Print

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

[HMS Invisible]

I think this is the situation....
Yes, for four channels you have five negative going ~10uS pulses at the anode rail so five diodes, four 47nF caps but one of the two astable transistors kicks off the sequence.
You end up with an extra transistor or feed it back into one of the two astable transistors.
I think that makes a fixed frame (say 20mSec for 7ch ) OR a variable frame with fixed syc pulse which needs to be not less than 2.3mSec.
If that is right then you have a decision about redoing the layout and some calculations to do.
This diagram of grouped timing components might help.

I think Futaba did variable frame, Hitec maybe did the other, and so Hitec IPD receivers didn't work with Futaba.- just a thought , and it may be wrong!
One 6channel radio I was given to look at had already been tweaked from 1.3mS to 1.5mS neutrals causing it to crash so one timing resistor needed changed.

[dodgy geezer]

Er... wow!  Redesigning a Micron kit is rather beyond my pay grade. It OUGHT to work as specified.  I see your circuit, which is rather like the situation I have got. I have put a drawing down below. Are you suggesting adding an extra 4124 - which, I suppose, could be done...


The 4124 outputs a pulse to the signal line as shown through the diode, and also along the line marked in red to the pot for the next channel. Only, for the last 4124, the components marked in red are just missing.  There is, however, a spare 4.7k sitting unconnected on the board at the end of the cascaded half-shots, and it looks as if the last 4124 for channel 7 should be connected to this as a sort of termination. If less than 7 channels are populated, I suspect that a fly lead should go to that resistor from the last 4124, as shown in green on the diagram. Does that make sense?
 

[dodgy geezer]

I connected the green link as above - got another pulse!  At that stage my skill (?) with using a scope took a dive, and the trim pot sliders started failling again, making it very hard to see what was going on.


It seems that altering the pots alters the gap between the pulses rather than the pulse width, but trying to investigate that collapses into noise whenever I move the trims. I have ordered a set of standard pots and will see if it gets better when they are connected.


Meanwhile - onto the RF board!

[Martin (Admin)]

All this is way over my head, ( I did learn how to repair HiFi and stuff ) but good to see reals skill / knowledge is still very much alive and kicking! 
 

 

[dodgy geezer]

The skill is all with Invisible - I probably know less about it than you do, Martin. But it's interesting to work on what looks like a new Micron kit that's 50 years old,  and it looks like this is an early design that no one has a schematic for, so publishing that will be useful...

[dodgy geezer]

Looks like the encoder is working. On with the RF board. 


I wanted to power it up, but as soon as I twiddled the ferrite slug it collapsed in a heap of dirt.  So I need to order another one - hope they are standard sizes...  In the meantime, here is my schematic drawing..

[roycv]

Hi DG, I think they must be interchangeable as I do not know if there is a measure of reluctance for them.
By the way I have all the plans for building the 6 ch. 27Mhtz FM Tx and Rx, which I did many years back.
Roy

[HMS Invisible]

https://en.m.wikipedia.org/wiki/Permeability_(electromagnetism)#Relative
μ_r

[dodgy geezer]

Hi DG, I think they must be interchangeable as I do not know if there is a measure of reluctance for them.
By the way I have all the plans for building the 6 ch. 27Mhtz FM Tx and Rx, which I did many years back.
Roy

Well, we shall find out!   


I now have 3 Micron sets here - all different. I did not realise that there was so much variation - I have a 1977 AM version, a 1979 am version converted to FM, and one of the unusual FM red ones with slider trims - unknown date. All seem to have different electronics. As well as several sets of blue construction notes, I have the following blueprint drawings:

am receiver assembly   jul77
charger unit assembly  aug77
fet/cmos am receiver    apl 78
am/fm transmitter circuit  jul 78
assembly  pl-7d am/fm tx   oct 78
open gimbal stick wiring   oct 78

I was going to make a package and send it off to SingleChannel, but I had no 1977 AM Tx circuit - which I am having to draw up at the moment... once I get it working...

[dodgy geezer]

https://en.m.wikipedia.org/wiki/Permeability_(electromagnetism)#Relative
μ_r

These look likely


DG

[HMS Invisible]

All I can tell you is you can exclude the lowest μr materials.

