Audio Project Amplifier Speaker Loudspeaker Kit
diyAudio.com diyAudio Forums Archive > Top > Other Stuff > Electronics and Parts
Pages: [1] 2 3 
[INDIA] Group Amp project - Click HERE for Original Thread
Rahul
Dear Friends,
Hi!
I have been following various threads at DIYAUDIO. I was very happy to find a keen group here making its presence felt worldwide.
I wanted a power amp, unfortunately Indian manufacturers are very few and the amps are available ex stock only at few places. Imported quality stuff was beyond my pocket, Chinese stuff I was not interested to put money in. Looking at various options I came across some excellent discussions on Diyaudio. It was noticed that we had limited options in case of drivers but when it came to amplifier, components and other hardware we were in much better position .
Tarun has already done most of the legwork in finding us a good source of cabinets, knobs and PCB. Net is full of excellent designs. We at DIY have one of the best brains available to discuss and seek guidance.
I feel a high end amp project will be of interest to most of us. A mention about the project on another popular thread has bought very enthusiastic response from many of my friends. A group project will be beneficial to all those involved mainly in sourcing of components, avoiding pitfalls and to sustain our interest to its end.
I have short listed my requirements and would be taking steps to fulfill them,
I seek your co-operation in creation of an equipment we all can be proud of having DIYed . The joy of having equipment bearing 'Diyaudio - INDIA' logo will be ours and ours alone.
Thanks and regards
Rahul
Rahul
If the posting needs to be under another heading I would request the moderator to kindly do the needful and excuse me for the error.
Rahul
Vivek
I am also thinking of building a good amp for my new Vifa-driver speaker project in the future. Exchanging ideas would be most welcome.
Rahul
Hi! Vivek

I have identified 3 broad categories that fullfill most of our demand:

1) High end :This broadly includes designs like Class A, excellent Pass lab schematics and tube amps. Outout power max 25W.

2) High power amp for our sub and other ocassions. Design like one from Aussie amp, Randy Salones designs and others in range of 200-250W.

3)Medium power designs of 60-80W, like the AKSA kit.

I suggest we pick one good design and move towards its completion.

I look forward to all your comments and possibly photocopy of documention of other good designs so that we may be able to get their boards made here in India.
Regards
Rahul.
sunil
Hi,

The project sounds good. To begin how are we going to protect the IP of the designers. The AKSA design is not available as a pcb. The only other I know which is reasonably priced for us here, in India & still does not infringe on IP is the Rod Elliot design. He sells the design & the PCB. So why don't we look at options that allow us to do such a thing.

All the other parts can be sourced locally, including the tranistsors, from rsindia.

cheers,
sunil
himanshuraval
hi friends,
i am also designing a car amplifier for my own car so if you any have a complate amplifier+smpsd designs with pcd design so please help me:car:
Rahul
Hi!
Good points Sunil.
As far as IP is concerned we need to take care while selection of the design. We go ahead only with those whose schematics are published and intended to be assembled by DIYers. This leaves us with few very good choices.
a) ESP and Nelson Pass.
b) Slone designs
c) Aussie amps
d) Elektor designs
e) JLH ckts.
Maybe few more. I seek more choices in this regard.

At present it is more important that we concentrate on one particular design. Important criteria while selection should also be the price to performance ratio as well as easy availability of components.
A single sided PCB. As it can be easily made in any city and I can always do the layout from schematics.

Least not we spend months collecting parts and project turn into 'Birbal ki Khichadi'.

My requiremnets have me leaning a bit towards Aussie amps N channel MOSFET Power amp design. Cause:
a)Simple single sided board
b)Good power, good references.
c)Cheap easily available MOSFET and rest of the components.
d)Guru Antony is there on Forum to help out. In fact he has lot of encouragement for people making their own layout of his designs.
Only thing lacking in this design is protection, an addon unit will be needed.

Himanshu you can always split the project into 2 parts
one make the amp and another do the SMPS. There are few designs of SMPS I would look for them. There are few sources of ferrites also.

It will be helpful if all my friends give their opinion on the subject so we can move ahead with the project.

Regards
Rahul

PS: Also checkout our thread:

http://www.diyaudio.com/forums/show...0124#post350124

For my post on MOSFET availibility.
corbato
Hi Rahul,

1. Most Pass design will require a 'scope to setup. Something that I suspect many of us don't have.

2. What is this Aussie Amp? Can I see a schematic somewhere?

3. Have you investigated the Leach Low TIM Amp? Lots of nice power.

4. JLH is a simple shot at building a nice high-end amp. All parts are easily available

Finally, how are we going to collaborate on this? A collective searching/buying of parts is one aspect. PCB's, as you have said correctly, can be made in any city or homebrewed. But it will be helpful to have a group PCB order.

If its tubes then I can organise some components.

Lets do it !!
Rahul
Hi!

Ashok as I was saying the choice would depend upon our requirements. Many Pass designs and JLH are biased for class A. No doubt for high end applications we would be hard pressed for better ckt. I remember studying JLH PCB sometime back, it is very nice and compact But for high power and possibilty of doubling up as sub amp we have to look at other designs.
When we go for high power levels Leach is very good but I felt it would be bit difficult to source NPN and PNP matched pairs specially considering group project. It is pretty difficult to get anything apart from TV, Tape ,CD,VCD and Invertor components in mkt. Similar is the situation at Delhi the quantity should be bulk only then shopkeeper will take some interest but forget about matched transistors.
Herein lies the beauty of N channel Mosfet design there are vast number of devices available. One type of transistor makes sourcing and matching easy also it gets good discount.
The site for this amp is:

http://www.aussieamplifiers.com/n-channe.htm

It appears manual downloading is disabled hence I am attaching a zip of schematics from my archive. Ashok you will also notice bare minimum of test instruments required.
While on topic I also suggest having a look at :

http://www.diyaudio.com/forums/show...goto=nextoldest

The layout by Pelle is something I am thinking in terms of.
Designer Anthony Holton appears to have a very constructive attitude towards people using their PCB layout for his ckt designs.

Contd.....
Rahul
contd...

Ashok you have rightly pointed that PCB should be done at one place this we can decide upon it can be either Mumbai or here North. My personal observation is that work quality is better at Mumbai.

On sourcing part I suggest we make a list and check out best deals in our respective town for whole lot. We then collect all stuff at one central place. Here we do the sorting and kitting and dispatch the complete package to its location by Speed post. ( Inter state often there are excise hassels with private courier.)
This way we will save time and also will not have to run after loose ends.
We can also go in for partial kit having only the PCB and non electrolytes, resistors. ie need to purchase MOSFET and arrange caps from ones stock. If we take Auusie amp as example

Components:

I suggest Master electronics at LRM ( Lajpat Rai Mkt) shop 345 for 1% MFR resistor cost is Rs12/ per 100 minimum quantity is 200pcs. If we take 25kit trial run we have 9 value that comes to Rs216 for 25 kits , Rs 9 per kit with lot of spare left.
Similarly ceramic cased wirewound can be ordered fron Vellohm. They are very reasonable and perfrmance is also good. In past I have placed small quantity orders and recived them by mail. Service is good.

For capacitors we can shortlist BC components ex Philips or Keltron both I feel are good. At Pune Mr Vilas Rabde at BC component is a helpful person specially if its hobby. Dealers are also there in all major cities. Critical values I find 1000uF160V and the main Filter caps.
Blue box type AC caps by philips can be used in zobel network and Yellow CTR and similar cap at input.
This now leaves us with the main power transformer now this I feel one should arrange locally ie get it custom made on EI cores or look for a torroidal ones. Sending them by post does not appears to be practical.

MOSFET from IR (International Rectifier) are easily available.

Looking back now I find this project quite feasible as well as enjoyable. BTW I have done some kitting in past too.

I look forward to your suggestion and co operation to get the project going.

Regards
Rahul
corbato
Hi Rahul,

Quite interesting.

Ok. Lets decide what goal we have.

I want to build at least 2 power amps. One to drive the upper band and the, more beefier amp, to drive the bass.

For the upper band I want a Class A. I'm looking at anything between 15 to 40 watts on this.

For the lower band a Class B or AB. I've built the Leach amp in the past and can vouch that its is simple enough design.

My past experiences with MOSFETS have not been very good. Both with commercial and homebrewed. The commercial was a Pioneer that I heard at my local high-end audio shop. The homebrewed was at guy's house whose skill I admire. But I've been reading on this forum that Pass MOSFET amps are really nice. Even Rod Elliot has given in to MOSFETS. So after all they are not a bad thing it seems.

Frankly, I was not giving much thought to MOSFET. But your views are changing that and if the group build veers to the Aussieamp then you may count me in too.

Parts:

While I agree that MOSFETS (BJT's) are available in plenty, I would be wary. I have been reading on net (Rod E has an excellent article) on the dubious nature of most of these devices. He points out China and India (naturally) as sources for these counterfeits. IR devices at Rs. 45 ring an alarm. But honestly I have no idea and things could be just fine. I will trust your views on these matters.

