Tenma 72-465 multifunction counter disassembly and repair

[LoyalServant]

Greetings everyone!

I have been lurking around the forums for quite some time and figured it was time to register and actually post something.
I work a lot with RF gear and spent 15 years working on consumer electronics.
When it started going towards module replacement I got disenchanted and jumped into software development 5-6 years back.

I am not good at this so bear with me. so... on to the Tenma...
As everyone probably knows this is a low end piece of kit sold by MCM 
It's quite old - probably mid 1990s.

I am not good with images... so if this doesn't work out I am sorry!
This is an image of the front of the counter:
http://imageshack.us/a/img29/9508/imag0026yy.jpg

When I fired it up and stuffed a signal into it I found that it did not work. 

Inside of the unit:
http://imageshack.us/a/img823/4339/imag0032qh.jpg

We have some decent components here from TI, Motorola, NEC, Harris, etc.
The Harris ICM7226 is the heart of this unit.
Take note of the metal can in the middle of the board - that is an OCXO. (oven controlled crystal oscillator)
I was semi surprised to see that it had any real attempt at accuracy.

Here is what the bottom of the board looks like:
http://imageshack.us/a/img822/914/imag0028jm.jpg

There has been some soldering and a cap bodged under the crystal oven.
It looked suspicious as did some of the soldering.

After I fixed a problem in the power supply the unit did fire up, but it was off frequency.
I took a look at the 10Mhz output and noticed that it was drifting around.
I grabbed an image with my scope:
http://imageshack.us/a/img35/6899/tenmaoff.png

At that point in time you can see that the 10Mhz reference was not quite 10Mhz....
Obviously there was an issue with the OCXO in this unit so I took the cover off for a peek.
I fed it a 10Mhz sine wave and this is what I got:
http://imageshack.us/a/img825/3028/imag0035p.jpg

Clearly not happy!

Here is the inside of the OCXO:
http://imageshack.us/a/img20/5972/imag0033mu.jpg

Someone unfortunately decided to poke around in there since there was evidence that the oven had been removed
and opened before I got to it so an inspection was in order to see if everything was in order.
The only thing that seemed out of place was the 4pF bodge on the other side of the board.
It looked like it was added after the fact.
I decided to remove it and give it a go again but it was off a little more.
The variable capacitor only adjusted the oscillator very slightly so it needed a little help getting nudged on frequency
and this is where I suspect the bodged cap on the bottom came into play.
The attempted repair OR factory addition must have been an attempt at getting the oscillator stable and on frequency.
I pulled out some caps and tried increasing the value and settled on a 15pF cap that got the oscillator stable and close to 10Mhz.

Here is the newly bodged 15pF cap:
http://imageshack.us/a/img803/1536/imag0038t.jpg

After putting the OCXO back together and a 30 minute warm up period I was able to get the 10Mhz reference
oscillator stable and damned close to 10Mhz:
http://imageshack.us/a/img100/2143/tenmastable.png

I really don't know what to make of the bodged cap on the underside of the board.
Was it added at the factory or by someone after the fact?
Whatever the reason it was obviously an attempt to resolve an issue with the oscillator.
I did check the other caps in the oven and they were fairly close to their rated values.
Both were less than 3% from the values stamped on the package.
I could probably do with putting some 1% caps in there but it's not worth it.

Here it is reading a 10Mhz sine wave from one of my function gens:
http://imageshack.us/a/img812/739/imag0037om.jpg

Spot on... amazing!
Now I am sure someone will ask... why bother with such an old unit?
I have 2 other frequency counters built into function gens plus a few scopes.. but I had time to kill and I wanted to see if I could
get it going.

There are some things that I think are a FAIL here like the mains switching.
They use this little coax cable at least in an attempt to shield the mains that is running to the front of the unit.
It's running parallel to the small coaxial cable that is carrying the 10Mhz reference input/output to the back.
The internal/external reference selector switch is one of those flimsy slide switches you would find in some $1 Chinese toy.
The method of grounding the case is a little questionable. All that is used is a copper clad spring.

