HP 8601A Multiple repair.

[radiobloke]

Apologies, as this is a really long post. However if you want or have one of these units it might actually save you a lot of hard work.



I got this quite some time ago, and it was about time to repair it ready for use.

So, first things first, the original mains lead had snapped at the plug end internally and the wires have turned hard, so it won't be long before they break short out internally causing all sorts of unpleasantness. I decided to upgrade the socket to a somewhat more modern design Looking at the service diagram the old socket has filter coils and capacitors moulded into it.Digging about in the spares box I found an old filter capacitor set from a computer power supply unit which had the same specifications as the old old one. I grafted this on to the socket and connected all of the outputs to their respective places, the only thing to do now was to very carefully remove a section of metal and file it down so it looked neat and tidy.

Knowing that the unit powered up to some extent I tested the output for a signal, nothing. Luckily on this model it only has three hybrid ic's and you can swap them around for each other temporarily for fault finding purposes. On doing so I got a strong signal but was unable to alter frequency. All of this meant one of the output preamplifier hybrids was faulty. So all got put back to their original places and I powered up the oscilloscope. The top preamplifier had lots of signal in but not much out.

If it was the final power amp I would have probably just bought one of this gentleman's fine replacement examples, but this was the preamplifier, so I decided to take the lid off carefully and have a look inside to see if there was anything obvious. Under the microscope you could just see that the transistors were looking a little sad for themselves, which was confirmed by very carefully testing with a meter confirming that the emitters were almost open circuit. The plan was to solder compatible transistors onto the hybrid directly with fine wires for base and emitters to the appropriate places. Borrowing from his research the BFE35 transistors seemed a pretty good match for what was needed, so I ordered a few to experiment with.



Using a fine pair of tweezers and a micro tip soldering iron I soldered on the new parts very carefully, so as not to damage anything else and powered up. Instant signal After grinning like a muppet for a while I discovered I was not completely done yet. Whilst it worked perfectly on the lower frequency scale, it had no signal on the upper scale. So, to rule out any spurious problems with switches I spent the next hour cleaning up all of the contacts on all of the switches and getting all of the dust out. Now it was much nicer to use, but the upper scale still didn't work. Grabbing the service manual and printing out the various flow charts and diagrams, it was obvious HP don't like engineers very much. The manual is a bit of a mess, and bits of it was scattered throughout the entire manual with no logic behind anything. However after working out how they wanted you to service the item things began to fall into place. First port of call was the divider assembly, there is a perfectly good signal on the 0-11 scale, but from the 11 to 110 MHz range there was absolutely nothing. After carefully removing the divider board and checking the signals on the chips something didn't add up. One of the chips had a random signal on its output, and even then it was barely there.



Naturally on checking, none of these chips are available any more, partly because they're HP branded versions of Motorola's chips, and also the fact that Motorola don't make them any more anyway. Thankfully some kind soul had put an equivalent table for the chips online which led me to find an equivalent to the faulty one which was numbered 1820–0102 and its replacement is an MC1013p. I managed to find someone selling some 'new old stock' of these online and quickly ordered up a few.

Now is the new chip installed and fitted the upper frequency was now reinstated. Another small dance ensued that was, until I tried to use the sweep function, The signal went dead. I was beginning to believe that this was made up of an entire bunch of these units that were all faulty and I,d ended up with all the broken bits from halfway around the planet. When the sweeper was switched from carrier wave to sweep the signal would die. Not quite sure where I was to actually go next I decided to work through the alignment procedure on each one of the packs to see if anything was grossly out of alignment. First couple of packs seemed to be okay, those being the power supply, and the crystal oscillator adjustments, I gave him a bit of a tweak whilst I was there to get them spot on. Next up was the voltage tuned oscillator, the first thing I noticed was when you disconnected the lead to do the initial adjustment, the signal went dead. So I removed this board and went around testing voltages around the various areas. Something didn't seem right, there was 75V on everything and whatever frequencies were there, were all over the place. so obviously something was wrong here to, on closer inspection like with all the other boards this one didn't seem to match anything to do with the serial number, so was yet another of somebody's broken lineup of products.