But MOSFET/BJT prices at RS India looks nice too. You need to order intelligently for this. Caps/Rectifiers etc are priced quite ok. All MFR that I've picked up in India have steel end caps. Throw them at a magnet and the stick. But the ones I rcvd from RS have no such issues. And they have clear leads to unlike the carbon coated one we get locally.

Aur ek source "phokat ka namoona" hai. Agar mei english pe likhun to iss forum ke finangi kaafi hulla karenge. Off list discussions karte hain iss barre mei.

Now coming to Chassis. Tarun has located an excellent source for the same. But if we are going to build a JLH or Aussi am then I believe we have to build our own chassis. Quite a simple thing actually. If required I can build them for this group.

So lets start by deciding on the following important issues:

1. Which amp or amps
2. How many builder
3. What skill set each of us can bring to this group effort.

Regards

NB: I'm also requesting amp guru Anshu to have a look at this thread.
tcpip
quote:
Originally posted by himanshuraval
i am also designing a car amplifier for my own car so if you any have a complate amplifier+smpsd designs with pcd design so please help me:car:
You're in the wrong forum. Go through the car audio forums; there are many of them around, and they are very active. Car audio is a specialised art+science.
tcpip
quote:
Originally posted by sunil
The project sounds good. To begin how are we going to protect the IP of the designers. The AKSA design is not available as a pcb. The only other I know which is reasonably priced for us here, in India & still does not infringe on IP is the Rod Elliot design. He sells the design & the PCB. So why don't we look at options that allow us to do such a thing.
I don't see any issue in IP infringement if we build our own amp with our own PCB, for our personal (i.e. non-commercial use) based on a schematic which has been published in a book or on the Internet.

I have been strongly influenced by Randy Slone's books and designs. And he has learned a lot from JLH and Doug Self. And I've been looking for a top-quality design which will have about as much power output as is needed for normal home audio, but no more. (If you go up the power curve, the absolute levels of distortion usually worsen. So an optimal power output figure is needed.)

I don't want to go for Class A, because I'm a beginner and I want to build an amp without messing around with all that heat dissipation. So my second amp will be the JLH 1996 (modified) design (with 2003 mods). But before that, I want to try my hands at a reasonable sized top-class amp design which even I can build. It is possible that many of you may have similar goals.

After going through all the sites that I've discovered or been referred to, I've finally come down to a BJT amp of modest power output (about 80W RMS), with ultra low distortion and output current limiting. The component count is low enough for me to hope that I'll be able to build it without too much trouble. The size of the PCB is about 6"x4" as given in Randy's book, and I'm quite sure we can even design our own PCB if we want to. It uses
  • 2N5551/5401 small-signal transistors
  • 2SB669/2SD649 driver transistors (high-voltage, low-capacitance versions of the old familiar BD139/140 class of devices)
  • 2SA1943/2SC5200 BJT flat-pack power transistors. Will work equally well with the Motorola MJ15003/15004 TO3 devices.
  • supply rails of +/-42V
This design is Randy Slone's improved version, with added current limiting and short circuit protection, based on Doug Self's very well known "Blameless Amplifier." Distortion specs are in the range of 0.0009% total, and second harmonic and third harmonic is actually at four zeros after the decimal point. Higher order harmonics are low, giving a benign frequency distribution of the harmonics. (And lest any of you feel like lecturing about how THD does not predict how an amp will sound sonically, please let me mention here that we all have read the same set of Web pages. :D )

Angshu has built the Doug Self Blameless amp, and it sounds outstanding. I dare say that these amps may be able to compete with Hugh's AKSA amp, sans the benefit of his fine tuning and premium grade components.

I'm totally in for this design. I want to make this one.... I've studied many, many alternatives, and for a variety of reasons, I feel that this is the one I want to try first, before I try anything else.

Those of you who want to build high-power systems may consider building a bi-amped or tri-amped system with this amp as a building block. For a high-power subwoofer, it may be necessary to use some other amp with 200W power output. Something like the Opti-MOS amp may be a better idea in that case. However, for the mids and highs, this BJT amp may give better performance than the Opti-MOS, in a multi-amped setup.

If you guys want to work on this amp, I can get the (single-sided) PCBs made here on 2-ounce FR4 board. Someone will have to design the PCB layout first, of course. And it may not be feasible to use the free version of Eagle that I have... it limits board size to 3.2"x4".

What do you think?
tcpip
quote:
Originally posted by corbato
While I agree that MOSFETS (BJT's) are available in plenty, I would be wary. I have been reading on net (Rod E has an excellent article) on the dubious nature of most of these devices. He points out China and India (naturally) as sources for these counterfeits. IR devices at Rs. 45 ring an alarm. But honestly I have no idea and things could be just fine. I will trust your views on these matters.
Ashok, I too had a lot of doubts about Rahul's pricing. So I did some R&D. You are welcome to go to www.findchips.com, type in "IRF250" and look for resellers who are selling it. The device is a 200V N-channel HEXFET device, which, in TO3 packaging, can deliver up to 30A. Rahul, please correct me if I'm wrong.

The price for this in USD 7.80/ea, in quantities of one hundred, from www.mouser.com. All other sources I checked have prices in this price range. It is totally unbelievable that any legal reseller of the genuine item will be able to import the device, pay Customs duty and shipping costs, and bring down the cost to USD 1.00/ea (i.e. Rahul's quoted price of about Rs.45/ea). I'm sure there can be differences of opinion on this, but when it comes to my personal choice, my mind is made up; I'll not trust these devices at these prices.

I have studied price variations and arbitrage opportunities in transistors and chips for almost the last two years as part of my hobby, just to see what's available where, how much volume discounts can one get if one has the budgets, etc. All my R&D tells me that a volume purchaser in India will be able to get the genuine items from overseas, cover his costs, and resell in India at about the same price as Digikey/Farnell/Mouser can sell in ones and two abroad. Any price advantage you get by ordering a thousand (no Indian reseller will order a million transistors) will be offset by the import duties and shipping costs. This is the rule of thumb I've arrived at after a lot of correlation of literally a hundred odd data points. Therefore, if US retailers sell this device in ones and two for USD 8.00, the Indian retailer will have to sell at about the same price or higher, even after accounting for large bulk orders.

This brings me back to a suspicion I've been having all along on this thread: we need to import the key components of our Indian group-build amp project. That's the only way to get genuine high-cost devices. If we can freeze on the amp, and the devices, soon, we can take advantage of some of our members' overseas trips. :D

BTW, the MJ15003/15004 devices, and the 2SA1943/2SC5200 devices are both widely used by a lot of high-end BJT amps. Even the AKSA uses the 2SA/2SC devices, I believe. And among MOSFETs, the 2SK1058/2SJ162 devices appear to be the consensus devices for high-quality amps. Rod Elliott and Randy Slone both use only these devices for all their MOSFET amps. So maybe we'll be safe if we stick to these devices... our investment will be protected, so to speak, whatever amp we build. We can even use these for the JLH Class A, or the Leach Low-TIM amp.

What do you say?
quote:
NB: I'm also requesting amp guru Anshu to have a look at this thread.
Yes, that'd be great, if you can bribe him to participate. His current bribery rates are very high. :)
corbato
Hi Tarun,

A Single Stage, Single Ended MOSFET design would appear to be an ideal Class A project for the first timer. MOSFET's are transconductance devices and Class A is the way to go. I think 15 watts from a suitable, single MOSFET rated at approx 175 watts/15A is a realistic thing. I'm sure there would be schematics floating around in the net with these goals.

But as I posted earlier, my experience with MOSFET has not been outstanding. I somehow like the sound of BJT's and therefore my first choice will be the JLH.

D.Self Blameless Amp has its own followers. I saw one guy's website (quite a hi end freak) who had built tri-amped rig around D.Self. And Angshu swears by his BA. So its is worthwhile to consider that. Yes.

Bi-amping is the ultimate goal here. At least for me. My Valve amps are fine. But I need another setup.

What are we thinking in terms of:

1. Linestage
2. Active XO (If Bi-amping)

Being a tube guy my choice obviously is that way. I'm not too sure if others would be that enthusiastic. However, if at all you guy's want to investigate this route then let me assure you that Tubes aren’t that dangerous or difficult as it would seem.
tcpip
quote:
Originally posted by corbato
A Single Stage, Single Ended MOSFET design would appear to be an ideal Class A project for the first timer. MOSFET's are transconductance devices and Class A is the way to go.
Do you think there are Pass designs which fit this description?

