Micron Circuits

[AndyS]

One more:

[Dodgy Geezer]

Andy,
That seems to cover set-up, which will be useful when (if) I get this working. Thanks!
The encider circuit description however mentions the 4017, so it's a later version than the one I have. Lets see what some powering up will do - though I'll have to wait until my wife lets me get out to the shed!!!

[Dodgy Geezer]

Well, I attached 4 pots as specified on the Norcim site and gave the encoder 5 volts. Couldn't find a pulse stream anywhere, even around the multivibrator...

[Pchristy]

Its almost impossible to diagnose by remote, and without knowing what test equipment you have. The encoder should run at 5V, thought it is almost certainly designed for 9.8V (8-cell NiXX pack).

First thing to check is current draw. Is it drawing anything at all? Do you have access to a 'scope?

[Dodgy Geezer]

Old Gould OS4020 DSO. Probably quite useful if I could understand the storage bit... Most useful thing would be drawing up the schematic...

[Pchristy]

Re: 'scope. Never mind the storage bit, or even if its calibrated accurately. All you need it for is to check for pulses.

Fault finding rule No1: Always check the power supply first! I see you are using a 5V supply. It should run on that, but it would have been designed for (probably) 10V. Check the supply voltage under load - ie: with the encoder connected. Does the voltage sag? Next check the current draw - should be a few mA.

Next step: Check the voltages around the transistors. One end of the stick pots should be wired to the +ve supply (I'm assuming its using NPN transistors, which would be the norm). Are the stick pots getting supply volts on one end? The transistors are normally "on", and turn off to produce the pulse. The voltage on the collector (other end of the pot track) should be zero, or nearly so. Voltage on the wiper should be around 1/3 of the supply voltage.

Next check with the 'scope, ideally set to DC in. Move the trace down to the bottom of the screen, and adjust the gain so that supply volts takes the trace to near the top. Set the rate (timebase) to 2 to 5 mS per cm. (That is so that you will see at least one whole frame of pulses, if the encoder is running). Set the trigger to "auto" or "line". The "line" setting will use mains frequency to trigger the 'scope, so you will get a display whether there are pulses present or not. If they are present, you should see them moving slowly across the screen. Check the collector of one of the multivibrator transistors. Easiest is the Ch1 stick pot end connected to Q2. You should see a square wave with a period of around 20 mS.

If you have this square wave, next check the collector of Q3, where you should see the Ch1 pulse.

The mutivibrator needs to be running for the whole thing to work. If it isn't, check voltages around Q1 and Q2. The collector of one will probably be low, and the other high. It is equally possible that both will be low or high, but in any of these cases, you need to check the PCB for short circuits (bridged tracks, or stray solder whiskers). Also check the component values, particularly the resistors. Collector resistors will typically be 4.7K (to match the stick pots), and the base resistors will be somewhere between 47K and 470K, depending on the capacitor values the designer has chosen.

That should keep you busy for a while!

[Dodgy Geezer]

Indeed it will! As it seems to have done for you - you have nearly written a book on the topic. Thanks particularly for taking the trouble to spec the scope settings - it's an old one of my father's, and I could never follow what all the knobs and buttons did.

I should see if the MV transistors pass a diode test as well I guess? Everything looks newly soldered, but it may still be faulty.

[Dodgy Geezer]

Voltage appeas on the rails when I plug things in. Circuit shows no mA drain or voltage sag. Transistors pass diode test. One end of stick pot is positive. I will go and check the rest. In the meantime, here is a rough schematic of the MV stage. Caps C2 and C3 are incorrectly marked at 047, when they should be .01... and I can't spell linear....

[Dodgy Geezer]

Some success!

The pic shows a couple of negative pulses from a 5v line along a 20 ms frame. But it ought to show 5...

I think I had a poor ground connection initially, coupled with poor pots. I bought a lightweight set for making into joysticks from China a little while ago, and I suspect thet the slider doesn't always connect...

With any luck now it's simply a job of chasing the signal through the components... Thanks for all the help!

[Pchristy]

Excellent! You are getting somewhere!

You should now be able to set the trigger on the 'scope to internal, auto and negative going. (Internal and auto may be on the same switch, if so just go with auto)

That will get the scope to trigger on the first negative edge it sees. You can now speed up the timebase to around 1 or 2 mS per cm. You should then be able to measure the width of the pulses, and see the space between them moving as you move the pot.

It looks like you are picking up the output from the encoder, so the negative going pulses should be around 300uS wide, and the space between them should vary with the stick pot. When finally aligned, that should be between 1 and 2 mS, at around a 20mS frame rate.

Now all you have to do is figure out what is stopping the other channels! If you are lucky, it will just be Ch2 that is faulty, because if that doesn't fire, neither will any of those after it.

Cheers,

--
Pete

PS: One other thing! All these early encoders were designed to use pots with 270 degrees (or so) of travel - of which we only use 90 degrees. It is possible you have got 90 degree pots, as used in some more modern radios. These may not work properly with a half-shot encoder without significant modification!