HP 53131A Hard Power Switch Modification

Following on from my previous post in relation to the HP 53131A Frequency Counter Teardown I have devised a reversible modification that replaces its soft power switch (where the switch mode PSU and the fan are running continuously when the unit is plugged into the wall outlet) to a hard power switch which properly turns the unit off from the front panel. The modification is simple to do, uses inexpensive parts and is completely invisible, it looks and feels like factory behaviour. It is will worth applying if you have one of these for bench use so you can leave it plugged into the wall without having the fan running and switch mode noise being generated when the unit is not in use. It’s hard to understand why HP did not include a hard power switch in the original design, you will see from my modification there is plenty of room, and it seems such an obvious thing to do. Anyway, I have created a video showing the details of the modification, if you have one of these frequency counters I hope you find the modification interesting and/or useful.

UPDATE: Having re-read the schematic again it appears that the unswitched +/-12v from the switch mode PSU are provided on the OCXO option connector J9. This would explain the design approach, the PSU will run the xtal oven even when switched off so when you are ready to use it, its instantly available. Thats fair enough except that many of these counters exist that do not have an OCXO option, and even with one fitted, I personally would be happy to know I have to switch it on in advance of using it.

NOTE: Please forgive the under-exposure on the first part of the video and the audio noise in a few places through it, I have a new camera setup and am getting used to it. The audio noise I believe comes from the phantom power not being liked by the wireless mic (I hope its that at least). The under-exposiure is just my misinterpretation of the zebra exposure indicator, you will see I get it right (or at least much better) in the second half of the video.

See you next time, and thanks for watching….

This content is published under the Attribution-Noncommercial-Share Alike 3.0 Unported license.


  1. Gerry, that’s a neat mod. My (older) HP5316B also has an “always on” linear power supply, I guess to keep the optional Crystal Oven nice and ready. Also, there is no fan, so when plugged it it’s “just” wasting juice, if not actually anoying me. Previously warm-up time from true power-off was maybe 15 minutes, but now I use an external Rubidium reference, which locks withing 2-3 minutes so counter warm-up is not a problem. I tend to keep the counter on a power strip which I shut off when I leave my workbench, but i might take a look inside to see if a hard power switch is an option.

    Somebody pickier than me might sugest heatshrink over the IEC socket and switch solder lugs… (Speaking as someone who’d picked up a home-made amp that was switched off, but still plugged in, when I was about 15 years old. I learned a lot about mains AC and the coduction properties of human skin, that day… )

    1. Hi Laurence, I did put heat shrink over the switch lugs, its clear so did not stand out on video, you will see it if you look hard enough. I did not put it on the IEC connector because it was not done originally – I was going to say “good enough for HP, good enough for me” but I guess the whole point of the mod would suggest otherwise 🙂 I should have used heat shrink on the IEC connector – its a fair cop…. Gerry

  2. Thanks for your video!

    BTW, the fan is 40x40x20mm, 12V, if anyone intends to swap out for a silent type like I do.

  3. Gerry,
    I commend you for the very nice as well as creative work! Thought it useful to suggest why this modification might not be best for everyone though. I hope you don’t mind. Some years ago I discussed this topic with an Agilent engineer and I’ll pass along how it was explained. .

    Many modern and higher-end Agilent/Keysight and other test equipment works this way – where what might have been an on/off switch is now a “standby” switch. There are several reasons for it. The most important being that a stable Calibration of these units cannot be done and the calibration cannot be fully relied upon until they are in a thermally stable condition which cannot happen until the unit has been powered up for a number of hours. Some of the calibration manuals spec as much as 24 hours. You’ll find the accuracy and stability specifications for this equipment ( including the 53131a) are spec’ at being powered up for some number of hours before this spec can be met. I believe the spec is 24 hours for the standard time base and less for the ovenized timebases. The Agilent manual also has a not that this time specification is inclusive of “standby” time. So, in order to hit the guaranteed specifications without having to wait many hours for the units to stabilize the 53131a would need to be powered up literally all the time. Related to this the 53131a uses an attractive VFD type display which look good but have a limited life span growing dimmer and dimmer as the age eventually becoming so dim you can’t read them. The “standby” switch configuration allows the appropriate circuitry to remain powered on in order to hit the specifications but shuts down the display to extend its useful life. If the standby switch is converted to an on/off switch the display will be on continuously even when it is powered up just to reach thermal stability and not in actual use. This will significantly shorten its useful life span. The last reason for not turning the unit on and off frequently is this is known to reduce the life of various components due to voltage variances and current inrush during power on initialization. How much is anyone’s guess however I’m sure it’s possible to find studies on this subject with an internet search.

    So anyway, the real question is how important it is to get as accurate and repeatable measurements as the unit is capable of without having to wait hours and hours for it to fully stabilize or is it more important to be able to individually control power to each piece of gear because each unit is infrequently used. On my equipment I have all the equipment that has a standby switch rather than and on/off switch connected to a power strip and I simply leave them in standby mode when not in use but I want them to be fully stable for immediate use However, when I know I won’t be using them for a few days I turn it all off all the power at the power strip switch. That way the displays aren’t burning out when I’m just keeping them warmed up for maximum accuracy and stability but can shut them down completely when not needed.

    I hope you find this info useful.


    1. Hi Ron,
      Thanks for the detailed explanation, yes I had concluded that pretty much myself too but the modification was more around using such equipment in a home lab. Having noisy fans running 24×7 is a minor annoyance and for my purposes the modification made sense, I think I pretty much said as much at some point in my waffling :). As far as stability is concerned you are absolutely right of course, that is a problem. However, firstly, the only critical component in a frequency counter is the reference clock. Now even if you leave the unit on for 24 hours or more, the stock on-board oscillator is terrible, its hard to imagine how one could make it any worse from a stability point of view, so getting it temperature stable might get you closer to calibration but its still unstable and horrible, if you only need the precision and accuracy of an AM radio then its ok, for everything else its hopeless. For me I have two options, first you will see that another project I blogged about was to make a clone of the ultra high stability option for these counters using an OCXO, these come up to working temp really within 30 minutes and for most practical cases thats more than good enough, and with that you get stability down to a couple of PPB. I also built another project which is a Rubidium 10Mhz reference clock and a distribution amp, and that comes up to spec from cold in about 1 minute, not as nice as the OCXO for short term stability but very very precise long-term.

      All in all, for my own purposes having the hard power switch proved to be the right choice. Thanks again for your comments and feedback.


  4. Gerry,
    I tend to agree with you with regards to your hobbyist/ home environment comment and ultimate accuracy not usually the top priority. However, I hope you didn’t miss the info on the limited life span of VFD displays though as I may have needed to make it more clear that extending the VFD life by powering it down while the unit iis being brought to operating temp is also a primary reason why Agilent designed the switch as a “Standby” switch.

    Some of my test equipment has a switch on the back that controls the standby state of the circuitry that needs thermal stability. That’s ideal and and nice convenience.

    Anyhow, thanks for all the interesting videos and articles and your friendly and kind responses!

    Best regards,

  5. Gerry,

    I know this is an old thread but I finally decided to replace the old, noisy fan on my 53132A. The original fan is a Fonsan (Delta) DFB0412M and new identical replacements are available for a few dollars. FYI the fan leads are soldered into the PS board as the lead spacing is not standard pin header. The new fan is much, much quieter.


Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.