And Class A or no Class A, I'd be interested in a simple amp with really good sonics, as a first high-end project for myself. If Rod had generously given away the values of the components for his new MOSFET amp, I'd have been willing to try that one too. Even the P3A is a good amp, but if I go towards the P3A, I don't see why one shouldn't go a step more and do either the Blameless or Randy's ultra-low-distortion amp.
quote:
I think 15 watts from a suitable, single MOSFET rated at approx 175 watts/15A is a realistic thing. I'm sure there would be schematics floating around in the net with these goals.
I would personally prefer this sort of power rating only for a tweeter in a 2-way or 3-way setup. I think the mid needs more power, specially if like me, you want to have one mid handling everything from about 150Hz to say, 3KHz. But then your intended use for the Class A is precisely the way I'd have done it... confine it to the higher reaches. Let the sheen of Asha's voice come out A-class. :D (****! I think today is the Day of PJs for me. Sorry!)
quote:
I saw one guy's website (quite a hi end freak) who had built tri-amped rig around D.Self.
Yes, I think that this Slone BJT amp too can be an excellent building block. Even for woofers, if you build more than one woofer per speaker, you can drive each driver with a separate amp, and you'll have more than enough power that way.
quote:
Bi-amping is the ultimate goal here.
Strangely, I too have reached a point where I want to build active xo instead of passive. (It's all Angshu's fault, really.) Hence the need for really high-quality but simple to build amps, which I can build half a dozen at a time. :D
quote:
1. Linestage
I guess some will want the minimalist volume-pot-only approach, and others will want a conventional preamp (I'm in the latter camp, with bypassable tone controls). If others want to go my way, my PCB design can be replicated rapidly. Roadkill of B'lore has a preamp based on the LM1036, I think, which he can probably make avaiable for another approach to linestaging.
quote:
2. Active XO (If Bi-amping)
Want to do do LR4 based on dual-opamp chips? (This is a question to non-tubers, not to you. :D ) (Hey! This is a good term for you. "Tuber."!! :D ) I think LR4 is about the most mainstream and reliable approach to active xo, specially for first-timers like me.
quote:
Being a tube guy my choice obviously is that way. I'm not too sure if others would be that enthusiastic. However, if at all you guy's want to investigate this route ...
My answer for now is very clear... I don't want to try them for my first couple of projects. Let me get a few reliable, inexpensive amps completed first. :)
Rahul
Hi!

Tarun I would like to point to a slight error on my part the MOSFET i was telling are IRFP250 these are Plastic packaged probably package is called TO247.

I confirmed again in mkt and one the reputed shop has quoted at Rs45/, IRFP460 Rs100/ and IRF540 Rs18/ all in genuine IR packing. You are right about IRF250 it is indeed TO3 metal pack and more expensive. The IRFP variant is very popular with UPS manufacturers. I suggest just giving a call to any local shop there and we all can have confirmed price of these MOSFET at Mumbai.I am sure they would be near about my quoted price.

Self and JLH are world class designers, the THD figures are realy a proof of their mastery of the art. I doubt that we require such figures for a raw power amp.

Tarun I would look for the numbers you have mentioned for outputs I have not used them before, I am not very confident about availibility of matched complimentry pairs BJTs for output. You will notice this was one of the reason I have even avoided complimentry MOSFET design.
Using IRFP250 appeared practical hence I suggested the same. I have now noticed that Aussie design has already been made by our another friend here:

http://www.diyaudio.com/forums/show...p?threadid=3533


Ashok you are right a single ended MOSFET design from PASS can be considered very good as a beginer project. I have made many SE amplifier using mosfet for RF and feel they work very well and can be aasembled very easily. Heatsink sourcing I feel would not be as difficult as sourcing exotic output devices.

As you have rightly pointed we all would need two type of amps. I fully agree to go with either a JLH or SE type from Pass for the highs and a hefty one for low I am keeping all options open but for now Auusie amp still appears attractive to me considering all factors for the beefy amp.

Tarun for PCB I can do unlimited size here on EasyPC an award winning software, many friends also use ORCAD can request them for layout.

Ah yes abt the phukat method a group of us are known over the radio as Bravo , Oscar , Lima (BOL) Beggars- on -line. We got good samples for very high end projects even a DDS (Direct digital synthesis) chips.

I request first we shortlist class and amp power before we proceed.

Regards

Rahul

PS: here I have scope and other instruments for testing etc.
Workhorse
Hello Fellow Indians,

AMPMAN WOULD FEEL VERY GRATEFUL IN HELPING YOU GUYS.

In INDIA THERE are only 3 Pro-amp fabricating companies.

1 . WorkHorse Audio Inc. [PUNJAB]
2 . Stranger Audio Ltd. [CALCUTTA]
3 . Ahuja Radios. [DELHI]

HAPPY AMPING
Kanwar Abhishek Singh
Chief Technical officer
WorkHorse Audio Inc.
Nawanshahar-Jalandhar-Punjab
TEL:01823-221436,01823-310225.
Email:amp_man_1@hotmail.com
tcpip
quote:
Originally posted by Rahul
Tarun I would like to point to a slight error on my part the MOSFET i was telling are IRFP250 these are Plastic packaged probably package is called TO247.
Hi, Rahul. :)

Now that you've made things clearer, I was expecting less confusion. But in fact, I'm even more confused by what I just saw. I found that IRFP250 is a device described by different manufacturers as being different devices. For instance, some claim it's a MOSFET, some claim it's a HEXFET. Some say it has 30A current rating, and others say 20A. I'm simply not comfortable dealing with any device till I clearly understand what people mean when they use this part number. If you want to see what I mean, just go to www.findchips.com, type in "IRFP250", and see what comes up. Digikey has three devices with this part number, one made by STMicroelectronics ($2.81), and two made by Int'l Rectifier (prices: $5.90 and $3.33). All have the TO247 packaging. Going down the page and looking at the other quotes from other sellers, I see prices ranging from $1.57 (if you buy in hundreds) to $7.57. And they are made by at least three manufacturers: STM, Fairchild, and IRF. Which one would you choose? Me beginner... I'll just stick to BJT.
quote:
I confirmed again in mkt and one the reputed shop has quoted at Rs45/, IRFP460 Rs100/ and IRF540 Rs18/ all in genuine IR packing. You are right about IRF250 it is indeed TO3 metal pack and more expensive. The IRFP variant is very popular with UPS manufacturers. I suggest just giving a call to any local shop there and we all can have confirmed price of these MOSFET at Mumbai.I am sure they would be near about my quoted price.
I'm quite sure you've asked a reputed vendor, but you will have to read (or perhaps re-read) Rod Elliott's article about counterfeit semiconductors to really understand that there's no such thing as "reputed" or "reliable" in this business. And this is not a comment on India alone; it's a worldwide racket. It needs sophisticated clean-room labs and fab houses to create the fakes.

Therefore, there's nothing to be gained by my checking Mumbai prices. It doesn't matter what the Mumbai prices are.... if they are as inexpensive as what you'd said, they're guaranteed to be fakes. If they are expensive, then there's no assurance that they're genuine, but at least one can try. If you ask me, I'd simply not buy even a single piece from any Indian shop.... I'd try to order either direct from the manufacturer (IRF sells directly to individuals) or from a big int'l reseller like Allied Electronics or Digikey. Rod Elliott tells you about fakes sold by Digikey Canada, incidentally.
quote:
Self and JLH are world class designers, the THD figures are realy a proof of their mastery of the art. I doubt that we require such figures for a raw power amp.
I'm now no more sure what you mean by "raw power amp." Can you please explain? Are there different kinds of power amps?
quote:
Tarun I would look for the numbers you have mentioned for outputs I have not used them before, I am not very confident about availibility of matched complimentry pairs BJTs for output. You will notice this was one of the reason I have even avoided complimentry MOSFET design.
This matching of OPS devices is a bit of a controversial topic, somewhat like going for expensive Hovland Musicaps for a "cleaner sound." Randy Slone categorically says he's never matched OPS devices for his amps, and all his super-low THD measurements have been made without any such matched pairs. He says that matched pairs are only needed when you make long-tailed pairs for input stages. His opinion is that any high-GNFB power amp will be able to totally override any differences in hfe of OPS devices because of the GNFB. So, if you're building a zero-feedback power amp, maybe you'll need to match the OPS devices.

Hence, don't worry about that, would be my suggestion.
quote:
Ashok you are right a single ended MOSFET design from PASS can be considered very good as a beginer project.
Can you please give me a URL to schematics and other info? I've briefly looked though www.passdiy.com but have not found any specific details. And if we are talking simple amps, what about the DoZ (Death of Zen) amp from Rod Elliott?
quote:
Heatsink sourcing I feel would not be as difficult as sourcing exotic output devices.
I fully agree on heatsinks... they're available in L.Road here. But even OPS devices can be easily brought in by any one of us who goes abroad. You can even directly order them from, say OnSemi (you'll get MJ15003/4 from there), ask them for courier delivery to India, and pay Customs. One of my friends got 25 each of these devices this way. And since this was OnSemi's sample program, he paid only for the shipping, not the devices.

The best thing about a group-build is that we can source the components through a group-buy and pool together our resources. Let's not be held to ransom by the L.Road class of resellers... let's do something better. (Better does not mean more costly.)
quote:
Tarun for PCB I can do unlimited size here on EasyPC an award winning software, many friends also use ORCAD can request them for layout.
Sigh.... I wanted to get Vutrax up and running (Vutrax and Eagle are the only two good eCAD packages on the OS we use here). Vutrax' free version has no limit on board size, so it would be ideal even for large power amps like the OptiMOS. But there's some system mismatch, and I'll have to upgrade my laptop OS first before Vutrax works on it. :( Till then, I'm jealous of you. :D
sunil
Hello,

I like Tarun's idea of going active X/O. I'm not very technical & can't understand the lingo you speak. I do know what class A sounds like.

I have put together a 3-way active X/O & am currently looking at a tweeter amp. I'm looking at the Aksa or a Rod Elliot design. If we are looking at a group thing why don't we all go active, low power, multi-amp & gently nudge the neighbours.

cheers,
sunil
Vivek
Has anyone considered the Crescendo or the mini Crescendo? This is however not as simple as Holton's N channel amp or the P3A.
One more thing all of us need to look into carefully is the speaker DC protection. We in India always have a doubt about the quality of the o/p devices. If they fail, your speakers are history. This should be looked into also.
But again, as I understand it, adding a relay-based protection circuit poses its own problems. We need a good relay in the first place and are protection circuits not known to introduce distortion.

Vivek
corbato
quote:
Originally posted by tcpip
[B]
Do you think there are Pass designs which fit this description?
I was just checking out the passdiy site that you gave. Mind boggling stuff ! I'm going to spend the night browsing that site. I can see some SE designs. Look at the 'legacy' section.
quote:
I would personally prefer this sort of power rating only for a tweeter in a 2-way or 3-way setup. I think the mid needs more power, specially if like me, you want to have one mid handling everything from about 150Hz to say, 3KHz..
Are you sure of that range? I think going down to 150Hz could prove daunting for a Midrange driver. I would ideally like to Xo at 300Hz.
quote:
Strangely, I too have reached a point where I want to build active xo instead of passive.
Both have their own share of pitfalls. Phase alignment is one issue that would worry me a lot. Easy to mess up even a seemingly good active design particularly if you choose haywire xo points.
quote:
I guess some will want the minimalist volume-pot-only approach, and others will want a conventional preamp.
Ideally I'd like no vol pot/attenuator at all :) Since that is not possible the broad thumb rule for me is just a pot if located in the amp body. Else a line stage with moderate gain, particularly as Phono is my main source. For those who are contemplating the simple pot/active line stage here is a quick rundown:

1. Simple Pot:
Things can’t be simpler. Your precious signal doesn’t need to traverse another maze of electronics/resistors/capacitors/solder joints/PCB tracks. This results in lower distortion and better linearity. Flip side is that you are looking at a high output impedance and capacitance. Your pot is away from the Amp and interconnect can play nasty capacitance games. The worst part is that a passive stage is well passive. As in dead. If you need gain to drive an amp then you are stuck. Not if you like a dynamically punchy sound.

2. Active preamp:
Usually an a pot followed by an amplifying device – valve, IC, etc. With or without feedback. It usually overcomes the negatives of the passive design. You get lower output impendence, your interconnects are less of a factor and you really have the wherewithal to drive a power amp to clipping. Flip side is you may loose clarity and may actually end up introducing noise to audio signal.

The third possibility is a gain stage followed by passive stage. Not my idea of a linestage because it incorporate all the failing of both the above topologies. So you need to decide what you want.
quote:
Want to do do LR4 based on dual-opamp chips?
Do you have the LR formulae? I want tweak my Tube XO.
quote:
My answer for now is very clear... I don't want to try them for my first couple of projects. Let me get a few reliable, inexpensive amps completed first. :)
Most sensible. Can’t see any reason why you won't be proceeding to Tubes in no time :)
quote:
…DOZ..
Forget it. This one is looser. I built just one channel with fairly standard devices. Sounds pathetic. It’s hard to set up and maintain the bias, which drifts like a drunkard. I was attracted by the simplicity and the fact that it is based on the JLH ’69 design, which I know sounds terrific. I really admire Rod Elliot but this one I didn’t like. Most regrettably.

Rahul:

JLH is my first preference as I’ve heard the ’69 version. FWIW Sonodyne used basic JLH Class AB design for their most successfully amp from the ‘70s. Direct Coupled. They also used the Leach Amp design for their LAB series 120+120 watters (1978) amp. I realised this recently when browsing Sonodyne Technician Manual at the local dealer. Maybe Navin can confirm this with his friend Anindyo of Sonodyne.

I think you have the most experience with MOSFETS in the group so I will take your word on that. If we are getting MOSFETS for that cost then nothing like it. I'm looking at the Pass DIY site (thanks to Tarun). I can see that Pass uses IRF devices. Will you check out availability of these in the local market? Of-course we can always ask On Semi to send any sample by paying the 20$ handling charge.

Ampman

Thanks for the offer. You will have to endure us clowns for what we are. I’m sure all of us are assured that professional advice is at hand. Tarun will particularly like it. He only likes professionals :)

Sunil:

You need to be fairly careful with phase alignment issues with an active XO. In the worse case situation you may end up shifting 180 deg at say 1kHz ! Yes its is possible. And do you really need to go tri-amp. Unless you are very skilled with theory and possess right equipment, I don’t think this is the right choice. I’d rather do a Bi-Amping project first. Read up Linkwitz-Riley papers if you can. You may find them on the net. Best of luck.

Vivek:

Do you have the Crescendo schematic? I’ve heard so much about it. A Mumbai online company is offering kits for around Rs. 1700 IIRC.

While your anxieties are real, I do not like the idea of adding a non-linear thing like a fuse or relay contacts in the output path. My best bet is to find the real devices and run them under the specified envelop. DC obviously is not a good thing for drivers :)
Rahul
Hi! Friends

Thanks Amp Man for joining what do you suggest for a high quality 60-80W amp and for a 200-250W amp.

Tarun I checked data on IRFP250 from different manufacturers. Hexfet is proprietry technolgy of IR for mosfet manufacture hence they refer to it as Hexfet and others as MOsfet . For all practical reason hexfet is Mosfet.

More commonly available one is IR it has a continous rating of 30Amp at 25deg C this falls as temp increases. St microelectronics gives their rating as 33amp.

By reputed vendor I refer to one who is able to give bills for what he sells ie stands behind his stock. I do not think it is just to consider all that is sold in local mkt as fakes after all each town has its share of electronic units. Ron elliot has many examples of fake BJTs ( Not bought from India) As you truly pointed its worldwide. This limitation we will have with any design.
As far as importing is considered I am sure, on MOSFETs the importars will give us better quotes on bulk purchase in comparison to mkt price. I somehow do not feel comfortable with foreign sourcing every time I decide to make a project.

What the heck I am willing to take a chance on 6 MOSFET purchased locally. It may prove fruitful.

By raw power I ment something to drive the sub I feel three or four decimal THD figure would be an overkill here.

I summarise my requirements as a pair of medium power high end amplifier ie 15-25W possibly a class A with lower power output. One high power amp 200-250W good sonic quality but not the ultimate.

Amp man as regular user what is your opinion about these devices, can we rely on our local mkts. Any suggestions on tried tested designs.

Looking forward to all suggestions

Regards

Rahul
Vivek
Ashok, I do not have the schematics but they are freely available on the net. I also remember people having posted it on this forum. A search should throw up something.

Vivek
Rahul
Hi ! Ashok

Crescendo is an acclaimed design . My friend Charu at Mumbai does sells kits online but it is mostly the VEGA/ VISHA stuff. I am sure many would have seen his site I can always contact him for details on any of his kits.

http://www.circuitsonweb.com/resour...wer_Amplifiers/

7-8 years back one of my friend made this 240W design. Yes N ch MOSFET and it was well behaved. This was originally published in Elektor and has lot of protection have a look.

http://www.geocities.com/pa_schemat...0Wpoweramp.html

While you are at Pass DIY Ashok also go thru the Citation 12 Mosfet version under legacy I have the boards ready with me.

I feel like going for JLH class A and the N Ch ausie as both are simple designs, lot of documentation and easily available components. If nothing else JLH can even work with 2N3055 from BEL while we await imported BJTs.

Open to suggestions

Regards

Rahul
corbato
Vivek, Rahul
Thanks for the crescendo info. Not something I will contemplate anytime soon.

Rahul,
I'm steadying my resolve toward the JLH revised. While I looked at all Pass amp project, I'm yet to commit myself to a MOSFET amp.

In a previous post you had mentioned ex-WD heat sinks. Can you give some details on them ?

Thanks
sunil
Ashok,

Thanx for the concern. Let me present my case & you tell me where I've gone wrong. I know about the phase alignment issues possible with a 3 way active X/O. I started out 2-way but quickly ran aground just getting the costs for a good quality 2way or 3way passive x/o. I had to pad the tweeter by 3db & put in a zobel to keep the impedance constant. I considered winding the inductors myself & even that did'nt exactly work out cheap, for 14gauge. Factoring in the time & effort involved I started thinking active. I already had baffle diffraction in the 2-way actice x/o so that was not an issue.

Then came the issue of the amp I had for the t/m. It is good for the mids but the highs can get harsh at higher levels. Since the average sensitivity of the speaker was about 88db, I thought ( please correct me here if I'm wrong ) with an average loss of about 5-10db in the average furnished living room with the speakers about 10ft away I was not going to get very good, clear, loud sound, around 90db. C'mon once in a while playing RUSH at insensible volumes helps......

