Replacement Power Switch for an HP Agilent 3101-2124 DPDT Push-Button Power Switch

I hadn't used one of my HP 3335A generators for a while, and when I turned it on a few weeks ago, the power-switch wouldn't stay latched ON.

In other words, it didn't act like a push-button power switch that toggles between OFF and ON states when pressed.  Instead, it acted like a momentary-contact push-button, and the generator would only remain ON if I kept my finger pressing on the button.  Not good!

I thought I'd try to pull off the key cap and squirt some DeOxit into the switch (if that were possible) to hopefully lubricate it.  Unfortunately, I accidentally pulled the switch apart while doing this, and it would not go back together again.  Below is a photo showing its parts (less the internal spring that I  lost).

Here's the switch shown on the HP 3335A's Keyboard Schematic Diagram:

This switch's HP part number is 3101-2124.  I searched the internet for an identical replacement but could not find one.  However, I did find a different HP / Agilent push-button power switch on eBay.  Its HP / Agilent part number is 3101-3226.

It has the appropriate number of pins (6), and, although the pin locations and spacing did not look identical to that of the original switch, I thought I'd order one and see if I could adapt it to my purposes.

Here are a couple of photos showing the differences between the original switch and the replacement switch:

As you can see, the pins of the original switch are offset from the vertical center-line of the switch, while the pins of the new switch are symmetrically placed about its center-line (see second image).

Also, the new switch is significantly shorter than the original switch.

But I thought I could make the new switch work.  Because the replacement switch is shorter than the original power switch, I spanned the gap between it and the PCB with stiff pins I cut from a standard header connector (e.g. the ones that have pins spaced 0.1 inches apart).  

These pins, in turn, I installed at a slight angle to align the new switch's pins with the off-center PCB pads.  And, being header pins, they are fairly stiff and less likely to bend when I press the button.

The image below shows the new switch mounted on the HP 3335A's front-panel PCB. 

The results:  It seems to work fine!  The "throw" of the switch (the difference in height between when the switch is ON versus when it is OFF) is a less than that of the original push-button switch, but that is an insignificant issue, in my opinion.

Standard Caveat:

As always, I might have made a mistake in my equations, assumptions, drawings, or interpretations.  If you see anything you believe to be in error or if anything is confusing, please feel free to contact me or comment below.

And so I should add -- this information is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.

Tektronix DC 509: Simple Mod for Adding an External Reference Clock

 

I recently needed a portable counter for some measurement work.  Looking at my shelf of test gear, I found a Tektronix DC 509 Counter Module that I could plug into one of my TM 503 Mainframe for transportation.

However, when I went to test it, I discovered, when measuring a 10 MHz signal from my HP 8640B (that is locked to my GPS frequency standard), that the counter was off by about 20 Hz, and that this error would change as the counter warmed up.

Very annoying.  If only I could synchronize the counter to my GPS frequency standard...

TM 500 Series Mainframe Requirements

Per the DC 509 manual, the counter does support an External Reference Clock via its rear connector pins, and this clock can be either 10 MHz, 5 MHz, or 1 MHz.

To provide, for example, an External Reference Clock to the counter, the TM500 series of Mainframes has an option (Option 2) to add BNC connector(s) to their rear panels, as shown below (image from eBay):

I have five TM 503 mainframes, but none of them have these BNCs.  Instead, their rear panels all look like this:

Rather than purchase a TM 503 from eBay with the appropriate BNCs, could I instead somehow make a simple modification to let me add a connector to the DC 509 for an External Reference Clock?

DC 509 External Reference Clock Circuitry:

The answer to the previous question is "yes," and it's actually a pretty simple modification.

First, let's look at the DC 509's clock circuitry:

I've annotated the schematic to show how to set the on-board jumpers to (1) select External or Internal Reference Clock, and to (2) select the frequency of the Reference Clock.  Note that the external clock's amplitude should be roughly about 4 VPP (into an open load), at a minimum, from my own measurements.

Adding an External Clock Connector to the DC 509:

The External Clock comes into the DC 509 via the following pins on the Rear Connector P1625:

     P1625 pin 14A:  External Clock Input

     P1625 pin 15A:  Ground

There are two side-by-side vias for these signals, next to the connector, as shown in the image, below:

These two vias are slightly more than 0.1 inches apart.  Never the less, a 2-pin header with 0.1 inch spacing can be inserted into them (after removing the solder in the vias) and soldered without too much difficulty:

A matching two-socket ribbon cable can now be attached to this header for the External Reference Clock connection to the DC 509.  In my case, I made a cable that adapts a short 2-pin socket cable to an SMA connector cable (all parts from my junk box).  See below:

Note that, assuming a 10 MHz clock signal, the length of the 2-pin ribbon cable is not too unimportant, given that a 10 MHz signal's wavelength is 30 meters long (in the ribbon cable this wavelength will be shortened somewhat from 30 meters due to the the cable's velocity factor).

The DC 509 can now be inserted into a TM 500 series mainframe and the External Reference Clock cable brought out the front of the mainframe (as long as the mainframe has at least one slot that can be left empty, to route the clock cable out the front of the mainframe), 

Setting the Jumpers:

On the DC 509 PCB, the 3-pin Header, J1621, selects whether the Reference Clock source will be Internal or External:

    Jumper J1621 pin 1 to pin 2 for Internal Clock (this is the default factory setting).

    Jumper J1621 pin 2 to pin 3 for External Clock

The 5-pin Header, J1511, should be set to match the Reference Clock frequency:

    Jumper J1511 pin 1 to pin 2 if the Reference Clock is 10 MHz (this is the default factory setting).

    Jumper J1511 pin 2 to pin 3 OR pin 3 to pin 4 if the Reference Clock is 5 MHz.

    Jumper J1511 pin 4 to pin 5 if the Reference Clock is 1 MHz.

The image below shows the 3-pin Header, J1621, set for External Clock:

Operation:

As a quick test I hooked up my "External Clock Header-to-SMA" cable to the DC 509, and then attached this cable to an HP 3335A signal generator set to 10.000000 MHz (itself locked to my GPS frequency standard).  The HP 3335A's amplitude was set to +10 dBm (i.e. 2 Vpp into a 50 ohm load, or 4 Vpp into an open load).

I connected the DC 509 to the output of my HP 8640B generator (the latter with its frequency display expanded to show resolution of  1 Hz), and compared the frequency displayed on the generator with the frequency displayed on the counter.  They matched (as shown, below)!

Standard Caveat:

As always, I might have made a mistake in my equations, assumptions, drawings, or interpretations.  If you see anything you believe to be in error or if anything is confusing, please feel free to contact me or comment below.

And so I should add -- this information is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.