advertisement


John Linsley Hood Class A or similar

Thanks gents, I see what you mean. Not sure though why the rectifier diodes need to be so heavy. High current I can see, obviously, if the thing is burning (say) 35W continuous then that's just over an amp, day in day out. However I don't get the 200V capability. If you have a 27V secondary I can see how you might need a 54V diode but not 200V. What's the maths on this one?

I'm also unsure how a 22-0-22V secondary will generate 30V rails. 22V is an RMS figure, yes? So the peaks will be over 22V. However they won't generate 30-0-30 rails, will they? Or are you referring to ignoring the centre tap and taking 2 x 22V which after rectification and smoothing might be 30-something V?

Sorry for the dim questions, it's been a long time since my electronics studies.
 
It's hard to get the terminology correct. Full wave rectified indeed, as a whole, but half from the pov of each part of the winding....
 
stevec67 said:
Thanks gents, I see what you mean. Not sure though why the rectifier diodes need to be so heavy. High current I can see, obviously, if the thing is burning (say) 35W continuous then that's just over an amp, day in day out. However I don't get the 200V capability. If you have a 27V secondary I can see how you might need a 54V diode but not 200V. What's the maths on this one?

I'm also unsure how a 22-0-22V secondary will generate 30V rails. 22V is an RMS figure, yes? So the peaks will be over 22V. However they won't generate 30-0-30 rails, will they? Or are you referring to ignoring the centre tap and taking 2 x 22V which after rectification and smoothing might be 30-something V?

Sorry for the dim questions, it's been a long time since my electronics studies.
Click to expand...

You're not trying to generate 30 0 30 rails!! Just a single 0 - 35 one. Trust me on the diodes;) 200V is a fairly low rating anyway and gives adequate safety margin.
As I said in last post you really want 27 0 27 which you should get from the TX of a 50 - 60W amp. This should give approx 0 - 35 for the schematic as posted by S-Man. 2-3V either way is not worth worrying about. Two diodes, anode end of each to the 27V of the TX, Cathodes connect together and go to the + end of the smoothing cap and the centre tap of the TX goes to the - end of the smoothing cap.
 
stevec67 said:
Not sure though why the rectifier diodes need to be so heavy. High current I can see, obviously, if the thing is burning (say) 35W continuous then that's just over an amp, day in day out.
Click to expand...

Worth a response for the record, even if obvious to the OP - the DC side of the supply is delivering somewhat over an amp, per channel. But recharging the reservoir caps happens only at peak input voltage therefore constrained to <2-3mS, every 10mS (at 50Hz): that means the peak current input through the rectifier diode is easily 4-5x the mean load current. Hence - 10A diode is a sensible suggestion. Also why reservoir caps with low ESR and a high temp rating are the way to go: ESR* (ripple current^2) quickly becomes a big number as wasted heat, that will cook the cap from the inside out.

Such an apparent over-spec is true for all simple dc supplies - but once you are into significant DC draw on the rectified output, the input requirement becomes hairy, quickly. Also why generously-dimensioned transformers are worthwhile, though not cheap: peaky 5-8A charging pulses through secondary winding resistance on a barely-sufficient secondary easily generates a lot of waste heat / a hot transformer.
 
Thanks Martin, hadn't thought of that. I am beginning to undersand why big Class A amps are so big and expensive compared to more efficient designs.

I didn't realise that transient currents could count for so much in a rectifier, I thought it all came down to linear Rms type calculations and the ripples could be neglected. Clearly not. There are more things in heaven and Earth, Horatio, than were ever taught to you in A level Physics.
 
Wow. That looks great. Where did we ever find this stuff before the Internet?

It's interesting, all these electrical engineering problems seem to have solutions. It's almost as if the problems have been encountered before and solved, by, say, engineers, who have then written up their findings.
 
stevec67 said:
Thanks gents, I see what you mean. Not sure though why the rectifier diodes need to be so heavy. High current I can see, obviously, if the thing is burning (say) 35W continuous then that's just over an amp, day in day out. However I don't get the 200V capability. If you have a 27V secondary I can see how you might need a 54V diode but not 200V. What's the maths on this one?
Click to expand...
You can use half a standard bridge rectifier, to keep things simple. Get one that bolts to the case and makes construction simple.

I'm also unsure how a 22-0-22V secondary will generate 30V rails. 22V is an RMS figure, yes? So the peaks will be over 22V. However they won't generate 30-0-30 rails, will they? Or are you referring to ignoring the centre tap and taking 2 x 22V which after rectification and smoothing might be 30-something V?
Click to expand...
The unloaded rails will be 1.4*22v, under load it gets complicated. The minimum voltage will be dependent on the size of the caps, the load and the maximum achieved, the max then on the ability of the transformer to charge the caps in the time the output voltage is higher than the current reservoir capacitor voltage.

Sorry for the dim questions, it's been a long time since my electronics studies.
Click to expand...
Hardly dim questions, but perhaps don't over-think it. Practical hands on experience makes things much clearer. Especially if you can source an oscilloscope and watch the shape of the voltage on the reservoir caps. And the worst that can happen is a bit of smoke.

Way long ago I built a phono preamp based around a pair of Naim MC cards and a couple of discrete voltage regulators built on Veroboard from a design somebody posted here. I knocked up a PSU with two caps separated by a small resistor, single ended. I thought a Schottky rectifier would be cool, found one with two in a package, so exactly as we're discussing here, centre tapped transformer, twin rectifier, single ended PSU. Everything seemed to work, but the rectifier got really, really hot. This is obviously a low current application. Eventually I read the data sheet for the rectifier and figured out what was going on, although the detail has been archived, changed the rectifier for something I already had and the phono amp has lived happily ever after. All you really need to know is not to touch anything that can kill you, and cover over all the mains wiring. After that you will figure it out.

