The title says it all, I answered a request from a member for a preamp circuit, no commercial intent at all.
Question for Scott (I know I should not ask since you are already retired.
)TI has apparently put the LM329 on EOL.
AD has of course taken over LT's LM329, so I wonder if they also plan to end ?
Is there really no demand ?
If I need a small 7V 2-terminal reference, what would you suggest ?
Seems a string of 4 red LEDs is the only option left ?
TL431 is too noisy, or require many additional components.
Cheers,
Patrick
Question for Scott (I know I should not ask since you are already retired.
TI has apparently put the LM329 on EOL.
AD has of course taken over LT's LM329, so I wonder if they also plan to end ?
Is there really no demand ?
I don't know if the review of the product portfolios for overlap, etc. has even started. The demand for legacy parts is all over the place, I have no idea about this part (references) of the business.
> There are >30,000 LM329DZ available at US distributors, how many do you need?
Ask me again in 5 years, and see how many there are left.
I guess I would still be DIY'ing then.
The GLED is essentially an improved Vbe multiplier.
But it is only 2.5V, instead of 6.9V.
Also not sure about tempco, -4mV/°C ?
Just for discussion, how would you do a version for 6.9V ?
Is there a SMD version of the ZTX951 ?
Patrick
Ask me again in 5 years, and see how many there are left.
I guess I would still be DIY'ing then.
The GLED is essentially an improved Vbe multiplier.
But it is only 2.5V, instead of 6.9V.
Also not sure about tempco, -4mV/°C ?
Just for discussion, how would you do a version for 6.9V ?
Is there a SMD version of the ZTX951 ?
Patrick
I have made some informal measurements of the TK of available LEDs
for an article "The Silence Of The Lamps", eternally unfinished at the
current rate of progress. The results seem to vary all over the place.
Couldn't AD make a decent RF pnp transistor, maybe in sot-23 & sot-89?
Now with NXP in state of vivisection, there is nearly nothing left.
RIP, BFT92/93, BFG31, BFQ149... Same with HEMTs like ATF50189
now that Avago has been swallowed by Broadcom.
Sometimes we really need comp. pairs for pulse generator outputs etc.
Friends of mine are in panic. That would not need to be jelly bean parts.
We are loosing capabilities.
OMG. I have bought a small 40mm fan for my new Red Pitaya board.
G4020H05B-RSR. (Digi-Key) That little thingie makes a storm on my table,
I never had thought that was possible.
for an article "The Silence Of The Lamps", eternally unfinished at the
current rate of progress. The results seem to vary all over the place.
Couldn't AD make a decent RF pnp transistor, maybe in sot-23 & sot-89?
Now with NXP in state of vivisection, there is nearly nothing left.
RIP, BFT92/93, BFG31, BFQ149... Same with HEMTs like ATF50189
now that Avago has been swallowed by Broadcom.
Sometimes we really need comp. pairs for pulse generator outputs etc.
Friends of mine are in panic. That would not need to be jelly bean parts.
We are loosing capabilities.
OMG. I have bought a small 40mm fan for my new Red Pitaya board.
G4020H05B-RSR. (Digi-Key) That little thingie makes a storm on my table,
I never had thought that was possible.
Last edited:
> BFT92/93
Still enough stock at Mouser. Just buy lifetime stock.
> BFG31
Available still from some vendors in China, if you are desperate.
Possibly a couple of thousands left.
> BFQ149
This is the most difficult.
Still 30 pcs left at RS.
I can't find it anywhere else.
> ATF-50189
670 at DK, 570 at Mouser.
NXP closed the 4" fab in Nijmegen because they used up all their 4" wafer stock, as Bonsai mentioned at the BT thread, post #94944.
So even the beloved BF862 is now out of stock.
I think someone active in this forum just cleaned them all off.
Luckily we bought enough for our own use a few years back.
Patrick
Still enough stock at Mouser. Just buy lifetime stock.
> BFG31
Available still from some vendors in China, if you are desperate.
Possibly a couple of thousands left.
> BFQ149
This is the most difficult.
Still 30 pcs left at RS.
I can't find it anywhere else.
> ATF-50189
670 at DK, 570 at Mouser.
NXP closed the 4" fab in Nijmegen because they used up all their 4" wafer stock, as Bonsai mentioned at the BT thread, post #94944.
So even the beloved BF862 is now out of stock.
I think someone active in this forum just cleaned them all off.
Luckily we bought enough for our own use a few years back.
Patrick
Last edited:
Running a design service and no production, I can see that quite relaxed.
If anything, that creates customers for me.
I have a nearly full reel of BFG31 and got some ATF-50189 delivered just yesterday
but my idea to use them as uwave ring mixers is probably dead.
BF862 have been made in HK and mainland China, with those from HK
sporting a better 1/f corner as someone has written here.
I think the new On Semi FETs are a worthy replacement, with an even lower corner.
Nevertheless, I would welcome a MAT03 on steroids, maybe in this
thermally enhanced SO-8.
Then there are still these transistor arrays by Intersil on a SOI process.
They are quite nervous in this space proof flatpack.
If anything, that creates customers for me.
I have a nearly full reel of BFG31 and got some ATF-50189 delivered just yesterday
but my idea to use them as uwave ring mixers is probably dead.
BF862 have been made in HK and mainland China, with those from HK
sporting a better 1/f corner as someone has written here.
I think the new On Semi FETs are a worthy replacement, with an even lower corner.
Nevertheless, I would welcome a MAT03 on steroids, maybe in this
thermally enhanced SO-8.
Then there are still these transistor arrays by Intersil on a SOI process.
They are quite nervous in this space proof flatpack.
> BF862 have been made in HK and mainland China,
As far as I know only assembled in HK (back end).
NXP has no fab in HK.
https://www.nxp.com/wcm_documents/n...emmer-china-press-event-final-2-16Mar2012.pdf
> I think the new On Semi FETs are a worthy replacement, with an even lower corner.
The 2SK3557 (Sanyo, now OnSemi) is the closest to the BF862.
But I think Dimitri measured them and found them noisier than the BF862.
Measurements Rate SMT Low-Voltage n-JFETs Under Consistent Conditions | Electronic Design
Patrick
As far as I know only assembled in HK (back end).
NXP has no fab in HK.
https://www.nxp.com/wcm_documents/n...emmer-china-press-event-final-2-16Mar2012.pdf
> I think the new On Semi FETs are a worthy replacement, with an even lower corner.
The 2SK3557 (Sanyo, now OnSemi) is the closest to the BF862.
But I think Dimitri measured them and found them noisier than the BF862.
Measurements Rate SMT Low-Voltage n-JFETs Under Consistent Conditions | Electronic Design
Patrick
ft = 325 to 350 MHz. That is nothing against the 20 or 30 year old BFT93,
well-tempered and 5 GHz.
Remember, ft means zero gain left.
Then I'm better served by these:
< https://www.intersil.com/content/dam/Intersil/documents/hfa3/hfa3046-3096-3127-3128.pdf >
for fast current mirrors, active loads and level shifters.
@Euvl:
I think I did read that here somewhere on diya.
IIRC, the stamp codes were given, with 2AW for China, with the worse 1/f corner of 1 KHz.
All my BF862 up to now had 2AW, the KHz could be true.
Anyway, they are history. 2SK2394 has the same gm/Cin ratio. Or was it the 3910?
Soon I'll know for sure.
Quick copy from my notes:
Source = Pin1
BF862 20V 35mS min 45mS typ ciss=10pF Crss=1.9pF S/D interchangeable
2SK2394 15V 6 to 32 mA 38 mS ciss=10pF Crss=2.9pF CP package
NSVJ 2394SA3 15V 10 to 32mA 38 mS ciss=10pF Crss=2.9pF soll der neue rauscharme sein.
NSVJ 3557SA3 15V 10 to 32 mA 35mS ciss=10pF Crss=2.9pF
CPH 3910 25V 20 to 40 mA 40 mS CPH3-Gehäuse
NSVJ 3910SB3 25V 20 to 40 mA 40 mS ciss=6pF Crss=2.3pF
MCH 3914 15V 16 to 50mA 29 mS ciss=4.9pF Crss=1.4pF MCPH3-Gehäuse, klein
Source = Pin 2
MMBF4391 IDss=50-150mA GateSource CutOff= -4 .. -10
MMBF4392 IDss=25-75mA GateSource CutOff= -2 .. -5
MMBF4393 IDss=5-30mA GateSource CutOff= -0.5 .. -3
On MMBFJ309L S/D interchangeable ciss=5pF Crss=2.5pF 10nV@100Hz 12 to 30 mA
On MMBFJ310L S/D interchangeable ciss=5pF Crss=2.5pF 10nV@100Hz 24 to 60 mA
On SMMBFJ309L S/D interchangeable ciss=5pF Crss=2.5pF 10nV@100Hz 12 to 30 mA
On SMMBFJ310L S/D interchangeable ciss=5pF Crss=2.5pF 10nV@100Hz 24 to 60 mA
Ganz anders
CPH5902 JFET with built-in bipolar Cascode
IF3601 30 to ? mA 750 mS ciss=300 pF Crss=200 pF Vgs_off=-0.35 to -2V
Last edited:
I'm confused, MAT02/03 are very ordinary ft. I was asking how the THAT versions of the very low noise diff pairs fell short since the MAT devices are obsolete.
EDIT - I see, by on steroids you meant 5GHz ft.