Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

Figuring out how the plosive keys work on the Bell labs / AT&T Voder and reproducing the keys

Michael Weaser

Member level 3
Joined
Aug 4, 2015
Messages
55
Helped
1
Reputation
2
Reaction score
1
Trophy points
1,288
Activity points
1,860
I have been trying to figure how the plosive keys work on the Voder. I have already figured out some of the stuff with the key.

I have a diagram from a Patent , that shows how the plosive key works but I don't exactly know for sure if its 100% correct. I am also guessing things could have been changed between the actual key circuitry and the circuit diagram in the patent, but though it also seems that patent for the Key was applied for and granted after the Voder came out?

The T/D plosive key is exactly the same as the K/G key except for different timings, and the P/B key is the same way except for there is no pop necessary for the beginning sound.

What I need help with :

- I need assistance on figuring out the filters for the pop in the beginning before the actual plosive sound? I tried using audacity to figure it out haven't been successful yet. In the plosive circuit diagram it seems the pops are generated by a 2 separate filters connected to the unvoiced/ voiced sound sources based on if the plosive is unvoiced or voiced. Also it seems like each filter are really 2 RLC bandpass filters combined in parallel? Obviously those might be not the real circuits used to generate the pop just representative of the circuits used.

- Theres a 3rd RLC bandpass filter on that plosive key diagram that seems connected to the output of the 10 keys and filters on the voder, which is labeled resonance controls , that I haven't figured out yet. and the patent doesn't really explain what it does. The explanation of what the plosive key is doing doesn't explain what it is doing either. I am guessing it could be representative of something. I am guessing the resonance controls could be the 10 keys and that filter is maybe representative of 1 of the 10 bandpass filters ? But though I have always thought the resonance controls meant both the keys and the filters in that block.

- How the plosive key can switch between the K and G plosive just by how fast you press it down? I know the key works by having multiple switch stages that it goes through when you press down on the key , but it seems like it is skipping stages based on how fast you press down on the key?

- Figuring out how to duplicate the key using modern switches and modern electronics. I was thinking of maybe using one shot timers to set the timing of each step ?

I have an explanation of how the key works, basically the key has multiple contacts and when you press down on it, it goes through the contacts as you press down the key.

How to make K sound using the K/G plosive ley to say the word "Key" :

Hold down wrist bar , press the K/G plosive key , and the 6th key on the Voder at the same time.

steps of the plosive key when the above is done :

1 ) removal of voiced energy from resonance controls and all filters

2 ) discharge capacitor in resonance circuit to make the pop sound shown at A

3) interval of silence

4 ) connect filter 6 to unvoiced source to produce noise shown at B

5) When key reaches bottom reconnect voiced source , so you can press the keys to make the "e" sound shown at C

If the K/G plosive key is pressed in 1/2 or 1/3 the time (faster?) , the key will produce the G plosive instead of the K plosive.

Timing for the K sound :

plosive-timing[1].png

information about the G plosive:
- It seems like the timing is backwards from the K plosive. as you seen above the timing is a 6 millisecond pop , 40 millisecond of silence , and 31 millisecond of white noise through the 6th filter, for the K plosive. Using audacity and finding where it says a G I found out , the timing is 31 milliseconds in the beginning, the silence, 6 milliseconds at the end. I am not sure though if the pop and the actual plosive sound is backwards , whenever the G is produced.

Plosive key circuit diagram :

plosive.JPG

actual photo of plosive keys :

if you look closely at the actual plosive key, it has the same amount of contacts on the key as the above circuit diagram , which is how it was figured out this circuit diagram in the patent is some version of this plosive key.

real plosive keys.JPG
 
seems like it is skipping stages based on how fast you press down on the key?

This might be achieved by a magnet attached to the key. The magnet moves past a coil, generating a waveform in the coil. Depending on speed the peak voltage is lesser or greater.
Read peak voltage using a sample-and-hold circuit.
Or install several reed switches or coils, to detect the magnet.

Or have the key block light to a photosensor, as you press the key down. The light comes from an led. The faster you press down, the faster the change in the photosensor.
 
This might be achieved by a magnet attached to the key. The magnet moves past a coil, generating a waveform in the coil. Depending on speed the peak voltage is lesser or greater.
Read peak voltage using a sample-and-hold circuit.
Or install several reed switches or coils, to detect the magnet.

Or have the key block light to a photosensor, as you press the key down. The light comes from an led. The faster you press down, the faster the change in the photosensor.
But also its possible that maybe nothing is being skipped, have been told maybe it is going through all the stages, but you just can't hear it? so all of the stages are being connected its just you can't hear them since the connection is so quick?
 
Comparing K and G...
These are breathy plosive sounds. I suppose the same key can be used to produce them if you press a shallow distance for one, and a full distance for the other.

The oscillogram shows K is formed with a sudden attack envelope. However when I imagine pronouncing G I believe it has a slower attack. So if you wish to produce G then the keypress should ignore the K switch (lest you mistakenly produce K). I don't see an easy method to bypass the K command, except by placing the K switch after the G switch.

I say this because it's an advantage if you can hear instant feedback as you press keys. You're in a position to improve on the patent illustration. It has a great many mechanical constructs. However you can do more tasks electronically, and get more immediate feedback to your ears.

Ventriloquists must study how our mouths pronounce words. No doubt their tutorials tell details about forming vowels and consonants.
 

LaTeX Commands Quick-Menu:

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top