Tuesday, September 29, 2015


I am going to indulge in some wanton and probably unjustified ranting here and, with it, cast slime and aspersions on two guys who post You Tube videos that drive me to drink. But before I go any further, I will post them here and at least give you all a chance to crack a bottle of Jack as well:

The first is by Dave Jones over at EEVBLOG:

..and the second is by M. J. Lorton:

So, a caveat and then my complaint: I am a complete ninny when it comes to electronics and these two gentlemen can run rings around me in terms of knowledge and experience. That said, both are off-putting because they do not construct too many videos that help the average twit like myself. In fact these two are nothing more than 45-minute brag tapes about how Dave Jones scored his voltage reference for a song from some poor, unsuspecting clod at an Aussie swap. Don't even get me started on Lorton who manages almost one hour lisping his way through minute-by-minute temperature changes in his workshop. (And this is the last of six similar videos on the subject. Get. A. Life.)

Paint, drying, etc.

So, that's this got to do with anything? Well, I scored an HP 3468A off eBay for a modest price and was looking forward to incorporating it onto the bench as a reference because it was more accurate than all of my other instruments. Or, as Juvenal once famously said:

Quis custodiet ipsos custodes?

In doing research on my scored treasure I cottoned to the fact that this unit uses parameters stored in CMOS memory to effect its accuracy. Basically, the instrument measures what it measures and then offsets or digital differentials are applied to the measurements to show up on the display. Again, these are stored in CMOS memory circa 1980s and that was volatile. If it lost power, the RAM contents vanished. In order to retain the memory, HP designed a 3V Lithium battery backup that lasted about 10 years.

Well, that's all well and good an idjits like myself open their treasures up, slap a DMM on the backup battery, read 3.12V and think all is well. Sadly, a Lithium battery has a life profile that looks like a ball rolling off a table. After ten years, you're living on borrowed time.

(This was the deal with my old Icom IC-745. When it lost its mind, Mother Icom had you by the short hairs to reprogram the radio. Fortunately, the Piexx folks came along with their non-volatile RAM and saved the day.)

So, casting about, I found a tremendous gentleman -- Mr Modem Head -- who did a treatise on restoring a 3468A and replacing the CMOS backup battery. Fearing a day of reckoning, I studied both posts assiduously, got a replacement battery form DigiKey, and last Sunday afternoon I did the surgery.

Well, the operation was a success but the patient died!  When I powered the unit up after the solder cooled, up came the infamous ERROR 1 message on start-up:

I swear by all that is holy, the backup battery was strapped with 3.25V at all times. I followed the procedure to the letter. But, alas, I crapped out, rolled snake eyes, busted, came up with three peach pits, etc. So now my treasure was only as good as the other DMMs I had lying around and I was off searching for how to calibrate it myself using a Voltage Standard DMM Check, the HP service manual, and my wits.

..well, the DMM Check and service manual at any rate.

By the way, Mr Modem Head's blog is everything that Wanker and Motor-Mouth Jones is not. It is artfully done, contains useful information; true wit and wisdom. Sadly, current events (life, responsibilities, etc.)  conspire to keep Mr Modem Head from to many recent posts but he is a roll-up-your-sleeves electronics hobbyist who shares his project efforts in a stunningly beautiful style superbly laid out on his blog. (See links above.)

..all for the common man without benefit of $20,000 instruments or being knee-deep in prime Fluke or HP surplus.

Anyway, the outfall here is that a calibration shop was located who would redo the parms for $65 and my HP is headed for Medford, Oregon as we speak. I will keep you apprised.


Monday, September 7, 2015

The Wreck of the old '485..

My Tek 485 had been sleeping on the shelf and I needed to take it to a friend's house to help him debug a problem. Checking it out before I left, I found it did not light up -- no power, to pilot light, no trace, nothing.

So, with the help of my friend, Dick, WB6JDH, I took it apart and did a preliminary visual on the back plane, the line board, the switcher board and the PS board. The two 430-ohm resistors (R1812 and R8113) were burnt out so I replaced them with two 2-watt 470-ohm resistors on the line board.

All else looked fine. After cleaning out some carbon and dust, pulling the "start" transistor and then the "stop" transistor on the PS board and bringing the unit up on a variac with each, we got the power supply to work -- but experienced some smoking and burning during the process.

Cleaning out further carbon and burned areas revealed no seemingly bad components so we brought the unit fully up under power and it seemed stable. However, attempting to view a trace on the scope revealed that the vertical position controls did not work and I was unable to get a dual trace. I was pretty careful with the setup and triggering so I do not thin thee was a cockpit error. Another anomaly was a distinct PS snapping when the "ADD" button was pressed.

Here's the video:

A tantalum cap on the A7 board (near U650 and the 15V test point) shorted and took out the associated RF choke. We swapped the A7 board with one from a donor 485. Turns out that one was bad as well. Q374 and Q384 were both bad and were replaced with matched 2N3904s. Was easy to do because most semiconductors are socketed on that board. But that was karma because removing and replacing the A7 board WAS a bitch! The delay line was soldered on and a semiconductor on the board was bolted on a heat sink to the frame below. Getting its leads through the two holes in the sub-chassis was a delicate operation -- especially on a 100-degree day with 90% humidity.

I deserve NONE of the credit for solving this problem! Dick, WB6JDH, [literally] sweated through the debugging process and suggested we swap boards from the old donor unit. I just dd the donkey work and took copious notes.

I also treated us to burgers and dogs afterwards.

Resolution Resolved
Actually, the resolution was a false positive and the malady returned. (And, truth be told, who's to say it won't return in the future?) But, in tracing the origin of the horizontal position controls back to their source on the power supply board, I found all of the components in each of the position controls to be within spec and performing correctly. However, it was noted that the cable plugs along the top of the PS board were unseated and, after re-seating them, the scope seemed to function correctly. So, as I said, the problem might re-occur but at least I'll know how to troubleshoot it.

Stay tuned..or don't!


Clearing out the shack and the bench..

I decided that I got too much stuff and need to "clear the bench" as they say in baseball. So below are pictures offered up in aide of a couple of scopes I have that I want to get rid of. They are a pair of Tek T9XX student scopes that have fallen into dis-use. Both are essentially in good working order. Here goes:

Tek T935 35 MHz Dual Trace
Immaculate inside and out. The trace is modestly bright and it generally seems to function o.k. although it may need some attention in the Dual mode.

Tek T922 15 MHz Dual Trace
Tektronix T922 Student Oscilloscope, every bit as clean and immaculate as the T935 above and it has a really bright trace. This scope is dual trace with a 15 MHz bandwidth and a bright trace and all seems to work o.k.

I discovered the on/off detent switch sometimes does no catch but hit it with a little contact cleaner and worked it a few times and it has returned to proper form. Other than that, it's a good unit.

..so there they are. Now, I'd be in great shape if I could get my 485 to work! (See next post up the road.)