Wavetek 75 (or 75A or 23) Rotary Encoder

[Hoi Wong]

The rotary encoder in my Wavetek 75 arbitrary waveform generator is
physically broken, and it's impossible to hook anything up to the broken
legs to probe the specs.

I'm totally new to encoders and I'd appreciate if anybody can offer any
advice on finding a suitable replacement. I couldn't find the service manual
for Wavetek 75, so all the information I have is the encoder itself:

http://www.stanford.edu/~wonghoi/DC/Wavetek75rotaryEncoderFront.png
http://www.stanford.edu/~wonghoi/DC/Wavetek75rotaryEncoderBack.png

Thanks in advance!

Cheers,
Hoi

 

[Hoi Wong]

"Hoi Wong" <wonghoi@stanford.edu> wrote in message
news:g7ga80$m3j$1@news.stanford.edu...

The rotary encoder in my Wavetek 75 arbitrary waveform generator is
physically broken, and it's impossible to hook anything up to the broken
legs to probe the specs.

I'm totally new to encoders and I'd appreciate if anybody can offer any
advice on finding a suitable replacement. I couldn't find the service
manual for Wavetek 75, so all the information I have is the encoder
itself:

http://www.stanford.edu/~wonghoi/DC/Wavetek75rotaryEncoderFront.png
http://www.stanford.edu/~wonghoi/DC/Wavetek75rotaryEncoderBack.png

Thanks in advance!

Cheers,
Hoi

Here's the cap for the rotary encoder, if that's useful at all:
http://www.stanford.edu/~wonghoi/DC/Wavetek75rotaryEncoderCap.png

 

[Dave Platt]

The rotary encoder in my Wavetek 75 arbitrary waveform generator is
physically broken, and it's impossible to hook anything up to the broken
legs to probe the specs.

I'm totally new to encoders and I'd appreciate if anybody can offer any
advice on finding a suitable replacement. I couldn't find the service
manual for Wavetek 75, so all the information I have is the encoder
itself:

http://www.stanford.edu/~wonghoi/DC/Wavetek75rotaryEncoderFront.png
http://www.stanford.edu/~wonghoi/DC/Wavetek75rotaryEncoderBack.png
http://www.stanford.edu/~wonghoi/DC/Wavetek75rotaryEncoderCap.png

Hmmm.

Based on the photos, it appears to be a mechanical rotary encoder,
with a two-bit quadrature output and a common terminal. By my count
there are 13 or 14 contact fingers per output per quarter-revolution
of the control... this would allow for 50-some pulses per revolution
for each of the two channels, or around 200 counts per revolution if
you're doing the usual sort of quadrature decoding.

I suspect that a direct replacement may be tricky to find. In looking
through the DigiKey and Mouser catalogs, it appears to me that
mechanical-contact encoders with this large a number of counts per
revolution are quite unusual. Mechanical encoders these days seem to
max out at around 32 counts per revolution, and usually have
detents. I'd guess that you'll probably want a detent-less encoder,
with a counts-per-revolution value not too different from that of the
broken one (which may very well be a custom part). That's probably
going to mean switching to an optical encoder, one way or the other.

I can think of at least two ways to do it, depending on how the
circuitry in the Wavetek 75 works:

- If the existing encoder's common wire (the one going to that
sliding contact) was grounded, and if the two quadrature leads were
of the "external pull-up resistor to +5 volts, and the encoder
pulls the quadrature output down to ground", then dropping in an
optical encoder could be fairly easy. Rotary optoencoders
generally need a +5v feed for their LED (many have build-in current
limiting resistors for the LED) and have phototransistor
pull-downs... so, substituting one of these might be as simple as
wiring it in with the common/ground and quadrature-outputs as
before, and adding a single lead for +5 for the LED drive.

- If the existing encoder was wired up any differently, you'd need an
interface board (transistors, relays, or some such) to take the
open-collector output of a rotary optoencoder and drive the Wavetek
circuit as it wants to be driven. Probably not hard to build on a
bit of perfboard once you know what you need to do.

In either case, having a schematic for the Wavetek 75 is likely to be
essential. There's a Wavetek test equipment special-interest mailing
list at YahooGroups - somebody there might have a schematic or even
the whole service manual.

As to specific encoder lines: Grayhill 61K or 61R series, Clarostat
600 series, or one of the Bourns encoders (the Bourns 14mm optos
carries in the new Digi-Key catalog are under $20) might all be
candidates.

