Hi,
I have salvaged an optical encoder from a digital tape measure. I would like to interface it with an optical quadrature decoder. The A and B inputs on the decoder are TTL and have 2.2 K pull-ups on them.
The encoder has 5 leads. Two of them are for the illuminator LED and the other three are connected to the two phototransistors. Through various tests I'm pretty sure that I've determined that one of the three leads is connected to the emitters of the phototransistors and the other two are the collectors (and that the phototransistors are NPNs).
Here's the problem: the phototransistors seem to conduct at most .1 mA when "on" (when the illuminator LED light is not blocked by the encoder wheel pattern.) This means that I would have to use > 50K ohm pull-up in order to achieve TTL logic level voltages. Since the decoder has much stronger pull-ups on its inputs, I'll have to buffer the outputs to make them compatible with the encoder inputs.
What would be your preferred method of adapting these encoder outputs? I don't know why, but I'm a little bothered at having to use such a large pull-up if only because the largest one I've ever seen in a phototransistor schematic is about 10K. This leads me to question whether or not I'm even dealing with phototransistors or perhaps just photodiodes. Or perhaps I've mis-identified the connections.
As for buffering solutions my first thought was to use second transistor in a Darlington pair arrangement. Unfortunately that's not possible since the emitters of both phototransistors are connected. Using the 50K pull-up with a second transistor configured as an open-collector inverter does work, although the large pull-up does affect the rise time of the signal when I look at it on a scope. Other solutions I've thought of are to use a CD4069 or a comparator. What I wondering is if there's another simple solution I've overlooked.
Well, thanks for any suggestions you might have!
Update: I guess I did overlook one simple solution: using a 50K pull-up but to the decoder input instead of the positive supply.
I have salvaged an optical encoder from a digital tape measure. I would like to interface it with an optical quadrature decoder. The A and B inputs on the decoder are TTL and have 2.2 K pull-ups on them.
The encoder has 5 leads. Two of them are for the illuminator LED and the other three are connected to the two phototransistors. Through various tests I'm pretty sure that I've determined that one of the three leads is connected to the emitters of the phototransistors and the other two are the collectors (and that the phototransistors are NPNs).
Here's the problem: the phototransistors seem to conduct at most .1 mA when "on" (when the illuminator LED light is not blocked by the encoder wheel pattern.) This means that I would have to use > 50K ohm pull-up in order to achieve TTL logic level voltages. Since the decoder has much stronger pull-ups on its inputs, I'll have to buffer the outputs to make them compatible with the encoder inputs.
What would be your preferred method of adapting these encoder outputs? I don't know why, but I'm a little bothered at having to use such a large pull-up if only because the largest one I've ever seen in a phototransistor schematic is about 10K. This leads me to question whether or not I'm even dealing with phototransistors or perhaps just photodiodes. Or perhaps I've mis-identified the connections.
As for buffering solutions my first thought was to use second transistor in a Darlington pair arrangement. Unfortunately that's not possible since the emitters of both phototransistors are connected. Using the 50K pull-up with a second transistor configured as an open-collector inverter does work, although the large pull-up does affect the rise time of the signal when I look at it on a scope. Other solutions I've thought of are to use a CD4069 or a comparator. What I wondering is if there's another simple solution I've overlooked.
Well, thanks for any suggestions you might have!
Update: I guess I did overlook one simple solution: using a 50K pull-up but to the decoder input instead of the positive supply.
Last edited:
