bipolar digital pot

[Martin Klingensmith]

I have an analog circuit that has a (normally) negative polarity DC
feedback signal. I'd like to change the gain of the circuit digitally,
so I'd like to use a digital pot. The problem is that a digital pot can
only take signals between VEE and VDD - most usually probably 0v to 5v.

My rails are +5/0/-5 and my feedback lies somewhere between 0 and -5.
Has anyone encountered a digital pot that can either take a 10v supply,
or has level shifters on the digital lines? I think I might be able to
use a CD4052 with discrete resistors, but then it's a 16 pin chip I have
to put on the board.

Any ideas?

-
Martin
--
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist

[Vasile Surducan-2]

On 5/8/08, Martin <martin@...> wrote:
> I have an analog circuit that has a (normally) negative polarity DC
> feedback signal. I'd like to change the gain of the circuit digitally,
> so I'd like to use a digital pot. The problem is that a digital pot can
> only take signals between VEE and VDD - most usually probably 0v to 5v.
>
> My rails are +5/0/-5 and my feedback lies somewhere between 0 and -5.

So? Who stops you to supply your potentiometer driver (assuming it's a
PIC) between 0 and -5V ?

> Has anyone encountered a digital pot that can either take a 10v supply,
> or has level shifters on the digital lines? I think I might be able to
> use a CD4052 with discrete resistors, but then it's a 16 pin chip I have
> to put on the board.
>
> Any ideas?
>
> -
> Martin
> --
> http://www.piclist.com PIC/SX FAQ & list archive
> View/change your membership options at
> http://mailman.mit.edu/mailman/listinfo/piclist
>

--
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist

[M. Adam Davis-2]

You have a few options outside your chosen solution, most of which can
be explored via this google search:
http://www.google.com/search?q=digital+gain+control

First, you might look at amplifiers with built int digital gain
control, such as these Linear Technology parts:
http://www.linear.com/pc/viewCategory.jsp?navId=H0,C1,C1154,C1009,C1121
If the link doesn't work, navigate to
Home  >  Products  >  Signal Conditioning  >  Operational Amplifiers
(Op Amps)  >  Programmable Gain Amplifiers

Second, you might look into CMOS digital switches - some have a 20v range:
http://schematics.blogspot.com/2005/01/digital-gain-control-of-opamp.html

Of course, if you really want to kludge something together, tie a
servo to a pot...  ;-D

Good luck!

-Adam

On 5/8/08, Martin <martin@...> wrote:

> I have an analog circuit that has a (normally) negative polarity DC
> feedback signal. I'd like to change the gain of the circuit digitally,
> so I'd like to use a digital pot. The problem is that a digital pot can
> only take signals between VEE and VDD - most usually probably 0v to 5v.
>
> My rails are +5/0/-5 and my feedback lies somewhere between 0 and -5.
> Has anyone encountered a digital pot that can either take a 10v supply,
> or has level shifters on the digital lines? I think I might be able to
> use a CD4052 with discrete resistors, but then it's a 16 pin chip I have
> to put on the board.
>
> Any ideas?
>
> -
> Martin
> --
> http://www.piclist.com PIC/SX FAQ & list archive
> View/change your membership options at
> http://mailman.mit.edu/mailman/listinfo/piclist
>

--
EARTH DAY 2008
Tuesday April 22
Save Money * Save Oil * Save Lives * Save the Planet
http://www.driveslowly.org
--
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist

[alan smith]

I'm reviewing a potential project for a client, where his device is inline with the telco tip/ring lines.
   
  Essentially is a speed dial/memory type where it takes the phone line when it goes the phone goes off hook and then returns it back to the telco line.
   
  So, he will want to have it pass Part 65 (thats telco...or is it 68?) and so there is the isolation issues, but he also wants to pull power from the telco line to run the micro and such.
   
  Now, we all know that Ma Bell hates it when something is attached to the telco line, other than her stuff.  So if a device is pulling power off the line, will it pass FCC or get dumped right away.
 
I want to squash his dreams right away rather than get down the road and then find its not going to pass.

       
---------------------------------
Be a better friend, newshound, and know-it-all with Yahoo! Mobile.  Try it now.
--
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist

[Spehro Pefhany]

Quoting Martin <martin@...>:

> I have an analog circuit that has a (normally) negative polarity DC
> feedback signal. I'd like to change the gain of the circuit digitally,
> so I'd like to use a digital pot. The problem is that a digital pot can
> only take signals between VEE and VDD - most usually probably 0v to 5v.
>
> My rails are +5/0/-5 and my feedback lies somewhere between 0 and -5.
> Has anyone encountered a digital pot that can either take a 10v supply,
> or has level shifters on the digital lines? I think I might be able to
> use a CD4052 with discrete resistors, but then it's a 16 pin chip I have
> to put on the board.
>
> Any ideas?