--
Dave Platt <dplatt@radagast.org> AE6EO
Friends of Jade Warrior home page: http://www.radagast.org/jade-warrior
I do _not_ wish to receive unsolicited commercial email, and I will
boycott any company which has the gall to send me such ads!

 

[Hoi Wong]

"Dave Platt" <dplatt@radagast.org> wrote in message
news:1c6vm5-uom.ln1@radagast.org...

The rotary encoder in my Wavetek 75 arbitrary waveform generator is
physically broken, and it's impossible to hook anything up to the broken
legs to probe the specs.

I'm totally new to encoders and I'd appreciate if anybody can offer any
advice on finding a suitable replacement. I couldn't find the service
manual for Wavetek 75, so all the information I have is the encoder
itself:

http://www.stanford.edu/~wonghoi/DC/Wavetek75rotaryEncoderFront.png
http://www.stanford.edu/~wonghoi/DC/Wavetek75rotaryEncoderBack.png
http://www.stanford.edu/~wonghoi/DC/Wavetek75rotaryEncoderCap.png

Hmmm.

Based on the photos, it appears to be a mechanical rotary encoder,
with a two-bit quadrature output and a common terminal. By my count
there are 13 or 14 contact fingers per output per quarter-revolution
of the control... this would allow for 50-some pulses per revolution
for each of the two channels, or around 200 counts per revolution if
you're doing the usual sort of quadrature decoding.

I suspect that a direct replacement may be tricky to find. In looking
through the DigiKey and Mouser catalogs, it appears to me that
mechanical-contact encoders with this large a number of counts per
revolution are quite unusual. Mechanical encoders these days seem to
max out at around 32 counts per revolution, and usually have
detents. I'd guess that you'll probably want a detent-less encoder,
with a counts-per-revolution value not too different from that of the
broken one (which may very well be a custom part). That's probably
going to mean switching to an optical encoder, one way or the other.

I can think of at least two ways to do it, depending on how the
circuitry in the Wavetek 75 works:

- If the existing encoder's common wire (the one going to that
sliding contact) was grounded, and if the two quadrature leads were
of the "external pull-up resistor to +5 volts, and the encoder
pulls the quadrature output down to ground", then dropping in an
optical encoder could be fairly easy. Rotary optoencoders
generally need a +5v feed for their LED (many have build-in current
limiting resistors for the LED) and have phototransistor
pull-downs... so, substituting one of these might be as simple as
wiring it in with the common/ground and quadrature-outputs as
before, and adding a single lead for +5 for the LED drive.

- If the existing encoder was wired up any differently, you'd need an
interface board (transistors, relays, or some such) to take the
open-collector output of a rotary optoencoder and drive the Wavetek
circuit as it wants to be driven. Probably not hard to build on a
bit of perfboard once you know what you need to do.

In either case, having a schematic for the Wavetek 75 is likely to be
essential. There's a Wavetek test equipment special-interest mailing
list at YahooGroups - somebody there might have a schematic or even
the whole service manual.

As to specific encoder lines: Grayhill 61K or 61R series, Clarostat
600 series, or one of the Bourns encoders (the Bourns 14mm optos
carries in the new Digi-Key catalog are under $20) might all be
candidates.

--
Dave Platt <dplatt@radagast.org> AE6EO
Friends of Jade Warrior home page: http://www.radagast.org/jade-warrior
I do _not_ wish to receive unsolicited commercial email, and I will
boycott any company which has the gall to send me such ads!

Thanks a lot. I'll give it a shot.

Cheers,
Hoi

 

[DaveM]

"Hoi Wong" <wonghoi@stanford.edu> wrote in message
news:g7ma9d$h63$1@news.stanford.edu...

"Dave Platt" <dplatt@radagast.org> wrote in message
news:1c6vm5-uom.ln1@radagast.org...
The rotary encoder in my Wavetek 75 arbitrary waveform generator is
physically broken, and it's impossible to hook anything up to the broken
legs to probe the specs.

I'm totally new to encoders and I'd appreciate if anybody can offer any
advice on finding a suitable replacement. I couldn't find the service
manual for Wavetek 75, so all the information I have is the encoder
itself:

http://www.stanford.edu/~wonghoi/DC/Wavetek75rotaryEncoderFront.png
http://www.stanford.edu/~wonghoi/DC/Wavetek75rotaryEncoderBack.png
http://www.stanford.edu/~wonghoi/DC/Wavetek75rotaryEncoderCap.png

Hmmm.