After a little while I managed to find a open resistor (shown above) and on the replacement of this and a quick set up of the now very grossly out of adjustment oscillator, I finally had a working piece of test equipment on every single setting.



Phew, this all took just about a week to achieve between various other jobs that came in needing to be done, But at least now I have a very nice piece of test equipment that I can use to set up the IF stages of radios. You can actually just sit a radio on top of the unit and it gives off a enough RF to hear the broadcast tuning past the radio's tuned point. If you see what I mean.

So that's it, I will probably refurbish the cabinet of this unit at some point, but for the meantime I'm just going to enjoy using it.

All the best out there in radio land, Radiobloke.

[Radio Wrangler]

Good work!

The 8601 is a useful little machine. They used to be the sweeper part of an early 110MHz network analyser. There was a mainframe with CRT (polar or cartesian) and gain/phase meter plug-ins.

David

[radiobloke]

Yep, it should make a good wobbulator, amongst other uses. I've accumulated a fair few RF Signal generators over the years, but this one will be interesting to have a visual peek at the response curve of the radio's in my collection.

It'll be fun to see if I did a good job of peaking them.

[Radio Wrangler]

A boardwith one of those analog devices inc. RF logarithmic detectors woulf be a good accessory for it.

David

[Station X]

Interesting write up. Rather you than me when it comes to repairing hybrids!

I have one of these units and find it very useful.

I hope you kept the old mains leads as the connector is difficult to find these days.

[Radio Wrangler]

The hybrids were originally made by hand. One operator using a microscope placed a splodge of silver-loaded epoxy and stuck transistor die down onto the metal package through holes in the substrate. Capacitors like modern SMT ones were hand soldered onto pads on the substrates. Resistors had been inked and fired beforehand. The epoxy was set by baking, then someone with a microscope and a ball-wedge bonder fitted gold bond wires.

A good microscope makes all the difference, and a steady hand of course.

Radiobloke did it all with solder and missed-out the bond-wire bit. It all shows it's quite possible.

David

[radiobloke]

Just need a steady hand really, it might not win awards for being pretty but it works well. (The photo shows it about 5x bigger than it actually is) the transistors were soldered onto the caps like that to enable a bit of heatsinking. It's been working now for about 10 hours so I guess it is probably fine.

Had a go with a basic AM IF stage today and it seems to work quite nicely. Bit fiddly setting up where on the scope the upper and lower sweep points are using the manual sweep, but once set on the screen it looks fine. No need really for a marker for AM as it tends to swamp the waveform, but FM will have a bit more bandwidth. Will try that tomorrow.

It's nice to know the other hybrids might be repairable if they go faulty. It's the only Achilles heel for these sweepers (cmos stuff notwithstanding - lack of availability) so it should be good to go for a few more years.

I guess I just don't like to be beaten by faults

As an aside, why would you need to sweep a network? Reflectance, or phase distortion, or am I not thinking of the right type of network?

Cheers everyone, Radiobloke..

[Radio Wrangler]

There are so many sorts of networks.

Swept measurements are pretty much essential for filter design and alignment. Youcan look at the magnitude and phase of the forwards path. You can use a directional coupler to look at the reflection of the port at either end of a network.

With just magnitude, all you can see is the VSWR or return loss (whichever form you prefer) and you need the phase to get that down from a circle on a Smith chart to a point on the Smith chart - a value of impedance with no ambiguity.

You can sweep antennae and plot VSWR/impedance.

And this is how the impedances and characteristics of RF transistors are arrived at.

And there are those famous filters in RA17s which need attention once the phantom twiddler has been.

David

[radiobloke]

Very interesting, thanks for that. Nice to learn something new every day, i'll have a look for more info on it.

All the best,

[Radio Wrangler]

and Look up hp 8407A and 8414A and 8412A which are the toys to go with your 8601

David

[John_BS]

The nice-looking grey resistor you found open was the exact same type which I found faulty in a Tek 454A horizontal amp. It may be they are susceptible to being stored damp.

John