There do exist digital pots which will operate from +/-5V supplies, with
level shifting for the digital bits. AD is one supplier, but they tend
to be more expensive (and higher performance!) than the low end ones. There
might be other suppliers with similar products, but I'm not aware of
any offhand.

Eg. http://www.analog.com/UploadedFiles/Data_Sheets/AD5290.pdf

Best regards,
Spehro Pefhany
--
"it's the network..."                          "The Journey is the reward"
s...@...             Info for manufacturers: http://www.trexon.com
Embedded software/hardware/analog  Info for designers:  http://www.speff.com


--
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist

[Mark Rages]

On Thu, May 8, 2008 at 11:45 AM, Martin <martin@...> wrote:

> I have an analog circuit that has a (normally) negative polarity DC
> feedback signal. I'd like to change the gain of the circuit digitally,
> so I'd like to use a digital pot. The problem is that a digital pot can
> only take signals between VEE and VDD - most usually probably 0v to 5v.
>
> My rails are +5/0/-5 and my feedback lies somewhere between 0 and -5.
> Has anyone encountered a digital pot that can either take a 10v supply,
> or has level shifters on the digital lines? I think I might be able to
> use a CD4052 with discrete resistors, but then it's a 16 pin chip I have
> to put on the board.
>
> Any ideas?
>

I assume the digital lines are 0->+5 V.

The digital pots I've worked with use three lines for their digital
interface.  These three are all inputs.

You can use a PNP transistor for a level shifter.

Instead of ASCII art, I'll try to describe the connection:  The PIC
output  is connected through a 10k resistor to the emitter, the base
is connected to ground, and the collector is connected to the digital
pot's input.  Also connected to the collector is a 100k pulldown
resistor to -5V.

Now when the PIC output is low, no current flows and the digital pot's
input sees -5V (low)

When the PIC output is high, current flows (the transistor saturates)
and the digital pot's input sees 0V (high). (well, maybe 0.2 or 0.3v
in practice)

The 10k could be dispensed with if you use the PIC's weak pull up and
switch from "output low" to "input", taking care to never go "output
high" with the pin.

Regards,
Mark
markrages@gmail
--
Mark Rages, Engineer
Midwest Telecine LLC
markrages@...
--
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist

[David VanHorn-2]

>  So, he will want to have it pass Part 65 (thats telco...or is it 68?) and so there is the isolation issues, but he also wants to pull power from the telco line to run the micro and such.

Part 68


> I want to squash his dreams right away rather than get down the road and then find its not going to pass.

You're allowed a few microamps or so.  You can also draw power when
you're off-hook, or when the line is ringing, to charge up a supercap,
if you can do it without upsetting the impedances.  Constant current
sources are your friend.
--
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist

[Bob Ammerman]

From: "David VanHorn" <microbrix@...>
> You're allowed a few microamps or so.  You can also draw power when
> you're off-hook, or when the line is ringing, to charge up a supercap,
> if you can do it without upsetting the impedances.  Constant current
> sources are your friend.

To make clear what David said:

When the line is on-hook you can draw _very_little current (otherwise the
Telco thinks you have gone off hook). This is a matter of microamps. The
good news is the on-hook (open loop) voltage is pretty high, somewhere
around 48V.

When the line is off-hook you can draw quite a bit more current (but of
course the voltage will be quite a bit lower).

You can also extract power from the ringing voltage (about 20Hz AC, maybe
100V or so). The source impedance is pretty high and the Telco only allows a
certain load on the line (measured in "Ringer Equivalences").

---- Bob Ammerman
RAm Systems

--
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist

[Vasile Surducan-2]

On 5/8/08, Mark Rages <markrages@...> wrote:

> On Thu, May 8, 2008 at 11:45 AM, Martin <martin@...> wrote:
> > I have an analog circuit that has a (normally) negative polarity DC
> > feedback signal. I'd like to change the gain of the circuit digitally,
> > so I'd like to use a digital pot. The problem is that a digital pot can
> > only take signals between VEE and VDD - most usually probably 0v to 5v.
> >
> > My rails are +5/0/-5 and my feedback lies somewhere between 0 and -5.
> > Has anyone encountered a digital pot that can either take a 10v supply,
> > or has level shifters on the digital lines? I think I might be able to
> > use a CD4052 with discrete resistors, but then it's a 16 pin chip I have
> > to put on the board.
> >
> > Any ideas?
> >
>
> I assume the digital lines are 0->+5 V.

If he is designing the whole stuff, you may assume digital lines are 0
-5V (if it's a clever designer) so he don't need any level shifter.