Based on the photos, it appears to be a mechanical rotary encoder,
with a two-bit quadrature output and a common terminal. By my count
there are 13 or 14 contact fingers per output per quarter-revolution
of the control... this would allow for 50-some pulses per revolution
for each of the two channels, or around 200 counts per revolution if
you're doing the usual sort of quadrature decoding.

I suspect that a direct replacement may be tricky to find. In looking
through the DigiKey and Mouser catalogs, it appears to me that
mechanical-contact encoders with this large a number of counts per
revolution are quite unusual. Mechanical encoders these days seem to
max out at around 32 counts per revolution, and usually have
detents. I'd guess that you'll probably want a detent-less encoder,
with a counts-per-revolution value not too different from that of the
broken one (which may very well be a custom part). That's probably
going to mean switching to an optical encoder, one way or the other.

I can think of at least two ways to do it, depending on how the
circuitry in the Wavetek 75 works:

- If the existing encoder's common wire (the one going to that
sliding contact) was grounded, and if the two quadrature leads were
of the "external pull-up resistor to +5 volts, and the encoder
pulls the quadrature output down to ground", then dropping in an
optical encoder could be fairly easy. Rotary optoencoders
generally need a +5v feed for their LED (many have build-in current
limiting resistors for the LED) and have phototransistor
pull-downs... so, substituting one of these might be as simple as
wiring it in with the common/ground and quadrature-outputs as
before, and adding a single lead for +5 for the LED drive.

- If the existing encoder was wired up any differently, you'd need an
interface board (transistors, relays, or some such) to take the
open-collector output of a rotary optoencoder and drive the Wavetek
circuit as it wants to be driven. Probably not hard to build on a
bit of perfboard once you know what you need to do.

In either case, having a schematic for the Wavetek 75 is likely to be
essential. There's a Wavetek test equipment special-interest mailing
list at YahooGroups - somebody there might have a schematic or even
the whole service manual.

As to specific encoder lines: Grayhill 61K or 61R series, Clarostat
600 series, or one of the Bourns encoders (the Bourns 14mm optos
carries in the new Digi-Key catalog are under $20) might all be
candidates.

--
Dave Platt <dplatt@radagast.org> AE6EO
Friends of Jade Warrior home page: http://www.radagast.org/jade-warrior
I do _not_ wish to receive unsolicited commercial email, and I will
boycott any company which has the gall to send me such ads!


Thanks a lot. I'll give it a shot.

Cheers,
Hoi

I've been loosely following this thread, trying to remember where I had seen a
software program that would produce the code wheel for rotary encoders. I
finally found it; it's called CodeWheel. The author has a web page at
http://www.mindspring.com/~tom2000/Delphi/Codewheel.html, which pretty much
describes the codewheel, design, theory, etc. It even describes how to build
one using the innards from an optical mouse.
Download the software from a link toward the bottom of the page. It's free.

From your photo, I don't see any score marks from the wipers on the wheel, so
I'm assuming that you have an optical encoder. The CodeWheel program might be a
great solution to your problem.
If you still have the crumbled pieces of the wheel, you should be able to piece
it back together well enough to determine the appropriate design of the wheel.
If the codewheel segments are conductive (a mechanical encoder), then you might
get enough conductivity from a laser printer. Alternatively, you might try your
luck at etching some very thin PCB material.

I tried to locate a manual for the model 75, but no such luck.

Good luck on the project. I would love to hear the results of your efforts.
Cheers!!!!
--
Dave M
MasonDG44 at comcast dot net (Just substitute the appropriate characters in the
address)

Experience: What you get when you don't get what you want

 

[DaveM]

"Franc Zabkar" <fzabkar@iinternode.on.net> wrote in message
news:10lu94p9tdq65ec27nq98g4odt3e1hufvq@4ax.com...

On Fri, 8 Aug 2008 17:52:49 -0700, dplatt@radagast.org (Dave Platt)
put finger to keyboard and composed:

The rotary encoder in my Wavetek 75 arbitrary waveform generator is
physically broken, and it's impossible to hook anything up to the broken
legs to probe the specs.

