I don't know about other brands, but Philips/NXP BC550's can only reach such low noise figures when the source resistance is in the kohm range or above. That is because they have a nice high current gain (little base current shot noise), but not a very low base resistance (which sets a lower limit to the obtainable equivalent input noise voltage).
One way around this is to buy a lot of BC550's (they are cheap anyway) and connect them in parallel for AC signals. For example, you can give each its own decoupled emitter resistor to ensure an equal current division. If you select them for equal VBE maybe you can even connect them directly in parallel.
One way around this is to buy a lot of BC550's (they are cheap anyway) and connect them in parallel for AC signals. For example, you can give each its own decoupled emitter resistor to ensure an equal current division. If you select them for equal VBE maybe you can even connect them directly in parallel.
I was just checking what transistor arrays are available at Farnell, and the THAT 300 series is also interesting, especially the THAT 320 PNPs: 25 ohm base spreading resistance. As there are four matched devices in one THAT 320, you can easily connect them in parallel and obtain 6.25 ohm base spreading resistance.
Here are few you may want to check out.
SOT23 :MMBT5089,PMBT6429,2N4401/03,2N2222
I lifted the list below from who knows where...(don't remember now!) anyway, see for yourself if any of these can be suitable for your needs!
NPN PNP
TO-92 SOT-23 TO-92 SOT-23
2N3904 MMBT3904 2N3906 MMBT3906
2N4401 MMBT4401 2N4403 MMBT4403
BC337 BC817 BC327 BC807
2N5089 MMBT5089 2N5087 MMBT5087
BC547C BC847C BC557C BC857C
MPSA14 MMBTA14 MPSA64 MMBTA64
ZTX618 FMMT618 ZTX718 FMMT718
PN2369 MMBT2369 2N5771 MMBT5771
2N5550 MMBT5550 2N5401 MMBT5401
MPSA42 MMBTA42 MPSA92 MMBTA92
MPS5179 BFS17
BFT92 BFT93
TIP142 TIP147
SOT23 :MMBT5089,PMBT6429,2N4401/03,2N2222
I lifted the list below from who knows where...(don't remember now!) anyway, see for yourself if any of these can be suitable for your needs!
NPN PNP
TO-92 SOT-23 TO-92 SOT-23
2N3904 MMBT3904 2N3906 MMBT3906
2N4401 MMBT4401 2N4403 MMBT4403
BC337 BC817 BC327 BC807
2N5089 MMBT5089 2N5087 MMBT5087
BC547C BC847C BC557C BC857C
MPSA14 MMBTA14 MPSA64 MMBTA64
ZTX618 FMMT618 ZTX718 FMMT718
PN2369 MMBT2369 2N5771 MMBT5771
2N5550 MMBT5550 2N5401 MMBT5401
MPSA42 MMBTA42 MPSA92 MMBTA92
MPS5179 BFS17
BFT92 BFT93
TIP142 TIP147
Excess noise, watch out for reverse vbe
Even if you choose the ideal transistor, design your circuit so that you NEVER have emitter base breakdown, not even for an instant.
You can, and should, use a diode from emitter back to base to protect your base emitter junction. Install this in your circuit FIRST, then add your low-noise device, observing ESD procedures. You'll thank me later.
The physics is that in base emitter avalanche, hot electrons create defect sites in the base emitter junction, and these can lead to excess (1/f) noise or even popcorn noise
. The effect is fast, and it is cumulative. Excess noise can be a problem in BJTs as high as a 1 kHz, well into the audio region.
The cardinal rule is: never reverse bias emitter base junctions into avalanche (usually at 3 to 5 volts).
Cheers.
Even if you choose the ideal transistor, design your circuit so that you NEVER have emitter base breakdown, not even for an instant.
You can, and should, use a diode from emitter back to base to protect your base emitter junction. Install this in your circuit FIRST, then add your low-noise device, observing ESD procedures. You'll thank me later.
The physics is that in base emitter avalanche, hot electrons create defect sites in the base emitter junction, and these can lead to excess (1/f) noise or even popcorn noise
. The effect is fast, and it is cumulative. Excess noise can be a problem in BJTs as high as a 1 kHz, well into the audio region.The cardinal rule is: never reverse bias emitter base junctions into avalanche (usually at 3 to 5 volts).
Cheers.
MPSA18
Has worked well for me in a range of projects from single tran regenerative receivers to the N half of a CFP paired with 2SA1837 pnp driver.
http://www.redrok.com/NPN_MPSA18_45V_100mA_0.625W_Hfe400_TO-92.pdf
Good luck
Terry
Has worked well for me in a range of projects from single tran regenerative receivers to the N half of a CFP paired with 2SA1837 pnp driver.
http://www.redrok.com/NPN_MPSA18_45V_100mA_0.625W_Hfe400_TO-92.pdf
Good luck
Terry
I second this. Long ago it was common knowledge that bipolar RF transistors were easily damaged by BE breakdown, causing 1/f noise and gain reduction (low capacitance and small feature size makes the transistor die more vulnerable)Even if you choose the ideal transistor, design your circuit so that you NEVER have emitter base breakdown, not even for an instant.....
The physics is that in base emitter avalanche, hot electrons create defect sites in the base emitter junction, and these can lead to excess (1/f) noise or even popcorn noise
Cheers.
How many transistors do you need?
I have genuine (bought in ca. 1983 - 1985) ROHM 2SD786 transistors.
They have a Rbb of 4 ohm, 0.55nV at 10Hz/10mA and are the most low noise transistor I have found.
@RayCtech: Hi - & thanks for making this offer ... Sounds interesting so I'll just send you a PM.
@Marcel: I often use the 2SC2240 & 2SA970 because - to my ears - they have a - not perfect - but more harmonious balance to their way of playing than the other transistors I've tried. And ASFAIK they're still available e.g. from Toshiba. If you are interested a search on chip1stop shows that they still carry them:
????????? - ??????????????
@teleman: ... Sorry, not my area of expertise ... maybe somebody else can help?
Greetings to all ;-)
Jesper
... Sounds interesting so I'll just send you a PM.@Marcel: I often use the 2SC2240 & 2SA970 because - to my ears - they have a - not perfect - but more harmonious balance to their way of playing than the other transistors I've tried. And ASFAIK they're still available e.g. from Toshiba. If you are interested a search on chip1stop shows that they still carry them:
????????? - ??????????????
@teleman: ... Sorry, not my area of expertise ... maybe somebody else can help?
Greetings to all ;-)
Jesper
Every month there are fewer and fewer parts available in TO92 packages (or any through-hole packages). In the "Super TIS" thread at http://www.diyaudio.com/forums/solid-state/209658-has-anyone-seen-front-end-before.html , it was recommended to use the KSC3503/KSA1381 pair. These are actually video driver transistors in TO126 and the noise performance isn't specified, but they have plenty of voltage capability, decent bandwidth, and low capacitance.
The KSC2690 and KSA1220 are similar devices.
The KSA916/KSC2316 are in "stretched" TO92 packages for higher power dissipation. They are lower bandwidth and higher capacitance than the TO126 devices I mentioned above.
Dale
I had never even heard of them until recently, when I was trying to repair a Philips FR675 tuner/amplifier. There they are used in the third stage of the power amplifier, dissipating some 400 mW to 500 mW while there maximum rating is 300 mW at 25 degrees C. Anyway, it is good to know that there are still some really low noise types available.
The KSC3503/KSA1381 pair can be recommended for low noise duty as they have a rbb of around 20 ohms and fairly high hfe. Thus and for their low capacitance they can be used as small signal devices. The KSC2690 and KSA1220 are nowhere similar and have much larger capacitances and their use is as drivers for output stages. The KSA916/KSC2316 are also not recommended for small signal duty.
First question to be asked is what the source impedance is going to be. This has a direct inpact as to what transistor is better suited for the application.
The best pair I have used are the ROHM 2SD786/2SB737 - 4 ohm and 2 ohm rbb and 0.55nV/10Hz noise. It is at 10Hz and not 100Hz or 1kHz.... The 1/F noise are remarkable low in these devices.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.