Vasle

>
> The digital pots I've worked with use three lines for their digital
> interface.  These three are all inputs.
>
> You can use a PNP transistor for a level shifter.
>
> Instead of ASCII art, I'll try to describe the connection:  The PIC
> output  is connected through a 10k resistor to the emitter, the base
> is connected to ground, and the collector is connected to the digital
> pot's input.  Also connected to the collector is a 100k pulldown
> resistor to -5V.
>
> Now when the PIC output is low, no current flows and the digital pot's
> input sees -5V (low)
>
> When the PIC output is high, current flows (the transistor saturates)
> and the digital pot's input sees 0V (high). (well, maybe 0.2 or 0.3v
> in practice)
>
> The 10k could be dispensed with if you use the PIC's weak pull up and
> switch from "output low" to "input", taking care to never go "output
> high" with the pin.
>
> Regards,
> Mark
> markrages@gmail
> --
> Mark Rages, Engineer
> Midwest Telecine LLC
> markrages@...
> --
> http://www.piclist.com PIC/SX FAQ & list archive
> View/change your membership options at
> http://mailman.mit.edu/mailman/listinfo/piclist
>

--
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist

[Martin Klingensmith]

Vasile Surducan wrote:

> On 5/8/08, Mark Rages <markrages@...> wrote:
>> On Thu, May 8, 2008 at 11:45 AM, Martin <martin@...> wrote:
>>> I have an analog circuit that has a (normally) negative polarity DC
>>> feedback signal. I'd like to change the gain of the circuit digitally,
>>> so I'd like to use a digital pot. The problem is that a digital pot can
>>> only take signals between VEE and VDD - most usually probably 0v to 5v.
>>>
>>> My rails are +5/0/-5 and my feedback lies somewhere between 0 and -5.
>>> Has anyone encountered a digital pot that can either take a 10v supply,
>>> or has level shifters on the digital lines? I think I might be able to
>>> use a CD4052 with discrete resistors, but then it's a 16 pin chip I have
>>> to put on the board.
>>>
>>> Any ideas?
>>>
>> I assume the digital lines are 0->+5 V.
>
> If he is designing the whole stuff, you may assume digital lines are 0
> -5V (if it's a clever designer) so he don't need any level shifter.
>
> Vasle
>

If I'm clever, eh? First product I designed, I'm the only engineer
working on it - I'm lucky it works at all. Digital has to be ground
referenced.

The input opamp is an LMC6042AIN. I need the extremely low input bias
current spec. so I can't use a controlled-gain amplifier i.e. LTC6910.

I found an SPDT analog switch in SSOP8: TC7W53
<http://www.toshiba.com/taec/components2/Datasheet_Sync//165/7442.pdf>

Should give me the gain range I need.

-
Martin
--
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist

[Harold Hallikainen-2]

The on-hook current limits are VERY low. Here are the US limits from
http://part68.org/SecureDocuments/TIA-968-A-Final.pdf .


4.7.2.1 On-hook resistance, metallic and longitudinal (up to 100 V dc).
The on-hook dc resistance between the tip and ring conductors of a loop
start interface, and between each of the tip and ring conductors and earth
ground, shall be greater than 5 MΩ for all dc voltages up to and
including 100 V.

4.7.2.2 On-hook resistance, metallic and longitudinal (100 V to 200 V dc).
The on-hook dc resistance between tip and ring conductors of a loop start
interface, and between each of the tip and ring conductors and earth
ground shall be greater than 30 kΩ for all dc voltages between 100
and 200 V.


Harold



--
FCC Rules Updated Daily at http://www.hallikainen.com - Advertising
opportunities available!
--
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist

[David VanHorn-2]

> 4.7.2.1 On-hook resistance, metallic and longitudinal (up to 100 V dc).
> The on-hook dc resistance between the tip and ring conductors of a loop
> start interface, and between each of the tip and ring conductors and earth
> ground, shall be greater than 5 MΩ for all dc voltages up to and
> including 100 V.

So at 48V that's almost 10uA, which is enough for memory retention and
limited operation.

> 4.7.2.2 On-hook resistance, metallic and longitudinal (100 V to 200 V dc).
> The on-hook dc resistance between tip and ring conductors of a loop start
> interface, and between each of the tip and ring conductors and earth
> ground shall be greater than 30 kΩ for all dc voltages between 100
> and 200 V.

I'm not sure under what condition this applies, both of these can't be
true in the same case.
But 30k allows 1.6mA

Proviso:  Although on-hook voltage is nominally 48, that will range as
high as 56 in real life.
24V is not uncommon, and I have seen 12V on-hook.


