[Elecraft] Linear PS Recommendations

Sat Jul 23 13:57:14 EDT 2016

 > If Alan is reading this, I'm wondering if I missed where that
 > 'strategically placed diode' is located?

One way to do it is just to put a pair of back-to-back diodes in 
parallel with the inputs.  That protects against over-voltage of either 
polarity.

One thing I always like to do when constructing a homebrew project is to 
paste a copy of the schematic inside the top cover.  That way I can 
never lose the schematic.  :=)

Here's an image of such a schematic of a 723-based, 10A, 7-15V bench 
supply that I built many years ago:

http://n1al.net/music/N1AL_10A.gif

It is very conservatively rated at 10A.  I probably could have designed 
it for a comfortable 15A, but the markings on the meter I wanted to use 
for an ammeter wouldn't easily accommodate that.  For a similar reason, 
I put a 5V zener in series with the voltmeter so it would read 5-15.

One issue with the 723 is that it needs a higher-voltage supply than the 
pass transistors.  If you make the unregulated input voltage high enough 
to avoid dropout, the pass transistors are dissipating more power than 
they need to.  So in this design I used a separate supply for the 723. 
One issue with doing that is if the transistor supply fails (it is 
separately fused) the 723 tries to supply all the output current.  In 
this case the 723 supply is derived from 65V (since that was the 
available winding on the transformer I used) with a dropping resistor 
and shunt zener.  The dropping resistor limits the current to a safe value.

This supply has fold-back current limiting.  The 22k resistor from pin 2 
causes the current limit to drop from 14A or so at 13.8V output to about 
11A at 7V output and 7A with a short circuit.  That's a good idea for a 
variable-voltage supply since pass transistor dissipation increases at 
lower output voltages.  But even for a fixed-voltage supply it's nice 
that the supply recovers automatically from a shorted output.  If you 
use a crowbar-type current limit, you have to cycle power to get it to 
turn on again.

Alan N1AL

On 07/22/2016 11:46 PM, Richard Fjeld wrote:
> Hello David, (and Alan), and thanks to John for the link showing the
> added parts.
>
> David, your PS sounds very similar to the one I built, also in the
> 70's.  Mine has four 3055's on very large heat sinks that are overkill,
> but I didn't know what to expect.  It was at a time when we were going
> from high voltage PS for tubes to Low voltage PS for solid state.  I had
> not seen a low voltage supply to power a 100 watt rig yet.  So I went
> heavy duty.  My diodes for the full wave bridge I made were stud mounted
> on 1/4 inch aluminum pieces for heat sinking, and mounted on wood to
> insulate them from each other. There were no full wave bridges in a
> package yet.
>
> I'm curious how your supply trips out, and if you used a relay?   It
> seems to me that I did that, and when the 723 clamps down, the relay
> drops to open the primary.  I have not had the time to review the
> schematic diagram, but as I recall, if a pass transistor shorts out, the
> crowbar may try to shunt the unregulated voltage, but the situation is
> not going to be good.  That is why I like the device ad5x is showing on
> his website.  It will protect the radio from over-voltage by blowing a
> fuse in the DC line.
>
> If Alan is reading this, I'm wondering if I missed where that
> 'strategically placed diode' is located?
>
> Dick, n0ce
>
>
> On 7/22/2016 10:16 AM, David Anderson via Elecraft wrote:
>> I have a couple of homebuilt DC regulated supplies that I built a
>> great many years ago, the first is a simple 3A 13.5V one that was the
>> very first supply I ever built and it used a 723 with a single 2N3055
>> series pass transistor. I remember I bought a lovely oil filled mains
>> transformer for it, but accidentally wired up the 120V windings in
>> parallel instead of series for our 240V mains and watched it start to
>> bulge when I switched it on. (after I put a bigger fuse in it).  I
>> ended up using another transformer and the power supply is still in
>> daily use, built in the early 70's.
>>
>> I later built a 25 Amp version with 4 3055s and a 723 with overvoltage
>> trip, and also short circuit protection. It can be shorted out and
>> with barely a spark it trips out and has to be reset before it comes
>> on again. No fuses to blow and replace, no destroyed series pass
>> transistors. I built it in a chassis that was lying around and 35
>> years later it is still in daily use and still not got a proper cover
>> made for the chassis.  I checked the output the other day on my scope
>> to see if there was any hum or ripple and on full load I had 10mV p-p
>> of noise. Regulation still excellent.
>>
>> Nothing much wrong with the humble 723.
>>
>> I have seen some truly awful "commercial" supplies with no short
>> circuit protection, that blow the series pass transistors like fuses,
>> or worse make them into short circuits that apply the full unregulated
>> supply on to the load.
>>
>> 73 from David GM4JJJ
>>
>>> On 22 Jul 2016, at 15:55, Alan Bloom<n1al at sonic.net>  wrote:
>>>
>>> The 723 regulator has some known reliability issues, but with proper
>>> design they can be mitigated.  In particular the differential voltage
>>> on the error amplifier inputs is only rated for 5v.  If one input is
>>> connected to the 7.15V reference, then if the power supply output is
>>> shorted the voltage rating is exceeded.  The solution is a
>>> strategically-placed diode.
>>>
>>> Alan N1AL
>>>
>>>
>>
> ______________________________________________________________
> Elecraft mailing list
> Home: http://mailman.qth.net/mailman/listinfo/elecraft
> Help: http://mailman.qth.net/mmfaq.htm
> Post: mailto:Elecraft at mailman.qth.net
>
> This list hosted by: http://www.qsl.net
> Please help support this email list: http://www.qsl.net/donate.html
> Message delivered to n1al at sonic.net