Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length.

[dodgy geezer]

I picked up what I thought was an old Micron combo the other day, and found that it was a part-finished kit. The RF board and the encoder board look fairly complete, but I would like to check them before I put power through.


The problem is that they look to be very early examples. The encoder board, for instance, does not have a 4017 IC chip. I have circuit diagrams for later Microns, but nothing that early. If anyone has a circuit diagram for a Micron encoder board without a 4017 chip, or an RF board with no Toko inductors, I would appreciate a copy....

[HMS Invisible]

Some descent photos will suffice. I'm familiar enough with that era of radio.
All the emitters of npn transistors are tied to 0volt and you don't have to worry so much about exceeding a voltage limit.

[dodgy geezer]

OK - if you want to have a go!


What turned up was a Micron receiver (looks like 27mHz AM, though no crystals), a Tx box assembled with sticks and unsoldered, an aerial which looks like a commercial radio receiver one (no Micron centre load), two PC boards populated and soldered apart from a few components which were threaded through and the wires bent back, and a box of oddments including a transformer, several resistors and a diode (parts for a charger?)


The RF board has +, - and signal wires attached, also an aerial-out lead (unscreened - later Microns were screened) pictures below. It has one PCB hole unpopulated. There are two copper sections on the board which look as if they are intended to take a wire connection.


The encoder board has no wires attached. It is populated with 4 'half-shot' transistors for four channels (no 4017 chip), with space for a further 3 channels. There is some electronics at one end that I do not recognise. I can work out the + and - lans, and the norcim site has a similar example encoder which I might be able to infer the stick pot connections from (though there are no obvious points on the lans to fix a wire...). There is one unpopulated hole in the electronics at one end, covered by a link.


I was planning to power up the RF board, connect to an aerial and see if I get any radiation.  Then I was going to wireup/power up the encoder and poke around with a scope until I found a stream of pulses. And then connect the two... But I could really do with a circuit diagram to see if there are any missing/incorrect components...

[HMS Invisible]

The extent of what I hoped to point out is what you seem to already understand.
As well as the half-shot transistors there is an astable timer. A recent thread about restoring a Digifleet transmitter showed one using inverting logic gates.
 The ring circuit is completed with a fixed or variable frame, and the output to the rf board will be obvious if you sketch one stage. Without referring to an old Tx or diagram, I think that will be npn collector 4 to 7.
 I'll look out an old set & search for the relevant thread about NiCds.

I could further add, which you may have guessed, that the polyester capacitors are the stable timing caps.

[HMS Invisible]

https://www.modelboatmayhem.co.uk/forum/index.php/topic,70532.msg764983.html#msg764983

[dodgy geezer]

The extent of what I hoped to point out is what you seem to already understand.
As well as the half-shot transistors there is an astable timer. A recent thread about restoring a Digifleet transmitter showed one using inverting logic gates.
 The ring circuit is completed with a fixed or variable frame, and the output to the rf board will be obvious if you sketch one stage. Without referring to an old Tx or diagram, I think that will be npn collector 4 to 7.
 I'll look out an old set & search for the relevant thread about NiCds.

I could further add, which you may have guessed, that the polyester capacitors are the stable timing caps.

Alas, I suspect that you think I understand more than I do!  I can see that there is a multivibrator feeding pulses into a series of transistors which will switch at variable times depending on the relevant pot setting.  But what happens to the signal after that I don't know - I suspect that the electronics I don't understand have something to do with that...  But, in the absence of any circuit diagram, does my proposed way ahead sound sensible to you?  I suspect that the original circuit worked off 9.6v nominal, and was proposing to use 7.2v NiAmh, since I don't have a bench power supply...

[HMS Invisible]

Disregard my earlier description as I was working from memory.
Fig 76 in scimitarjohn's post shows the collector outputs fed into 4001 logic gates but your much simpler circuit will probably use diode logic.
 The function is the same as Fig 76. Each one-shot triggers the next one and also triggers the r.f.
In your board it is possibly done via a diode. So r.f. input should be a copper rail with diodes connecting to each channel's transistor collector.
 You'll begin to see a similarity with Fig76 where your colour coded polyester caps equate to C_a and C is a cheap ceramic disc cap.

* I just had a closer look at the photo and can see the 4148 diodes.

[HMS Invisible]

Here is an RC circuit link you have possibly found. http://sm0vpo.altervista.org/use/rc_enc_02.htm
The 2 and 6 channel diagrams show the 'wired-OR logic' by common-cathode connected diodes. The output is buffered by a BC557 to make a positive pulse.

[dodgy geezer]

Disregard my earlier description as I was working from memory.
Fig 76 in scimitarjohn's post shows the collector outputs fed into 4001 logic gates but your much simpler circuit will probably use diode logic.
 The function is the same as Fig 76. Each one-shot triggers the next one and also triggers the r.f.
In your board it is possibly done via a diode. So r.f. input should be a copper rail with diodes connecting to each channel's transistor collector.
 You'll begin to see a similarity with Fig76 where your colour coded polyester caps equate to C_a and C is a cheap ceramic disc cap.

* I just had a closer look at the photo and can see the 4148 diodes.

I can see the diodes along the output rail - that rail leads into the transistor pair in the mystery electronics. A bit of smplification? 

[roycv]

Hi, I made up the transmitter and receiver way back.  But although 27Mhtz my set is FM not AM.  I also have a Futaba 27Mhtz FM Tx but use the Micron rx as it is far more stable and reliable.
Roy

[HMS Invisible]

I can see the diodes along the output rail - that rail leads into the transistor pair in the mystery electronics. A bit of smplification?

Three copper lands for each pot are adjacent to the holes.
The next task is to identify the output. I use a phone camera to read semiconductor numbers, but you can assume you have two normal  signal transistors then determine npn or pnp polarity from connection or by diode testing.
 If you place the pcb over a diffuse light source you should be able to sketch the encoder circuit and it should closely match the 6-channel one hosted on  altervista.
 I'd happily power up the encoder board but not the rf board. I'd look at the Rx components to see what band it was operating in.

[dodgy geezer]

Three copper lands for each pot are adjacent to the holes.
The next task is to identify the output. I use a phone camera to read semiconductor numbers, but you can assume you have two normal  signal transistors then determine npn or pnp polarity from connection or by diode testing.
 If you place the pcb over a diffuse light source you should be able to sketch the encoder circuit and it should closely match the 6-channel one hosted on  altervista.
 I'd happily power up the encoder board but not the rf board. I'd look at the Rx components to see what band it was operating in.

To hear is to obey!

I use a head-mounted loupe to read components - by now I need it for resistors as well as semiconductors!  The encoder section looks in reasonable nick, though I cannot tell whether the component values are correct.
I enclose a schematic of the 'mystery electronics'. It looks like what I think they call a 'darlington pair' to give the signal a bit more 'oomf', but you can see that the items in red do not make sense, and may be mistakes, or partial assemblies for some added circuitry...  I do not know whether the suporting resistors/caps make sense, or why there are two parallel caps on the right hand side, one an electrolytic. The rectangles are points on the copper lands which look as if they could take a connected wire..


Why not power up the RK board with a 27MHz Futaba crystal?  The Rx looks like a Micron 27Mhz, but it has no crystal, and I can't tell the band from the components. I assume that the RX board has not been tuned to oscillate, but there is only one adjustable coil, and a little fiddling should get it to radiate something which can be detected using a diode/multimeter...


P.S.  Just noticed that the 3906s are PNP, so the arrows on the emitters should be going the other way.... 

[HMS Invisible]

  That diagram makes perfect sense with the arrows changed to reflect 2N3906.
An Internet search of "capacitor frequency response" will turn up many  diagram hits of impedance Vs frequency graphs. The resonace point of electrolytic caps is possibly lower than you think. At high frequency a parallel connected ceramic capacitor has lower impedance and takes over decoupling. 
  Testing the Rx with a working 27MHz crystal is a logical task. I didn't dabble in rf circuit design enough to recognise what I'm looking at at a finger click. All I can say is the rf board has the wrong copper layout for 459MHz vhf, ferrite & crystal socket.
 I'm just overly methodical. You'll have a set that doesn't conform to the revised regulation for 27AM but you won't come to any harm!

[dodgy geezer]

  That diagram makes perfect sense with the arrows changed to reflect 2N3906.
An Internet search of "capacitor frequency response" will turn up many  diagram hits of impedance Vs frequency graphs. The resonace point of electrolytic caps is possibly lower than you think. At high frequency a parallel connected ceramic capacitor has lower impedance and takes over decoupling. 
  Testing the Rx with a working 27MHz crystal is a logical task. I didn't dabble in rf circuit design enough to recognise what I'm looking at at a finger click. All I can say is the rf board has the wrong copper layout for 459MHz vhf, ferrite & crystal socket.
 I'm just overly methodical. You'll have a set that doesn't conform to the revised regulation for 27AM but you won't come to any harm!

Thanks for that!  I doubt that I'll be using it for real - 2.4Ghz is so much more convenient - but I would like to get it completed and functional after around 50 years, and perhaps add some more channels as an exercise. Learning all the time...      

 

[dodgy geezer]

Well, I attached 4 pots to the board and gave it 5v, which should have been enough to wake it up. And got nothing. No pulses at the output end and, as far as I could tell, no pulses around the multivibrator....