The active x/o is consistent with designs available on rane, linkwitzlabs or ESP. It is a 2-way kit from John Pomann which I have changed for a 3-way. It uses good components, atleast what most people say is good. For the baffle diffraction I had some help from Alex Megan. The linkwitz transform is yet to be wired. I know I'm risking it big here but a lot of people have good things to say about 3-way x/o's as well. The x/o is crossed at 3500 & 200. I've tried to keep the baffle diffraction compensation within one driver, the midrange. The baffle width is 7 inches. These crossover points don't fall in the critical range of hearing, or so I hear.......

The tweeter & the midrange are not time-aligned. They are on the same front plane. Even if they were time-aligned would'nt there be phase alignment errors with the cone of the midrange having to move a lot more than the tweeter. Is this a valid argument ? Isn't this a problem by itself ?

Since the power is now available for the individual drivers especially during dynamics, are'nt the drivers going to be operating at their optimum levels. They are not going to contend with each other for power on loud passages or during dynamics. The music I listen to is mostly like that. Tribal Tech, Rush, Zappa, Coryell

The drivers are D27, P13wh & P21wo-39-08

This is why I am keen to get a low power amp for the tweeter, something that is sweet & won't be harsh at the volumes my hearing-impaired brain likes to listen to.

Any corrections, thoughts, comments are more than welcome.
I have access to stuff from down-under & we can do a group-buy, if you like.

Cheers,
sunil
corbato
Hi Sunil
quote:

I had to pad the tweeter by 3db & put in a zobel to keep the impedance constant.
Thats fine I'd think.
quote:
I considered winding the inductors myself & even that did'nt exactly work out cheap, for 14gauge.
Not to mention sore fingers :)
quote:
Factoring in the time & effort involved I started thinking active.
Quite sensible.
quote:

..with an average loss of about 5-10db in the average furnished living room with the speakers about 10ft away I was not going to get very good, clear, loud sound, around 90db.
On the contrary, with separate amps you may get uncomfortable that the 1kHz to 2kHz range. Our ears are most sensitive at this range to any perceived peak in dB and/or phasing.
quote:
The x/o is crossed at 3500 & 200...These crossover points don't fall in the critical range of hearing, or so I hear.......
Yes you are right. Messing around with the so called golden band is not advisable. But my only worry with this thinking is how to find squawkers that handle things less than 700Hz nicely. Back in my younger days I had a wonderful book called "Build and Test your own Speaker System" or something like that. It proffered a thumb rule of xoing at 800Hz for an 8" woofer and 500Hz for 12". I built many speakers with that in mind. They sounded quite nice at least those days, when we mostly listened to loud Led Zep or Deep Purple things.
quote:
Even if they were time-aligned would'nt there be phase alignment errors with the cone of the midrange having to move a lot more than the tweeter. Is this a valid argument ? Isn't this a problem by itself ?
Yes and No. The time alignment theory by voice coil placement at best is erratic. Not a problem by itself when you consider what bigger mistakes one can make by aiming for that socalled symmetrical placement idea relative to the other enclosure.
quote:
They are not going to contend with each other for power on loud passages or during dynamics.
Actually a bigger problem occurs within your amp, where the entire spectrum of music is fighting for a fixed power handling capacity and looking to out put the same level.
quote:
This is why I am keen to get a low power amp for the tweeter, something that is sweet & won't be harsh at the volumes my hearing-impaired brain likes to listen to
Since you like to listen to music at a robust level consider the following power bands amps for tri-amping:

High: 15 to 20 watts approx
Mid: 40 to 55 watts
Low: 100 to 120 watts.

I'm sure your active XO has gain setting for individual freq band. You mentioned that you have changed that 2 way XO to a 3 way. Did you cascade one stage or did you add an additional stage to the existing design. Quite easy actually either way.

I'm sure your tri-amp project will be come out with flying colours. Keep posting the details so that we can all learn.
Rahul
Ashok

Agra has COD (Central Ordinance depot) any thing that is withdrawn all over the country is sent here. The local mkt Chipitola is the main surplus mkt with lot of disposal dealers. Range of stuff available is amazing be it auto spares, clothings , communication equipment, electronic stuff, binoculars test instruments etc etc... all used. Even parachutes, laser sights just amazing.

Heatsink is integral part of any equipment and here it is not the local stuff we are accustomed too but quality stuff mostly all imported and mil spec. Something on the lines of what we see in foriegn projects and high end commercial equipment. There is lot of variety, plain ones for diode mounting and also huge ones 5-10 kg each range or more, usually back panel of high apearage charger and other power supples. Then again heatsink from transmitter is also there now you can imagine a 1KW transmitter heatsink, amp there is usually Class AB and only 50% efficient. Then again HS from telephone exchanges, troposcatter units and number of things that even I could not make out head or tail. Generally these heatsink may have the transistors mounted on them genuine Motorola devices mil grade, as to junk dealers used components are not worth keeping. 3055 and 3773 as well as diodes are generally removed as it can be sold but anything else stays. There is lot of Russian stuff all hevy solid metal job just at scrap value.

Ashok of your interest there are lot of tubes miniatuire 9 pins, octal, ceramic tetrode just name it.and transformers from simple ones to 2KV or more from even in KVA ranges again this is all Mil spec like Hypersil transformers. Sorry not PP audiophile garde but old PA system yes.

All sort of test instrument is also available eg AVO 7 or 8 multimeter goes for Rs300/, scope in range of 1.5-3K, moving coil meter is also availble by tonnes good stuff SIEMENS etc goes for Rs20/ each. I recently purchased a Tektronix 2465B scope 4 channel 400MHz BW, for 2k. (Of course I had to repair it). You also get get function generators. Electronic equipment price is approximated at Rs50-100 per kg depending upon its condition, weight and appearance. New plastic panel stuff goes cheap> Main reason I got Tek scope for 2K, an old heavy Aplab 50Mhz was priced at 3K.

Of interest to DIY community, cooling fans are there , cetrifugal blowers, brushless, muffin from 3inches to more than 12inches dia.

Heatsink goes slightly higher than scrap rate last i bought they were Rs100-130/ per kg depending upon the condition and holes less holes more expensive but max around Rs130/. Amazing part since all the dealers are lincesed all stuff is sold on BILL. Pricing much depends upon your bargainig skills. My experience in dealing with dealers has been very good as most of them are Jains and quite honest once committed they stick to their words and do not take you for a ride. Much of the availibity also depends upon what has been auctioned lately but heatsinks are year round stuff.

For any thing even close to an Class A amp I highly recommend these used heatsink as they are much more efficient and offer bettar cooling than the usual stuff available in mkt. Cost apart.

A word of caution there is no guarantee that thing would work if you invest in complete electronic stuff after all it has been scrapped for a valid reason but generally I have managed to get them going.

I have given reference of other stuff so that DIYers of India can benefit at large. Hope this will be of help.

Regards

Rahul
Rahul
WARNING:

I have visiting this mkt for last 15 years. After sometime you will find your room starting to resemble the junk mkt.
sunil
Rahul,

The heatsink stuff sounds really good. Good stuff is expensive & with class A its going to be big. Can we all agree on what we are going to put together, Class A, class AB, 25w, 50w, 100w. This project sounds really good & we should make it happen. I'll leave it to the experts to decide which design is best.

Ashok,

Thanx for that. So with a 3-way active x/o I'm going to have phase errors in the 1khz & 2khz anyways. Is that it ?

The power band you've spec'd for tri-amping hit the nail on the spot. For the tweeter, that is exactly what I'm looking at, 25-50w. The rest is what I've already got. The x/o has gain adjustment for the tweeter only. Although padded by 3db, if I need more I want the flexibility. The 2-way was cascaded to a 3-way. The outputs are taken from the 1st & 3rd stages. I based the design on the stuff put up by Rod Elliot.

Cheers,
sunil
tcpip
quote:
Originally posted by Rahul
Agra has COD (Central Ordinance depot) any thing that is withdrawn all over the country is sent here. The local mkt Chipitola is the main surplus mkt with lot of disposal dealers. Range of stuff available is amazing be it auto spares, clothings , communication equipment, electronic stuff, binoculars test instruments etc etc... all used. Even parachutes, laser sights just amazing.
Man! This was one post of the kind which makes all the time I spend on this site worthwhile. Don't know how to say "Great stuff!!" loudly enough so that you can hear. :) It's not just how useful this info is for me (don't know when I'll ever go to Agra... I've never been there yet and half my life is over), but how interesting and evocative your description was.

Imagine! Building arrays of amps with big heatsinks and transformers (the two costliest items) all retrieved from disposal sales. At most the heatsinks may need to be re-anodised, for appearance sake. Slurpp! And finally, one day, I may even be able to afford a 'scope. The first thing I'll do with it is to put one probe in each ear of my wife, and actually see what goes on between those ears. Any other married member here felt this way? :D
tcpip
quote:
Originally posted by Rahul
Tarun I checked data on IRFP250 from different manufacturers. Hexfet is proprietry technolgy of IR for mosfet manufacture hence they refer to it as Hexfet and others as MOsfet . For all practical reason hexfet is Mosfet.
Got it. Thanks.
quote:
By reputed vendor I refer to one who is able to give bills for what he sells ie stands behind his stock. I do not think it is just to consider all that is sold in local mkt as fakes after all each town has its share of electronic units.
Rahul, I don't think I really should spend any more time arguing this point, because I think I fully understand your perspective, and any more arguments can lead to misunderstanding.

