This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

TLC7226: Is it possible to cascade several TLC7226 for higher voltage output?

Part Number: TLC7226
Other Parts Discussed in Thread: DAC8728, DAC7728

I try to simulate different accumulator voltage level. The system I try to test has 3 accumulators connected in series but controlled separately (each has a nominal voltage of 12V). To test all parts like charging, overvoltage etc. I need at least 15.2V per accumulator. My idea was to use 2 TLC7226 as a cascade to simulate one of the accumulators. (see file for circuit diagram).

BatterySimulator.pdf

AGND of one TLC is connected to OUTA of the second. The REF source is a simple 12V battery, one for each TLC. If I use only one TLC it works fine, but if I add the second one (for cascading) the output voltage level goes down, and it looks like the TLC gets damaged. If I switch back to one TLC after that I get the full REF voltage at the output, independend of the data settings. AGND and DGND are not connected to eachother. Do I have to conntect the lowest AGND with DGND? Does this kind of cascading work in principle or is it not supported?

Sunny Regards

Mario

  • I forgot to mention:
    VDD is 12V
    Vss and DGND are 0V
  • Mario,

    In the case of a single TLC7226 with 12V VREF, 12V VDD, 0V VSS / DGND it looks like you're operating well within the defined bounds of the datasheet and it makes sense that you have success in this configuration. The only caveat is maybe to point out that the Recommended Operating Conditions table suggests 4V headroom between VDD and the reference voltage, probably to give headroom for the output amplifier to avoid any non-linearity associated with being so close to the rail.

    When the cascade approach comes in to play it does sound like devices are starting to be damaged given the behavior you've described independent of input code. Fundamentally there will be some problems as the datasheet absolute maximum ratings table describes a maximum of +/-17V difference from AGND to DGND. With this many devices cascading the VDD / AGND analog domain but a shared DGND domain this will eventually be violated.

    I also do not believe that these devices have the sinking capability to regulate their outputs with another DAC with no load Iq at 16mA maximum. The load regulation curves for sinking current seem to fall off pretty quickly.
  • Hi Duke of DACs,

    Thank you for the fast response. Looks like I have to find another solution.

    Sunny regards

    Mario

  • Mario,

    Absolutely - it's what we're here for :) Thank you for the positive feedback.

    Perhaps if you can explain your exact requirements we can help find a solution. It sounded like 15.2V maximum output and probably 10 or so bits of resolution would be sufficient. What kind of current source capability does the output need to have? Any other unique challenges could help us find a solution.
  • Hi Kevin,

    ok, I will try to explain the required functionality :)

    I have to test the software for an operating table. The table works with mains, but it has also to work for at least 1 usual working day with accumulators. For that the table has 3 12V accumulators connected in series (so you get 36V). The voltage of each accumulator (and via this the state of charge) is measured, and the software shows the state of charge to the user. Additionally the software starts recharging the accumulators if the current state of charge falls below a specified value. And, of course, the recharging stops if the accumulator reaches the end-of-charge voltage (which is 15.1V in this case). 8 Bit resolution should be enough.

    My idea was to simulate each accumulator with 2 cascaded DAC's, which does not work because of the voltage difference between analog and digital ground. I could switch to another DAC with higher voltage, then I only need one per accumulator. But I have the same problem because the accumulators are cascaded and so the 3 DAC's will be cascaded too. Wit this solution I still have the problem that I need to connect the digital ground with the gound of the table because the programming of the DAC's is not decoupled.

    My current idea now is to isolate the DAC completely, also decouple the data input, so that the DAC is completely independend (in case of GND potential) like the accumulator is too. But for this I need a DAC that is able to create at least 15.2V (to test the end-of-charge voltage). For me it looks like the DAC8728 or DAC7728 are good candidates to try it out.

    Sunny regards

    Mario

  • Hi Mario,

    In general I think I agree with your appraisal of the situation and the concept to include the isolation., the problem seems to be unique to your application with these cascaded accumulators so there's not really a straight-forward path ahead. The devices you have pointed out are some candidates, also pretty well any MDAC would have similar capability though you would need to include some more external supporting circuits to get it done since output amplifiers are not included in MDACs.
  • Hi Kevin,

    I already figured out that my challenge is quite unique :)  But if it would be easy then everyone can do it :)

    As soon as the DACs arrive I will try my approach. Then I will see if it works.

    Thank you a lot for your help.

    Sunny regards

    Mario