Schematic for Heathkit/SEI RM-4 radiation alert? / Electromagnetic Radiation / Forums

Forums

4hv.org :: Forums :: Electromagnetic Radiation

« Previous topic | Next topic »

Schematic for Heathkit/SEI RM-4 radiation alert?

Move Thread

LAN_403

klugesmith

Sun Dec 13 2009, 09:18AM

Registered Member #2099 Joined: Wed Apr 29 2009, 12:22AM
Location: Los Altos, California
Posts: 1716

Just picked up my first radiation measuring instrument, an original Heathkit RM-4 "geiger counter" meter. I had fun with a borrowed specimen 13 years ago, when my cat was temporarily radioactive (after treatment for hyperthyroidism).

Unfortunately, it has a problem. The flashing LED and audible beeps/clicks work fine, but the analog meter pointer does not move except in battery-test mode. Probably easy to fix, but would be much easier if I could find a schematic. Downloaded a schematic-free users manual from the company which took over the product from Heathkit:

.

I am not eager to pay some online dealer $10 to $20 for a scanned copy of the original Heathkit instructions. I doubt that money goes to cover copyright royalty payments. Can anyone here provide a schematic for less?

Then: Calibration step 1. When analog meter is functional, we will see what actual event rate brings the pointer to full scale, which is labeled 500CPM and 0.5 mR/hr.
The manual says: "MONITOR 4: Halogen-quenched uncompensated GM tube. Thin mica window is 1.5-2.0 mg/cm2 thick. Approx. 1,000 CPM/mR/hr for Cesium 137. "
and "Detects gamma and x-rays down to 10 keV typical through the window".
There is a chart showing spectral sensitivity wrt CS-137 reference, but only for sidewall exposure. It shows a strong peak at 60 keV, substantially suppressed in the compensated version with 2mm tin filter.

Proud Mary

Sun Dec 13 2009, 10:44AM

Registered Member #543 Joined: Tue Feb 20 2007, 04:26PM
Location: UK
Posts: 4992

Hi Rich,

Routine GM tubes have little value in X-ray work, since they can detect about 1% of the photons impinging upon them, but not measure them.

The energy response of uncompensated tubes always peaks around 60-70keV, so a small increase in kVp can give a large increase in counts on the forward side of the peak. There are also two points on the graph where the count rate will be the same - one on the forward slope, and one on the reverse.

Economy GM counters with analogue display usually use capacitor integration to drive the meter, which means that at low radiation levels the needle will kick now and again. They tend not to have dead-time compensation, or the means to prevent pulse pile-up, such that at high dosage levels found in the beam - say 3 - 6 Gy/hr at `100cm from a dental tube @ 75kVp/1 - 2mA - the readings will be totally meaningless.

1260700994 543 FT80936 Gm Energy Response 2

klugesmith

Sun Dec 13 2009, 12:06PM

Registered Member #2099 Joined: Wed Apr 29 2009, 12:22AM
Location: Los Altos, California
Posts: 1716

Proud Mary wrote ...
Routine GM tubes have little value in X-ray work, since they can detect about 1% of the photons impinging upon them, but not measure them.

Understood & respected, Stella.
I have never expected this unit to be quantitative for x-rays. Still it would be nice if I can fix the pulse-averaging meter circuit, without having to reverse-engineer part of the 1-layer PCB. So still hoping someone offers a schematic.

It can be part of a safety ensemble, having been shown to be sensitive to low energy x-rays. Tickling the dragon's tail, I got this meter to buzz at a rate comparable to point-blank measurement of a smoke detector source or a permissibly radioactive cat. Did that with a plain glass x-ray tube, half of a common NST secondary, and less than 1 microamp of anode current. So lack of buzzing, when meter is with operator and tube is driven harder, will be -a- sign that the shielding and distance are working.

That exercise also serves as a safety reminder. With only 7.5 kV (RMS) and a relatively minuscule current, the x-ray intensity next to the tube has reached a level where permissible exposure time is a consideration.

Proud Mary

Sun Dec 13 2009, 12:56PM

Registered Member #543 Joined: Tue Feb 20 2007, 04:26PM
Location: UK
Posts: 4992

Check for the presence of HT using a high-impedance meter.

If HT is present, inject a pulse train into the signal take off from the tube. Bad designs usually take this from the anode, and good designs from the cathode. If the meter starts counting the pulse train, you will know that the tube is most likely unserviceable.

It would help if you could give the tube number, so we can look up its plateau voltage, which must appear somewhere in the circuit plus/minus about 10-15% for the tube to function.

If you have an oscilloscope, look for the presence of a regular waveform, be it pulses, square waves, or sine, that one would expect to see as the basis for the voltage generator.

Remember that the HT to the tube is at very high impedance, and may only be able to source a few hundred uA - or less - such that shunting by a typical 10M multimeter will pull the voltage right down. In this case, you can use a high value resistor to charge a capacitor of 100nF or so, and then measure the voltage on the capacitor.

Appendix added later: it may be that I have misunderstood your post, and you meant that the tube is responding, and that the LED blinks,and the beeper beeps with each Townsend avalanche in the tube, but this is not shown on the meter.

If this be such, and your instrument has switched counting ranges, (which generally add more C to extend the time constant of the integrator) then you should inspect the switch and its wiring as being likely culprits.

If you have a function generator, disconnect the tube, and find the point in the circuit which actuates both the LED and the beeper in response to a low voltage pulse.* The meter feed to the integrator must also come from this point, so it's just a question of tracing the circuit between this take off node and the meter itself.

* set function generator o/p volts after inspection to see whether CMOS or TTL level is required.

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob