I've run out of small signal diodes with >300V rating, and I need 6 of them now... But I have some small signal transistors such as FZT558 (pnp), FZT458 (npn), KSC3503 (npn), MJE340/350, and some others, rated at Vceo = 300V and above.
Can I use the collector-base diode? I know I can create a diode by connecting base to collector, but the reverse voltage it then handles only 5-10V in reverse. So what if I only use the C-B diode?
Thnx
(It is to protect grids on start-up where I have conditions where the grids get pulled positive before the cathodes get warm and they properly conduct. So I need to limit the positive voltage for just a moment when it is turned on).
Can I use the collector-base diode? I know I can create a diode by connecting base to collector, but the reverse voltage it then handles only 5-10V in reverse. So what if I only use the C-B diode?
Thnx
(It is to protect grids on start-up where I have conditions where the grids get pulled positive before the cathodes get warm and they properly conduct. So I need to limit the positive voltage for just a moment when it is turned on).
Connect B & E together and you will obtain the equivalent of a good epitaxial diode having a Vrrm equal to Vcb (larger than Vceo)
That should work. The series resistance may be a bit higher than for a normal diode, but otherwise I don't see any problem. In fact you could go up to the VCBO rating, which is always >= VCEO. Elvee's variant probably has less series resistance than the variant with open emitter, because reverse transistor action helps to conduct the current when the base-collector junction gets forward biased.
Last edited:
Your real problem is capacitance there, should check datasheets but very much doubt a BC junction has lower capacitance than a plain high voltage diode, say 1N4007
And if not, the kludge is not worth it.
And if not, the kludge is not worth it.
Last edited:
Thanks for answers.
Yes I need a little over 250V so I will use 300V devices. It is a DC coupled amp with 6AS7s and the voltage swing can go to almost -200V, and then some headroom of about 50V, referenced to cathode. On startup the grid voltage will go to positive 60-70V before the tubes start to conduct. The tubes will probably handle it, but why chance it. I should build in a b+ delay instead, but that requires a few more parts.
Regarding junction capacitance, thanks but I am aware. I have some rectifier diodes like BYV29 and SiC types, but the capacitance on those beasts!!!
I do look for low C when selecting.
Yes I need a little over 250V so I will use 300V devices. It is a DC coupled amp with 6AS7s and the voltage swing can go to almost -200V, and then some headroom of about 50V, referenced to cathode. On startup the grid voltage will go to positive 60-70V before the tubes start to conduct. The tubes will probably handle it, but why chance it. I should build in a b+ delay instead, but that requires a few more parts.
Regarding junction capacitance, thanks but I am aware. I have some rectifier diodes like BYV29 and SiC types, but the capacitance on those beasts!!!
I do look for low C when selecting.
Hi Semper, you can use the CB junction as a diode, but don't expect a great diode. To get the 300V ratings, the collector is lightly doped so the series R will be high. Not sure if it is spec though. I would do an IV curve trace before using it to see if it is adequate for your application.
I believe this is "highly doped". (But it is all relative.)
FZT558's Vce(sat) spec suggests just a few Ohms, "zero" for most tube circuits.
That cannot be, if it was highly doped, then you would have too many free carriers and the high electric field at 300V will lead to avalanche currents that would destroy the device. The physics don't add up.
I would curve trace it to make sure.
I would curve trace it to make sure.
How about 3 or 4 - 75V small signal guys in series; then high value (haven't worked it out, but maybe a meg or two) resistors paralleled with each, just to keep 'em honest on their sharing?
Bonus: Lower capacitance. Yay.
Bonus: Lower capacitance. Yay.
sandrohv is correct, it is lightly doped. With high doping you get a narrow depletion region and a low breakdown voltage, like in a Zener diode.
I think I have some 4148s so thats a good idea placing them in series. But there are 3 grids per ch and I am not sure I want to use up that many of them...
Regarding the Vf of a 300V CB junction. Is it any better with a 300V diode? Surely that too must be doped according to intended break down.
I will test on one tube and measure, as I can just not install all tubes.
Regarding the Vf of a 300V CB junction. Is it any better with a 300V diode? Surely that too must be doped according to intended break down.
I will test on one tube and measure, as I can just not install all tubes.
I have decided to do this more proper and insert a time delay on the drive ckt.
Thnx for replies. It is still interesting knowing the pros and cons of using a bjt as diode.
Thnx for replies. It is still interesting knowing the pros and cons of using a bjt as diode.
Exactly, a transdiode is much more perfect than a real one. The reverse beta might not be very large, but it suffices to improve matters considerably
In a BJT the emitter is extremely highly doped, the base moderately doped and the collector very lightly doped. This is essential to good specs, and the ratios are around 100 or so between the electrodes - vast differences.
This is often represented as N++ P N- (or for PNP P++ N P-) - the signs represent concentrations, confusingly!
In normal transistor operation the CB junction is reverse biased, but carriers from the emitter diffuse across the base into the depletion region and are immediately pulled across by the electric field, the collector acts as a resistor that depends on the density of injected carriers - more injected carriers, less resistance. Since the emitter has 10000 higher density of charge carriers than the collector, this means the collector resistance during normal operation can become very low when high currents flow, despite being high when quiescent.
As a BC-diode, you only have the base to inject carriers into the collector, so it won't be nearly as conductive as during transistor forward operation, since the carrier density is orders of magnitude lower. Also base-spreading resistance can be a factor too.
In transistor saturation the charge carriers only diffuse across the collector (no electric field to accelerate them), so the density of charge carriers in the collector can become even higher as its limited by recombination - no field to pull them clear - so the collector resistance can be much less again. As a diode this isn't going to happen.
In reality collectors have a highly doped layer where they contact the electrode metal, and a BC diode will perhaps behave a bit more like a PIN diode, since the lightly doped part of the collector is so much less doped than the base or the collector electrode contact.
So the device is actually N++ P N- N+.
The devil is always in the details...