BJT for plate CCS load
Tom Schlangen
as it seems suitable BJTs for plate CCS load usage are
hard to find at the parts shop at the corner here
in Germany. Biggest problem is the needed high maximum
emitter-collector voltage of, say, 500 volts (at say, a
dozen mA).
So far I have used PNP type MJE350 in several circuits,
but this only is rated up to 300 Vce max.
Any hints on truly obtanium types that will do okay at
500 volts?
Tnx2u,
Tom
--
Ian Iveson
> ...So far I have used PNP type MJE350 in several circuits,
> but this only is rated up to 300 Vce max.
>
> Any hints on truly obtanium types that will do okay at
> 500 volts?
Looking at http://rswww.com, I see some PNP up to 800V but they are
probably too big and clumsy for your kind of current.
Some small NPN get up to 400V but higher voltage are again designed
for higher power than you need. see eg 2SC3150
Why choose a BJT? Many would use a MOSFET, such as the STD1HNC60 I
have used for a mu stage.
I know almost nothing about transistors, as perhaps you can tell.
cheers, Ian
Patrick Turner
Tom Schlangen wrote:
I have used many MJE350/340 for CCS in circuits where the B+ was 300v,
and never blown one.
I suspect at 3 mA, they are very quiet, and never have I noticed noise
increases with a bjt
CCS.
The high hfe of the bjt and their high Gm results in a CCS with a very
high finite value,
so any noise emanating from the CCS is attenuated by the low anode
resistance of the triode,
since an effective resistance divider exists.
In cathode CCS circuits, they last even though B+ maybe 480V.
But in normal triode gain circuits, such as at the input V1 of a power
amp,
they allow a lower B+, because usually only a few vrms of output is
needed,
and those few volts are a lot more linear than if you had a resistor to
bring DC to the anode, since the distortion of a triode rapidly
increases with decreasing
resistance values as the loading.
The only significant RL on the V1 of an amp with a CCS DC supply is the
following
cap coupled biasing resistor, and if that is 220k, then its a far better
load than
say 47k *and* the 220k in parallel.
The CCS allows a higher than normal plate current, and a lower Ea, and
you still get
splendid linearity.
I have tried a few mosfets in regs and CCS, and found they were all too
easily zapped.
Patrick Turner.
Tom Schlangen
> Why choose a BJT? Many would use a MOSFET, such as the
> STD1HNC60 I have used for a mu stage.
> I know almost nothing about transistors, as perhaps you can tell.
Well, no hard facts from me either (I surely know less about transistors
than you do) but the BJTs (mostly MJE... series) I used so far just worked
fine in such an app, be it as (PNP) CC plate load or (NPN) CCS for
a tubed LTP or diff amp.
I am just too stupid, I am afraid, to set up, say, an IRF8xx TO220
type or something for this task (CC plate load) correctly ("biasing"
considering depletion mode, gate protection) and so on.
If someone would be so kind to give a working example using such
a beast as CC plate load, I probably could modify the circuit myself
as needed.
Tom
--
Tim Williams
news:jmid82-...@unichf002.frueh.de...
> Biggest problem is the needed high maximum
> emitter-collector voltage of, say, 500 volts (at say, a
> dozen mA).
Rip apart a few computer or other switching supplies.
Otherwise, they are cheap and plentiful at say, Mouser and Digi-Key. I
ordered 10 x MJE18008 (800V 8A) from Digi-key, for $25. That switches
6.4kW!
Tim
--
"I've got more trophies than Wayne Gretsky and the Pope combined!"
- Homer Simpson
Website @ http://webpages.charter.net/dawill/tmoranwms
Tom Schlangen
> But in normal triode gain circuits, such as at the input V1
> of a power amp they allow a lower B+, because usually only
> a few vrms of output is needed,
Yes, sure. Problem is - unless one uses active voltage regulation
instead of the "usual" RC-chain B+ decoupling, the full PSU B+
might be present at the device until tubes start conducting and
voltages stabilize.
A RC-chain-fed V1 tube might (and will) take this if things are
not grossly out of order, but I am not sure a solid state device
would be so forgiving ...
> I have tried a few mosfets in regs and CCS, and found they were
> all too easily zapped.
That is what I fear, too. BJTs seem to be a bit more forgiving.
Not as much as tubes, but probably more than FETs, I suppose.
--
Yves
"Tom Schlangen" <t.sch...@ndh.net> a écrit dans le message de news:
ncee82-...@unichf002.frueh.de...