I'm totally new to encoders and I'd appreciate if anybody can offer any
advice on finding a suitable replacement. I couldn't find the service
manual for Wavetek 75, so all the information I have is the encoder
itself:

http://www.stanford.edu/~wonghoi/DC/Wavetek75rotaryEncoderFront.png
http://www.stanford.edu/~wonghoi/DC/Wavetek75rotaryEncoderBack.png
http://www.stanford.edu/~wonghoi/DC/Wavetek75rotaryEncoderCap.png

Hmmm.

Based on the photos, it appears to be a mechanical rotary encoder,
with a two-bit quadrature output and a common terminal. By my count
there are 13 or 14 contact fingers per output per quarter-revolution
of the control... this would allow for 50-some pulses per revolution
for each of the two channels, or around 200 counts per revolution if
you're doing the usual sort of quadrature decoding.

The cap is stamped with a "50". Could that be the number of pulses?
BTW, that's a Matsushita logo. I can't guess at what "6L" might mean,
though. Could it be a part number series?

FWIW, there are some Wavetek manuals here:
ftp://bama.sbc.edu/downloads/wavetek/

Would this wavegen have a similar encoder?
ftp://bama.sbc.edu/downloads/wavetek/275/

- Franc Zabkar
--
Please remove one 'i' from my address when replying by email.

I misinterpreted the image of the broken rotary encoder that the OP posted. I
saw the broken part as the codewheel, not the wiper part with the contacts
attached. Apologies.

I forgot that the subject line mentioned the model 23. Here is a link to the
manual for the model 23:
http://www.ko4bb.com/cgi-bin/manuals.pl?dir=Wavetek/Wavetek_23-32MHz_Synthesized_Function_Generator.
It gives you a pretty good description of the encoder operation, and a picture
of the code wheel. Not much else about the encoder itself, though. The
description of the encoder in the model 23 manual doesn't seem to fit the
images of the model 75 encoder.

The only info given about it in the parts list is the Wavetek part number, which
is 5109-00-0026. Wavetek is listed as the manufacturer.

The schematic shows two contacts in the encoder. That would seem to agree with
the OP's thinking that it is a 2-bit encoder. The model 23 manual indicates
that the encoder outputs 50 pulses per revolution.

Since there doesn't seem to be an indexing contact on the wheel, I think you
could use an encoder with a lower number of steps per rev without any problem
(electrically). It might, however, make the adjustment less sensitive due to
the reduced resolution of the encoder. You'd have to make more turns of the
knob with the lower resolution encoder.

Find one at Mouser or Digikey with the appropriate physical characteristics and
highest resolution and give it a try. Nothing to lose at this point.

--
Dave M
MasonDG44 at comcast dot net (Just substitute the appropriate characters in the
address)

Experience: What you get when you don't get what you want

 

[Hoi Wong]

"Dave Platt" <dplatt@radagast.org> wrote in message
news:1c6vm5-uom.ln1@radagast.org...

The rotary encoder in my Wavetek 75 arbitrary waveform generator is
physically broken, and it's impossible to hook anything up to the broken
legs to probe the specs.

I'm totally new to encoders and I'd appreciate if anybody can offer any
advice on finding a suitable replacement. I couldn't find the service
manual for Wavetek 75, so all the information I have is the encoder
itself:

http://www.stanford.edu/~wonghoi/DC/Wavetek75rotaryEncoderFront.png
http://www.stanford.edu/~wonghoi/DC/Wavetek75rotaryEncoderBack.png
http://www.stanford.edu/~wonghoi/DC/Wavetek75rotaryEncoderCap.png

Hmmm.

Based on the photos, it appears to be a mechanical rotary encoder,
with a two-bit quadrature output and a common terminal. By my count
there are 13 or 14 contact fingers per output per quarter-revolution
of the control... this would allow for 50-some pulses per revolution
for each of the two channels, or around 200 counts per revolution if
you're doing the usual sort of quadrature decoding.

I suspect that a direct replacement may be tricky to find. In looking
through the DigiKey and Mouser catalogs, it appears to me that
mechanical-contact encoders with this large a number of counts per
revolution are quite unusual. Mechanical encoders these days seem to
max out at around 32 counts per revolution, and usually have
detents. I'd guess that you'll probably want a detent-less encoder,
with a counts-per-revolution value not too different from that of the
broken one (which may very well be a custom part). That's probably
going to mean switching to an optical encoder, one way or the other.

