How to digitally adjust output voltage of Taylor Edge HVPS 1363

237 views
Skip to first unread message

jb-electronics

unread,
Sep 7, 2014, 12:52:40 PM9/7/14
to neoni...@googlegroups.com
Hi folks,

I want to adjust the output voltage of the Taylor Edge HVPS 1363 [1]
digitally via a microcontroller. I drive the HVPS with 12V to have
maximum output current.

I know that I could just select the voltage with a couple of switches
and resistors, but is there a way to do that digitally?

Thanks for your input,
Jens

[1] http://tayloredge.com/storefront/SmartNixie/PSU/index.html


David Forbes

unread,
Sep 7, 2014, 1:04:16 PM9/7/14
to neoni...@googlegroups.com
Jens,

You can use an analog multiplexer chip to switch in different resistors
using logic signals. The ADG series form Analog Devices is good, if a
bit expensive. There are other lower-cost options if you want to design
a bit of circuitry.


--
David Forbes, Tucson AZ

H. Carl Ott

unread,
Sep 7, 2014, 1:16:08 PM9/7/14
to neoni...@googlegroups.com
Digital Potentiometer?
Lots of different types depending on what kind of digital interface / resolution you need.



carl
--------------------------------------------------------
Henry Carl Ott   N2RVQ    hcar...@gmail.com


jb-electronics

unread,
Sep 7, 2014, 1:30:42 PM9/7/14
to neoni...@googlegroups.com
Hey David, hey Carl, thanks for your input. I would like to avoid a resistor approach due to the 1/R characteristic of the output voltage...

Jens


--
You received this message because you are subscribed to the Google Groups "neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email to neonixie-l+...@googlegroups.com.
To post to this group, send email to neoni...@googlegroups.com.
To view this discussion on the web, visit https://groups.google.com/d/msgid/neonixie-l/CAP_hP_QwwgBhhr5M_zd7fD%2BVPtAyyrUiQUsx0tcxzGmvJANSTQ%40mail.gmail.com.
For more options, visit https://groups.google.com/d/optout.

David Forbes

unread,
Sep 7, 2014, 2:09:43 PM9/7/14
to neoni...@googlegroups.com
Jens,

Two things:

1. Make sure you know the maximum voltage on that resistor adjust pin,
and select the control chip accordingly.

2. The control will be non-linear no matter what, so use a lookup table
to convert desired voltage to the resistor switch or digital pot value
or whatever.


On 9/7/14 10:31 AM, jb-electronics wrote:
> Hey David, hey Carl, thanks for your input. I would like to avoid a
> resistor approach due to the 1/R characteristic of the output voltage...
>
> Jens
>
>
>>
>>
>> On Sun, Sep 7, 2014 at 12:53 PM, jb-electronics
>> <webm...@jb-electronics.de <mailto:webm...@jb-electronics.de>> wrote:
>>
>> Hi folks,
>>
>> I want to adjust the output voltage of the Taylor Edge HVPS 1363
>> [1] digitally via a microcontroller. I drive the HVPS with 12V to
>> have maximum output current.
>>
>> I know that I could just select the voltage with a couple of
>> switches and resistors, but is there a way to do that digitally?
>>
>> Thanks for your input,
>> Jens
>>
>> [1] http://tayloredge.com/storefront/SmartNixie/PSU/index.html
>>
>>
>>
>>
>> Digital Potentiometer?
>> Lots of different types depending on what kind of digital interface /
>> resolution you need.
>>
>>
>>
>> carl
>> --------------------------------------------------------
>> Henry Carl Ott N2RVQ hcar...@gmail.com <mailto:hcar...@gmail.com>
>>
>>
>> --
>> You received this message because you are subscribed to the Google
>> Groups "neonixie-l" group.
>> To unsubscribe from this group and stop receiving emails from it, send
>> an email to neonixie-l+...@googlegroups.com
>> <mailto:neonixie-l+...@googlegroups.com>.
>> To post to this group, send email to neoni...@googlegroups.com
>> <mailto:neoni...@googlegroups.com>.
>> <https://groups.google.com/d/msgid/neonixie-l/CAP_hP_QwwgBhhr5M_zd7fD%2BVPtAyyrUiQUsx0tcxzGmvJANSTQ%40mail.gmail.com?utm_medium=email&utm_source=footer>.
>>
>> For more options, visit https://groups.google.com/d/optout.
>


jb-electronics

unread,
Sep 7, 2014, 2:25:11 PM9/7/14
to neoni...@googlegroups.com
Thanks. I will do some experiments.

