General Radio Random Noise Generator 1390-A

[Excavatoree]

I'm not sure if this can go in the "test equipment" category, it might be better suited for "repair."

My favorite surplus place is going out of business.  It seems the owner's land is worth more than he can get from the business, and at 77,he's ready to retire.  I had to purchase this from the person who bought much of his inventory.  I've considered buying it many times, but always declined.  Well, given "one more chance" I took it.

This is a General Radio random noise generator.  Made in the <Dave Voice>  United States of America.    I know absolutely nothing about it.  If Dave and the moderators don't object, I'll be posting some pictures.

I have not applied power, as there are some potential problems I want to address first.

[Excavatoree]

Inside pictures:

I'm not concerned about the large, vertical capacitors, unless someone believes I should be.

[Excavatoree]

I've never seen this before.  A valve/tube mounted in a magnet.  "Keep Steel Away"

[Excavatoree]

I'm a bit concerned about the capacitors.  The first capacitor obviously needs to be replaced.  The second one, the paper Mallory, I'm not sure about, but I believe I should replace it.  I'm most unsure about the two, black, cylindrical capacitors.   They aren't leaking, but I've read that they should be replaced.  Does anyone have any advice?

[Excavatoree]

Finally, I am also concerned about the selenium rectifiers.  It seems to be in good shape, but I know they've been problematic.   There are two, but I only tool pictures of one, as they are both alike.


You don't see much with the GE logo that's made in the US anymore.

[T3sl4co1l]

6D4, I believe?

You've got the cheap version it would seem -- glow discharges have some auto-correlation, which can be removed by using two independent sources, inverting one, and summing them.  Reason being, like zener/avalanche diode noise (at low currents), the waveform is rampy, jumping down randomly due to avalanche discharges and rising slowly with bias current until the next event.

I have a one of these laying around,









looks to be 60s vintage, hybrid construction.  I think most of the transistors are Ge mesa or Si types.  I think it boasts 3 sigma performance, which isn't too bad as Gaussian noise goes (for the bandwidth that it has, you'd probably be waiting hours or days for more sigmas, if it could produce them).

Tim

[Excavatoree]

Very nice.

I'm surprised that a small liberal arts college has (had?) a physics department. 

[T3sl4co1l]

I'm a bit concerned about the capacitors.  The first capacitor obviously needs to be replaced.  The second one, the paper Mallory, I'm not sure about, but I believe I should replace it.  I'm most unsure about the two, black, cylindrical capacitors.   They aren't leaking, but I've read that they should be replaced.  Does anyone have any advice?

Yeah, probably better off replacing them all.  The "Black Beauties" can go leaky or shorted over time.  The case cracks, oil leaks out or moisture in; etc.  Paper-in-oil caps (non-hermetic) never last, they always pick up moisture.  Also beware "Micamold" types, if you find any; the larger values were often PIO, not mica.

Tim

[edavid]

I've never seen this before.  A valve/tube mounted in a magnet.  "Keep Steel Away"

"The Type 1390-B Random-Noise Generator uses a gas-discharge tube as its noise source. A transverse magnetic field is applied to the tube to eliminate the oscillations usually associated with a gas discharge and to increase the noise level at high frequencies."

http://www.ietlabs.com/pdf/Manuals/GR/1390%20Noise%20Generator.pdf

[Excavatoree]

"The Type 1390-B Random-Noise Generator uses a gas-discharge tube as its noise source. A transverse magnetic field is applied to the tube to eliminate the oscillations usually associated with a gas discharge and to increase the noise level at high frequencies."

http://www.ietlabs.com/pdf/Manuals/GR/1390%20Noise%20Generator.pdf

Thank you very much. 

[andy2000]

You definitely need to replace all of those black cylindrical caps.  They are paper in oil and are almost always very leaky. 

The electrolytics (including that Mallory in the paper tube) may well be fine, but they are over 50 years old, so it wouldn't be a bad idea to replace them.  It all depends on how much time you want to put into it now to avoid future problems as those caps eventually fail.  Personally, I usually check them for ESR and make sure none are getting hot from leakage current and hope for the best.  Unless you're going to use it a lot, it's probably not work replacing every single cap.

[wn1fju]

I've got the GenRad 1390B which is based on the 6D4 vacuum tube.  I didn't find the "Gaussian" noise to be particularly white when looking on a spectrum analyzer.  I tried a couple of replacement 6D4s to no avail.  Oh well, I guess they don't make them like they used to - which, in this case, is a good thing!

[T3sl4co1l]

The datasheet indicates bandwidth is not flat: http://www.mif.pg.gda.pl/homepages/frank/sheets/137/6/6D4.pdf

There should be amplification and equalization to make it look "whiter".  But given the small amount of stuff inside the chassis, perhaps they skipped this.  They do at least provide a voltage scale.

Note that the frequency response of a Gaussian distribution over time need not be white, or pink or even lumpy; all that matters is that the likelihood of the voltage being some level at any given point in time be Gaussian.  There will be a strong correlation over short time scales, due to the limited bandwidth (it simply can't change more than some fraction of a sigma within so-and-so nanoseconds), but other than that, it should be pretty good.

Tim

[joeqsmith]

I'm not sure if this can go in the "test equipment" category, it might be better suited for "repair."

This is a General Radio random noise generator.  Made in the <Dave Voice>  United States of America.    I know absolutely nothing about it. 

I like it! 

[SeanB]

Selenium rectifiers can be replaced with a 1N4007 and a series 100R 5W resistor in most cases. If in good condition just leave them, I did so an an old battery charger ( GE made circa 1947) until I replaced it with a newer 20A bridge rectifier and a charge controller (thyristor charge control) so I could do SLA batteries with it.