches and rumbblings and erratic operation in case of dirty or loose contact, after all this is what keep companies like WD40 alive.
In this design I feel it is practical if we provide some sort of DC protection , being a series pass regulator supply it easily lends itself to modification. While designing this we can easily incorporate power on delay also without introducing something in series with signal path.
As for the stuff available all is true, just think in terms anything used by the various services, Its all there. Possibly as condemmed, with a slight chance of restoration but its there.
It is a gamble its all your luck. He who plays stand to win.
Now Ajju few heatsink I too gladly procure for my friends but think of what I have picked up over years just to satisfy my own whims. Tough job convincing my wife that all the disposed stuff is realy worth many times over. If you have anything specific in mind drop a line direct I will make all efforts to arrange it for you.
Regards
Rahul
PCB for JLH ready . PDF is slightly more than 600Kb. Finding it difficult to convert to JPEG kindly suggest suitable solution for its posting on forum.
Regards
Rahul
Regards
Rahul
Tarun ,
Keep forgetting to ask you how has the tone control come out and its performance. Can the ckt drive an amp directly or a pre amp is needed.
Regards
Rahul
Keep forgetting to ask you how has the tone control come out and its performance. Can the ckt drive an amp directly or a pre amp is needed.
Regards
Rahul
Rahul, I too don't know what's true and what's irrelevant, but I tend to take a perspective similar to yours.Rahul said:
I won't put fuses in the signal path or the power amp output, partly because they act too slowly to protect the OPS devices, and partly because they may be non-linear, with a non-zero tempco. I'll gladly put fuses in the supply rails.
But I don't have any problem putting a (good) relay in the speaker output. The problems people have pointed out about relays is true, but those problems can be mitigated by using carefully selected good relays. In this area, my practical experience is limited, but do you get good high-current relays which have gold-plated contacts and are hermetically sealed? These should last long and perform well, without any of the non-linearities mentioned. Randy Slone happily uses relays for all his high-power amps for DC protection, but detects no problems either in audible or measured performance. Similarly, for signal handling, he's recommended that I use reed relays, and he's made it clear that he's never seen any problems with the signal quality due to them. He recommends mercury-wetted reed relays, but he's taken pains to point out to me that the reason for this choice is reliability, not any problems with the signal, i.e. distortion or noise. With all relays, corrosion at the contacts is always a problem, and if that can be addressed by choosing the appropriate type of relay, I think we're home free.
And as a sanity check, let's remember one thing: if we are able to hear sonic differences due to relay contact quality, maybe we will be able to hear differences due to selection of resistors and capacitors (e.g. Caddocks and Black Gates and Hovland Musicaps and so on). How many of us are willing to spend for those expensive passives? And if we are not, then will we really be affected by the alleged non-linearities of a carefully chosen no-nonsense relay? I'm not sure. This reasoning is the clincher for me.... basically, I'm not rich enough to be able to afford to hear the subtle audiophile effects of super-high-end component selections. 😀
Send me PDF, I'll generate whatever image you want by using screen capture from the Acrobat Reader window. (You can do it too, in MS Windows. No big deal.)Rahul said:
I'm still building it. 🙁 😀 When it's finished, I'll let you know. And yes, it'll drive an amp directly. After all, it's an opamp-based circuit (see here) so whatever a normal opamp can drive, this tone control circuit will drive. I don't know how well it'll drive 20-foot-long interconnects, but that, as they say, is another story. 🙂
Here are some thoughts.
First, about the output protection issue: fuses will have a measureable tempco, and if they are to be useful, expect lots of "nuiscance" blowing of fuses. Relays on the other hand, don't have too much of a problem. Sealed relays are already widely available, esp. in b'lore. There are PCB mounting hermetically sealed relays for telecom applications (rated to 10A at 30VDC) from (possibly Taiwanese companies) "Goodsky" and "Leone". I've used them for two of my amps for output thump protection. Work quite well...
One method I've always wanted to try is to use a controlled bridge (half or full), with SCRs. Then, by shutting off the gate drive, you can turn off the amp power in case there are problems such as overcurrent or heatsink overtemp. I wonder if this will be a useful method, esp coz it doesn't do anything about the filter caps.
I just got an idea: you can control the conduction angle of the SCRs to get a variable supply rail voltage. Then, the supply rails can be adjusted to be around 5-6V above the required minimum, yeilding very low heat dissipation... essentially what Bob Carver achieved with his "Sunfire" amp, although he used an SMPS instead of a controlled bridge. Yes, I know the SCRs turning off will introduce a large amount of noise, but this could be filtered out easily. Hmm, unless someone can conclusively prove this won't work or be beneficial, I'll try it. Someday... 😀 Comments?
First, about the output protection issue: fuses will have a measureable tempco, and if they are to be useful, expect lots of "nuiscance" blowing of fuses. Relays on the other hand, don't have too much of a problem. Sealed relays are already widely available, esp. in b'lore. There are PCB mounting hermetically sealed relays for telecom applications (rated to 10A at 30VDC) from (possibly Taiwanese companies) "Goodsky" and "Leone". I've used them for two of my amps for output thump protection. Work quite well...
One method I've always wanted to try is to use a controlled bridge (half or full), with SCRs. Then, by shutting off the gate drive, you can turn off the amp power in case there are problems such as overcurrent or heatsink overtemp. I wonder if this will be a useful method, esp coz it doesn't do anything about the filter caps.
I just got an idea: you can control the conduction angle of the SCRs to get a variable supply rail voltage. Then, the supply rails can be adjusted to be around 5-6V above the required minimum, yeilding very low heat dissipation... essentially what Bob Carver achieved with his "Sunfire" amp, although he used an SMPS instead of a controlled bridge. Yes, I know the SCRs turning off will introduce a large amount of noise, but this could be filtered out easily. Hmm, unless someone can conclusively prove this won't work or be beneficial, I'll try it. Someday... 😀 Comments?
- Status
- Not open for further replies.