Discrete shift register to divide one signal with pulses into two signals?

[Refrigerator]

So i've been wanting to make myself a little RX and TX for a while now. The TX is pretty simple, just a radio with some signal to shoot out but the receiver seems a bit more complicated. I want to make a TX/RX that could control a servo but i want everything to be very simple, preferably discrete transitors and all that good stuff.
I tried looking up discrete shift register and other circuits but couldn't find any to fit, i want the circuit to take the received signal that would have several pulses in it, sort of like the pulses that control a servo but two in one signal that i would then need to split up into two outputs although i'm not sure if such a receiver is possible.
I thought you guys could help me out.
Thanks.

[Simon]

How descrete do you want it to be ? shift registers made up of transistors will be pretty big, there are plenty of relatively simple chips on the market that can do it.

[Refrigerator]

How descrete do you want it to be ? shift registers made up of transistors will be pretty big, there are plenty of relatively simple chips on the market that can do it.

Maybe not a shift register but a circuit that would divide the signal, i thought of using a flip-flop and maybe making it switch between signals but i'm not sure how i would go about it.

[Simon]

I think you are look at this from the wrong perspective. What do you mean by split the signal in two ?

[Refrigerator]

I think you are look at this from the wrong perspective. What do you mean by split the signal in two ?

I thought of taking the received signal that would have two pulses with independently variable on time about 0,5mS to 1.5mS and making the first pulse go through one transistor into one output and the other going to the other output through transistor as well.
Just today i took apart a vintage Sanwa transmitter and it sends a signal just like that although the radio transmitter doesn't work since the board had battery acid all over it but the signal generator part still works perfectly and it only uses like 7 transistors to do all that, i was quite impressed with that and thought it was about time i make my own.

[Simon]

I think you are refering to PCM signals where there is a start pulse and then a train of pulses one for each function.

I am not sure of what you are describing and I don't know how such pulses are decoded. If I had to do a receiver i would probably use a micro controller to detect and time the pulses to establish the content.

[rfeecs]

Look up RC PPM.
The pulse spacing determines the servo position for each channel.
This can be decoded by detecting the sync period between pulses and then feeding the pulse train into a shift register.

I had an RC receiver in the 1970's that did exactly that.

[rfeecs]

Here's some schematics from an ancient Heathkit RC system (first post):
http://www.rcuniverse.com/forum/golden-age-vintage-antique-rc-196/11132312-heathkit-gd-47-gd-19-schematics.html

This is the decoder, with a discrete shift register that uses some odd-ball SCRs, maybe?:
http://www.rcuniverse.com/forum/attachment.php?attachmentid=1775903&d=1375603437

You might want to search around for old RC schematics for inspiration.

[Refrigerator]

In case anyone's interested here's the signal, i might not have a big scope but without this one i'd be lost. 

[Refrigerator]

Here's some schematics from an ancient Heathkit RC system (first post):
http://www.rcuniverse.com/forum/golden-age-vintage-antique-rc-196/11132312-heathkit-gd-47-gd-19-schematics.html

This is the decoder, with a discrete shift register that uses some odd-ball SCRs, maybe?:
http://www.rcuniverse.com/forum/attachment.php?attachmentid=1775903&d=1375603437

You might want to search around for old RC schematics for inspiration.

Thanks.
I actually saw a discrete receiver schematic very similar to the one with the odd looking SCR but i thought that trying to get them would be too much of a hassle.

[rfeecs]

Looks like those are Silicon Controlled Switches:
http://www.allaboutcircuits.com/textbook/semiconductors/chpt-7/silicon-controlled-switch-scs/

An SCR with an extra terminal so you can turn it off.  I guess you could make one with and NPN and PNP.

Personally, I think the microcontroller solution is a good option these days.

[Refrigerator]

Looks like those are Silicon Controlled Switches:
http://www.allaboutcircuits.com/textbook/semiconductors/chpt-7/silicon-controlled-switch-scs/

An SCR with an extra terminal so you can turn it off.  I guess you could make one with and NPN and PNP.

Personally, I think the microcontroller solution is a good option these days.

Thanks.
A µC would be a more elegant solution but i want to make it all discrete, not only to make something practical but also to learn.

[Simon]

Well there are different things to learn. Although all too often people use a micro controller when it's not needed there is a lot of value in using them and a whole new world of programming to learn there. And i don't mean learning C I mean learning to structure programs and how to solve a proglem in software.

[Refrigerator]

Well there are different things to learn. Although all too often people use a micro controller when it's not needed there is a lot of value in using them and a whole new world of programming to learn there. And i don't mean learning C I mean learning to structure programs and how to solve a proglem in software.

I just ordered a new HDD for a laptop that i will use solely for programming and other electronics design stuff. Will be about time I jump on the C/C++ bandwagon.

[Refrigerator]

It works ! Thanks guys for all the help. 
I made the circuit on a breadboard and as i noticed the SCSs were pretty much cascaded ( one SCS per channel ) so i only used two SCSs in my circuit as it only has 2 channels and the circuit worked almost flawlessly, although it can't drive a servo directly as the signal is too weak and the servo tends to induce some amount of noise which makes circuit go a bit unstable and makes servo jitter, so a simple emitter follower might just fix it.
And it's quite cool how i can pretty much cascade as many SCSs and get as many channels as i want as long as i give it the proper signal so i might even reverse engineer the circuit in the remote and make a new transmitter circuit for the remote with more channels.
Although the circuit took up about 3/4 of my 400-point breadboard i think it would be absolutely tiny if made it with SMD parts.
Here's some photos of the setup:
The whole thing connected to the butchered remote.

A close up of the breadboard circuit.

And the final signal.

Again thanks for the great help !