Just wanted to make one last attempt to clarify my point. The key point to understand here is that even the most reputed dealer, paying his Sales Tax and selling with a pucca bill, can be selling fakes. This was my reason to cite the case of Digikey Canada. The reseller does not know that these are fakes. He is just the last man in a chain of international distribution, and his immediate upstream chap may not know about the fakes either. It has been found that merchandising personnel, doing bulk deals for international distribution markets, often strike "very competitive one-off" deals from parties they do business with, and such one-off deals can sometimes be fakes, re-routed (laundered?) through "regular" channels to look respectable.

Let me explain a bit about how fakes work. We in India think that fakes are cheap knock-offs made by one sleazy guy with a monocle, sitting in a 5'x5' tenement in one corner of a slum, and shipped off at night in covered pick-up trucks. In the semiconductor world, fakes don't get made this way. Fakes are made by very large factories who have spare capacity, who identify that there are high-priced transistors (usually power transistors, both MOSFET and BJT) which can be "re-made" at a lower price and sold for high margins. The real margins in legal production of authentic stuff is wafer-thin, hence this constant search for "opportunities."

So, these guys will pick up, let's say, the very-highly-faked 2SC3281, which sells for a dollar in the wholesale market. They'll pick up the semiconductor core of the 2N3055 and package it in a TO247 case, with proper electrical connections, and label it "C3281". They'll sometimes omit the manufacturer name, or represent the label in some other format, so that they can even escape legal action at the crunch. This product, if sold as a 2N3055, would have cost perhaps ten cents in the wholesale market.

Now, what the end-user gets is actually a fully-functioning 2N3055 in a 2SC3281 package. The difference between the 2SC3281 and the 2N3055? Very big for some of us, but not at all that big for many of us. How many of us push a 2SC3281 to its limits of Vceo or the limits of its operating current? What we'll find is that the fake transistor will blow when you play your amp at high volumes but will behave quite okay at lower volumes. There are hundreds of circuits where a 2N3055 put in place of a 2SC3281 will not cause any problems most of the time. In audio power amps in particular, specially in OPS of Class B amps, you'll find that the OPS devices are stressed very little when you listen at "normal listening levels", but are stressed very heavily at full volume. For such apps, the fake devices will fail at full volumes, and the end user will simply conclude that he'd "pushed his amp too hard." The truth may be that he'd not even scratched the limits of a real 3281. And how many of us can tell the difference in linearity between a 2N3055 and a 2SC3281 unless we do careful A-to-B comparisons? Angshu made his first Doug Self Blameless amp with 2N3055s, and he said they sounded excellent. And let's face it: a 2N3055 can push out perhaps 6A of current. How often do have 6A of current flowing through our speakers? Our home use systems don't need the high rating of the top-class power devices. Yes, we'd love the linearity and extra reliability, but we are incapable of noticing the difference if we are given a fake.

So our end-user goes to the market, gets a replacement device (another fake), and again starts using the amp, believing it to be as good as new. Once again, if he plays the amp hard, it blows. And the cycle continues. If on the other hand he had these devices in a Class A amp, it would probably blow every evening.

I believe that many Indian (and non-Indian) end-users, right now, may be using fake devices in audio power amps without realising it. This is particularly true of those who buy Indian-made PA amps for "sound systems" for hotels and renting out for festivals and parties. Such systems are always manufactured under tremendous cost pressures.

Now I hope I've been able to explain that the reputation of the dealer and the presence or absence of a pucca bill has no connection with whether the transistor is a fake or not. The only way you can depend on a dealer is if that dealer procures his parts straight from the manufacturer, i.e. NatSemi or Toshiba. Digikey, Allied, RS, and a lot of the other big players do. I am sure not a single Indian retailer (not distri, but retailer) does. They don't have the volumes.
quote:
What the heck I am willing to take a chance on 6 MOSFET purchased locally. It may prove fruitful.
Fully agree. After all, you're talking of a risk of Rs.250. My problem, which does not apply to you, is that I am very poor at troubleshooting and repair. Hence, I want to take steps that problems don't arise at all, rather than take a risk and then have a part blow up in my face.
quote:
By raw power I ment something to drive the sub I feel three or four decimal THD figure would be an overkill here.
I now understand. You want a subwoofer amp. Yes, for such amps, you can use much more simple designs. Cool! :)
tcpip
quote:
Originally posted by corbato
It proffered a thumb rule of xoing at 800Hz for an 8" woofer and 500Hz for 12". I built many speakers with that in mind.
A good rule from a more modern source (Vance Dickason's Loudspeaker Design Cook book) says that if you build a 3-way, you should either have one decade or at least three octaves as the minimum gap between the lower and upper xo frequencies. This means that one-decade gap translates to 300Hz-3KHz, or 200Hz-2KHz, and so on. Three-octaves is a little less.
quote:
High: 15 to 20 watts approx
Mid: 40 to 55 watts
Low: 100 to 120 watts.
I find this diagram useful to understand music power distribution across frequencies.
tcpip
quote:
Originally posted by corbato
Are you sure of that range? I think going down to 150Hz could prove daunting for a Midrange driver. I would ideally like to Xo at 300Hz.
Aha! The trick I intend to use is to not use a traditional midrange driver at all, but use a so-called full-range driver. Something like a Fostex or a Jordan JX92S. That can handle a decade-wide range quite well.
quote:
Ampman

.... Tarun will particularly like it. He only likes professionals :)
I'll get you for this one day. Watch out! :D

Tarun
Rahul
Thanks! Friends

I sincerly hope we are able to come up with something useful.

Tarun

I realise often vendors are unaware of the electrical charecteristics of stuff they are selling. Sadly we have to live with fake components for now. For a one off unit purchasing direct or from suppliers in foriegn countries can be done but it becomes difficult when we have to compete in our local mkt. For our Amp use we can go in for foreign sourcing.

I just remembered about PCB layout I suggest you tryout "Sprint Layout" ver4 a German software, demo version is available for download, you will be able to master it in an hour I am sure, it is very simple and very good. Then download its Russian version4 from:

http://ra3ggi.qrz.ru/file.shtml

This has all features enabled saving , printing , two point autorouting, block area fill, groundplane and even appearance of complete finished PCB.

On subject of amps my opinion, a single sided PCB for the OPTI-MOS design is not an easy job it will require lot of trial with the placement.
As far as complexity is concerned class A JLH is very simple to assemble, it is only heatsinking and power supply we have to pay more attention to. Given our limitations JLH appers to be attractive. If we make any other design it still would keep nagging me that I have try out a class A solidstate amp.
Options still open for high power amp.

Sunil and thers I have made the PCBs for Pass modification of Citation12 and presently assembling them. Just check out about this design for our use. 45W /ch into 8ohms is my conservative guess it will push out. I will put up a pic of assebled unit and PCB in few days, you all will be able to see one of the heatsink purchased from disposal stuff.

While working along the lines I also peeked into one of the chinese amp 5ch from 'Horror'. In case there was a possibilty of modifications and much to my horror I found it was only 2 ch unit inside.final transistors are C4466 and A1693 one each per channel, volume control PCB has a BA5053 chip and tone control is based on 4558 dual OP amp.Power transformer is torroidal and supply caps are 10,000uF/50V 2nos. It appears 5 inputs are summed up and output for sorrounds derived from main L&R thru resistors in series. All in all a waste of money i feel. Price 4K.

Recently I was at IIT Kanpur conducting workshop have a look at the pictures:

http://www.indiahams.com/techkriti.htm

My radio call ia VU3WJM.

Regards

Rahul
tcpip
quote:
Originally posted by Rahul
I just remembered about PCB layout I suggest you tryout "Sprint Layout" ver4 a German software, demo version is available for download
Does it run on Linux? If not, there are plenty of other Windows-based packages too, but I use Linux.
quote:
On subject of amps my opinion, a single sided PCB for the OPTI-MOS design is not an easy job it will require lot of trial with the placement.
There's the PCB design with component placement diagram that he's given in his book. One can always try to replicate that.... it's single-sided. About 6"x9". I'm not suggesting that OptiMOS is the only way to go, but in case you want to, we have lots and lots of info from the designer's book.

This is getting interesting... good you started us all on this Indian group build thing. Now if only the professionals would leave us alone.... :D
corbato
Rahul:
Awesome info on the COD. I'm sure you may even be able to pick up Manpacks for your use too. Keep an for AF modulating tubes such as 845.

Do you think it is feasible to collect Heat Sinks from COD for JLH group build? i think I'll give those used Transformers as pass for a Class A project.

Ok. So how many of us are in this group build. I'm going to Germany this month end and I can collect WIMA MKP capacitors for the JLH or any other project. Just let me know your requirements. I can despatch the same to whoever needs 'em by 1st week of April. But then any one interested will need to tell me know as I will ask my contact in Germany have them ready.

I'll again emphasis that Semi should either be requested as "phiokat ka maal" or bought from RS. Resistors to should be sourced from RS. They sell Caddock too. And I offer to build and ship the chassis for those keen.

