Maker Pro
Maker Pro

Help to fix HP5308A counter

Hi,
As surplus from my old job I have an old HP5308A counter/timer that does'nt work and I'd like to fix it.
I have manual and schematic for the HP5308A part of the instrument.
Testing around, it seems the problem is that the internal 10 MHz clock signal is dead.

Unfortunately, the clock is generated in the top section of the instrument, the measuring unit HP5300B,
and for that I have not been able to find a manual or schematic.

Or would you know of any forum that is more specifically dealing with this kind of stuff?
Any suggestions appreciated,
Sverker
 

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
First thing to check is that there's not an internal/external switch for the 10MHz signal that is in the wrong position.

Next thing is to provide an external 10MHz, if that is possible, to check the operation of everything else.

If you confirm that it's the 10MHz signal alone, then there are various options available to you.
 
*steve*,
Your advice was spot on, thanks!
When I separated the two halves of the instrument, on a side otherwise not visible, there was indeed an OScillator INT/EXT switch, and on the back panel there's an OSC-BNC connector .
Set to EXT and with a simple function generator (a UDB1210S) at 10 MHz plugged in,
the instrument came alive and is working all right. Great!
You mention there are various options now, to fix this permanently.
What may they be?
I am not looking for very high precision and stability.
This cute 75 MHz timer/counter is of 1965 design vintage, built mostly with SN74S-series chips and a few ECL.
The internal osc is crystal discrete comp type on the board, including an 1820-0806 ECL gate package,
but I don't figure trying to repair it, I don't have schematics for that half of the instrument.
I've been told about TCXO but know nothing about them, how could it be used here?
Are any examples of installing it available, or application notes?
Are there other reasonably simple and cheap ways to go?
Further advice appreciated,
Thanks a lot,
Sverker

First thing to check is that there's not an internal/external switch for the 10MHz signal that is in the wrong position.

Next thing is to provide an external 10MHz, if that is possible, to check the operation of everything else.

If you confirm that it's the 10MHz signal alone, then there are various options available to you.
 

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
There are several options, and they are all worth considering.

The various options will all vary in cost, short term, and long term stability.

Short term stability generally included changes in frequency due to temperature, noise in the signal, and short term fluctuations. This would affect readings taken 5 minutes to a day apart.

Long term stability is often dominated by aging of the crystal (for crystal oscillators), temperature drift in ovens (caused by aging of the temperature regulation components), etc. This would affect readings taken a week, a month, or a year apart.
  1. Keep using your function generator. This is simple, available, and almost no cost. However, it deprives you of the function generator while using the counter, and is possibly limited in accuracy (depending on your function generator).
  2. Make your own Crystal oscillator. You may even have a suitable module, if not it's a simple and cheap thing to do. However there accuracy won't be really good and it will drift with temperature.
  3. Buy a TCXO from eBay. This is simply a crystal oscillator using a temperature stable crystal. Chances are, the unit has one of these internally and you may even be able to fit it in place. You get a 10MHz signal that is more stable with temperature.
  4. Get an OCXO. An ovenised crystal oscillator has an inbuilt heater and temperature control (the oven) used to keep the crystal at a constant temperature. These are even more easily available on eBay, and second hand units are often reasonably priced. Some of these units have a double oven. This means that the crystal is inside one oven, and that oven is inside a second oven. There oven can draw a significant amount of power (I have one that draws in excess of 15W when started from cold). The power requirements drop once they heat up. Beware of OXCO's that don't draw a lot of power, the oven may be broken. Also realise that these may require 30 minutes or more to come up to temperature. I have a counter with one of these fired internally and it is always peered (even when the main switch is in standby -- it doesn't have an off position).
  5. GPS 10MHz signal. There are a heap of cheap gps modules. Find one with a 10MHz output. Their long term stability is excellent, being similar to that of the hydrogen maser on the satellite. However, they suffer from short term drift, and require the GPS to be able to see at least 3 satellites. Aside from not working in really nasty conditions when GPS signal is lost, and potentially needing an external antenna, this is pretty much set and forget (there is nothing to adjust). It is also becoming very cheap.
  6. Rubidium oscillator. These are generally units removed from decommissioned cell towers. They have really good short and long term stability. Again, these use a lot of power on startup. I think mine consumes 25W or so. They also have a limited lifetime. These can take up to 10 minutes to come into lock, but they do have an output that tells you they are locked.
  7. GPS disciplined oscillator. This is a combination of one of the other oscillator types with a GPS to improve the long term stability. Mine has a double oven crystal oscillator. These can take 30 minutes for the oven to warm up, and maybe 10 minutes for the GPS to lock, but it's still not over, because the units need to know their precise position to correct the signal they receive from the satellites. They may need to be powered up for several weeks before they achieve their ultimate accuracy. Mine is designed to run from DC. This is not uncommon because you would normally have it run from batteries so it could remain powered during outages. If there is a GPS loss (really heavy rain can do it) the internal oscillator will continue to give you an output that is as good as it's short term stability until the GPS signal is recaptured.
  8. Hydrogen maser. Good luck getting one of these :)
Depending on your needs, 2, 3, 4, or 5 will probably be what will work for you.

Some googling will give you indicative comparisons of short and long term stability of these options.

This doesn't compare everything (and it does list a few more options), but it may be useful.

Oh, I also should mention that a second hand crystal oscillator may actually have very good long term stability because frequency changes caused by aging slow down with time. However, the device may have been removed from service for a reason (you takes your chances).

I compare my GPS disciplined oscillator against the unknown 10MHz source using my oscilloscope. When locked to the GPS source you can watch the other source drift (and adjust it where possible). Once you get them accurate to a fraction of a Hz, you can leave everything running for a few days to monitor how the signal drifts (although you have to make assumptions as to which signal is doing the most drifting).
 
Last edited:
Thanks *Steve* for your very useful overwiev of the reference oscillator field, a few weeks ago.
To do some trying out, I have obtained a quite cheap TCXO (I suppose that's what it is), called H14.
It is on a 14pin grid, with #1 N/C, 7 Ground, case, 8 output, and 14 supply +5V.
However, I have never used any crystals or crystal oscillators, and I haven't been able to find any application example. With a 100 nF ceramic over the supply, does it require any more surrounding components?
Like some capacitor or resistor on the output?
Once I get it to run I mean to feed it to a HCMOS buffer.
Does that seem right?
Advice appreciated -- as always,
Sverker
 

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
In general, you hook them up to power and the output does its thing.

I went looking for the service manual and only found this. If you can find a more complete service manual it will often have schematics and a discussion of the operation of the various parts of the circuit. This would be invaluable in doing the mod you're planning.

Ah, go to the source... Not googlable, but keysight have a lot of manuals available even for *really* old equipment. This is what you want.

I don't have time right now to try to figure out more detailed instructions, but I did see it mention the optional high stability oscillator. If it goes into more detail of how this is fitted, you should be able to locate the power and ground (hopefully 5V is available) and where to connect the output. If not, there's a lot more information that should make the process easier.

It's interesting to note that the 5308A seems to be part of a system. I wonder if the clock is external?

Hope this helps.

edit: check out section 4-36 on page 4-10. It may point to where the fault lies.

Page 7-5 has the procedure for adding the TXCO -- this isn't a simple plug-in mod, so don't start doing this!

The pages following 7-6 show how the internal/external switching works, and also show the crystal oscillator. This should give you a good indication of where you can slot in the new oscillator. The main issue though, is that the oscillator you have isn't adjustable. I'd try to fix the existing one first.
 
Last edited:
Top