In the meantime I relaized that a logarithmic 47k potentiometer can be
used to mimick a reasonably linear behaviour between 170V..200V in terms
of the adjusted potentiometer angle. This is not digital of course, but
fun to imagine.

Jens

Arne Rossius

unread,
Sep 7, 2014, 6:55:32 PM9/7/14
to neoni...@googlegroups.com
Hi,

jb-electronics wrote:
> I want to adjust the output voltage of the Taylor Edge HVPS 1363 [1]
> digitally via a microcontroller. I drive the HVPS with 12V to have
> maximum output current.

If I understand it correctly, the output voltage is a function of the
resistor R_Adjust and the voltage it is connected to, V_Radjust. It
seems that chosing the resistor for maximum output voltage (0 Ohms) and
then increasing the voltage decreases the output voltage, so it might be
possible to use a well-filtered PWM output to generate a voltage for
adjustment. For an adjustment range between 150V and 200V, it seems that
the necessary voltage is between 0V (for 200V output) and 1.27V (for
150V output), so well within the limits of what can be generated from a
PWM output. Depending on the input current of the adjust pin, an opamp
buffer might be necessary as well.


Best Regards,
Arne

John Rehwinkel

unread,
Sep 7, 2014, 7:05:29 PM9/7/14
to neoni...@googlegroups.com
> If I understand it correctly, the output voltage is a function of the
> resistor R_Adjust and the voltage it is connected to, V_Radjust. It
> seems that chosing the resistor for maximum output voltage (0 Ohms) and
> then increasing the voltage decreases the output voltage, so it might be
> possible to use a well-filtered PWM output to generate a voltage for
> adjustment.

I suspect that wouldn't work the way you want. I think that resistor parallels
part of the voltage divider in the feedback loop. By introducing a separate
voltage, you would reduce the sensitivity of the feedback loop and thereby
produce poorer regulation.

- John

Arne Rossius

unread,
Sep 7, 2014, 8:27:48 PM9/7/14
to neoni...@googlegroups.com
Hi,

John Rehwinkel wrote:
> I suspect that wouldn't work the way you want. I think that resistor parallels
> part of the voltage divider in the feedback loop. By introducing a separate
> voltage, you would reduce the sensitivity of the feedback loop and thereby
> produce poorer regulation.

Possibly, yes. But it looks like this is an "official" way to change the
output voltage, especially for voltages below 150V, so I expect the
regulation won't get too poor. I don't own one of the modules in
question though, so OP might want to try if it works or not by supplying
an external voltage from a lab supply or similar.


Best Regards,
Arne

taylorjpt

unread,
Sep 8, 2014, 1:50:24 PM9/8/14
to neoni...@googlegroups.com
With a processor the best way to adjust the output voltage is via a PWM (Or a DAC if you want to get fancy).

Since a voltage into the adjustment resistor results in a linear change in current (The other side of the resistor is maintained at a constant 1.24V) in the feedback network, the output voltage changes linearly as well with a a change in duty cycle.  Unlike control with a resistor which can only drive the output down to 150V, a PWM at 5V can drive the output to zero:  Voltages below 1.24V result in a increase in the output where voltages above 1.24V result in a decrease.  The main point to remember is that the cutoff frequency of the PWM filter must be well below the operating frequency of the switcher which can be as low as 10kHz when operating in PFM mode during light loading.

This spreadsheet illustrates the circuitry and relationship between duty cycle and output voltage. 
http://www.tayloredge.com/storefront/SmartNixie/PSU/1363_PWMAdjust.xls

jt

Jon

unread,
Sep 8, 2014, 1:56:28 PM9/8/14
to neoni...@googlegroups.com

Hi Jens,

At the risk of stating the bleeding obvious... Have you asked John Taylor directly? He is a member of this group, and may well stop by shortly anyway, but if not there's a contact route through his website (http://www.tayloredge.com/). He's always been responsive when I've had questions / needed support. And I'm sure nobody knows the circuit better than him.

Jon.