I can think of at least two ways to do it, depending on how the
circuitry in the Wavetek 75 works:

- If the existing encoder's common wire (the one going to that
sliding contact) was grounded, and if the two quadrature leads were
of the "external pull-up resistor to +5 volts, and the encoder
pulls the quadrature output down to ground", then dropping in an
optical encoder could be fairly easy. Rotary optoencoders
generally need a +5v feed for their LED (many have build-in current
limiting resistors for the LED) and have phototransistor
pull-downs... so, substituting one of these might be as simple as
wiring it in with the common/ground and quadrature-outputs as
before, and adding a single lead for +5 for the LED drive.

- If the existing encoder was wired up any differently, you'd need an
interface board (transistors, relays, or some such) to take the
open-collector output of a rotary optoencoder and drive the Wavetek
circuit as it wants to be driven. Probably not hard to build on a
bit of perfboard once you know what you need to do.

In either case, having a schematic for the Wavetek 75 is likely to be
essential. There's a Wavetek test equipment special-interest mailing
list at YahooGroups - somebody there might have a schematic or even
the whole service manual.

As to specific encoder lines: Grayhill 61K or 61R series, Clarostat
600 series, or one of the Bourns encoders (the Bourns 14mm optos
carries in the new Digi-Key catalog are under $20) might all be
candidates.

--
Dave Platt <dplatt@radagast.org> AE6EO
Friends of Jade Warrior home page: http://www.radagast.org/jade-warrior
I do _not_ wish to receive unsolicited commercial email, and I will
boycott any company which has the gall to send me such ads!

Hi Dave,

The part# on the service manual is no where to be found, neither does it
contain much information:
http://www.stanford.edu/~wonghoi/DC/4600-05-0211.png

I just probed the connections to the Wavetek 75 board, and it seems like one
wire is grounded while two are hooked up to +5V (My mechanical encoder was
too broken to tell which lead is which). Is there any opto-encoder that
you'll recommend for this case?

Thanks.

Cheers,
Hoi

 

[Hoi Wong]

"Franc Zabkar" <fzabkar@iinternode.on.net> wrote in message
news:kh24a4lqus9cr0ahab02ura0h6rgruea5b@4ax.com...

On Wed, 13 Aug 2008 08:00:00 +1000, Franc Zabkar
fzabkar@iinternode.on.net> put finger to keyboard and composed:

The above encoders are optical types, so you will need to wire up a 5V
DC supply.

I notice that there are three unused wires in the J12/J13 harness in
the Model 23. If the Model 75 also has unused conductors, then maybe
you could utilise them for bussing the +5V supply to the optical
encoder.

- Franc Zabkar
--
Please remove one 'i' from my address when replying by email.

Thanks a lot Franc. I found a convenient +5 from CR2 (left part of battery
test schematic on p.126) to steal for optoencoders too. By the way, am I
supposed to wire the switch terminals 1,2 to the ground in this case?

Cheers,
Hoi

 

[Dave Platt]

In article <MPmdnW-DR-oVvz_VnZ2dnUVZ_sninZ2d@earthlink.com>,
Michael A. Terrell <mike.terrell@earthlink.net> wrote:
X-Newsreader: trn 4.0-test74 (May 26, 2000)

http://search.digikey.com/scripts/DkSearch/dksus.dll?Cat=1114163;keywords=encoder
shows 549 encoders to chose from.

the mecanical encoders use pullup resistors, and the switch pulls the
+5 VDC lines low as the shaft is turned. The interface may need to be
modified to use the electronic encoders. The electronic encoders I've
used had a direction output, and a pulse output which will not work with
that Wavetek design.

Interesting... those would be ones with a decoder IC on board. You're
right, those would not be appropriate for repairing the Wavetek.

The rotary optical encoders I've used myself have a direct two-bit
quadrature output, and use open-collector phototransistor pulldowns
rather than electrical-contact pulldowns. They don't have on-board
direction/pulse decoders - they were/are designed to be almost direct
replacements for similar mechanical encoders.

--
Dave Platt <dplatt@radagast.org> AE6EO
Friends of Jade Warrior home page: http://www.radagast.org/jade-warrior
I do _not_ wish to receive unsolicited commercial email, and I will
boycott any company which has the gall to send me such ads!

 

[Hoi Wong]

given the costs of an optical encoder, and the complexity (changing the
pullups), I might try out the grayhill mechanical encoders first (if dialing
becomes unbearable, I'll open the AWG up again for another surgery) then.
Thanks for your advice.

