I am having trouble with a phono stage where the gain stages are Sziklai pairs. I have built these before without trouble but this time I am getting bursts of oscillation. There was no problem initially, but developed after running it in for about a week, which I find very confusing. The transistors in each pair are bc337 and bc327: perhaps these are unsuitable? I can find no consensus about how to address the problem; maybe a small cap to ground from Q2 base or an RC in the same place? There is maddeningly little information on Sziklai pairs in textbooks!
Complementary Darlingtons are quite common but have an extremely high Hfe.
Sight of your schematic, may shed some light on your issue.
Sight of your schematic, may shed some light on your issue.
Most effective method would be to insert resistor in emitter of first transistor in Sziklai pair. Start with 10Ω and go up in value till oscillation is gone, then increase value another step for the safety margin. Another way is to insert ferrite bead on emitter leg of first transistor in pair.
https://www.diyaudio.com/community/...compensate-for-stability.389509/#post-7102778
https://www.diyaudio.com/community/...compensate-for-stability.389509/#post-7102778
Oh and by the way, the reason for the differential pair is to allow for a likely change of tonearm wiring in the future. The current tonearm cable wiring is coaxial, unfortunately.
You don’t have to increase value of 62Ω resistors. New resistor from the emitter of first BJT to the collector of second BJT in pair should be added. That reduces only internal NFB and combined hFE of a CFP.
It appears to be an inverting phono stage with single ended input, which is a little unorthodox.
The base of the undriven pair of the input diff stage is drawn as not being bypassed to ground. Is the drawing correct?
All good fortune,
Chris
All good fortune,
Chris
You could take 1 Sziklai pair out in the first stage, return the preamp to non inverting and reduce noise.
Bonsai, The preamp is for my brother and the reason for the input diff pair is that he may replace his tonearm in the near future with one which has twisted pair and overall shield. The inexplicable thing is that everything worked perfectly when I was running the thing for a whole week, but when I brought it to my brother’s place there was immediate loud and continuous noise which is now pulsing in nature rather than continuous and the pulsing is exactly coordinated and identical on both channels. I can’t see anything that would change with time except for the electrolytic caps and cannot see why this change could cause a problem. I know that Sziklai pairs are notorious but I have used them frequently without this difficulty.
Two suggestions have been made and I will try them in order of difficulty when I return from work today.
Hey. Looking at schematic, i see inputs biasing resistors connected directly to power supply positive, without any filter, that would be first resson to oscillate or just amplify any power supply noise. I would recommend disconnect both resistors 330k from +48v, connect them thru 1k and 100uf filter. Also , at least on schematic, no supply bypass capacitors present, are they present on test board ?
Yes, you can see in my photo two RC filters per channel and one for the LED chain for the cascode current sinks.
Filters must be close to amplifier itself, otherwise they may have no effect, also possiblu make ground loops . If oscillation is present at output, try to block ac signal to gnd at variuos places with a capacitor, maybe 10uf , as your stages are dc coupled. Is power supply clean and no ripple ?