In a previous post I briefly talked about a 6GHz (002, 006, 010, 011) HP8753D that I had recently purchased on eBay. And along the way I had also acquired all of the other bits and pieces (mainly from Eric Haskell in Texas) to turn an 011 unit into a non-011 (with integrated test set) unit. Well, it's finally done - fully cal'd and ready to go! The hardware conversion only took a couple of hours. The full cal took longer. 
 
 

Posted by: Lab Coat

I purchased this Tektronix 2782 SA a few years ago. And for the most part I've been very happy with its performance. There were a few things that did bother me though. One was a cold start issue where the unit couldn't detect one of the YIGs. After a warm up (and an unplug / replug / restart cycle) the unit would work fine. The second issue was a little more mundane, but annoying none the less ... the Real Time Clock was stuck! How could that be?! 
 
So, a trip to eBay to find some NOS Statek RTC-58321 real time clock IC's (really only needed one, but bought a few). And after a bit of minor surgery to remove and replace the part, all is well (at least for the clock). The startup issue was a little harder to fix. I had to replace the internal 500Mhz oscillator, as the module that was in there was previously (incorrectly) replaced from an older rev. So that's 2 for 2. In the end, I now have a fully functioning 2782. 
 
Perhaps (as John Miles KE5FX warns me) I might need to replace the electrolytic caps someday soon ... apparently they come from an era where they leak electrolyte and can cause lots of issues. 
 
 

Posted by: Lab Coat

So we ran into an issue working with a client using this module. Specifically, we couldn't figure out why we were NOT getting full output power (RSSI with Lightblue from only a few feet away was -90dBm). Long story short, Laird has a special whisper mode. If you have similar issues, see this blog post on the Nordic Semiconductor Dev Forums --- https://devzone.nordicsemi.com/question/4629/laird-bl600/ 
 
Lazy? ... to paraphrase: 
 
"It seems Laird uses pin 20 to enable whisper mode, so if it is not explicitly set to low, then you will get terrible RF performance. 
 
Response from Laird: 
 
If you are using the Nordic SDK, the output of GPIO 20 must be set as a low (0v) to enable the RF transmit power circuitry. Your reference to a 10m range is the most commonly reported symptom for this issue. Setting GPIO 20 low is done via the Nordic application that you are using. We instituted GPIO 20 programmatically with smartBASIC however if you are indeed using the Nordic SDK you must set it manually." 
 
Here's an image of what's under the can. You can see a three terminal device in the lower right that goes to GPIO 20 and the PA Enable pin. They are probably throttling the amount of power available to the output stage on the nRF51822, thus limiting the range. 
 
 

Posted by: Lab Coat