Ballantine 3205B Microhm meter

[Vgkid]

-----Ballantine 3205B Teardown-----
All images are linked to larger versions.
So I always wanted a milliohmmeter, ever since Robbenzs teardown of his LOM-510.  The commonly available milliohm meters on ebay that are relatively compact include the Keithley 580(possibly the most high specked milliohm meter, also several multiples more expensive than the other options). Next up is the Simpson 444(these go for a couple hundred, these use an ultra low current,5-ma). Various Valhalla models, each one did something different, look into Valhalla site for more info. The Ballantine 3205B, these are rather uncommon(I thought that it looks very much like a Valhalla instrument), then lastly the awesome LOM-510. Then there are the much newer hioki's that are on the internet. To bad more companies did not make microhmmeters. Then you have the newer Tonghui's/Applents/Extechs, and those nifty microhm meter based on the hx-711 chip, eg DannyF's.
Back to the meter.

Now we can see the front of the unit. If you have ever seen any of Valhalla's test gear, you will see why I thought that this was a rebadge. The measure button is only needed for non pulse resistance measurements. The Zero trimmer is used to null out the thermal EMF's. The slightly newer, and much more exspensive hp3456A actually does this by taking a measurement with no current, then with current, and subtracts the thermal emf's in processor.
   On top of the unit is nothing but discolored beige plastic, with a ring of epoxy on it.

Flipping the meter over we are greeted with a  placard detailing the user instructions, and the accuracy specs. I wish these were still included, heck my Digibridge feature a bunch of info relating to impedance calculations .

Taking off the top of the unit we can look at the display board. The to-18 transistors are bc109's. I did not make note of the to-92 transistorsThe solitary ic is a DM7447, a BCD to 7-segment decoder. The only difference between the A-B is the additional digit of resolution.

Looking at the mainboard we see a rather full board.
The large filter capacitor is a Uptron branded capacitor. The only information I can find is that this was an Indian firm that went out of business in 2002. The bulk of the electrolytic caps are Indian branded Websen produced by license from Rubycon. The non elco caps are made by Desai Electronics PVT.

Taking a look at the back half of the mainboard reveals an interesting dip-8 ic. The icm755 is a high performance 555 timer. I think this is for the pulse mode. The intersil 7103 is a 4. digit a/d
The gold lidded cerdip is another 4.5 digit chip, Not sure what that one is for. I have only had time to do some basic testing/ teardown of the unit.

Looking at the front of the unit, we are greeted by my LEAST favorite type of switch...
The dip-16 by the input is a ICL7652 Az opamp. At least that is really nice.  The red resistors are wirewound resistors by the indian firm PEC, datasheet here: http://peccomponents.com/psp.htm
Every range has its own current set by the resistors.
I will add some more tomorrow.

[Vgkid]

Now that the unit is home, I will do a quick test before going to bed. Milliohms will be tested with the L&N 4385, Ohms with the ESI-655.
Milliohm     Recorded     Error
2000            1999.5           +0.025
1000            1000.2           -0.02
500               500.0              0.00
100                99.98             0.02
050                 49.99            0.02
020              19.986             0.07
010              9.997               0.03
Ohms          Recorded        Error
2000             1999.7            0.015
200                199.21            0.395
20                    19.104           -5.47
Unfortunately I'm not sure if there is a problem with the meter, decade resistor, wiring, or everything.
Will investigate later.

[Vgkid]

Well when I was putting everything away last night, the leads were still hooked up, I noticed that the readings became rather jumpy, I assumed it was the connection. Well I quickly discovered that the conection from the coaxial resistor to the back soldered connection was flakey, even coming loose.
Grab the soldering iron, and 1 hour later, it works. Ok that hour was exaggerated, but it took me longer than I wanted to.
Now back to testing.
Ohms      3456A      Recorded      Error
1990      1989.9      1990.9         -0.05
199      199.03      199.20         -0.08
19      19.026      19.054         -0.14
For fun I hooked up my ESI SR1,1k
1000      1000.06   1000.9         .084
Something is definitely wrong with the ohms ranges. I will update this in a few.

[Vgkid]

Haven't started troubleshooting it yet, but had a few musings. I ordered the 3205A manual off epay, BL wants $95.00 for it, that is the price of the meter...
I had a few musings over the source of the error. The 2000mOhms, and the 20 Ohm ranges share the same current source.At that current level, if the measured voltage is 2v, there must be another gain 100 amplifier that is off, as there is only 200mV(max) across any resistor.  I will fire up the 3456a, and the 87, and test some rails/voltages.   Mainly this is a placeholder text, and ramblings.

[Vgkid]

Testing resumes
3456a      3205b         voltage
19.036      19.054         .19053            
199.04      199.20         .199183
1989.9      1991.0         .199087
Checking the highest milliohm range
1.0288      1028.8         .010294      
It looks like the voltage sensing portion is ok, but the current generation is out. If this used the voltage reference from the A2D then it is fine.   
So I went and measured the reference resistors in circuit.
1-.9923
10-9.9236
100-99.2468
1000-992.572
10000-9.92694
Mini update, messed around with the 10k range setting resistor. Desoldered it, and its parallel higher valued resistor. Hooked up the DB-655, and noticed that the output reading was 40 counts higher than it should be, Lets dial back the resistance, ended up with the decade set to 9,936K when the reading read 1000 .0Ohms, perfect, now to repeat this later, with the other ranges.

[Vgkid]

Getting back to this, I’m still on the fence about keeping the(current) 10k resistor, and adding the external trim resistor network1.74M + 20k trimpot. Or cobble together a better 10K and a string of better resistors+ trimpot.
Back to the unit…
Since the 20Ohm range was out, yet that range uses the same current that is used in the 2 Milliohm range. Since that was good, the current source is OK. I will go into the current range trim later.