Memory retention by microcontrollers is of course commonly done in
phones with speed dial memories.
--
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist

[Harold Hallikainen-2]

>> 4.7.2.1 On-hook resistance, metallic and longitudinal (up to 100 V dc).
>> The on-hook dc resistance between the tip and ring conductors of a loop
>> start interface, and between each of the tip and ring conductors and
>> earth
>> ground, shall be greater than 5 MΩ for all dc voltages up to and
>> including 100 V.
>
> So at 48V that's almost 10uA, which is enough for memory retention and
> limited operation.
>
>> 4.7.2.2 On-hook resistance, metallic and longitudinal (100 V to 200 V
>> dc).
>> The on-hook dc resistance between tip and ring conductors of a loop
>> start
>> interface, and between each of the tip and ring conductors and earth
>> ground shall be greater than 30 kΩ for all dc voltages between 100
>> and 200 V.
>
> I'm not sure under what condition this applies, both of these can't be
> true in the same case.
> But 30k allows 1.6mA

I think the DC voltages between 100 and 200 is MAYBE during ringing (but,
of course, that's AC on top of DC). I don't know when else it would apply.

Harold


--
FCC Rules Updated Daily at http://www.hallikainen.com - Advertising
opportunities available!
--
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist

[Bob Ammerman]

> I think the DC voltages between 100 and 200 is MAYBE during ringing (but,
> of course, that's AC on top of DC). I don't know when else it would apply.
>
> Harold

I think that they don't want significant current to flow during some kinds
of 'fault' conditions (like a short to a power line maybe?)

This is probably here because many Telco-interfaced devices include MOVs or
similar devices which can present a very low impedance to a voltage spike.

I am guessing that he milli-amp level current allowed at 200V is still below
the real trip current for an off-hook condition.

-- Bob Ammerman
RAm Systems

--
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist

[David VanHorn-2]

> I am guessing that he milli-amp level current allowed at 200V is still below
> the real trip current for an off-hook condition.

You have to draw at least 20mA to guarantee being detected as off-hook.
--
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist

[Bob Ammerman]

>>Bob Ammerman:
>> I am guessing that he milli-amp level current allowed at 200V is still
>> below
>> the real trip current for an off-hook condition.


>From: "David VanHorn" <microbrix@...>
> You have to draw at least 20mA to guarantee being detected as off-hook.

I should have said below the 'maximum guaranteed stay-on-hook current.'

-- Bob Ammerman
RAm Systems

--
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist

[Sean Breheny]

Hi Martin,

On Fri, May 9, 2008 at 10:11 AM, Martin <martin@...> wrote:
>  The input opamp is an LMC6042AIN. I need the extremely low input bias
>  current spec. so I can't use a controlled-gain amplifier i.e. LTC6910.

Why can't you use the 6042 to buffer the signal and then use a
variable-gain amplifier after that?

Sean
--
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist

[Vasile Surducan-2]

On 5/9/08, Martin <martin@...> wrote:

> Vasile Surducan wrote:
> > On 5/8/08, Mark Rages <markrages@...> wrote:
> >> On Thu, May 8, 2008 at 11:45 AM, Martin <martin@...> wrote:
> >>> I have an analog circuit that has a (normally) negative polarity DC
> >>> feedback signal. I'd like to change the gain of the circuit digitally,
> >>> so I'd like to use a digital pot. The problem is that a digital pot can
> >>> only take signals between VEE and VDD - most usually probably 0v to 5v.
> >>>
> >>> My rails are +5/0/-5 and my feedback lies somewhere between 0 and -5.
> >>> Has anyone encountered a digital pot that can either take a 10v supply,
> >>> or has level shifters on the digital lines? I think I might be able to
> >>> use a CD4052 with discrete resistors, but then it's a 16 pin chip I have
> >>> to put on the board.
> >>>
> >>> Any ideas?
> >>>
> >> I assume the digital lines are 0->+5 V.
> >
> > If he is designing the whole stuff, you may assume digital lines are 0
> > -5V (if it's a clever designer) so he don't need any level shifter.
> >
> > Vasle
> >
>
> If I'm clever, eh? First product I designed, I'm the only engineer
> working on it - I'm lucky it works at all. Digital has to be ground
> referenced.
>
> The input opamp is an LMC6042AIN. I need the extremely low input bias
> current spec. so I can't use a controlled-gain amplifier i.e. LTC6910.
>
> I found an SPDT analog switch in SSOP8: TC7W53
> <http://www.toshiba.com/taec/components2/Datasheet_Sync//165/7442.pdf>
>
> Should give me the gain range I need.

Martin, this is not different from any other analogic switch except
maybe the two channels...
There is a mistake on the first page of the datasheet, for switching
analogic signals between -5V to +5V you need Vcc=+5V and Vee=-5V.
On the other hand, if you really need inputs of 2-5fA then TAKE CARE
about any digital feedback into the analogic path.
Working alone is easier than working in a team because all faults are
just yours...you can't blame anybody and no one will blame you...
except the customer if there is any.
--
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist