North Hills CS-152 Current calibrator

[eas]

Question: Is there a way to reduce settling time/ repeatability?
Meaning that this unit is +\-  50 ppm after 5 minutes but takes half an hour to be in the +/- 5 ppm range.

Short the power switch?

[Vgkid]

I could just switch the polarity switch to zero, and get a more stable zero...
 

[Vgkid]

All hope is not lost, just needed a more stable current shunt, currently performing the stability tests on the 100mA range.
now to find a higher power precision 10 Ohm resistor.

[Vgkid]

The 24 hour 100mA stability test are in, a 80 ppm Peak to peak difference...
I know you are thinking that is horrible, but...
99.9993 max, 99.9913min, 99.9954avg with a 00.007315E-9 variance.
Considering the about 7-8deg C temp swing over that 24 hours, not bad at all.
**I will assume my shunts tempco is around 10ppm/C, don't have a manual.

[Vgkid]

Necro update.
I was running this over 48 hours, and it can either be semi stable 80ppmp-p, or really stable 5ppm over 18 hours. Both are set tk 1mA output. Im not sure if the resistor is really playing into this, or if the oven is not the best. I will try insulating it somewhat, from the controller board.

[pigrew]

I ended up purchasing a similar unit on eBay. It's circa 1965, and needs repair. It has a calibration sticker from 1985. I'll post more exhaustive photos the next time I do some disassembly. The major differences I saw is that my cabling is tied together with twine, whereas the newer unit used zip-ties. Also, the clips to hold the cables on the sides of the unit were clear plastic that disintegrated.

Every board is shielded which two or three metal plates. To disassemble the unit, one must first take the top, sides, and back off the case, and then unbolt the modules from the bottom. Then, the module will lift up, and access is provided to the screws on the sides of the module. Most of the module have shielded sub-modules, too.

The shipping weight of the unit was 54 lb.

As mentioned, the unit needs repair. With the switch off, the ovens (both reference oven and resistor ovens) heat up. The reference provides close to 10 V (I didn't accurately measure it). The ovens used about 60W of power. But, with the power turned on, there's a loud 60 Hz hum. My guess is that some (or all) of the capacitors are leaky.

I'm thinking about attacking the problem one module at a time. I'll disassemble the module, catalogue the components and maybe make schematics, order replacement capacitors, and perhaps test the module if I can figure out what it does. During the process, I'll post photos. I hope that nothing will be damaged if I power it up without modules in place....

I did find one small vacuum tube (shielded, of course). Everything was much more transistorized than expected.

A photo of board C is attached. I'm going to try to find film capacitors and aluminium electrolytics to replace the caps.

I recognize most of the components, but what is the yellow epoxy coated component in the lower right? 1.250k resistor? Capacitor?

-Nathan

[Vgkid]

Looks like a film capacitor. 
What is the module, I might be able to help with the manual, that I have.

[tautech]

Looks like a film capacitor. 

I think so too.


1000pF 250V
Measure to be sure.

[wn1fju]

I bought one of these a few months ago.  Mine is labeled as a CS152-7.  Does anybody know what the -7 means?  My unit was in reasonably good shape when received, but a bit noisy.  The problem was a dirty front panel output switch that defied my attempts at cleaning.  I ended up replacing the 2-pole, 3-throw rotary switch with a 4-pole, 3-throw switch, doubling up on the contacts for extra reliability.  Another question:  The two neon bulbs on the front panel for the heaters don't ever seem to go off.  One of them will flicker a bit eventually, but the other one is solidly on.  Is my unit going to burn itself up?

[Vgkid]

That is normal behavior of those lights.

[pigrew]

That is normal behavior of those lights.

My unit has test points to measure the 10 reference oven heater current (that and the 10 V reference are the only labeled testpoints in the entire unit. The voltage across the test points at startup was about 1V. Unless I should be measuring using an ammeter instead, I'd expect the voltage to decrease as it reaches steady state.

Sent from my LG-D850 using Tapatalk

[Vgkid]

I ended up replacing the 2-pole, 3-throw rotary switch with a 4-pole, 3-throw switch, doubling up on the contacts for extra reliability.

That switch caused issues with my unit as well, mine was partially cleanable though. Still need to turn it a bunch before use.
No idea on the "-7" It looks like North Hills abandoned their T&M division a long time ago...
They made some interesting products.

[pigrew]

I'm attaching photos of the chassis of my unit (SN 44D, or maybe 440).

