Re: TIP50 for HT regulator
- From: Patrick Turner <info@xxxxxxxxxxxxxxxxxx>
- Date: Fri, 25 Jul 2008 12:19:22 GMT
flipper wrote:
On Thu, 24 Jul 2008 13:27:45 GMT, Patrick Turner
<info@xxxxxxxxxxxxxxxxxx> wrote:
Ian Thompson-Bell wrote:
flipper wrote:
On Wed, 23 Jul 2008 10:00:28 -0500, John Byrns <byrnsj@xxxxxxxxxxxxx>
wrote:
In article <4prc84lsgbk0dvoorirkcnrk888gemis2v@xxxxxxx>,
flipper <flipper@xxxxxxxx> wrote:
On Tue, 22 Jul 2008 23:34:17 +0100, Ian Thompson-BellI'm looking for a capacitance multiplier design to replace the inductors
<ruffrecords@xxxxxxxxxxx> wrote:
At a local ham fest I picked up 100 of 150V zeners so I thought I wouldWatch out for SOA and turn on surge into downstream caps. They're good
build myself a stabilised 300V HT supply - not big current, just enough
for my mic pre projects. So, I have been looking at high voltage
transistors and came across the TIP50 which are rated 400/500V. I picked
this oas it the only one whose number I vaguely recognise.
for about 45mA DC at 300V.
Has anyone used these for HT supplies or what else do you use?STP2NK60Z is my sort of 'all purpose' N-MOSFET. 600V with built in
gate zener protection and relatively inexpensive. A few pennies less
than a TIP50 but more current and no base drive.
For power supplies I tend to use them as cap multipliers (note that a
cap multiplier is not a reservoir) and screen regulators. I've also
used them as source (cathode) followers when short a triode.
in the power supply for a Single Ended amplifier, are you willing to
share your capacitance multiplier design?
Sure. I used it on the headphone amp breadboard and, as fortune has
it, it's posted on the website.
http://flipperhome.dyndns.org/headphone%20amp.htm
Two ways to do the resistor divider. One is from input V to ground but
with large voltages you can end up with difficult values so I run it
from input to output.
If you combine it with a regulator, using zeners in the grounded leg of
the pot divider, you avoid difficult resistor values and get a fixed
output voltage too.
You want enough drop to cover MOSFET gate threshold, input ripple, and
supply droop. If I remember correctly that one has about 10V across
it.
Regards,
John Byrns
Cheers
Ian
Hi all,
Instead of 6CY7 I would have thought a pair of 6SN7, 6CG7, 12BH7, 12AU7
would be a lot "stronger" than 6CY7, which is Ra = 52k, µ = 68,
so rather like 6SL7.
That's the vertical oscillator section but the 6CY7 is a dissimilar
double triode. The second section is for vertical deflection service
and is a 5.5 Watt, 920 ohm Rp triode.
Sort of a 'baby 6EM7'.
I stand corrected. Busy day yesterday.
The "capacitance multiplier" is fine, but usually the R divider
is between the input Vdc and 0V.
And I made that note.
Or as JB says, have a zener diode string, or a voltage reg tube to
fix the gate voltage. BUT, zeners are noisy, so you should have a
series R from top of the zeners to the gate, say 10k, then 10uF to 0V to
filter the noise at the gate.
And then you need a seroes R of say 100 ohms between C7 reservoir cap
33uF and the mosfet drain.
I didn't want a regulator, it's a cap multiplier.
The capacitance mulitplier is ok in a class A amp.
The B+ won't vary if the output resistance at very low F is high.
If you short the source output to 0V, the mosfet is turned on fully and
tries to discharge
C7 too fast, so POOF goes it, and you cannot get the smoke back in.
Good point. It was a breadboard, though, so I didn't bother. In an
actual I either fuse or series R.
Hmm, SS devices have a nasty habit of blowing to protect the fuse.
So you also need another series R from the gate filter to the gate, say
1k, and a transorb
or string of diodes or zener to pull the gate voltage down with the
source voltage without relying
on the little internal zener inside the mosfet.
The internal zener is sufficient and is why it's there. With a
'normal' MOSFET there isn't one.
It's also why I have a series 10K on the gate.
As long as you are happy....
If the input C37 is shorted to 0V, current tries to flow backwards to
the short and
the mosfet is also allergic to this, so a 3A diode needs to be placed to
allow the flow,
but not through the mosfet.
That MOSFET has a 1.4A reverse recovery diode built in.
Fair enough.
That's pretty common and is, I think, a natural artifact of the
structure.
After blowing a few fragile mosfets, I learnt all about protection
measures everywhere.
I know what you mean but the built in protection is one of the reasons
why this particular MOSFET is one of my favorites.
I still like to add external protection around all SS regulators,
and output current limiting, which means a sensing R between the source
output
and the load, or cap to where the load is connected.
Excessive Iout produces a voltage across this sensing R and a string of
series diodes
from gate to load shunt the gate voltage into the output.
This can be set up to give very definate overload protection
against excessive Iout in adition to a series R between reservoir cap
and drain.
One should be able to short the output or input before the series drain
R and
the SS device should not be perturbed.
In my regulated temporary bench supply for up to +600Vdc output,
I blew up a few BU208 and finally the penny dropped about SS regulators;
they were not to be fully trusted.
In that supply I paralleled a pair of 6080 and used a 6BX6 for the gain
element.
No more fused SS parts. I have a floating 12V supply and relay to
turn off the supply if the output current goes high; there is an SCR
which latches
on after an excessive Iout incident. That's never missed a beat for
10 years now.
A regulator has all the same problems as a capacitance multiplier,
but no attempt is made to regulate the Vdc, so it floats up and down
with mains variation, and fortunately, all these transients
are so well filtered the artifacts are never audible in the output.
Its a worry in a phono stage with huge LF gain though.
And where you have high global NFB, one must watch out for increased
LF gain and a tendency to instability if there is LF rail movements.
In my preamps I just use lots of CRCRC and that is usually enough
unless there is an MC amp stage involved.
One good thing about active regulation or cap multpliers
is that you can escape the effects of LC resonances at below
10Hz due to an LC filter being undamped well enough by the load.
The R that should load the LC is usually 1.41 x ZL or ZC at the Fo.
Since the load ( Vdc / Idc ) is often way above the required R
which might work out to be say 200 ohms, then one should place a 200 ohm
resistance
in series with the choke. Voila, the PS resonance dissappears.
What ever you do though, you should test the amp at 1/2 max Vo
with a load and without a load, and plot the F response between 1Hz and
50Hz.
And take a look at the rail signals.
Just my 3.1428 cents worth.
Patrick Turner.
.
Patrick Turner.
- References:
- TIP50 for HT regulator
- From: Ian Thompson-Bell
- Re: TIP50 for HT regulator
- From: John Byrns
- Re: TIP50 for HT regulator
- From: Ian Thompson-Bell
- Re: TIP50 for HT regulator
- From: Patrick Turner
- TIP50 for HT regulator
- Prev by Date: Re: TIP50 for HT regulator
- Next by Date: Re: Cruelty to a 12AU7 ?
- Previous by thread: Re: TIP50 for HT regulator
- Next by thread: Re: TIP50 for HT regulator
- Index(es):
Relevant Pages
|