Milliohms  Voltage adjust:
Set meter to 19.0 milliOhm range.
Attach leads to calibrator with output shorted/off shorted.
Adjust offset trim pot to read Zero(or as close to zero as possible).
Apply 19mV with a stable voltage source. I used the MV-106 in conjunction with the 3456A.

Use R16(trim pot) and turn it until the display reads correctly.
Set meter into Ohms range

Ohms Voltage adjust:
Disable the output to the calibrator/short the leads.
Adjust offset trim pot to read zero.
Set calibrator to read 190mV.

Use R6(trim pot) to adjust reading to be correct.
Ohm Range Current calibration.
This is pretty straight forward.
Any Range:
Attach Standard resistor(or resistor of known value) I only felt comfortable calibrating the ranges above 2000mOhms.
Null out emf voltages with offset trim pot.
Depress pulse switch.
Put meter into proper range.
Press Measure button
Repeat as necessary, adjusting appropriate trim pot on range resistor. For convenience I left the meter in continuous mode, then switched to pulse as I got closer.
Continue for every range.
Getting back to the meter…

On the display board I noticed the name Motwane, and didn’t think anything of it. I figured it was a pcb company.
Upon further disassembly I noticed the same name on the pcb’s and case. This struck me as odd.
Researching this, it is an Indian firm, still in business. That produces related T&M gear. This explains the large number of Indian companies, as parts manufacturers.
I’m now curious if Ballantine Labs designed the meter, and it was made by Motwane. Final assembly was by Ballentine labs, or what.
Questions/Comments welcome.

[don.r]

Guess what I just picked up? Expect some incoming questions. 

[Vgkid]

Cool.
My unit is in storage, the manual is in a box somewhere (3205a, close enough. They calibrate the same way), but the 3205b schematic is on my computer .
Take some pictures of it.
You are collecting milliohm meters     

[don.r]

Its in fair condition in need of some glue and a good cleaning but one of the binding posts is disintegrating so I'm thinking of putting in shrouded plug jacks. First, however, I have to get it to make sensible readings as it looks somewhat out. Pics soon.

and, yes, it does appear to be that I'm collecting these....    ...but at $80 I could not pass it up.

[don.r]

I already see some major differences. My A/D chip is an ICL7135CPI and there is no gold covered chip. My display board has no part number, no to-18 only to-92, and R116 on the left is not populated.

[don.r]

Here is some pics

[Vgkid]

Yours is a bit different than mine. Have you had a chance to function test it?

[don.r]

Only a cursory glance. It measures resistance when none is present and seems to vary by range but given how dirty it was I need to clean the switches thoroughly before looking into it further. Front label is looking sad. Case has some serious yellowing so I might use it as a test of my hydrogen peroxide skills.

[Kleinstein]

The more modern ICL7135 combines most of the functions of the older 2 chip set (ICL8068+ICL71C03).

[Shock]

Oh I must have missed this thread. I have a LOM510A as well, and you don't want to know how long I had to wait patiently to get one. Robrenz and Highvoltage have one each also unless things have changed or I'm missing someone.

That Ballantine 3205B has a Fluke or Keithley vibe to it.

[don.r]

That Ballantine 3205B has a Fluke or Keithley vibe to it.

The Ballantine screams Keithley with the red LED display and rectangular bush button front. The handle even fits the Keithley 485 that I'm working on atm perfectly and has a similar Keithley brown tone to it.

[don.r]

Finally had time to revisit this. After cleaning the switches ... it works like a champ! Rock solid and accurate on all ranges with a nice bright LED display. In mine the cap on the far right of the display board is a tantalum so that will have to go. Other than that its a case of case restoration and some new shrouded sockets to be retrofitted.

[Shock]

Finally had time to revisit this. After cleaning the switches ... it works like a champ! Rock solid and accurate on all ranges with a nice bright LED display. In mine the cap on the far right of the display board is a tantalum so that will have to go. Other than that its a case of case restoration and some new shrouded sockets to be retrofitted.

C9 looks Tantalum where is the far right one?

[don.r]

Finally had time to revisit this. After cleaning the switches ... it works like a champ! Rock solid and accurate on all ranges with a nice bright LED display. In mine the cap on the far right of the display board is a tantalum so that will have to go. Other than that its a case of case restoration and some new shrouded sockets to be retrofitted.

C9 looks Tantalum where is the far right one?

Good catch. C9 needs to go too. The cap in question is the large blue electrolytic on the far right in Vgkid's third picture. Mine is a tant.

[Vgkid]

Finally had time to revisit this. After cleaning the switches ... it works like a champ! Rock solid and accurate on all ranges with a nice bright LED display. In mine the cap on the far right of the display board is a tantalum so that will have to go. Other than that its a case of case restoration and some new shrouded sockets to be retrofitted.

Good job. My case is yellow, unfortunately epoxy was spilled on the top case.

[don.r]

Finally had time to revisit this. After cleaning the switches ... it works like a champ! Rock solid and accurate on all ranges with a nice bright LED display. In mine the cap on the far right of the display board is a tantalum so that will have to go. Other than that its a case of case restoration and some new shrouded sockets to be retrofitted.

Good job. My case is yellow, unfortunately epoxy was spilled on the top case.

Sometimes paint thinner will soften epoxy. Test on an inconspicuous area first as it might affect the plastic too. My case isn't in the best of shape which is why I though about using it as a tester for H2O2 restoration.

[Vgkid]

Necro bumping this to attach the schematic.