I'm really not sure how to go about working on the unit, other than trying to reverse engineer each board. Two boards seemed to have been burnt, one of which was reworked. The back right module has been replaced (Replacement was from SN11).

I'm guessing that the yellow poly film capacitors should still be good, and that I should replace all the electrolytic capacitors, as a first step?

As I investigate the individual modules, I'll draw out some schematics. Can anyone identify the function of the various modules?

On the left side, there are 6-pin connectors going to each module. Are these power connectors? And signals are the cables on the right side?

[Vgkid]

Date wise my manual should be really close your version, but it doesn't seem that way.
The serial number on my unit is 376R
module wise:
CO-10 : Resistance module. A heated box containng the current setting resistors.  Values are 60,100,1K,10K,100k.
RS-1 : Contains the 10V reference in its own heated enclosure. Yours is quite a bit different from mine. The internal reference(solidcell) contains an ovenized zener reference outputting slightly over 10v(10.5 in your case, most likely 2 "5.6v" zeners in series ) This is divided down by that resistor network(The binary one+trimmer) to 10V.
TR-1 = Contains a +/_ 20v power supply, chopper amplifier, and is used to produce currents up to 10mA
TR-2 = used to boost current up to 150mA, contains +/_ 41v suplies.
Front most can shields the KVD.
If you want to tear down the modules more, please do.
To actually remove the modules from the chassis, flip the unit over, and unscrew them. It will make sense.

[tautech]

I'm guessing that the yellow poly film capacitors should still be good, and that I should replace all the electrolytic capacitors, as a first step?

It's very unlikely the film caps will be crook, focus on the e caps.

[pigrew]

Attached is the power supply and the adjustment network for the 10V reference (probably the simplest circuit in the calibrator).

The reference module (and all of my modules) have their own transformer, and all the modules are fed the 120V power line (without any filtering). The 10V reference module has a tap for the Zener reference, at about 33 VAC. This is rectified, filtered, and regulated to be 29.5 V. It uses some General Electric A13D2 rectifier diodes (can't find a datasheet), and three 9.5 V zener diodes in series (only marked with a date code). It is filtered using a 2N2035, and a second unknown transistor (which seems to also be a NPN BJT).

The reference adjustment board is fed with that 29.5V.  The "I monitor" ended up being the voltage across a 55 ohm precision WW sense resistor in series with the zener reference. At startup, it's about 0.98 V, but decreases to about 0.92 V over half an hour (as the oven warms, or maybe the power supply voltage settles?). This means that the Zener ends up having about 16 mA through it. The big pot controls the current. My zener produces about 10.3V. I don't know what current is best for the Zener. I wonder if it should be adjusted so that IMonitor has 1 volt (a nice number), or maybe so that the 10.5V node is 10.5V? For now, I won't adjust it.

Next is an adjustable voltage divider. The board has reconfigurable jumpers to bridge various series resistors. The printed values are the resistances of wire-wound precision resistors. Then, there is a trim-post which adjusts the next series element. I don't know the resistances of those resistors (I have not figured out their labeling); but the pot is labeled as a 500 ohm pot. I measured 178 ohm across these elements.

The other side of the divider is the polarity knob on the front panel. The polarity knob connects the 10V to the KVD when in the + or - setting. But, it connects to a non-precision 10k resistor when in the "0" mode! So, the 10V reference must be adjusted when the polarity switch is NOT in the 0 mode. (Or maybe the reference shouldn't be adjusted to output 10 V, and only adjusted such that the output current is correct? I only have found three trim pots in the whole unit.).

The reference oven controller will be figured out another day. It's proving really weird (and has a sealed module on it, which perhaps is an amplifier or SCR?).

[pigrew]

And the reference oven controller PCB. I hate diode color-bands, and still need to figure out what that sealed black thing is. There's evidence of the 5W resistor overheating, and someone did some rework on the back of the board there. It still runs quite hot. Schematics are in progress, though I can't identify a few of the components. The board is fed with 120VAC.

[Vgkid]

Nice job    You went a lot further than I did on my first try...
--------------
I made it to the actual voltage reference. Holy shit, that is impressive.

[SeanB]

Black block would likely be an opamp, and there are some seriously overrun resistors there on the right of that board, to have charred it like that.  I would replace all those resistors in that area, and the Mallory capacitors as well, and uprate the power ratings of the replacements, so they run cooler. Replace the grey ceramic one with a 10W wire wound one to get a larger cooler surface, and the carbon composition 1.5k 1W one with a 3W one, and the other 2 carbon composition ones with 1W metal film ones, provided you can fit them in the space.

