HVPSFB kit short current at primary

Hi John,

Can you please explain your input-output connections and the start-up procedure that you are following? Are you running the default configuration (boot from Flash) and then connect to the controller using the GUI or are you running from CCS? Are you using an external 12V bias supply or relying on the on-board aux power?

Which user guide are you following - the GUI quick start guide or the CCS user guide?

Thank you.

Hrishi

Hi Hrishi,

Thanks for your prompt reply.

Input: 390v DC from a Lab power supply;

output: open or with 25ohm load;

Boot: from flash;

not using GUI,just let the demo runs itself when testing the start-up short ciruit at beginning, later we tried using CCS5 to change SR_mode and rebulid and program the chip;

No extern 12v supply, using on-board 400-12v module;

following HVPSFB User's guide v1.0

it was actually discovered in our our prototypes developed from your kit,then we found this short ciruit at start-up issue, so we consult our golden samples of this demo unit and we found similar short ciurcuit problems.

everything we found in our prototypes was reproduced in the demo kit: short circuit at start-up(mostly the primary which blows one of the legs),shoot over 18v in snchrounous mode 0), no voltage regulation in synchorouss mode 0, etc.

pelase let me know if more inofrmation is needed.

Thank you again for your help.

Hi John,

I understand your test conditions now. When I say low load, I mean less than 0.3A on the HVPSFB kit (it could be lower, but this number is a bit conservative). When you are operating under no load or very low load condition, the operation should change to voltage mode control or duty/burst mode. This allows the output voltage to be regulated to ~12V. With PCMC, there is a need for blanking. This blanking implies a minimum duty value for the power stage. Under very low load conditions, this minimum duty value might be enough to make the o/p voltage go very high (like 18V in your case). This high voltage can be disastrous for the power stage.

Unfortunately, this change from PCMC to VMC/duty mode and back has not been implemented in the default software. This has been done before but here it is left up to the users to implement their own algorithms for this function.

I hope this helps.

Hrishi

Hi Hrishi,

1)what you said is just applicable to mode 0 or for all modes, 1 and 2 as well?

2)changing from PCMC to VMC needs hardware changes( jumper changes), which is impossible for a unit to behave so differently between low and heavy load.  how can we implement two modes for the same unit without hardware changes?

3)the short current observed is not only in mode 0, but in mode 1 and 2 in the demo kit as well; what;'s the possible casue for this?

Thank you,

Hi,

What I said is applicable to all SR modes. However, it also depends on the power stage and switching frequency. For the HVPSFB kit, SR mode 2 operation applies enough load for the output to be regulated at ~12V, even under no load.

Changing from PCMC to VMC is possible without the hardware changes, but involves re-configuring PWM modules correctly. This along with need to ensure smooth mode transitions (PCMC <--> VMC/duty) gets very tricky to implement. The actual implementation is also depended on the complete system implementation.

I believe that running the kit in SR_mode 0 under no-load could have damaged the secondary side MOSFETs and drivers (especially with the output voltage going to 18V).

Hrishi

Hi Hrishi,

Please see if I understand your answer correctly:

1)the minimum load for both mode 0  and mode 1 is 0.3A;

2)no minimum load requirement for mode 2,i.e. output could be open;and you saidi mode 2 applies enoug load for the output, why?

3)you said changing from PCMC to VMC doesn't need hardware changes, however, VMC requires capacitors to block DC current while PCMC doesn't not, so this does requires a jumper to swap between PCMC and VMC modes.

with the capacitors in the circuit, it cannot work at PCMC mode.si this correct?

Thank you,

John

Hi John,

1. Yes.

2. In mode 2, part of the PWM cycle is spent in freewheel mode. This provides a low impedance path for the current and reduces the voltage build-up on the transformer secondary. This is enough on the HVPSFB board running at 100 KHz.

3. Voltage mode control at very low load should not cause any substantial flux imbalance of much consequence. As soon as you go above a certain (say 0.3A) load, the system will start operating in PCMC again. If this is still a concern for your system, you may use an inner average current loop in addition to the outer voltage loop (only when operating at very low load conditions).

Hrishi

Hi Hrishi,

Thank you very much for the help.

In your demo unit, to swap between PCMC and VMC, both hardware and firmware are needed to be changed as PWMs configuration are very different, as you said we cannot use VMC mode firmware to do VMC with the actual hardwired configuration for PCMC.

I feel a bit strange for this , why not making the same PWMs confuguration in circuit in order to switch the two modes easily by just swapping firmware then?

Does TI have demo code for doing VMC using the hardwired connection configued for PCMC?

Does TI have the demo code  for swapping bewteen VMC and PCMC modes according to load current?

Thank you very much,

John 

Hi John,

I am really sorry for the late reply.

The two different configurations for VMC and PCMC make sure that the inductor 'L2' placement does not change relative to the leading and lagging legs of the full bridge. Unfortunately we do not have a demo code (to switch between VMC and PCMC) that can be shared at this time.

Hrishi

Hi Hrishi,

thank you for your reply.

I mean, by reconfiguring PWMs in firmware, you could always get what you want, say, in this case, you want the same legs, right?

why need hardware jumpers to reconfigure those PWMs then?

Thanks,

John,

The difference being PWM1 can synchronize PWMs down the chain (PWM2, PWM3…) but PWM2 cannot synchronize PWM1. Yes, I agree that in the firmware you can overcome this but not without adding some code. It was easiest for us to make this work with the hardware configuration change without needing to change our power library (or adding additional library modules).

Hrishi

Hi Hrishi,

Thank you for your reply.

what about my first post in this thread, short legs at startup?did you ever see it?

Thank you,

Hrishi,

Could you please have a look at another post:

http://e2e.ti.com/support/microcontrollers/tms320c2000_32-bit_real-time_mcus/f/171/p/226693/802942.aspx#802942

VMC mode never works.

Are there any typoes in the documents I followed ?user guide 1 and quick user guide.

Thank you,

John,

Can you look at this post? http://e2e.ti.com/support/microcontrollers/tms320c2000_32-bit_real-time_mcus/f/171/p/221673/784389.aspx#784389

It's possible that you have the same problem. Make sure that C7 on the controller card does not suffer from bad solder. Also make sure that there is a jumper populated on pins 1 and 2 of J4.

Hrishi

Hrishi,

I don't have this problem.

My two TI demo kits work in PCMC mode but not work in VMC mode.

thank you,