Cheers,
Hoi

"Franc Zabkar" <fzabkar@iinternode.on.net> wrote in message
news:cib6a4dgeq27hl9k8l5o4d4s8makiql19e@4ax.com...

On Tue, 12 Aug 2008 16:47:40 -0700, "Hoi Wong" <wonghoi@stanford.edu
put finger to keyboard and composed:

I found a convenient +5 from CR2 (left part of battery
test schematic on p.126) to steal for optoencoders too.

I notice that the optical encoder's output is rated to sink about
20mA. However, the pullup resistors are 221K which would result in a
current draw of only 20uA. Perhaps it would be advisable to change the
pullups to 2.2K in order to bring the collector current into the 2mA
range ???

- Franc Zabkar
--
Please remove one 'i' from my address when replying by email.

 

[Franc Zabkar]

On Fri, 8 Aug 2008 17:52:49 -0700, dplatt@radagast.org (Dave Platt)
put finger to keyboard and composed:

The rotary encoder in my Wavetek 75 arbitrary waveform generator is
physically broken, and it's impossible to hook anything up to the broken
legs to probe the specs.

I'm totally new to encoders and I'd appreciate if anybody can offer any
advice on finding a suitable replacement. I couldn't find the service
manual for Wavetek 75, so all the information I have is the encoder
itself:

http://www.stanford.edu/~wonghoi/DC/Wavetek75rotaryEncoderFront.png
http://www.stanford.edu/~wonghoi/DC/Wavetek75rotaryEncoderBack.png
http://www.stanford.edu/~wonghoi/DC/Wavetek75rotaryEncoderCap.png

Hmmm.

Based on the photos, it appears to be a mechanical rotary encoder,
with a two-bit quadrature output and a common terminal. By my count
there are 13 or 14 contact fingers per output per quarter-revolution
of the control... this would allow for 50-some pulses per revolution
for each of the two channels, or around 200 counts per revolution if
you're doing the usual sort of quadrature decoding.

The cap is stamped with a "50". Could that be the number of pulses?
BTW, that's a Matsushita logo. I can't guess at what "6L" might mean,
though. Could it be a part number series?

FWIW, there are some Wavetek manuals here:
ftp://bama.sbc.edu/downloads/wavetek/

Would this wavegen have a similar encoder?
ftp://bama.sbc.edu/downloads/wavetek/275/

- Franc Zabkar
--
Please remove one 'i' from my address when replying by email.

 

[Franc Zabkar]

On Mon, 11 Aug 2008 06:52:50 +1000, Franc Zabkar
<fzabkar@iinternode.on.net> put finger to keyboard and composed:

Would this wavegen have a similar encoder?
ftp://bama.sbc.edu/downloads/wavetek/275/

Or this one (Model 23)?
http://www.bitsavers.org/pdf/wavetek/23_Jun90.pdf

Other Wavetek stuff:
http://www.bitsavers.org/pdf/wavetek/

- Franc Zabkar
--
Please remove one 'i' from my address when replying by email.

 

[Michael A. Terrell]

Hoi Wong wrote:

"Dave Platt" <dplatt@radagast.org> wrote in message
news:1c6vm5-uom.ln1@radagast.org...
The rotary encoder in my Wavetek 75 arbitrary waveform generator is
physically broken, and it's impossible to hook anything up to the broken
legs to probe the specs.

I'm totally new to encoders and I'd appreciate if anybody can offer any
advice on finding a suitable replacement. I couldn't find the service
manual for Wavetek 75, so all the information I have is the encoder
itself:

http://www.stanford.edu/~wonghoi/DC/Wavetek75rotaryEncoderFront.png
http://www.stanford.edu/~wonghoi/DC/Wavetek75rotaryEncoderBack.png
http://www.stanford.edu/~wonghoi/DC/Wavetek75rotaryEncoderCap.png

Hmmm.

Based on the photos, it appears to be a mechanical rotary encoder,
with a two-bit quadrature output and a common terminal. By my count
there are 13 or 14 contact fingers per output per quarter-revolution
of the control... this would allow for 50-some pulses per revolution
for each of the two channels, or around 200 counts per revolution if
you're doing the usual sort of quadrature decoding.

I suspect that a direct replacement may be tricky to find. In looking
through the DigiKey and Mouser catalogs, it appears to me that
mechanical-contact encoders with this large a number of counts per
revolution are quite unusual. Mechanical encoders these days seem to
max out at around 32 counts per revolution, and usually have
detents. I'd guess that you'll probably want a detent-less encoder,
with a counts-per-revolution value not too different from that of the
broken one (which may very well be a custom part). That's probably
going to mean switching to an optical encoder, one way or the other.