[Vgkid]

I would do what SeanB says. On my board there are 2 5W wirewound resistors mounted on a heatsink. A 3.3k,and a 3.5k. I didn't check how they were wired up  . But I can check on the resistance board. No sign of excess heat on those boards 

[pigrew]

I ran the board overnight, and the hysteresis of the controller is AWFUL. Maybe the op-amp is dead/dynig, or something like that. The reference voltage has about 5 mV variation with oven temperature. I'm wondering if I should just replace the whole board with the newer-style board. Mostly-complete schematic attached.

I _think_ that it wants the thermal sensor to output 15V, but I'm not sure. Unijunction transistors are weird.

[pigrew]

I ran the board overnight, and the hysteresis of the controller is AWFUL. Maybe the op-amp is dead/dynig, or something like that. The reference voltage has about 5 mV variation with oven temperature. I'm wondering if I should just replace the whole board with the newer-style board. Mostly-complete schematic attached.

I _think_ that it wants the thermal sensor to output 15V, but I'm not sure. Unijunction transistors are weird.

I was measuring the voltage across the white wires going to the sensor. Assuming I drew the schematic correctly, it it should be held at 15V... Which it wasn't at all. I'm also not so sure about the use of the 15V rail, as opposed to the 100k/100k divider, as an input to the amplifier. Maybe time to hack in a LM301, like the newer board has, or just design a new board.

Sent from my LG-D850 using Tapatalk

[SeanB]

Mistake in your hand drawn schematic, there is a 22k resistor in series with the one diode connected to the mains. As well there is some errors in the schematic as drawn, especially the clamp to the 15V rail by the UJT. Very likely your UJT is dead or has suffered massive parameter shift, the regulation is basically bang bang, as the opamp saturates. It should be doing pulse width modulation of the heater temperature as it gets to setpoint, and not oscillating like that.

However first change out all those carbon composition resistors, and for the 1k5 ones on the mains side use 10W versions as they will run cooler, but if you cannot fit them in use at least one in the first position. Replace the 14uF 125V capacitor with a 22uF 105C 200V version, as it will need the high temperature rating where it is. Replace the 15V zener diodes with 3W ones, and then try again, it should, if the UJT is still functional, work better at regulation. Change out the input 100k and 4M7 resistors as well, and that 10n ceramic cap, using 1% resistors and a 10n 100V film capacitor there.  An upgrade will be to add a 100n 100V film capacitor across the 30V rail, will give some better noise immunity.

[Vgkid]

Here is the RS1 Oven controller for my unit.

Here are the load resistors.

I can get more pics of the oven controller If you want. I will be putting it back together later today /maybe Monday.

[pigrew]

However first change out all those carbon composition resistors, and for the 1k5 ones on the mains side use 10W versions as they will run cooler, but if you cannot fit them in use at least one in the first position. Replace the 14uF 125V capacitor with a 22uF 105C 200V version, as it will need the high temperature rating where it is. Replace the 15V zener diodes with 3W ones, and then try again, it should, if the UJT is still functional, work better at regulation. Change out the input 100k and 4M7 resistors as well, and that 10n ceramic cap, using 1% resistors and a 10n 100V film capacitor there.  An upgrade will be to add a 100n 100V film capacitor across the 30V rail, will give some better noise immunity.

Thanks for the suggestions. I just pulled the power supply resistors off of the board, and measured them. The smaller 1.5k was actually 980 ohm, and the 100 ohm was actually 130 ohm. I'll put together an order for replacements.

For the 15 V zeners, It looks like they're 1N418A. 1N418 is 250mW, 15V@12mA, Z_ZT=13ohm. I'll replace them with 1n4744A which are 1W, 15V@17mA, Z_ZT=14ohm. The board does look slightly toasty under them, too.

The other two diodes (one transparent glass, the other not) are confusing me since they have the same bands (yellow,green,violet,brown). 1N4571 is a 6.3V temperature compensated Zener, whereas 1N457A is a 60V small signal diode. Maybe I should pull them off the board, and measure if they're Zeners or not.... I'd guess that the upper one is a zener, since it looks like the other Zeners, and the lower one is a small signal diode?

Once I figure out what the components are, I'll post the corrected schematic.