advertisement


John Linsley Hood Class A or similar

It's a current-feedback set-up, and the top output device works more like an active, steered current source. So the setup controlling the output current are the loops involving the input and 'phase-splitter' transistors, which control what that 'upper' output device is doing.

It's not actually a push-pull output stage with quiescent current controlled by the voltage between the bases of the output pair; hence - no emitter resistors required for stability etc.
 
Having another look - the quiescent current is set by the choice of R1+R2. So assuming the voltage Ve is set up correctly to be ~ V/2 (where V is the supply voltage), the current through these resistors is roughly 0.5 x V/(R1+R2), which is much larger than Vbe/2.2k (the current through the emitter resistor of Tr3). So most of the current through R1+R2 is split equally between the bases of the output transistors (assuming they are matched in hfe). So looks to me as if the quiescent current is ~ hfe x this current, i.e, ~ 0.25 x hfe x V/(R1+R2).

Hence for given values of R1 & R2 the quiescent current depends directly on the hfe of each of the output devices.

Not a big deal, but happy now I've worked it out. Also means this current likely to vary with temperature, and between different instances of the same circuit. And something to bear in mind if changing transistor type. (Maybe already covered elsewhere.)
 
Yes, and therefore there are views you'll find about optimum split of hfe for the transistors you might have available for positions Tr3 and Tr4 :)
 
I've started on this project, going well so far, I had the resistors in the parts box, nice 1/8W metal film jobs. I am assembling a BOM to take to Maplin next week, one cap value escapes me. It's the input coupling cap. The output calls for 220uF electro, is the same recommended at input?
 
I think the output needs to go up by an order of magnitude. 0.5u should do for the input.

(Think of making a high pass filter with the capacitor and, at the input, the input impedance of the amp and at the output the impedance of the load. For the same low frequency cut off you need a bit capacitor with a low impedance and vice versa. The capacitor has an impedance at frequency 'f' of 1/(2*Pi*f*C) I think that means that at 20Hz a 1000u cap is about 8 Ohms equivalent.)
 
Thanks Paul,
I will use 0.5uF on the input. I might even have a polyester cap in that size, I'll have a look in the parts bin. Never much liked electros. Thanks also for the maths re impedance vs frequency, I'll have a think about that. I'll go with the original 220uF for now. I'm not sure of what impedance I'm looking for, other than input impedance is normally desired to be high and o/p low. I'm a bit away from having something to play with just yet, the trannies are proving to be awkward (the small signal jobs, not the 2N3055s, they are currency) and if I can't get an equivalent in Maplin or Farnell I may end up rummaging through the bits box for something with the right hFE and hoping for the best. Examination of the PCB has made it clear how I make the TO3 packages work though, there is a pickup on each bolt, I'll use a star washer for good measure.
 
Found a pack of 470nF polyester 63V jobs for the inputs, the original calls for an electro, as Paul says 0.5uF is the best size. Isn't film reckoned to be an upgrade on electro in the signal path?

I also need to find a name for the connector pieces, they are like a blue chocolate block connector used to connect the input and power supply leads. Any clues?
 
Right, more bits are in and it's going together. One problem I am having is with dry joints, or more specifically the solder not wetting out on the board. It's as if it's oily, but I have degreased it with meths. My soldering technique is OK (it certainly used to be back in the day) but I am finding the solder is going onto the lead but isn't wetting onto the board. It sits in a round bead on the lead instead. The solder is old stock Altai or similar, so it's good stuff. Am I missing something WRT board cleaning?
 
Got a spare old PCB to test the solder with? Otherwise I’d try emery cloth on one of the new board pads and then see if it’ll take solder...seems a bit strange though.

Richard
 
Just had a play with an old PCB and bit of scrap vero, no problems. Solder flowing like water. I'll try giving the new PCB a further clean tomorrow and see if I can tin the solder points then.

Edit - used an acetone based cleaner and given it a bit of a scrub, it's better but still not as easy as an old bit of Vero. I'll get it done, but I can do without any dry joints that aren't due to me. Of course the ideal would have been to solvent clean it before starting, but I didn't think of that.
 
Last edited:
stevec67 said:
Just had a play with an old PCB and bit of scrap vero, no problems. Solder flowing like water. I'll try giving the new PCB a further clean tomorrow and see if I can tin the solder points then.