I can think of at least two ways to do it, depending on how the
circuitry in the Wavetek 75 works:

- If the existing encoder's common wire (the one going to that
sliding contact) was grounded, and if the two quadrature leads were
of the "external pull-up resistor to +5 volts, and the encoder
pulls the quadrature output down to ground", then dropping in an
optical encoder could be fairly easy. Rotary optoencoders
generally need a +5v feed for their LED (many have build-in current
limiting resistors for the LED) and have phototransistor
pull-downs... so, substituting one of these might be as simple as
wiring it in with the common/ground and quadrature-outputs as
before, and adding a single lead for +5 for the LED drive.

- If the existing encoder was wired up any differently, you'd need an
interface board (transistors, relays, or some such) to take the
open-collector output of a rotary optoencoder and drive the Wavetek
circuit as it wants to be driven. Probably not hard to build on a
bit of perfboard once you know what you need to do.

In either case, having a schematic for the Wavetek 75 is likely to be
essential. There's a Wavetek test equipment special-interest mailing
list at YahooGroups - somebody there might have a schematic or even
the whole service manual.

As to specific encoder lines: Grayhill 61K or 61R series, Clarostat
600 series, or one of the Bourns encoders (the Bourns 14mm optos
carries in the new Digi-Key catalog are under $20) might all be
candidates.

--
Dave Platt <dplatt@radagast.org> AE6EO
Friends of Jade Warrior home page: http://www.radagast.org/jade-warrior
I do _not_ wish to receive unsolicited commercial email, and I will
boycott any company which has the gall to send me such ads!

Hi Dave,

The part# on the service manual is no where to be found, neither does it
contain much information:
http://www.stanford.edu/~wonghoi/DC/4600-05-0211.png

I just probed the connections to the Wavetek 75 board, and it seems like one
wire is grounded while two are hooked up to +5V (My mechanical encoder was
too broken to tell which lead is which). Is there any opto-encoder that
you'll recommend for this case?

Thanks.

Cheers,
Hoi

<http://search.digikey.com/scripts/DkSearch/dksus.dll?Cat=1114163;keywords=encoder>
shows 549 encoders to chose from.

the mecanical encoders use pullup resistors, and the switch pulls the
+5 VDC lines low as the shaft is turned. The interface may need to be
modified to use the electronic encoders. The electronic encoders I've
used had a direction output, and a pulse output which will not work with
that Wavetek design.

The scan from the manual indicates that the part was customized for
Wavtek. We used a similar design in a product, 20 years ago. I have no
idea how many steps the detent has, per turn. The damaged encoder has
the Panasonic logo, but that body style is no longer made.


<http://www.panasonic.com/industrial/components/electromechanical/elec_encd.htm>
shows the current body styles, and something might fit, if you can
determine the shaft length, style and number of detents,

Personaly, I would see if any of the CTS encoders would fit. They
are a lot higher quality, and were used in some of our military grade
designs. The only bad ones I saw were damaged in shipping, when some
idiot returned equipment in a wood crate, and no packing material. The
equipment had over $2000 physical damage after being shipped half way
around the world so it was no wonder the shaft was pushed back through
the encoder's body.


--
http://improve-usenet.org/index.html

If you have broadband, your ISP may have a NNTP news server included in
your account: http://www.usenettools.net/ISP.htm


There are two kinds of people on this earth:

The crazy, and the insane.

The first sign of insanity is denying that you're crazy.

 

[Franc Zabkar]

On Tue, 12 Aug 2008 16:47:40 -0700, "Hoi Wong" <wonghoi@stanford.edu>
put finger to keyboard and composed:

"Franc Zabkar" <fzabkar@iinternode.on.net> wrote in message
news:kh24a4lqus9cr0ahab02ura0h6rgruea5b@4ax.com...
On Wed, 13 Aug 2008 08:00:00 +1000, Franc Zabkar
fzabkar@iinternode.on.net> put finger to keyboard and composed:

The above encoders are optical types, so you will need to wire up a 5V
DC supply.

I notice that there are three unused wires in the J12/J13 harness in
the Model 23. If the Model 75 also has unused conductors, then maybe
you could utilise them for bussing the +5V supply to the optical
encoder.

- Franc Zabkar

Thanks a lot Franc. I found a convenient +5 from CR2 (left part of battery
test schematic on p.126) to steal for optoencoders too. By the way, am I
supposed to wire the switch terminals 1,2 to the ground in this case?

Cheers,
Hoi

AIUI, if you use one of the Bourns EM14 series optical encoders, then
you would wire the original 1 & 2 terminals of your mechanical encoder
to the Channel A and Channel B output pins of your optical encoder. If
you get the direction wrong, then you just swap the A & B terminals.

As for the 1 & 2 switch terminals of your optical encoder, I would
think that you would leave them unconnected.

- Franc Zabkar
--
Please remove one 'i' from my address when replying by email.

 

[Franc Zabkar]

On Tue, 12 Aug 2008 16:47:40 -0700, "Hoi Wong" <wonghoi@stanford.edu>
put finger to keyboard and composed:

I found a convenient +5 from CR2 (left part of battery
test schematic on p.126) to steal for optoencoders too.

I notice that the optical encoder's output is rated to sink about
20mA. However, the pullup resistors are 221K which would result in a
current draw of only 20uA. Perhaps it would be advisable to change the
pullups to 2.2K in order to bring the collector current into the 2mA
range ???

- Franc Zabkar
--
Please remove one 'i' from my address when replying by email.

 

[Franc Zabkar]

On Tue, 12 Aug 2008 11:47:29 -0700, "Hoi Wong" <wonghoi@stanford.edu>
put finger to keyboard and composed:

The part# on the service manual is no where to be found, neither does it
contain much information:
http://www.stanford.edu/~wonghoi/DC/4600-05-0211.png

I just probed the connections to the Wavetek 75 board, and it seems like one
wire is grounded while two are hooked up to +5V (My mechanical encoder was
too broken to tell which lead is which). Is there any opto-encoder that
you'll recommend for this case?

Thanks.

Cheers,
Hoi

Page 121 of the Model 23 manual shows that pin 3 of SW1 is grounded
and pins 1 and 2 are switched to ground. The switched pins connect to
IN1-0 and IN1-2 of the uProc/Synthesizer PCB (page 126) via pins 1 and
2 of J13/J12. These inputs are pulled up to +5V via R21 and R22.

- Franc Zabkar
--
Please remove one 'i' from my address when replying by email.

 

[Franc Zabkar]

On Wed, 13 Aug 2008 07:42:14 +1000, Franc Zabkar
<fzabkar@iinternode.on.net> put finger to keyboard and composed:

On Tue, 12 Aug 2008 11:47:29 -0700, "Hoi Wong" <wonghoi@stanford.edu
put finger to keyboard and composed:

The part# on the service manual is no where to be found, neither does it
contain much information:
http://www.stanford.edu/~wonghoi/DC/4600-05-0211.png

I just probed the connections to the Wavetek 75 board, and it seems like one
wire is grounded while two are hooked up to +5V (My mechanical encoder was
too broken to tell which lead is which). Is there any opto-encoder that
you'll recommend for this case?

Thanks.

Cheers,
Hoi

Page 121 of the Model 23 manual shows that pin 3 of SW1 is grounded
and pins 1 and 2 are switched to ground. The switched pins connect to
IN1-0 and IN1-2 of the uProc/Synthesizer PCB (page 126) via pins 1 and
2 of J13/J12. These inputs are pulled up to +5V via R21 and R22.

How about this 2-bit, 64 pulses/rev rotary encoder:
http://www.digikey.com/scripts/DkSearch/dksus.dll?Detail?name=EM14C0D-E28-L064N-ND

Datasheet:
http://www.bourns.com/pdfs/em14.pdf

Digikey have 2 other suitable encoders in that same series that I can
see.

The above encoders are optical types, so you will need to wire up a 5V
DC supply.

- Franc Zabkar
--
Please remove one 'i' from my address when replying by email.

 

[Franc Zabkar]

On Wed, 13 Aug 2008 08:00:00 +1000, Franc Zabkar
<fzabkar@iinternode.on.net> put finger to keyboard and composed:

The above encoders are optical types, so you will need to wire up a 5V
DC supply.

I notice that there are three unused wires in the J12/J13 harness in
the Model 23. If the Model 75 also has unused conductors, then maybe
you could utilise them for bussing the +5V supply to the optical
encoder.

- Franc Zabkar
--
Please remove one 'i' from my address when replying by email.