Looking forward to seeing results. I remember going to the library to get back copies of Wireless World and photocopying the JLH article. It's easy nowadays...
 
Paul R said:
You can use half a standard bridge rectifier, to keep things simple. Get one that bolts to the case and makes construction simple.


The unloaded rails will be 1.4*22v, under load it gets complicated. The minimum voltage will be dependent on the size of the caps, the load and the maximum achieved, the max then on the ability of the transformer to charge the caps in the time the output voltage is higher than the current reservoir capacitor voltage.


Hardly dim questions, but perhaps don't over-think it. Practical hands on experience makes things much clearer. Especially if you can source an oscilloscope and watch the shape of the voltage on the reservoir caps. And the worst that can happen is a bit of smoke.

Way long ago I built a phono preamp based around a pair of Naim MC cards and a couple of discrete voltage regulators built on Veroboard from a design somebody posted here. I knocked up a PSU with two caps separated by a small resistor, single ended. I thought a Schottky rectifier would be cool, found one with two in a package, so exactly as we're discussing here, centre tapped transformer, twin rectifier, single ended PSU. Everything seemed to work, but the rectifier got really, really hot. This is obviously a low current application. Eventually I read the data sheet for the rectifier and figured out what was going on, although the detail has been archived, changed the rectifier for something I already had and the phono amp has lived happily ever after. All you really need to know is not to touch anything that can kill you, and cover over all the mains wiring. After that you will figure it out.

Looking forward to seeing results. I remember going to the library to get back copies of Wireless World and photocopying the JLH article. It's easy nowadays...
Click to expand...

Oh yes, been there seen it done it worn the T shirt etc:)
 
Well, the PCBs are here, they look great. Very nice quality etc. However no schematic, no parts list. The PCBs are partly labelled on one side but no indication of transistor type or pin out. Today's stupid question, I've never worked with a 2 sided PCB, do the components go on the labelled side?
 
stevec67 said:
Well, the PCBs are here, they look great. Very nice quality etc. However no schematic, no parts list. The PCBs are partly labelled on one side but no indication of transistor type or pin out. Today's stupid question, I've never worked with a 2 sided PCB, do the components go on the labelled side?
Click to expand...

Generally, yes. It should be pretty obvious in most cases...
 
As a general rule you only put components on one side and hopefully there is some form of silk screen ID telling you roughly what goes where.
You would normally put the components on the IDd side of the board.

You may have to find the appropriate version of the schematic that they have produced the board to.
 
Thanks gents, that makes sense. In the absence of any other info I would have put them on the side with the silk screen printing. I'm not used to seeing copper on both sides of the board, that's all.

I will most certainly need the appropriate schematic, otherwise I am lost. There are 3 sockets for TO3 cases, I think that the version I am using uses one as a voltage reg.
 
Point to which one you bought or post some pcb pics and we will probably be able to work out the schematic, or something close.
 
Thanks all, I'm back in circulation now and resuming this. I've had a busy few weeks, "stayed outta circulation till the dogs get tired" as someone once said.

I have a schematic from the vendor, it's a pdf, how do I post it here? Screenshot?

There is also a thread on DIY audio, I have posted on #4436 (!), it's been running for a long time and my input is minimal, but it's worth reading. There is a post at #4325 that shows the populated boards.

Link here: http://www.diyaudio.com/forums/solid-state/3075-jlh-10-watt-class-amplifier-442.html
 
Can anyone explain how what determines the value of quiescent current in the output devices?

I know you can adjust it by altering the values of R1+R2 in the original circuit, but I can't see the explicit calculation that shows the dependence, since there are no resistors in series with the emitters.
 
I don't know the calculation but in the schema I have adjusting the variable resistor KT2 will turn on the top power transistor that makes up the upper half of the voltage divider that drives the speaker. This is connected to the base and emitter of Q2 (the small NPN transistor that makes up the small end of the Darlington pair) and turn this on. This is balanced by adjusting KT1, the variable resistor that shifts the voltage on the base of the PNP signal transistor. As this turns on or off it will shift the voltage on the base of Q2, via voltage divider theory, thus shifting the degree to which Q3 (the lower power transistor and the big end of the Darlington pair) is turned on. I think so, anyway.
 
OK thanks, but I was after something explicit. Not sure the degree to which it depends on detailed transistor parameters. Otherwise how would you know the effect of differences due to tolerance of different transistors of same type, or of substituting different types. (My University professor would say you don't understand a circuit unless you can reasonably accurately predict quantitatively how it will behave - certainly for static values of voltage and current.)
 
If you want that degree of analysis then I suggest you have a read over at DIY audio, where there are any number of engineers who spend more time modelling circuits on software than they do building them. There are 444 pages so far on the JLH, discussing all manner of arcane details ranging from the effect of power supply via transistor case styles and heat sinking to the possibilty of using one steel and one plastic mounting screw for the 2N3055s. No, I am not making it up. It's here:
http://www.diyaudio.com/forums/solid-state/3075-jlh-10-watt-class-amplifier-444.html
In the unlikely event that it hasn't already been discussed someone will most definitely give you chapter and verse.
 
Thanks - I'll give it a try. If not already covered in the existing 444 pages, it may lead to another 300-odd ;-)
 
You must log in or register to reply here.


advertisement


Back
Top