Edit - used an acetone based cleaner and given it a bit of a scrub, it's better but still not as easy as an old bit of Vero. I'll get it done, but I can do without any dry joints that aren't due to me. Of course the ideal would have been to solvent clean it before starting, but I didn't think of that.
Click to expand...

Are you using lead free solder?.
 
No, proper stuff that's been in the electrical toolbox since before the RHS regs were written. It flows well on an equally old bit of vero.
 
Hooray! As you suggested paul a bit of a rub has sorted it, I was being too delicate. 600 grit glasspaper has the solder flowing nicely.

A bit more thinking - I am measuring up for the heat sink. All OK, but what is the protocol for insulating the board from the bracketry and heat sink? I have bought some mica insulators, got the wrong ones, TO3P not TO3, so that's an error. i will buy some TO3 kits. The thing has 3 x 2N3055 per channel, 2 setting up the voltage divider for the output, one on rthe PSU capacitor multiplier. I am attaching it to a 3mm ally angle attached to a heatsink. Am I right in thinking that I need (as seen from above) the trannies sitting on a mica pad, with thermal grease, then the ally bracket, then another mica pad, then the PCB (which has copper on the top surface so needs insulating with the second mica pad) then the thing held together with M4/M3 nuts and bolts, a star washer where the nut meets the PCB to pick up the collector, and sleeves on the 2 legs and 2 bolts to keep the electrical connectors away from the bracket? Is this best practice?
 
A transistor question, the 1969 design I have specifies 2 off D667 transistors per channel. One is in the capacitor multiplier as the upper half of the darlington pair, the other is the upper transistor in the darlington pair formed with the lower of the 2 power transistors that make up the voltage divider. I only bought 2, didn't realise that this is per channel. I measured the gain on each, they are 180. The spec says hFE min 60 max 320 so hardly a precision item. I have 2 x NPN transistors in the bits box, gain measured at 180-200, no ID code visible. They are in a metal can with tag by emitter, like the good old BC107/109 were. Can I substitute these for D667?
 
No. Either get more 2SD667 (there is a 2S in front of all Japanese transistor numbers but it's not written on the device) or use BD139-16.
 
Thanks Jez, that's clear then. I don't hink Farnell offer D667 but BD139-16 is a possibility. I need to get some TO3 mounting kits anyway so a trip to Farnell is probable.
 
stevec67 said:
Thanks Jez, that's clear then. I don't hink Farnell offer D667 but BD139-16 is a possibility. I need to get some TO3 mounting kits anyway so a trip to Farnell is probable.
Click to expand...

Well you won't find it by searching for D667 as it's 2SD667...

Watch for pin out differences. Both will be either E-C-B or B-C-E so it's not too bad as one way round will be correct without needing to bend and insulate any of the leads.
 
The orig
stevec67 said:
A transistor question, the 1969 design I have specifies 2 off D667 transistors per channel. One is in the capacitor multiplier as the upper half of the darlington pair, the other is the upper transistor in the darlington pair formed with the lower of the 2 power transistors that make up the voltage divider. I only bought 2, didn't realise that this is per channel. I measured the gain on each, they are 180. The spec says hFE min 60 max 320 so hardly a precision item. I have 2 x NPN transistors in the bits box, gain measured at 180-200, no ID code visible. They are in a metal can with tag by emitter, like the good old BC107/109 were. Can I substitute these for D667?
Click to expand...


The original circuit used a 2N697, a TO5 packaged device, if you can't get one, a TO126 or TO220 100 volt 1 amp upwards npn device would do.
 
Arkless Electronics said:
Well you won't find it by searching for D667 as it's 2SD667...

Watch for pin out differences. Both will be either E-C-B or B-C-E so it's not too bad as one way round will be correct without needing to bend and insulate any of the leads.
Click to expand...
Yes, it's the 2SD667 that I'm looking for, no dice at Farnell. They do have the alternative though. Good news re the pinouts, thanks.
Edit - thanks Russell. I may have a couple of 2N697 in the kit box. Certainly got some trannies in that case style, I was idly testing them for functionality the other day.
 
You must log in or register to reply here.


advertisement


Back
Top