| |
G4EBT > TECH 19.09.03 17:38l 154 Lines 7443 Bytes #999 (0) @ WW
BID : B01546G4EBT
Read: DB0FHN GUEST OE7FMI
Subj: Re: Power supply fan mod?
Path: DB0FHN<DB0RGB<DB0MRW<OK0PKL<OK0PPL<DB0RES<ON0AR<GB7FCR
Sent: 030919/1438Z @:GB7FCR.#16.GBR.EU #:15561 [Blackpool] FBB-7.03a $:B01546G4
From: G4EBT@GB7FCR.#16.GBR.EU
To : TECH@WW
Here you go then, something to do with amateur radio from the pen of 'EBT:
The topic of the popular twin meter PSUs sold under the Manson EP925 and
other brand names (Palstar?) has arisen on packet from time to time -
recently in connection with the fan. There must be thousands of these
PSUs throughout the world that have given, and continue to give, giving
trouble free service.
From personal experience having repaired several, I would alert users to
certain design shortcomings which will be apparent on taking the lid off.
They're clearly built down to a price but represent good value for money,
so long as too much isn't expected of them.
The most important shortcoming is the lack of overvoltage protection,
which isn't uncommon in "budget" PSU's. This means that if a
collector/emitter short arise in one or more of the power transistors, the
full rectified DC output of the mains transformer (>25V) would be plonked
on any equipment connected to the PSU. The adverse effects of this would
depend on the ability of that equipment, (multiband transceiver or
whatever), to withstand that sort of overvoltage.
An overvoltage crowbar circuit could have been incorporated into the PSU
with just a few components, but for some reason the makers left that
feature off, yet included a fan control circuit using 13 components,
including one IC, two transistors, two diodes, and a thermostat.
I'd say that his fan control circuitry is needless sophistication, and
the wisest thing to do is to disconnect the wires from the thermostat and
insulate them, which will leave the fan on 100% of the time.
If you take the lid off you'll note the following:
The fan:
The fan doesn't shift much air - the term "forced air cooling" being
overly optimistic. The flow of air needs to pass over the heat sink onto
which five 2N3055 power transistors and the bridge rectifier (rectum
frier?) is mounted.
However, the airflow doesn't do that. The mains transformer blocks the
way, so the air just blows at the transformer and out through the vent
holes in the lid. Any air that passes over the heat sink does so more by
luck than design.
The heatsink:
In the ones I've repaired the heat sinks have been dull silver alloy.
Elementary physics ( black body/non-black body radiation etc), dictates
that the ideal radiator is smooth, black, and shiny. In the EP925's
I repaired, the heatsinks had the opposite of these qualities.
If heatsinks aren't fan-cooled, (and that must include the EP925 as
the fan does very little), ideally, they should be large, black, and
mounted externally and upright to enable convection currents to rise
past the heat sink causing a flow of cooler air to pass over it.
The heatsinking on the EP925 is internal, horizontal and compact. It
consists of two stacked above one another, carrying the five 2N3055's
and the bridge rectifier between them. Thus, the upper heat sink shields
the lower one and little heat can be dissipated from the upperside of the
lower heatsink, and the underside of the upper one.
The fan switch is a bimetalic device mounted on the heatsink. When the
fan is off, the switch is closed circuit, when the heatsink temp rises
to approx 70 degrees C, the switch goes open circuit and the fan should
cut in. When the temp drops to 40 degrees, the fan should cut out.
In the ones I repaired I simulated a 10A load using two car headlight
bulbs. The fan came on within ten minutes and never went off again,
so with a load of more than say 5A the fan will stay on continuously.
How well the PSU copes depends on how heavily loaded it is. I suppose
the average users will need 13.8V, and maybe 5 amps intermittent. It's
claimed that the PSU will supply 25A "continuous", at 15 V.
A glance at the small size of the transformer will show that this rating
is optimistic, but it depends on what's meant as "continuous". 13.8V at
25A = 345W. I doubt that many - if any, amateurs are going to use one of
these PSU's for that sort of duty.
It would be interesting to see what would happen if, say, six 55W car
headlight bulbs were paralleled to the output terminals to simulate a 330W
(24A) "continuous" load. Looking at the small smoothing cap, which doesn't
look as though it could cope with the ripple current that such a demand
would induce, I can guess what would happen, so best not to try
that experiment!
The other thing about "continuous", is that by the nature of amateur
radio, we spend at least as much time on receive as on transmit, so the
PSU is lightly loaded for much of the time, even on lengthy "overs". It
might be an oxymoron, but I suppose that amateur use in these
circumstances can best be called "continuously intermittent".
All of the EP925 PSU's I've repaired had the same fault. For some reason
all five 2N3055 power transistors had gone open circuit, which was lucky
in a way, as had just one gone short circuit, the consequences for any rig
attached to the PSU could have been dire. (25V + across input to the rig).
Why all five should have failed is a mystery, as they are generously
rated. None of the PSU's were heavily loaded, just powering 10 watt
transceivers, albeit one had been in 24 hour use for a BBS for some years.
As I recall, the bridge rectifier wasn't very highly rated either. In
a bridge, there are always two diodes in series, and it's ordinarily
recommended that the bridge should be rated at 3 x the RMS output of
the transformer. So, if the offload output of the transformer is, say,
30 Volts, that's roughly 45 Volts RMS x 3 = 135 Volts. Thus, a rating
of 150V for am margin of safety would be wise, which admittedly seems
high for a PSU which will mostly be used at 13.8V.
It must be said that these PSU's are neat, compact and attractive, with
their two meters on the front, and low current as well as high current
output terminals. As with most things in life, you get what you pay for,
and for anyone wanting no more than say 5 - 7 amps at 13.8V I'm sure
they're up to the job, despite the design shortcomings referred to above.
They do at least have shortcircuit protection, to protect the PSU from
damage if the output terminals are shorted out.
Due to the lack of overvoltage protection I wouldn't personally connect
one to an expensive piece of equipment. Having said that, a little
outboard overvoltage protection crowbar circuit could be added, and would
give peace of mind.
The TIP31C (TR7) mounted on the little PCB next to the smoothing cap is in
the fan circuitry; the TIP31C mounted on a little PCB at the front of the
heatsink is TR6, and drives the five 2N3055's. If PSU fails, and there's
voltage at the output of the transformer, and if the 2N3055's are OK,
suspect the TIP31C. If the 2N3055's fail, it's a labour of love to replace
them. (About three hours work).
If to do any work on the PSU you have to remove the PCB, look at ZD1 (1W
15V zener diode), and if the PCB looks scorched, I'd suggest you replace
it with a 5W one.
I hope these wordy notes are of interest, if only to keep as a reference.
If anyone has a duff EP925 and need any advice on repairs, I'll be happy
to help as best I'm able.
73 - David, G4EBT @ GB7FCR
Eddystone User Group Member
G-QRP Club Member No: 1339
QTH: Cottingham, East Yorkshire.
david@crofters89.freeserve.co.uk
Message timed: 15:36 on 2003-Sep-19
Message sent using WinPack-Telnet V6.70
(Registered).
Read previous mail | Read next mail
| |
