[time-nuts] TSC5120A and 5125A heat issue

[Tom Knox]

Hi Bob;
A great product you designed. I wish Microchip would have continued evolution in that (standalone instrument) direction.
After many hours researching the heat and time-out issues on the 5125A and 5120A I found the solution was adding a larger 60mm fan and an airflow deflector near the front panel to create a circular airflow through-out the instrument. There are a number of additional details to this upgrade but It completely solved the issues and the final mod had no negative effect - perhaps actually improving performance by reducing measurement artifacts. To accomplish this, I tried over 10 fans, and in the final REV placed the fan outside the chassis using gaskets and neoprene washers to reduce electrical noise and isolate mechanical vibration. At the time I also added an external monitor output.
Another upgrade that helps protect the 5120/5A series from front end damage ( I believe is standard on many later 5120A/5125A analyzers ) are DC blocks on the inputs. I did this differently than Microchip by adding metrology grade stainless TNC to SMA front panel connectors and installed Mini -Circuits BLK-89-S+ DC Blocks directly to those adapters with SMA-m to SMA-m RA cables between the front panel DC blocks and P/N module inputs.
I also added additional RAM doubling the original (256MB?) to 512MB.
I attached a few photos.
All these minor improvements made an exceptional product better achieve its full potential.

Feel free to contact me for details.

Enjoy;
Tom Knox
SR Test and Measurement Engineer
Phoenix Research
4870 Meredith Way Apt 102
Boulder, Co 80303
Formerly of:
357 Fox Lane
Superior Co 80027
303-554-0307
actast at hotmail.com

https://www.thedenverchannel.com/news/marshall-fire/superior-man-moving-forward-after-losing-dream-research-lab-during-marshall-fire

"Peace is not the absence of violence, but the presence of Justice" Both MLK and Albert Einstein
________________________________
From: Bob Martin via time-nuts <time-nuts at lists.febo.com>
Sent: Saturday, August 5, 2023 10:19 AM
To: time-nuts at lists.febo.com <time-nuts at lists.febo.com>
Cc: Bob Martin <aphid1 at comcast.net>
Subject: [time-nuts] TSC5120A

Jim,

   I was involved in the design of the TC5120A and can confirm the
issue.  There is a single board computer (SBC) in the unit which
runs off a compact flash.  The flash can and does occasionally get
corrupted if the power ends during a write operation.

   Since support has gone away for the product, I suggest you remove
the cover on the unit, find the SBC, remove the compact flash and
make a copy of it as a backup.  If that works you will be covered in
case of failure.  Others in this list may be able to provide CF
images. I haven't tried this as I no longer have a TSC5120A and am
not sure if the CF is protected or not. I'd try and use the same
size and make CF when making an image.

  I'm retired now but could wander over to:
  Timing Solutions -> Symmetricon -> Microsemi -> Microchip and ask
the original software developer whether the CF can be copied if
it becomes an issue.

   Running the unit with the cover off is something else to try as
the unit runs "warm". The internal signal paths are all shielded
and if the results with cover on vs cover off are the same in your
working environment, I'd leave it off as heat is the enemy of long
life for electronics. The enclosure was designed originally for the
TSC5110 which drew less power than the TSC5120A.

Good luck and hopefully many years of good use,

Bob Martin
_______________________________________________
time-nuts mailing list -- time-nuts at lists.febo.com
To unsubscribe send an email to time-nuts-leave at lists.febo.com
-------------- next part --------------
A non-text attachment was scrubbed...
Name: knox-5120-42C7659B-6F13-46F7-9B20-696CF5FF5A21.jpg
Type: image/jpeg
Size: 227689 bytes
Desc: not available
URL: <http://febo.com/pipermail/time-nuts_lists.febo.com/attachments/20230806/8af1f442/attachment.jpg>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: knox-5120-D28821A4-1969-4C54-87DE-9D6316A2E94A.jpg
Type: image/jpeg
Size: 267437 bytes
Desc: not available
URL: <http://febo.com/pipermail/time-nuts_lists.febo.com/attachments/20230806/8af1f442/attachment-0001.jpg>