> Hi Patrick,
>
> > But in normal triode gain circuits, such as at the input V1
> > of a power amp they allow a lower B+, because usually only
> > a few vrms of output is needed,
>
> Yes, sure. Problem is - unless one uses active voltage regulation
> instead of the "usual" RC-chain B+ decoupling, the full PSU B+
> might be present at the device until tubes start conducting and
> voltages stabilize.
No, nothing accross the CCS until the tube starts to conduct !
Uh, except perhaps some link cap to load ?
A TIP50 sustains 500V VCE as long as the base is not left open and the risk
for second breakdown is null at such lo current.
Anyway, if you have some volts to vaste, you may use an higher value for the
base reference voltage thus reducing the voltage accross the BJT. The emiter
resistor must be adjusted according to !
CCS as anode load provides so good psrr that regulation is not a must.
But ... use a VDR ( or any kind of shunt regulation) to void B+ rises to a
too hi value before being loaded ;)
Yves.
Patrick Turner
Tom Schlangen wrote:
> Hi Patrick,
>
> > But in normal triode gain circuits, such as at the input V1
> > of a power amp they allow a lower B+, because usually only
> > a few vrms of output is needed,
>
> Yes, sure. Problem is - unless one uses active voltage regulation
> instead of the "usual" RC-chain B+ decoupling, the full PSU B+
> might be present at the device until tubes start conducting and
> voltages stabilize.
I try to have a string of 4 x 75v zeners to regulate the B+
of V1 rail, since it also offers LF stability.
>
>
> A RC-chain-fed V1 tube might (and will) take this if things are
> not grossly out of order, but I am not sure a solid state device
> would be so forgiving ...
When you use a pnp device, and the turn on voltage goes
way above the rating, all 3 electrodes, emitter, base, and collector all
are at the high B+
until the triode starts conducting. Then the only way the bjt could be
destroyed is if
its collector is grounded while the emitter & base is at say 450v.
The triode comes on slowly, and pulls the collector voltage down to
where it wants it, say
+150v, and even if the B+ stays at 450v, the V rating isn't exceeded.
Patrick Turner.
Jeff Goldsmith
comes up very high voltage at no load. Zero failures in twenty years.
Jeff Goldsmith
Ian Iveson
> When you use a pnp device, and the turn on voltage goes
> way above the rating, all 3 electrodes, emitter, base, and
> collector all
> are at the high B+
> until the triode starts conducting.
Not quite true, Patrick. The valve anode will initially be at 0V. If
500V is supplied instantly to the other end of the current source,
then it has 500V across it.
The anode capacitance must be charged. The current source may charge
this at its designed current, which won't take very long, so you
will get an initial short pulse of voltage across the source. Or the
source may not be fast enough and you will get a current pulse
instead.
Then there is the capacitance into the following stage which is
probably fed from the anode. That will take much longer to charge at
a much lower current, being a larger capacitance and a much higher
resistance.
Also stray inductance, etc, that will tend to resist current pulses.
Also the capacitance of the source device electrodes, and the rest
of the source circuit.
>> That is what I fear, too. BJTs seem to be a bit more forgiving.
>> Not as much as tubes, but probably more than FETs, I suppose.
>>
Perhaps the mosfets you smoke are more vulnerable to all these
start-up transients?
Plenty mosfet mu-stages seem to work. It is commonly argued that
they make better current sources than BJT (similar to pentode/triode
difference?). Problem however with electrode capacitance.
cheers, Ian
bill ramsay
<IanIves...@blueyonder.co.uk> wrote:
>"Patrick Turner" <in...@turneraudio.com.au> wrote
>
>> When you use a pnp device, and the turn on voltage goes
>> way above the rating, all 3 electrodes, emitter, base, and
>> collector all
>> are at the high B+
>> until the triode starts conducting.
>
>Not quite true, Patrick. The valve anode will initially be at 0V. If
>500V is supplied instantly to the other end of the current source,
>then it has 500V across it.
>
Hate to disagree with you Ian, but if the valve is not conducting it
is not conducting, and therefore the anode will float up to the HT
level as no current is flowing through the load resistor/ccs.
Ian Iveson
> Hate to disagree with you Ian,
You're not, as far as I can see, Bill.
> but if the valve is not conducting it
> is not conducting, and therefore the anode will float up to the
> HT
> level as no current is flowing through the load resistor/ccs.
