Hypex SMPS1200 A700

[AP2]

Originally Posted by ekaerin

Yes, when I looked at the board before I assembled it I remember there was a very standard PWM controller. So it should be quite easy to go from there and locate the SS cap. Fiddling with the current limit directly is more risky since it probably took a lot of testing, including thermal, to get it right.
Even if the SS cap is involved once the current limit trips "hard" making the recovery a bit longer seems OK for what I know. As long as it can be discharged fast enough that is.
Also there are margins in time before the amp boards enable.

Regards / Mattias

Hi,
I do not know this "PSU" but from experience I can deduce cke startup circuit is simply a resistor that feeds the chips directly on high-voltage cell. this is not wrong even if the charging time (flashing) depend on the load (including capacitors). problem can occur in off, this circuit is not able to immediately turn off. continuously generates on / off until the cell is discharged (if the chip does not have a high UVLO) (but if have a high UVLO,then can not generate the ramp to the pseudo soft-start)
A good soft-start (fixed time) with current ceck while start-up, required different technique.

Regards

[Jan-Peter]

Originally Posted by ekaerin

Hi All,

While we have Hypex representatives on the line.
I would like Hypex to specify, in the datasheet, the maximum external capacitance allowed.

Although you don't need to add other caps than the HV+ output storage on the PSU there will always be people (like me)
adding their favorite caps.

In my case cause I had them already and could squeeze them in...
What happened is, my guess, that the current limit kicks in
(fold back or hic-up) making the start up ramp non monotonic.

See my latest Hypex project:
Hypex UCD400 OVP problem

Thanks / Mattias

Hi Mattias,

You don't have to add more caps. The SMPS has a synchronical rectifying...., so our caps are on the primairy side...

If you add too much caps external due to charge current the SMPS will go in current protection.

Jan-Peter

[ekaerin]

Hi Jan-Peter,

Thanks for your reply. I know the main storage is on the primary, regardless of sync rectification or not. I also can assume the thru-put impedance in the PSU powertrain is low.
(The sync FET rectifier adds sink capability preventing rail pumping, and may improve output impedance).
Voltage drop will be corrected by the amps NFB. SO, YOU ARE PROBABLY RIGHT.

Nevertheless, you know how people are. Adding whatever cap they have at hand or can fit in the case.
I had hard times to fit mine in the case but did so anyway....

Therefore adding a note in the datasheet about external caps should save you some questions.

Regards / Mattias

[Jan-Peter]

Originally Posted by ekaerin

Hi Jan-Peter,

Thanks for your reply. I know the main storage is on the primary, regardless of sync rectification or not. I also can assume the thru-put impedance in the PSU powertrain is low.
(The sync FET rectifier adds sink capability preventing rail pumping, and may improve output impedance).
Voltage drop will be corrected by the amps NFB. SO, YOU ARE PROBABLY RIGHT.

Nevertheless, you know how people are. Adding whatever cap they have at hand or can fit in the case.
I had hard times to fit mine in the case but did so anyway....

Therefore adding a note in the datasheet about external caps should save you some questions.

Regards / Mattias

Actually in our experience adding too much caps will not improve sonic performance. Increasing caps is also increasing short peak charge currents, for best sonic performance you don't want to have those small high peak currents in your system....

[ekaerin]

OK,
Worth a try. Stored energy is proportional to voltage squared....
I'll take the time to test and listen without the caps. Without the extra caps I'll be able to add some clever cooling architecture.
It does run a little bit warmer than I would like it to.

I've still have to figure out the OVP problem though: Hypex UCD400 OVP problem

Regards / Mattias

[Jan-Peter]

Originally Posted by ekaerin

OK,
Worth a try. Stored energy is proportional to voltage squared....
I'll take the time to test and listen without the caps. Without the extra caps I'll be able to add some clever cooling architecture.
It does run a little bit warmer than I would like it to.

I've still have to figure out the OVP problem though: Hypex UCD400 OVP problem

Regards / Mattias

Please contact Frank (support (at) hypex (dot) nl), he can advise you...

[ekaerin]

I will. Thank you. / Mattias

[Eva]

Originally Posted by AP2

Hi,
I do not know this "PSU" but from experience I can deduce cke startup circuit is simply a resistor that feeds the chips directly on high-voltage cell. this is not wrong even if the charging time (flashing) depend on the load (including capacitors). problem can occur in off, this circuit is not able to immediately turn off. continuously generates on / off until the cell is discharged (if the chip does not have a high UVLO) (but if have a high UVLO,then can not generate the ramp to the pseudo soft-start)
A good soft-start (fixed time) with current ceck while start-up, required different technique.

Regards

Improve your english. We were talking about PWM soft start, not about controller bias supply.