
news
news archive
SEND NEWS!
articles
FORUMS!
links
contact
PC Database ([an error occurred while processing this directive] entries)
Folding Team
SETI@HOME Team
RC5 Team
Genome Team
OCAU chat
|
Radiator
Zen SCR325-2F - Page 3
21-Oct-01 - Review by Chainbolt
page 1 2
3
The Final Run
We first tested the Zen Radiator, then the PAL 6035, and finally
the MC462. After the first run we thought it a good idea to run
the test again with the Zen Radiator in order to verify our results.
This turned out to be a bad idea. The installation was again rather
painful. Again a lot of pressure and time was necessary until
we finally managed to connect the Zen Radiator with the socket.
After powering up, the OS did not boot. To make it short: when
mounting the unit on the socket, we had damaged the CPU. Chunks
from all 4 corners of the die were broken off.

click to enlarge - broken
beauty
We then had a closer look to the
Zen Radiator's housing and found that probably a combination of
several shortcomings in the design and the material of the holding
mechanism and the base plate caused the damage to the CPU:
- The holding arms are too short
in relation to the protruding nickel base plate: too much force
is necessary to bring them down so that it is possible to push
them over the socket extensions.
- The holding arms are made of
plastic, this material is too soft and too easily tiring/wearing
out. We doubt very much that these arms will keep their grip
like metal clips after repeated installations.
- The radiator itself is not firmly
fixed in the housing by screws. It is just held back by a nose
and therefore moving inside of the housing. This makes it difficult
to find the right position when mounting the radiator on the
CPU.
- A spring like mechanism in the
base of the housing (to improve the pressure of the nickel base
onto the CPU) is not centered. This means the nickel base plate
is not evenly pressed onto the CPU die (picture).
- The Zen Radiator's nickel base
plate is measuring only 30 x 30 mm. This is too small to utilize
the AMD plastic pads which are standing at 35 x 35 mm. That means
that when mounting the unit (which anyway has already too short
holding arms) the full pressure without any cushion effect through
the pads is extended on the die.

click to enlarge - bottom
spring not centered

click to enlarge - base
plate: uneven position

click to enlarge - CPU
Plate: too small for AMD TBirds
We then plugged in a spare Tbird
1.33 GHz in order to complete the second test run and found out
that the OS still would not boot. At that moment our frustration
was immensely growing. Another inspection of all components unveiled
that the motherboard itself had been damaged during the second
installation process. A deep scratch through several lines of
PCB printing obviously had ended the life of this Gigabyte 7DXR.
Again the question, why? And again we have to blame the Zen Radiator's
design and material: The end-part of the holding-arm with the
spared-out opening, which takes in the socket extension is too
thick. It has to be this way, because it is made of plastic and
not metal. If it were thinner, it would bend and could never hold
the massive housing. But unfortunately it's so thick, that it
almost touches the PCB and almost inevitably scratches the printed
lines on the PCB. That is what happened in our case and killed
the motherboard. Plastic is an improper material for a holding
mechanism of this kind. It wears out too easily, and it does not
keep the crisp grip which is necessary to hold the arms firmly
and safely in the socket extension.

click to enlarge

click to enlarge
We have never come across a more
flawed holding mechanism and base-plate design. Unfortunately
TS Heatronics's inexperience in the PC market shows here. Maybe
a shim would have prevented at least the damage to the CPU. But
we expect any cooling device to work without such a help, which
is not part of the AMD specifications for socket A. We later checked
with the vendor other Zen Radiators to make sure that we did not
receive accidentally a malfunctioning unit. But it turned out
that all units were 100% identical with the one we had used for
our test.
Conclusion
The Zen Cooler delivers excellent cooling performance, only slightly
behind the MC462 with the 4,600 rpm Sanyo Denki fan. It does however
not match the MC462 (and probably also the PAL 8045) with the
stronger 5,600 rpm Delta EHE fan. It is obvious that the manufacturer's
experience with radiator technology pays off. We also think that
this concept has a lot of potential. We are sure to see more cooling
devices based on this technology in the future. It is however
also obvious that the Zen Radiator's manufacturer does not have
experience in the production of CPU heatsinks and that this device
was not properly tested with socket A motherboards. The base-plate
and bottom design of the device does not suit AMD CPU's. Furthermore:
the holding arms are too short, the material is not sturdy enough
for such a mechanism, there is no support mechanism to push the
arms over the socket extensions, and the end part of the holding
arms with the spared out openings is so thick, that scratching
the PCB is very likely. Design and material are under no circumstances
acceptable. With or without a shim there is a high probability
to damage the CPU and the motherboard.
If we compare the Zen Radiator
with the current performance leaders MC462 and PAL 8045, we would
have to vote for the later ones. The Zen Radiator does not reach
their performance even with the weaker Sanyo Denki fan, in particular
for heavy duty overclocking the massive MC and PAL sinks seem
to be the better choice. The space requirements are basically
the same. The MC and the PAL have a far superior holding mechanism
utilizing the 4 PCB holes around the socket . We expect the Zen
Radiator outside of Japan to be more expensive than the MC and
the PAL. We see as only remaining advantage the substantially
lower weight of the Zen Radiator.
Zen Radiator Pros
:
- Good performance, only slightly
behind the MC462 with the 80 mm Sanyo Denki fan
- Innovative cooling concept (liquid
+ radiator) with great potential for further improvements
- Good airflow through due to
2 fans (blowing and sucking)
- Detailed, well written manual
(Japanese language only)
Zen Radiator Cons:
- The unit-to-socket A holding
mechanism is badly flawed by design and material
- The design of the base plate
does not match AMD Athlons and is likely to damage the CPU
- The 2 x 4,600 rpm fans are relatively
noisy due to turbulences in the unit
- Upgrade to regular size (and
stronger) 60 mm fans is not possible
- With roughly the same dimensions
as the MC462 and PAL 8045, the Zen will not fit on all motherboards
- Expensive: with a street price
of 10,500 Yen in Tokyo (including tax) = 85 USD/170 AUD more
expensive than a MC462.
- Manual only in Japanese language
at present
Heatronic should go back to the
drawing board and redesign the holding mechanism. The current
form is not acceptable. They have a potential winner in their
hands, but not in the existing form. Due to the very serious design
shortcomings, we recommend the Zen Radiator in its current form
only for the brave overclocker, who boldly wants to go where our
CPU and motherboard have already gone.
Have
you added your PC to the PC Database yet? Other Recent Reviews:
|