[dodgy geezer]

All I can tell you is you can exclude the lowest μr materials.

I'd be happy to if I knew what they were  At the moment simply getting the right (or similar) size  is proving tricky...   

[HMS Invisible]

I'd be happy to if I knew what they were  At the moment simply getting the right (or similar) size  is proving tricky...   

I think I see see a ferrite cored transformer from your schematic and your photo.
Here is an inductance calculator for a solenoid that you can key in your six Neosid optional values of μ_r. http://hyperphysics.phy-astr.gsu.edu/hbase/electric/indsol.html
The coil length is that of the copper winding of N turns, not the ferrite length. The coil diameter is of the ferrite core because the magnetic flux in the ferrite predominates over the bit in the air and the former, where μ_r =1
The Neosid catalogue shows six grades of ferrite material so you'll get six values of specific inductance ( L_sp for one turn.
If you had a solenoid of 8 turns, and high μ_r permeability grade of 10k , self inductance of the 8 turn will be 64,000 L_sp.
  So matching size isn't your only problem, and it isn't the biggest problem.
If I had two or three pieces to glue together I'd be in with a chance of finding material grade from an inductance measurement.
 Realistically, you've got a ton of old published circuits to track down and digest unless you have a radio ham friend.

[HMS Invisible]

I keyed in some values to see the ferrite grade would be low μ_r. Show's how much I know!
High μ_r came to mind because I've used ferrite material, rather than iron, for high L_sp in larger, medium frequency transformers.

[dodgy geezer]

......
  So matching size isn't your only problem, and it isn't the biggest problem.
If I had two or three pieces to glue together I'd be in with a chance of finding material grade from an inductance measurement.
 Realistically, you've got a ton of old published circuits to track down and digest unless you have a radio ham friend.

...and tracking circuits is where I came in....  Though I suspect that there will not be a precise specification of the inductance, even on the original design. A coil former which was standard at the time would probably have been used. I have the dust if you want to compact it, but the largest piece is 3mm dia x 2mm!!


Unless the different ferrite specifications have an impact on the saturation time-to-rise and the timing is therefore impacted, I suspect that they will mainly influence the strength of the electro-magnetic response. If this is the case, a different grade slug might simply need screwing less or futher in? I have ordered a Toko 10k slug of about the right size (very slightly smaller in diameter, from my measurements) and we will see what that does.

[HMS Invisible]

You still need to make a 27MHz parallel resonant oscillator with a 15pF capacitor.
F29 has a μ_r around 10, about the same as the 10mm dia rod inside mw/lw/sw AM radios. An F25 core in the same wound former used for F29 material results in ten x5 times the inductance.
  The hyperphysics calculator uses the k constant rather than μ_r.

[dodgy geezer]

You still need to make a 27MHz parallel resonant oscillator with a 15pF capacitor.
F29 has a μ_r around 10, about the same as the 10mm dia rod inside mw/lw/sw AM radios. An F25 core in the same wound former used for F29 material results in ten times the inductance.
  The hyperphysics calculator uses the k constant rather than μ_r.

Hmm...quite a difference, then?  Even though the Neosid pamphlet shows that F25 and F29 overlap in suitability for 27MHz...


Calculations are a bit moot, however, without any base data to work on. In the absence of that, I went looking for a core that would fit. I made the original slug to be 8mm length, 3.73mm diameter over the screw threads (probably nominal 3.75mm).  The best I could find was a Toko 10K Ferrite Screw Core Slug L=8mm Dia=3.65mm 1g ET08. ET08 does not seem to be enough to identify the item. 

I only have a simple diode/cap/multimeter to test for RF, so I'm not too sure how to determine whether it works or not. I suspect that I might get some RF reading from the aerial even if there was no slug in the coil at all, since this component just seems to be impressing the pulse signal on the carrier wave...

[HMS Invisible]

Hmm...quite a difference, then?  Even though the Neosid pamphlet shows that F25 and F29 overlap in suitability for 27MHz...

Outside the suggested area is not impossible, but less practical due to parasitic effects.
You see the cost, size and performance penalties and design constraints when you've had to design examples.  Capacitor dielectric material is analogous to core material and easier to understand the same effect where you can have the same overlapping values in capacitance with different dielectrics.

