Hi Guys
"I find it it interesting that the design is more than one pair of output devices in an inverted CFP arrangement, such that the emitter resistors couple to the power rails, not the output rail. I haven't seen that in a medium power BJT amplifie..."
Bryston's 50W 2B was built that way since 1974, although with a single output pair of TO-3s originally, later TO-3Ps.
In a CFP, there are pros and cons to having the current-sense resistors be on the output device emitter or its collector. The latter places the R at the emitter of the driver and there is an obvious feedback link between the output current and the control voltage at the driver base, which keeps the output device under control. When there are paralleled outputs, one has to add distinct Re's to help current sharing. This also gives a convenient rail-referenced monitoring point for overload circuitry.
With R in the collector of the output BJT there will be a limit to how low z-out can be. This is the most popular reason for placing the R in this position (or is lethargy the most popular reason?). The raw collector output impedance of a CFP is quite a bit lower, which is what can lead to oscillation with certain loads. An output inductor will usually suffice in keeping z-out controlled.
The amps I've built using CFPs all have R at the emitter, on the rail. No issues so far and very simple protection circuitry.
Have fun
"I find it it interesting that the design is more than one pair of output devices in an inverted CFP arrangement, such that the emitter resistors couple to the power rails, not the output rail. I haven't seen that in a medium power BJT amplifie..."
Bryston's 50W 2B was built that way since 1974, although with a single output pair of TO-3s originally, later TO-3Ps.
In a CFP, there are pros and cons to having the current-sense resistors be on the output device emitter or its collector. The latter places the R at the emitter of the driver and there is an obvious feedback link between the output current and the control voltage at the driver base, which keeps the output device under control. When there are paralleled outputs, one has to add distinct Re's to help current sharing. This also gives a convenient rail-referenced monitoring point for overload circuitry.
With R in the collector of the output BJT there will be a limit to how low z-out can be. This is the most popular reason for placing the R in this position (or is lethargy the most popular reason?). The raw collector output impedance of a CFP is quite a bit lower, which is what can lead to oscillation with certain loads. An output inductor will usually suffice in keeping z-out controlled.
The amps I've built using CFPs all have R at the emitter, on the rail. No issues so far and very simple protection circuitry.
Have fun
Last edited:
"It just fried when I turned it on.
Magic smoke."
That made me laugh. Magic smoke - sort of really annoying but strangely interesting at the same time...
Magic smoke."
That made me laugh. Magic smoke - sort of really annoying but strangely interesting at the same time...
First electronics lesson we all learn - Magic Smoke is inside every component (especially active components) and is what makes them work.
When you connect +/- wrong way round or similar, you allow the Magic Smoke to escape and you cannot put it back therefore device no longer works.
When you connect +/- wrong way round or similar, you allow the Magic Smoke to escape and you cannot put it back therefore device no longer works.
Magic Smoke is inside every component (especially active components)
Yes, yes, and Magic Bang inside capacitors........
Yes, yes, and Magic Bang inside capacitors........
I appreciate the detailed comments, having built them with a mixed bag of results, yet some quite docile whilst sounding impressively good. 'always interested to hear other's wisdom on CFP - thanks indeed Kevin, for yours.
Amplifier arrived today.
Assembled it and it didn't work.
I found a piece of solder shorting 2 tracks.
I un-shorted it but it still didn't work.
I tested every component and a ZTX304 was blown.
I replaced it and the amp burst into life.
Its a bit sad it didn't work first time but the pcb doesn't have solder resist on it so I had to be extra careful not to get solder bridges.
Assembled it and it didn't work.
I found a piece of solder shorting 2 tracks.
I un-shorted it but it still didn't work.
I tested every component and a ZTX304 was blown.
I replaced it and the amp burst into life.
Its a bit sad it didn't work first time but the pcb doesn't have solder resist on it so I had to be extra careful not to get solder bridges.
An externally hosted image should be here but it was not working when we last tested it.
The amp seemed very sensitive to noise.
So I compared the amp with the schematic.
Seems the parts list is wrong in a couple of places.
2 off 22R resistors should be 2K2.
Also R22 has 2 different values. The second R22 should be R24.
One of the 2k2 resistors sets the current for the LTP so the current was totally wrong.
The other 2k2 sets the RF rejection on the input. Replacing that fixed the noise problem.
So I compared the amp with the schematic.
Seems the parts list is wrong in a couple of places.
2 off 22R resistors should be 2K2.
Also R22 has 2 different values. The second R22 should be R24.
One of the 2k2 resistors sets the current for the LTP so the current was totally wrong.
The other 2k2 sets the RF rejection on the input. Replacing that fixed the noise problem.
Last edited:
Interestingly the last capacitor I 'popped' from reverse bias was a chemicon and it emitted a decidedly controlled fine mist of tea coloured electrolyte.
I remember on my electronics course we had to build a circuit.
One chap put his electrolytic in the wrong way around.
He turned it on and the cap exploded hitting the ceiling.
On this 50 watt amp one of the electrolytics hasn't got the + marked on the pcb. I had to trace the tracks to work out which way around it went.
There have been a few mistakes in the parts list and on the pcb, seems a shame when the kit was meant for amateurs.
Consider most of those 'amateurs' back then didn't know one end of the soldering iron from the other. I reckon Maplin should have supplied the PCB blank without any screening, supply the schematic only (no component placement diagram) and a bag of parts. That way if people couldn't read a schematic, work out the actual PCB routing and solder properly, people had no business building the thing in the first place.
I have fond memories of the late 70s/ early 80s designs over here (EA/ETI and post 87 Silicon Chip) and still have stacks of PCBs I picked up cheaply. Projects from a bygone era, but fun nonetheless.
How does the little amp perform?
It sounds very good.
Seems a bit touchy towards noise.
Any clicks on the mains go straight through it to the speaker in pops.
On the circuit there is a capacitor between B+ and the VAS transistor base and this will pass any B+ noise straight into the VAS.
One of the resistors that was wrong was the input RF filter resistor, it was 22R on the parts list but 2K2 on the schematic.
Hi Nigel,
Pick up some liquid flux for electronic soldering. It will eliminate your problem with solder bridges unless you really work hard to create one. I always use the stuff. Lacquer thinners will clean up nicely afterwards. Use that with a toothbrush on the solder side, or a "Q-Tip", okay for either side. Some components react strongly and negatively to some cleaning agents or fluids in general.
-Chris
Pick up some liquid flux for electronic soldering. It will eliminate your problem with solder bridges unless you really work hard to create one. I always use the stuff. Lacquer thinners will clean up nicely afterwards. Use that with a toothbrush on the solder side, or a "Q-Tip", okay for either side. Some components react strongly and negatively to some cleaning agents or fluids in general.
-Chris
I use quite a big soldering tip so I must have accidentally bridged the two tracks.
As the PCB is about 36 years old it was a little touchy to solder and needed quite a bit of heat to get a good joint.
The area that got bridged is very well packed in with thin tracks.
I do have flux but never thought to try it.
As the PCB is about 36 years old it was a little touchy to solder and needed quite a bit of heat to get a good joint.
The area that got bridged is very well packed in with thin tracks.
I do have flux but never thought to try it.
Hi Nigel,
I also use a 3mm screwdriver tip (the big one) as well. The flux will allow the solder to wet on those old PCB tracks with ease. Rub the oxides off first, then use the flux. Life is so much easier with that stuff.
-Chris
I also use a 3mm screwdriver tip (the big one) as well. The flux will allow the solder to wet on those old PCB tracks with ease. Rub the oxides off first, then use the flux. Life is so much easier with that stuff.
-Chris
I have been a bit spoiled by pcb's I buy in from China.
They have solder resist and solder very easily.
I also design my own pcb's with larger gaps between components to reduce risk of solder bridges.
They have solder resist and solder very easily.
I also design my own pcb's with larger gaps between components to reduce risk of solder bridges.
My tip is more like 5mm.
I sometimes have quite large components so need the extra heat from a bigger tip.
Its just a cheapo Maplin soldering iron but it has been much more reliable than previous soldering stations I have bought.
I got a good deal on a metcal a few years ago. Best upgrade to my soldering ever. Couldn't justify list price for one tho!
This iron is an absolute deal. Works very well with adjustable temp, wide assortment of Hakko style tips.
Online Shop DIY Brand 60W Adjustable Electric Temperature Gun Welding Soldering Iron Solder Tool 5Tips 110V-220V 18.5cm Black And Blue|Aliexpress Mobile
Online Shop DIY Brand 60W Adjustable Electric Temperature Gun Welding Soldering Iron Solder Tool 5Tips 110V-220V 18.5cm Black And Blue|Aliexpress Mobile
Ouch !
I turned up the amp tonight and it sounded terrible.
No volts across the zener diode reference.
Q2 was faulty so I replaced it.
Still sounded terrible and Q5 wasn't working right, no constant current circuit volts.
So replaced that too and now it sounds good even with plenty of volume.
That solder bridge I had did more damage than I first realised.
Lesson learned, check visually very carefully before switching on for first time.
I turned up the amp tonight and it sounded terrible.
No volts across the zener diode reference.
Q2 was faulty so I replaced it.
Still sounded terrible and Q5 wasn't working right, no constant current circuit volts.
So replaced that too and now it sounds good even with plenty of volume.
That solder bridge I had did more damage than I first realised.
Lesson learned, check visually very carefully before switching on for first time.
- Status
- Not open for further replies.