Lets freeze on the JLH. Its simple. It sounds good.
corbato
Posted by tcpip
quote:
..Something like a Fostex or a Jordan JX92S
Yes I'd thought of a FR. But do you get these in Mumbai?
quote:
I'll get you for this one day. Watch out!
You threating to let loose a certain Ape on me?? I'm wetting my pants already :hot:
corbato
quote:
Originally posted by Rahul
Recently I was at IIT Kanpur conducting workshop have a look at the pictures:
What? so many OM's and not a single YL ??? :dead:

My 73's 2 u.
tcpip
quote:
Originally posted by corbato
Yes I'd thought of a FR. But do you get these in Mumbai?
Strictly speaking, I got my pair of JX92S in Bombay. I asked a friend to order them from E J Jordan UK, and he paid full courier charges. The Customs guys decided that only two drivers means it's a sample, and charged no duty. My friend (a hardware reseller and audio keeno) thus procured the pair through absolutely proper channels, and gave me a bill with sales tax added. And I got them in a week's time, delivered to my doorstep. Any hardware reseller who occasionally imports items will be able to do this for you, provided you are willing to pay shipping costs. The drivers cost me UKP 150 or so for the pair, and another UKP 50 or so for courier charges.

Jordan is one of the few speaker driver makers who sell retail quantities directly to customers anywhere in the world.
Pixo
These are some of the IRF distributors in India i pulled off the IRF site
I hope it helps

Hi-Rel Products Sales Rep
International Rectifier
#407, Prestige Centre Point
Cunningham Rd.
Bangalore 560052, India
Tel: (91) 80-2208 8025/26
Fax: (91) 80-2208 8028

Arrow Electronics India Ltd
5th Main, II Floor
Pusa Road
New Delhi 110 005, India
Tel: (91) 11-578 4629 / 3845
Fax: (91) 11-578 5751

Arrow Electronics India Ltd
311/312, Chintamani Plaza
Adheri-Kurla Road
Adheri (East), Mumbai 400 099, India
Tel: (91) 22 5692 5196/8/9
Fax: (91) 22 5692 1415
E-mail: rajeev.mhashelkar[at]arrowasia.com

Arrow Electronics India Ltd
No. 49, I Floor, Rama Archade
Hospital Road, Shivajinagar,
Bangalore 560 001, India
Tel: (91) 80-558 1520
Fax: (91) 80-558 0909

Future Electronics (New Delhi)
No. 1308/9/10, Ansal Tower
38 Nehru Place
New Delhi 110 019, India
Tel: (91) 11- 646 1414/1417
Fax: (91) 11-646 1418

Future Electronics
#710/711, Gateway Plaza
Hiranandani Gardens
Mumbai 400 076, India
Tel: (91) 22-570 2745
Fax: (91) 22-570 2743

Future Electronics
No. T 8&9, Gem Plaza
66, Infantry Road
Bangalore 560 001, India
Tel: (91) 80-558 7885/7886
Fax: (91) 80-558 7890

Kudamm Corporation
D-17 Kalkaji
Opp. Nehru Place Terminus
New Delhi 110 019, India
Tel: (91) 11-622-9093
Fax: (91) 11-647-9097

Intraco Limited
No. 579 2nd Floor
11th Main Road
5th Block Jayanagar
Bangalore 560 041, India
Tel: (91) 80-665 4545
Fax: (91) 80-699 1001

Cresta Instrument
9-A, Sitladevi Indl Estate
140, Sitladevi Temple Road
Mahim, Mumbai 400 016, India
Tel: (91) 22-446 4946/ 446 4801
Fax: (91) 22-445 7230

Advent Electronics Pte. Ltd.,
R-283, Greater kailish-1
New Delhi 110 078, India
Tel: (91) 11-629 4370/4372
Fax: (91) 11-629 4398

Advent Electronics Pte. Ltd.,
# 202, Lok Centre
Marol Maroshi Road, Andheri (E)
Mumbai 400 056, India
Tel: (91) 22-5675 6584
Fax: (91) 22-5675 6583

Advent Electronics Pte. Ltd.,
No. 85, Ground Floor
5th Main, 1st Cross
Domlur 2nd Stage
Bangalore 560 075, India
Tel: (91) 80-535 0357
Fax: (91) 80-535 1869
tcpip
Remember, I was talking about Randy Slone designs? Well, I found this thread with lots of war stories about his amps. And apart from the OptiMOS design, which people seem to be really raving about, there's also the "Figure 11.4 design" mentioned in this post, which is the ultra-low-distortion schematic I'd mentioned earlier, with clear similarities to the Doug Self Blameless amp. The design is covered in Figure 11.4 of his book, "High Power Audio Amp Construction Manual."

Just thought I'd tempt you towards the Class B devil, seeing how all of you seem to be veering towards room heaters ... oops, Class A amps. :D
Pixo
Oops! went a little too long....... sorry! :rolleyes: :)
Rahul
Hello! friends,

Sunil

Tnx a lot for info on IR distibuters. I am sure it will be of help if we go in for Amps based on these devices. I do remember giving a call to this Pusa road unit long time back but it might have to do with something else.

Tarun

For Linux PCB software keeps popping up every now and then I have not used it but its free from here:

http://bach.ece.jhu.edu/~haceaton/pcb/

Sprint is windows based.

This is really helpful that OPTI-MOS has a single sided layout already done this will be one big help if in future we decide to go ahead with this design. Will it be possible to mail me a scan of this PCB at
lkorahul@yahoo.com
I would be highly obliged and will try to learn something from this layout.
Tarun since I am already in middle of an amp project I will be grateful if you will just go thru the MOSFET mod of Harman Kardon Citation12 at Pass DIY under legacy section and give an opinion. I have been facing a very embarrising situation, every now then friend would request me for a good system for occassion and parties now I simply do not have anything to give them hence a powerful amp is also needed.

Ashok

I too feel JLH class A is one we should freeze for reasons
1) It is simple
2) Common availibility of components.
3)Simple PCB
4) Ultimate class of scheme.
5)Good performance and documentation.
PCB I can do at par any on the forum and even get them made here.
It is fully practical to go in for bulk purchase of heatsink from disposal. I frequently visit Agra and can select one suited to our needs. How else do you think I collected all these check here:

http://www.qrz.com/callsign.html?callsign=vu3wjm&bs=vu2

Same cannot be said of transformers as no two of them are similar one off its OK. Secondly our specs would vary from what is available.
Again we would need to decide on type of transformer torroidal or EI. If EI then I suggest looking for old 24V type defuct UPS 500VA are good ones for our class A project, multiple taps at mains input gives us some control over output voltage and yes they can be had for free. ( I already have one).

I did the magnet check on the MFR they failed all stuck, so did all CFRs and surprisingly even the Philips blue resistors.

All Ok abt the YLs ie young lady from 6-116 yrs young. Only ones I saw were at CR the campus restaurant. During Fox hunt we suggested giving hints like I see YLs all around , students said it will mislead as all will take it as Fox is definately out of campus.

Tnx for the 73 ( Regards) My 73 to you and family and 88s ( Love and kisses) no porn here all international telegraphic lingo.

Best wishes

Rahul
Rahul
Bunchful of extra 88s for the YL on boat. OOOOH LALA LA OLE O OOOH LA LA!

Good Times

Rahul
corbato
quote:
Originally posted by Rahul
[B]I too feel JLH class A is one we should freeze for reasons
1) It is simple
2) Common availibility of components.
3)Simple PCB
4) Ultimate class of scheme.
5)Good performance and documentation.
And it ain't uses no MOSFETS :) I've set my eyes on this version:



There is another, more powerful version which I’m not keen because of the following reasons:
1. Multiple gain stages
2. Dissipating all that heat from 8 output + 4 PS devices will render my listening room very hot indeed.

Your opinion pls.
quote:

PCB I can do at par any on the forum and even get them made here
How about the above schematic?
quote:
Again we would need to decide on type of transformer torroidal or EI.
EI. No doubt about that. A dual sec 25volt, 5A transformer wont be very expensive. My last transformer was 350-ct-350 (250mA), 5 Volts (2A) and 6.3 V (3A) cost me around 550 bucks.
quote:
...as Fox is definately out of campus.
Yeh kya bandhu?? IIT K kya marubhumi mei parivartit ho gaya hai ?? Yeh to atyant hi chintaa ka vishai hai :xeye: Sub apne karmo kaa fal hai bhai.
corbato
quote:
Originally posted by Rahul
Bunchful of extra 88s for the YL on boat.
None that I can see in the India threads :( We do have a LID though. Thankfully he is back in his tree.

Hoping for eyeballs soon.
Rahul
Dear Ashok and others,

Yes this is the same ckt I also had decided upon. Let us hear from others. At this power level we need not worry much about heat and rightly you have pointed a 125VA EI will not be too big. While on subject of PSU I missed telling you from disposal we also get very good quality main filter caps I have usually bought 33,000uF -68,000uF 25-30V for my radio PSU but other ratings are also available ie lower uF and higher V. Again they are of best quality like Sangamo, Farnell, Cornell-Dublier, Sprague Italy etc. Cost in range of 30-45Rs. Ideal high end stuff.