Calculations are a bit moot, however, without any base data to work on. In the absence of that, I went looking for a core that would fit. I made the original slug to be 8mm length, 3.73mm diameter over the screw threads (probably nominal 3.75mm).  The best I could find was a Toko 10K Ferrite Screw Core Slug L=8mm Dia=3.65mm 1g ET08. ET08 does not seem to be enough to identify the item. 
 
I only have a simple diode/cap/multimeter to test for RF, so I'm not too sure how to determine whether it works or not. I suspect that I might get some RF reading from the aerial even if there was no slug in the coil at all, since this component just seems to be impressing the pulse signal on the carrier wave...

The quartz crystal is just a capacitor with a known, high-Q resonance points.
It's only one part of a unity-gain tuned circuit that manages to sustain oscillation.
 Internet hits will describe the function, but you have a transitor, crystal and tuned LC.
You can find L through an improvised waveform source, dropper resistor and known-value substituton inductors.
X_L= 2.Pi.f.L so, ideally, you'd want a bit higher than 3kHz from an esc.
You could rewind primary and secondary on an open bobbin cored inductor for scrap breadboard experiments. 100 microhenry with ten turns is L_SP OF 1 microhenry.

[dodgy geezer]

A lot of work, building a signal generator!  Perhaps it would be easier to use a circuit simulator - something like LTSpice? I assume there is one where you can just alter the inductance...

Looking at the Micron set-up notes, it does not seem that the coil setting is critical.  The notes say:

1 - wind the core nearly out
2 - set up an RF detector multimeter and power the Tx on - there should be no RF
3 - wind the core in until you get an RF signal
4 - wind the core in 1.5 turns more for safety
5 - turn the Tx on and off a few times to check that the circuit starts properly...

which suggests to me that so long as it's oscillating the set-up is OK?


P.S.  CircuitLab looks fun...

[HMS Invisible]

A lot of work, building a signal generator!  Perhaps it would be easier to use a circuit simulator - something like LTSpice? I assume there is one where you can just alter the inductance...

Looking at the Micron set-up notes, it does not seem that the coil setting is critical.  The notes say:

1 - wind the core nearly out
2 - set up an RF detector multimeter and power the Tx on - there should be no RF
3 - wind the core in until you get an RF signal
4 - wind the core in 1.5 turns more for safety
5 - turn the Tx on and off a few times to check that the circuit starts properly...

which suggests to me that so long as it's oscillating the set-up is OK?

A 27 MHz crystal is actually a 9MHz crystal, btw. The resonator covers the whole band but the turns they refer to means a mm of lateral movement. You'll find out how tricky it is to get any range at all.
 For an improvised sig gen I used a 555 circuit to hand, but all you need is a frequency high enough (eg esc at 3kHz) for  X_L to be high enough compared to your dropper resistor.

[dodgy geezer]

Are you saying that the resonant circuit will probably work over a wide spread of inductances, but will not provide much power unless it is optimised quite precisely?  I can't see how that matches with the circuit notes, which seem to indicate that so long as the oscillator is running the precise core position is not critical.


Doing a range check would give a definitive answer, of course. If this Toko 10k produces detectable RF I am inclined to put the PCBs into the Tx and see if I can get the set to talk to a receiver...

[dodgy geezer]

The ferrite core arrived this morning, so I put it in and powered up the RF board with a Futaba AM crystal. I used the diode/cap/multimeter circuit to detect RF.


The first thing that I noticed was that I seemed to detect some mV on pretty much any bit of metal, which varied along it's length, making any precise measurement difficult. Next I tried a working Micron Tx, and could definitely see a jump when I turned that on.


Then I tried the RF board, with a crocodile clip to a partly extended aerial. I was running off 6v rather than 9.6v, but I could detect a small jump on power-up, even when the slug was almost completely out. I got a bigger reading (though still much less than the working Tx) as I wound the slug in, and there seemed to be a slight peak at a point where the slug was 1.5mm out of the coil.


I'm not sure how to interpret this. The inductor seems to be working, and the low output can be explained, at least in part, by the lower voltage and non-resonant aerial length. Since this is the only slug that remotely fits the coil that I can easily source, (it's a little bit loose, but the threads do connect) I'm inclined to assemble the whole Tx and do any further mods on the complete system...