The anode indeed floats up, we all agree on that. I am saying that
it will take time to do so, and it will start from 0V, so the
instantaneous initial voltage across the source will be the full HT.
The time taken to float up depends on the time constants in the
circuit attached to the anode.
Just simulating a simple driver and typical output stage, with a
perfect CCS set at 10mA between a perfect 500V HT supply and the
driver's anode. Coupling is via a 150nF cap into 100k grid leak.
The anode voltage takes 1.5ms to reach 100V, 5ms to get to 200V,
10ms to 300V, and nearly 20ms to get to 85% of its final value of
380 volts.
So the CCS drops 500V to begin with, falling to 400V after 1.5ms,
300V after 5ms, etc. The valve uses a cathode resistor for bias, so
the valve is hard on to begin with (but with a high resistance until
the anode voltage rises).
A simulated piece of wood instead of the valve takes pretty much the
same time. Most of the delay is due to the coupling caps to the next
stage it seems.
But the real picture would be much more complicated due to reactance
in the CCS circuit, etc, etc.
cheers, Ian
in message news:78vbr0l9nkvqtqj18...@4ax.com...
robert casey
> Gentlemen,
>
> as it seems suitable BJTs for plate CCS load usage are
> hard to find at the parts shop at the corner here
> in Germany. Biggest problem is the needed high maximum
> emitter-collector voltage of, say, 500 volts (at say, a
> dozen mA).
>
Transistors intended for horizontal output duty in
TVs and computer monitors can handle up in the thousands
of volts. Unless you are designing for production
quantities, try salvaging one out of a discarded
monitor. Or visit repair shops that will sell small
quantities to people.
Fred Gilham
Toshiba makes a bunch of high voltage high power transistors.
Here's a link:
http://www.semicon.toshiba.co.jp/ eng/prd/pdf_presen/tr_hv_e.pdf
There's also another link whose author is messing around with high
voltage switching transistors that gives some part numbers and specs:
http://www.hills2.u-net.com/electron/ignit.htm
--
Fred Gilham gil...@csl.sri.com
In 1981, an OSHA epidemiological study of 2,500 workers, half of whom
had been exposed to PCBs for over 17 years, found that the number of
deaths from cancer was 10 percent lower than what would be expected
for a group with the same profile in the general population.
Yves
"robert casey" <wa2...@ix.netcom.com> a écrit dans le message de news:
_Wqtd.9840$714....@newsread2.news.atl.earthlink.net...
Be aware they usually exhibit very lo beta, down to 5 !
Patrick Turner
Ian Iveson wrote:
> "Patrick Turner" <in...@turneraudio.com.au> wrote
>
> > When you use a pnp device, and the turn on voltage goes
> > way above the rating, all 3 electrodes, emitter, base, and
> > collector all
> > are at the high B+
> > until the triode starts conducting.
>
> Not quite true, Patrick. The valve anode will initially be at 0V. If
> 500V is supplied instantly to the other end of the current source,
> then it has 500V across it.
But no current flows, and the anode voltage whizzes up to whatever
voltage the
CCS collector wants it to be.
>
>
> The anode capacitance must be charged.
Its only a few pF.
> The current source may charge
> this at its designed current, which won't take very long, so you
> will get an initial short pulse of voltage across the source. Or the
> source may not be fast enough and you will get a current pulse
> instead.
It depends on how the CCS is set up.
I have an R and C between the base and B+, then an R from base to 0V,
so the CCS B+ to base voltage takes some time to establish, the anode
capacitance is low,
no problems so far....
>
>
> Then there is the capacitance into the following stage which is
> probably fed from the anode. That will take much longer to charge at
> a much lower current, being a larger capacitance and a much higher
> resistance.
Still only pF quantities....
>
>
> Also stray inductance, etc, that will tend to resist current pulses.
>
> Also the capacitance of the source device electrodes, and the rest
> of the source circuit.
>
> >> That is what I fear, too. BJTs seem to be a bit more forgiving.
> >> Not as much as tubes, but probably more than FETs, I suppose.
> >>
>
> Perhaps the mosfets you smoke are more vulnerable to all these
> start-up transients?
I'm not the one to smoke mosfets, because I don't use then for CCS
when there is a cheaper easier locally available bjt.
>
>
> Plenty mosfet mu-stages seem to work. It is commonly argued that
> they make better current sources than BJT (similar to pentode/triode
> difference?). Problem however with electrode capacitance.
I am not sure.
Patrick Turner.
>
>
> cheers, Ian