I will start working on PCB just improvement on other designs. It can be made on vero board or point to point wiring but again we cannot expect every one to be equally deft with the soldering iron, a PCB would be helpful.

BTW 88s were for the kingfisher pic going between George and you. LID has been shut tight for time being.

Soory Pixo Thanks to you for the distributer list.

Regards

Rahul
corbato
quote:
Originally posted by Rahul
Cornell-Dublier
Didn't know that Dublier guys are still around ! Have some NOS Dublier oil filled caps waiting for a proper tube project.

Resistors from RS have non-magnetic lead/end caps. But can we have a head count here. Just who else, besides you and me, is keen on JLH project.
Rahul
Hi! Friends

Its my majboori to make a good amp, Ashok here willing to join now I sincerely look forward to all your participation.
We now have inputs and some excellent sources that would enable us to achieve our target not only at a low cost but provide us with an ultimate high end equipment of a standard which is not available in mkt. Also the satisfaction of having done it with our own 2 hands.

I say lets do it We can make it work.

Regards

Rahul

PS: I still dont see any better options at comparable cost.
Rahul
Dear Ashok,
This one specially for you.
I have to make one. I will be making a couple of boards for my self and feel there would would be spare enough for few friends.
BTW I have to collect 2 HF sets that I have purchased so if things work out I would procure sufficient stuff for those interested in this project ie the heatsink and good quality caps.
Cornell Dublier is still in mkt, it appears it supplies to only top of the line companies like Tektronix, Hughes, Fluke, TRW Dallas and possibly NASA etc I did recieve some samples last year in relation to a defence project.

Ashok my judgement as Secy of UP radio org you qualify fully as a HAM how abt just fulfilling the formality of giving the exam and getting the ticket ie license.

regards

Rahul
corbato
quote:
Originally posted by Rahul
[B]Its my majboori to make a good amp[/qoute]
Mailing something for your on this.[QUOTE]not only at a low cost but provide us with an ultimate high end equipment of a standard which is not available in mkt. Also the satisfaction of having done it with our own 2 hands.
Amen to that :)
Boss. I think even two of us will make it a group build. So do you concurr that we should start ordering parts from RS? I may be having some of the 3055 or MJE devices from On Semi samples.
quote:
Originally posted by Rahul
[B]I will be making a couple of boards for my self and feel there would would be spare enough for few friends.
Chalo boss jald kaam shuroo karte hai :)
quote:
I would procure sufficient stuff for those interested in this project ie the heatsink and good quality caps.
Heatsinks will be particularly helpful, But do you think caps would be in good condition?
quote:
Ashok my judgement as Secy of UP radio org you qualify fully as a HAM how abt just fulfilling the formality of giving the exam and getting the ticket ie license.
Janaab aap to bahut bade fish ho !. I was very attracted to Ham as kid. Unfortunately, my family was not financially well off to support those kind of activities. By the time the cash was more plentiful I had other interests. But yes! Me and pal of mine (60yrs young) have been thinking of HAM for a while. Jsr had a thriving Ham community in younger days. Perhaps some of them are still around.
tcpip
quote:
Originally posted by corbato
Boss. I think even two of us will make it a group build.
Hey guys, I'm in on it too, and so will Angshu be, I think. He's been keen on this one for some time. I have some questions:
  • Is this the schematic of the most recent, refined version with the 2003 mods? I would like the most recent mods, for obvious reasons.
  • Are we going to make the PSU PCB too? I hope so...
  • Can we design the PCB, get two fabricated, build them, test them, before we make the mass fabrication? I think there'll be a demand for mass building, because PCBs are inexpensive (I'm getting rates of less than 50p/sqcm in Bombay, with solder masking and legend printing, for quantities greater than 25). At that rate, and for an expected smallish PCB, I'll probably opt for half a dozen myself, "just in case"
  • I hope we'll use the MJ15003/4 devices? I'd really like that. Maybe George can even get us some devices if we order them right away and ask for them to be shipped to his brother's address. We'll have to decide amongst ourselves first, then check with George. He's agreed to get some other chips etc for me from there.
Angshu will be back in the loving embrace of his family today. From tomorrow, we can turn up the pressure on him and make him join in. He can even do PCB design and the like...he did the Blameless amp PCB in an afternoon, and it works! :)

Of course, I know the JLH will not satisfy the needs that all of us have for a more powerful amp, but I guess we'll decide that one as time goes by. Maybe we can even do the OptiMOS yet. :)

Tarun

PS: We should now demand that all of us Indian builders read the entire super-long thread on the JLH Class A which is there in the Solid State Forum... I'm sure it'll help to build a common ground for our discussions/designing.
corbato
Tarun:

Good u ain't watching the stupid match. 73/3 :( I give up.

Back to the amp.
  • Yes this the most recent revision. Their is another, higher power version too. But not being thought of at the moment. Reasons are listed in an earlier post
  • Yes
  • Why not. A vero board job. I've heard the '96 version rigged up that way.
  • Those MJ devices always make a better sub for the 2N3055. I got MJ's as sample from On Semi. And why should we trouble George? At least to me On Semi ships regularly. Now they have $20 handling fee of-course.
I trust Angshu had a close inspection of those Soft Dome Twits. Will be interesting to know his views.

In addition to that JLH thread, do visit this website.
tcpip
quote:
Originally posted by corbato
  • Yes this the most recent revision. Their is another, higher power version too. But not being thought of at the moment. Reasons are listed in an earlier post
  • Yes
  • Why not. A vero board job. I've heard the '96 version rigged up that way.
  • Those MJ devices always make a better sub for the 2N3055. I got MJ's as sample from On Semi. And why should we trouble George? At least to me On Semi ships regularly. Now they have $20 handling fee of-course.
Great. I'm certainly in on this, then.

BTW, I just got a quote from Balaji Hertz. If you want to build 50 PCBs, 2-ounce copper, 9cm x 4 cm, he'll charge less than Rs.25 per PCB, including legend printing and solder masking, and after including all applicable taxes. The prices that guys in the US and Europe pay are astronomical in comparison. We should build arrays of Gainclones just for decoration pieces, at these prices. :D

The JLH should be good! What power output can we hope to get with these devices and the latest mods? About the prototype, I was wondering whether we should actually make a couple of PCBs, and actually test that the PCB design itself is okay, before making the bulk set of PCBs.

Tarun
roadkill
Hi all,
MOSFET fans, here's a nice design to consider. I found it in Electronics World, while in Pune. The schematic's at my website, with a small change: the original asked for a single IRF640/9640 pair, the schematic shows two without any split gate drive resistors or source resistors. My bad... anyway, the version I've built used only one driver pair per channel. I later pulled them out and used an IRFP240/9240 pair (cost me Rs. 250 per pair). The amp works just GREAT, it's my preamp which gives me trouble (ground loop, and it's only marginally stable). I know, it's a mess, but I made it while in college, so I had to take a lot of shortcuts. Consider this design for something in the 50-100W range. Probably we can increase the output power by paralleling output drivers. IRFP240 seems to be genuine, and, well, it hasn't blown on me yet. And I do wake up the neighbours some time with some Massive Attack.
In case anyone's interested, I can put up a board for this amp, but it has holes only for IRF640/9640. I used this same board with the 240/9240 through soldered wires.
Rahul
Hi Friends,

Call it a gift from almighty I feel good vibrations wether I see a good design, a good project or a group of very dedicated and nice people. It is my good fortune that I find all of the above here.

Tarun :

Rs25 per board is very reasonable. You dont even get a kiddies project PCB for the price and here we talk of high end amp. I offer to pay up for all the 50 PCBs. Others joining simply pay the actual cost incurred and postage to their respective destination as when they join this way we will have something available for Indian DIYers for a long time. Tarun as far as PCB goes there are already many tested layouts it is simple matter to make one along the lines and of such a size will take only few hours. Other experienced members can guide us to suitable changes I for one fully willing to do the artwork. No trouble at all in providing the gerbers for the manufacturer. I do not recommend PCB for the power supply board as we already have the bridge rectifier and fiter caps off the board. Now it all leaves us with the two transistors and regulator if we go in for regulated PSU version , again a point we need to decide before we proceed ahead.

Roadkill:

I am happy the neighbours do wake up least, you may have to change ur ID here but more important that u are not whisked to the big guest house with many guards, huge gates .
Again I find this amp design very good (my vibrations) CCS for all stages, complimentry and not overtly complex. I am sure the performance is also good.
I have made Citation12 Mosfet PCB but yet to fire them up, hopefully in few days. It is much simpler and maybe in performance department too.

Ashok :

Hey we WON. Thanks for ur offer, more on that in direct mail.
The caps Ashok I had no trouble in radio use but again that is no audiophile stuff, basic communication quality 2.7Khz and that also in class AB.
Today I do the final check of all available layouts and then we proceed with our own. What do you say should I try to go in for a plug and play type layout ie RCA socket for input, SIL connectors 6 pin for power devices all on board. Those willing still have the option to solder all direct.
Even I have couple of original Motorola 3055s but we will hav