[HMS Invisible]

If you have a suitable resistor, like 10 to 100 ohm you can put that in series for safety as you up the voltage.
If it doesn't burst into life there is something wrong in the circuit or the soldering. It's pretty basic stuff that I've used.
I'd be drawing ( by hand) the astable multivibrator.

Double check the function then centralise pots without desoldering. If the combined length of the pulses > the frame it goes haywire. However I think it's just component values that don't work at low voltage or an error.

I'd be working off a full schematic, double checked!

[dodgy geezer]

Haywire would be nice! All I can detect on the scope is 0v, going up to 5v occasionally as I try each land. Do you think that more volts are needed? It would have been run at 9.6v originally, but the transistors ought to start at about 3v...


A full schematic would be great. Maybe I'll draw one...

[HMS Invisible]

Haywire would be nice! All I can detect on the scope is 0v, going up to 5v occasionally as I try each land. Do you think that more volts are needed? It would have been run at 9.6v originally, but the transistors ought to start at about 3v...


A full schematic would be great. Maybe I'll draw one...

No, the voltage was an accidental misdirection and you should be OK at 5.
The voltages at the astable part will point you right to the problem if you use the diagram in the link below for number reference.
The astable timing capacitor colour codes, diode check of the transistors, and dmm measurements of its resistors is enough.
If there is a component missing it won't work.
https://www.electronics-tutorials.ws/waveforms/astable.html

[dodgy geezer]

I will need to do the schematic. I wonder why there are a pair of what look like diodes in the MV circuitry....

[HMS Invisible]

The two extra diodes conduct to clamp the negative kick to -1 volt.
The clue to what is going wrong will come from the Vc, Vb state at which the oscillator transistors settled at, and if diode tests say the transistors are OK.

[dodgy geezer]

'Sort' of schematic for the MV done. I thought I would copy Terry Tippett's style a bit, and 'draw' the components - but I only managed the ttransistors...  I think that the PCB is very elegant.  And I notice that I can't spell 'linear'. The transistors pass the diode test...Caps 2 and 3 are actually 0.01, i think - I'll update later

[HMS Invisible]

'Sort' of schematic for the MV done. I thought I would copy Terry Tippett's style a bit, and 'draw' the components - but I only managed the ttransistors...  I think that the PCB is very elegant.  And I notice that I can't spell 'linear'. The transistors pass the diode test...Caps 2 and 3 are actually 0.01, i think - I'll update later

Notes
-The effect of C4 to C7 (1nF) caps is to soften the switching for rf suppression.
- The extra diodes are series connected to block the -Vbe kick, not anti-parallel to conduct and clamp at -Vf.
That's the alternative way to do the same job.
Remarks
I earlier suggested reading off voltages thinking it always stopped at the same condition where proper circuit function is flipping between two.
The value of C4-C7 is too uncomfortably high, to near to the timing caps for my liking.
If it was me I'd snip these out if the 10nF polyester caps and resistors fit the timing formula.
If it suddenly worked, 100pF is more than enough & there are other ways to soften switching.
So check timing component values.
Still check if it stops in one state to pinpoint a solder error or damaged component.
Snip out the extra suppression caps.

[dodgy geezer]

Some success....   I found a poor ground connection (mine) and up came some pulses. Only two. I tried to upload a scope pic, but for some reason the site wouldn't take it...


Looked at the transistor where the pulses stopped and found a poor solder connection (theirs). Mended it and got three pulses.  Can't see anything obviously wrong with the next transistor stage, but there must be something somewhere... 


I'm unimpressed with the trimmer pots I put on to do the testing - the sliders seem unreliable. I will get a set of (slightly) better ones. That could be the problem.


I should have a lot of 101 caps somewhere, but it will be a fiddle to replace them, so I'd like to see if I can get that last pulse first...  moving forward(with lots of help!!)

[HMS Invisible]

If the astable is working there is no need to change it.
* It's repetition period (frame time) needs to be the maximum width of six channels plus another ~ 2.5msec for a receiver to sync.

Each one-shot triggers the next so you know where to look if your not exceeding the time limit.

[dodgy geezer]

Well, I now have 4 nice negative-going pulses (after hitting the trim pots a lot), and I can vary the three gaps between them smoothly using the first three trim pots. But I do not have a fifth pulse, and the fourth pot does nothing. A pulse arrives at it from the third transistor, but the fourth 4124 does nothing. It passes the diode test, and all the soldering seems fine - I redid it anyway.


I note that the collector of each 4124 passes a pulse to the next. But the fourth 4124 collector is just hanging in the air. All the others are connected to the +ve rail via the 5k pot. Shouldn't the last 4124 also be connected this way? I note that there is a 'spare' 4.7k resistor on the darlington pair schematic which would perform this task for the last 4124 if there were 7 channels. Perhaps connecting the last 4124 collector to this resistor would wake up the fourth channel?