Other than a few nit picky things it's actually well built considering the target audience.
The use of a crystal oven was rather nice of them.
I have seen some really dodgy counters out there that had a resistor glued to the crystal and that was it.
Here they clearly cared about having a more or less constant temperature for the crystal and it shows.
After a 3 or 4 minute warm up period it's quite stable.

Hope this was of interest to someone.
I have a few Hitachi CROs that need a little love so perhaps I will tear one of those down.

This was my first time so go easy on me guys!

[PA0PBZ]

Hi,

Interesting, thanks for the story and the pictures.
Since it is only the crystal inside the can it is not really an OCXO, the oscillator is not in there but it is in the ICM7226.
According to the datasheet there should be a 22M resistor across the crystal. I can see a resistor between the can and the plug but it looks more like 1M in the picture. That could be a good reason that the oscillator has a bit of trouble.

Paul

[LoyalServant]

I have always known those as crystal ovens since the purpose of them is to keep the crystal at a constant temperature.
The TIP31 has a thermistor mounted next to it that pulls the base to ground as it warms up.
There is a resistor between the base and collector that is on the underside - 220 ohms.
The thermistor starts at 700k and when I touch a soldering iron to it the resistance drops to a few hundred ohms effectively
shutting the TIP31 off.
The collector is tied to the 5V rail and the emitter is tried to ground.
It heats up quite quickly and seems to maintain it's temperature quite well.

I have a feeling that the crystal is a bit sketchy.
This thing has been banged around a lot and it's quite old so I am sure things have drifted off.
They only used 5% tolerance components in there...

This is nothing more than me playing around....
It actually appears that I got this thing dead on the money.
It takes less than 5 minutes to stabilize and it agrees with 3 other counters.

I know there is some time gods out there that are probably cringing at the crudity of this reference oscillator

[Fraser]

I have a 2.7GHz frequency counter that looks suspiciously lik eyour unit.

Mine has some interesting 'bodges' added by the manufacturer, such as copper foil very badly installed and soldered around the prescaler and such like.

I have to admit that I was unimpressed with the general build quality of my unit. Surprisingly these are still available new and demand a price of around GBP80 on ebay.  I used mine for occasional rough checks on oscillators running at 2.5GHz, but nothning more demanding than that.

Mine will soon be heading to ebay for conversion into something more useful.....money

Fraser

[LoyalServant]

I thought about tossing it on ebay but these don't go for much - at least not enough for me to care about.
Not really sure that a standalone frequency counter is really that important to me nowadays.
Good function generators have one built in and so do most modern DSOs that is good enough for my purposes.

[bingo600]

I have a 2.7GHz frequency counter that looks suspiciously lik eyour unit.

Mine has some interesting 'bodges' added by the manufacturer, such as copper foil very badly installed and soldered around the prescaler and such like.

I have to admit that I was unimpressed with the general build quality of my unit. Surprisingly these are still available new and demand a price of around GBP80 on ebay.  I used mine for occasional rough checks on oscillators running at 2.5GHz, but nothning more demanding than that.

Mine will soon be heading to ebay for conversion into something more useful.....money

Fraser

I also have one of those 2.7Ghz $80 counters , in my summerhouse.
Mine had a "13Mhz chinese diy ocxo" in there , and it was rather bad (mildly said).

I opened it , and found an AVR inside , and the ocxo with an adjustable ferrite core thingy ....
Being unable to adjust the ocxo , i decided to replace it with a decent ocxo.

I had some $5 Pletronics 26Mhz OCXO's (eB #130718871982) - Stability: ±100ppb
So i used one of those , a 10K 1-turn for EFC (had no 10 turns) and an AC74' (to divide the 26Mhz to 13Mhz), and got a rather decent/stable Freq cntr.
Even with the "lousy 1-turn pot for EFC" , after 30 Sec turn on time. The counter just wanders ±2 on LSD , when connected to my 10Mhz Trimble Thunderbolt GPS reference.

I did wire up a separate OCXO 5v supply (using a MIC2940-5 and a decent Cap) , grabbing the V(in) from the existing diode bridge.

So i'm keeping my $80 China counter in the summerhouse , but have later added a PM6674 with a tcxo (and ext. 10Mhz input for the Tbolt) , as i got a PM6680 counter @home.


Edit: Ohh ... No funny tinfoil etc. in my China Counter.

/Bingo