Jon

unread,
Sep 8, 2014, 2:14:37 PM9/8/14
to neoni...@googlegroups.com

On Monday, September 8, 2014 6:56:28 PM UTC+1, Jon wrote:

At the risk of stating the bleeding obvious... Have you asked John Taylor directly? He is a member of this group, and may well stop by shortly anyway


Well, it seems that was prophetic! At least at the time I was typing it, anyway :)

Jon.

taylorjpt

unread,
Sep 9, 2014, 2:43:50 PM9/9/14
to neoni...@googlegroups.com
You can also build your own DAC with an R2R network controlled by GPIOs that will give you a linear Radj = Vout voltage vs binary code.  You can buy R2R ladders off the shelf:  http://www.digikey.com/product-detail/en/4610X-R2R-103LF/4610X-R2R-103LF-ND/3787967

jt

taylorjpt

unread,
Sep 9, 2014, 4:44:50 PM9/9/14
to neoni...@googlegroups.com
When choosing your network resistors for an R2R ladder, remember that the input impedance is always R, and this R vs the 30.1K series resistor in the module sets the minimum and maxim adjustment voltages.  1K is a good value to use for R since you can drive the voltage from 5V to 199V with a 5V GPIO rail and 2R=2K=2.5mA is well within the capability of most GPIOs.   The part linked below with a 10K R would allow adjustment from 37V to 188V at 1/10 the processor current as a 1K network.

The next choice is how many GPIOs do you want to use, where the granularity is simply Span/(2^N):  A 1K R2R network and 6 GPIOs gives you 3volts/step from 5V to 199V.

jb-electronics

unread,
Sep 9, 2014, 4:48:30 PM9/9/14
to neoni...@googlegroups.com
Hello John, thanks for your detailed input. And Jon, yes, you were quite prophetic. Eerie...

So I can set the output voltage with PWM, that is convenient. I don't know enough about switch mode power supply feedback loops to understand why it works like that exactly, though, but for now I will take it. Thanks for your spreadsheet!

I will probably build an 8bit R2R network and use a shift register to create a small DAC that way.

Now that I have you here, can you answer me another question? I was wondering if the circuitry couldbe tweaked as to allow a maximum output voltage of 250V or even 300V. I will only need a maximum current of 10mA since I want to build a testing stand for single Nixie tubes.

Jens




Am 09.09.2014 20:43, schrieb taylorjpt:
You can also build your own DAC with an R2R network controlled by GPIOs that will give you a linear Radj = Vout voltage vs binary code.  You can buy R2R ladders off the shelf:  http://www.digikey.com/product-detail/en/4610X-R2R-103LF/4610X-R2R-103LF-ND/3787967

jt

--
You received this message because you are subscribed to the Google Groups "neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email to neonixie-l+...@googlegroups.com.
To post to this group, send email to neoni...@googlegroups.com.
To view this discussion on the web, visit https://groups.google.com/d/msgid/neonixie-l/fbad3290-2b97-4c81-bd96-3bf778595768%40googlegroups.com.

jb-electronics

unread,
Sep 9, 2014, 4:51:04 PM9/9/14
to neoni...@googlegroups.com
Addendum: 8bit is of course a bit optimistic for R2R, so maybe I will just use a chip with built-in DAC.
Message has been deleted

taylorjpt

unread,
Sep 9, 2014, 9:15:28 PM9/9/14
to neoni...@googlegroups.com
>>Now that I have you here, can you answer me another question? I was wondering if the circuitry could be tweaked as to allow a maximum output voltage of 250V or even 300V. I will only need a maximum current of 10mA since I want to build a testing stand for single Nixie tubes.<<

If you replace the 30.1K Radj resistor on the module with a 15K resistor (Or put another 30K resistor in parallel) this will change the max feedback current from 165uA (Rbottom = 10K || 30.1K = 7.5K) to about 200uA (Rbottom = 10K || 15K = 6K) or Vmax goes from 200V to 250V.  The main output capacitor is rated 250V and since it is a good quality part you should have no problem operating it at 250V.  300V is pushing it but for a test fixture you should be OK for short periods.  To get 300V, replace the 30.1K resistor with 10K (Rbottom = 10K || 10K = 5K) to push the feedback current to 240uA.  In all cases, the Radj=Open voltage will be about 150V.

At 300V the output diode will be seeing about 420V (600V limit) for a 12V input and the main switch will be seeing about 42V (60V limit).  At voltages above 200V both the transistor and diode will incur higher switching losses (The diode because of itself and the transistor because of the diode) so keep an eye on the dissipation of both those devices.

jb-electronics

unread,
Sep 12, 2014, 8:07:24 AM9/12/14
to neoni...@googlegroups.com
Thanks for the detailed description, John. I will do some experiments with the module I have here and report back. I am quite busy at university at the moment so that may take a couple of weeks.

Thanks again!

Jens
--
You received this message because you are subscribed to the Google Groups "neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email to neonixie-l+...@googlegroups.com.
To post to this group, send email to neoni...@googlegroups.com.
Reply all
Reply to author
Forward
0 new messages