@jean-paul
But that's very exciting, I assumed for almost 40 years that the BC5xx form declared as low-noise is really the low-noise form /device. With the ZTX versions it is clear. Now you've given me a brain teaser. Ice_max 1A vs. 0.2A is one of the most striking fixed parameters. What are the main differences between the two PSPICE models?
thx
But that's very exciting, I assumed for almost 40 years that the BC5xx form declared as low-noise is really the low-noise form /device. With the ZTX versions it is clear. Now you've given me a brain teaser. Ice_max 1A vs. 0.2A is one of the most striking fixed parameters. What are the main differences between the two PSPICE models?
thx
My guess is that the base track resistors of both types differ significantly and that therefore, depending on the complex source impedance, sometimes the BC3x7 is the more advantageous choice.
Since “3/4 + 3/4” must be driven with low impedance, the input (in our case) will see approximately a 600Ohm source. In the original MF-A1 it is 3k3 ..! Thanks for the tip, CLG is less than 19dB (very strong negative feedback!), nothing should go wrong with the thermal noise.
Since “3/4 + 3/4” must be driven with low impedance, the input (in our case) will see approximately a 600Ohm source. In the original MF-A1 it is 3k3 ..! Thanks for the tip, CLG is less than 19dB (very strong negative feedback!), nothing should go wrong with the thermal noise.
Yes it are assumptions based on the datasheet but reality really is different. Or was different as it was measured long ago and AFAIK only those by Philips as these were good parts. Funny thing is that I still have the BC327/BC337 from that lot.
No idea about PSPICE. Never simulated anything and never will.
BC550C is worst as expected. Too bad BC560C was not tested. Still the preferred combo of many clever designers and this for decades. Some even proudly stating they used BC5xx "low noise" transistors 🙂
Audio is mainly consisting of assumptions backed up with nice graphs.
Audio is mainly consisting of assumptions backed up with nice graphs.
All the components are now in place, I just need free time.
But the drought is taking its toll on us; since mid-February, you'd think we were already in summer.
There is no rain!
But the drought is taking its toll on us; since mid-February, you'd think we were already in summer.
There is no rain!
Hi again!
At last the BC327/BC337 arrived...
I didn't have the 4u7 input cap... how high can I go for Ri to use, say 1uF? Is 33K or 47K acceptable?
Today's status:
R5 and C4 are part of the overall frequency compensation, the OLG is extremely high due to the “utopian THD”, we have to be careful here - but R5 also determines our input resistance with less than or equal to 5k6Ohm. C1 should not be less than 2µ2F, if the driving preamplifier also has a coupling capacitor of this size, we would have a resulting lower cut-off frequency of just under 30Hz.
Stay with the (suggested) values for the time being, 4µ7 MKT (or try a 22µF/100V electrolytic capacitor).
Stay with the (suggested) values for the time being, 4µ7 MKT (or try a 22µF/100V electrolytic capacitor).
For BJT input, 10K is the max I would go.
With 10K Ohm, a 10uF coupling cap would do the job.
BC327 has about 3dB lower noise assuming the source output Z is 50R, with a 600R output Z the difference is slim,
also the 327 has higher OLG distorsion at equal gain and output amplitude, so all in all it would be better only as
MC preamp FI.
Last time I tried BC547 (the cousin of BC550) in the input stage with +-40v supply. It was a failure. They start to leak before the specified 45V.
Not sure about BC327/BC337. I only use them in low voltage environment such as the current mirror for the long tail pair.
Hi, I take you mean recently bought BC327/337?
I thought I made it clear that is was about old Philips BC327/337 which were then significantly better than the most used "low noise" specified types by Philips. Despite that everyone used the "low noise" types just because they were called "low noise". I totally don't do anything with such anymore for decades but it is parked in memory, that is all. What changed in the following decades with BCxxx transistor production... I did not keep track 🙂 Then we used BC547 for switching and BC327/337 for audio as they indeed performed better.
Like never using serrated washers directly on plastic power transistors as in picture 129. Once taught right with explanation why it simply stays in memory.
Last edited:
@jean-paul
That s the result of the simulator with Phillips s Spice models for those devices, i didnt check with other manufacturers
models since the Phllips are the one which have the longer list of characterised parameters and as such are the most usefull.
Also i ll point that the BC327 is actually a plastic packaged version of the BC140/160.
On a side note the only devices that seems comparable are the 2SC1775/2SA872 from Hitachi, they have both ultra low noise
and a gain comparable, and often better, than the BC550C.
That s the result of the simulator with Phillips s Spice models for those devices, i didnt check with other manufacturers
models since the Phllips are the one which have the longer list of characterised parameters and as such are the most usefull.
Also i ll point that the BC327 is actually a plastic packaged version of the BC140/160.
On a side note the only devices that seems comparable are the 2SC1775/2SA872 from Hitachi, they have both ultra low noise
and a gain comparable, and often better, than the BC550C.
Thank you for good comments!
1 - Yes, recently bought from Farnell. My supply of BC550/BC560, BC546/BC556 are also mostly new ones. I have no equipment to measure noise etc, so I just go with the author's advice! (You better be right, hbtaudio!)
2 - Caught red-handed! I will change those. They were the only ones I could reach without raising from the chair...
🙂 morten
OK, we never simulated anything nor did we use software. We just measured Philips transistors in batches in a test setup.
Last edited: