[SOLVED] Transformer based linear power supply output behavior

[ankitvirdi4]

Hello,

I have designed a linear power supply using a step down transformer, bridge rectifier and a filter capacitor , there on it goes to a shunt regulator TL431 to generate 3V3.

Ratings: Transformer 6V 230mA at 230V AC Input.
Capacitor: 2200uF/16V

The load is a 16 Digit Seven Segment Display module with each segment drawing about 8-9 mA.

There are 8 CCs multiplexed so that at any instant only 1 CC is on which means it sinks about 130-145mA

While validating the supply my observation is that the secondary AC voltage waveform appears to be clamped sinewave instead of a pure sinewave. Because of this Vdc at lower auxillary supply (185VAC) is about 4.36 with a 400mV ripple thus 3V3 has a huge 300mV ripple across it if the secondary waveform were pure sine Vpp would have been higher thus Vdc would been higher enabling operation at rated voltage 185VAC considering 20% tolerance on the power supply of the product.

I have clearly mistaken somewhere, can somebody help by pointing it out.

 

[KlausST]

Hi,

A lot of information, but still some is missing.

I assume your problem is with 5V supply voltage. But you don't tell how you do this.

If there is a ripple on the 3V3 reference, then most probably the current is too low.

About clamped sine voltage: small transformers have relatively high source impedance. Because of the rectifier there is only current, when voltage is high. The current causes the voltage to drop, resulting in the flat top.

Please measure secondary current.

****
Please show us your schematic.

Klaus

 

[chuckey]

Your reservoir capacitor takes a pulse of current only at the peaks of the incoming sine wave. This high current is saturating the transformer core. You need a better transformer or a choke input DC filter.
Frank

 

[ankitvirdi4]

Hi,
About clamped sine voltage: small transformers have relatively high source impedance. Because of the rectifier there is only current, when voltage is high. The current causes the voltage to drop, resulting in the flat top.

Please measure secondary current.

****
Please show us your schematic.

Klaus

Yes, I want to deal with the flat top, how can I correct this , The secondary current is 97mA at 185VAC; 127mA at 230VAC and 138mA at 270VAC.

There is no ripple (20mV) on the (3V3) Vcc after I have about 4.6V at the Vdsp. Earlier the ripple is natural due to voltage drop across the Q1 junction as Vdsp then is 4.36 +/- .4V ripple.

if the correct the transformer core the problem can be taken care of ? This will have to be informed to the transformer supplier.

- - - Updated - - -

Your reservoir capacitor takes a pulse of current only at the peaks of the incoming sine wave. This high current is saturating the transformer core. You need a better transformer or a choke input DC filter.
Frank

Understood, although what considerations should I make while designing the DC filter? Thanks.


Correction in schematic: C41 is not fitted; its only in the schmatic.

 

[KlausST]

Hi,

I wonder why you don´t use a simple voltage regualtor IC..
(because of precision?)

Change R1 to 0R.

Lower R2 value.


Klaus

 

[d123]

Hi,

In my opinion, whatever other problems you have with that circuit - if any, I am not certain devices like the TL431 are at all suitable for wide fluctuations in load. Maybe it's insufficient capacitance or other related, but what stands out to my eyes is the TLC431.

I understood these types of reference (I'm sure it's better to use it and understand it as more a reference than a regulator, it isn't a 7805 etc.) are more usful for setting a reference voltage (or tiny reference current) with ADCs etc. ref pins, or other device ref. voltage pins where there will be no change in load. You could try a Zener in its place.

In my limited experience of experimenting with the 431, I concluded that it really doesn't like changes in load, but is good for ADC refs. And if you think of the datasheet formula, it's for a static condition, and where you can expect x% change with y% change in load as stated in one of the "electrical characteristics" fields (10%, or something like that I think).

Don't remember any "typical application/application hint" schematic for wide change in load, maybe datasheet has one though, you never know.

Good luck!

 

[ankitvirdi4]

Hi,

I wonder why you don´t use a simple voltage regualtor IC..
(because of precision?)

Change R1 to 0R.

Lower R2 value.

Klaus

For Precise Vcc generation; yes I have taken trials with reducing the R2 and R1=0E as per your suggestion which have improved results but they aren't enough.

it is definately the reservior because I just checked the current waveform and concluded that peak current is 300mA this
must be saturating the transformer.

- - - Updated - - -

Hi,

In my opinion, whatever other problems you have with that circuit - if any, I am not certain devices like the TL431 are at all suitable for wide fluctuations in load. Maybe it's insufficient capacitance or other related, but what stands out to my eyes is the TLC431.

I understood these types of reference (I'm sure it's better to use it and understand it as more a reference than a regulator, it isn't a 7805 etc.) are more usful for setting a reference voltage (or tiny reference current) with ADCs etc. ref pins, or other device ref. voltage pins where there will be no change in load. You could try a Zener in its place.

In my limited experience of experimenting with the 431, I concluded that it really doesn't like changes in load, but is good for ADC refs. And if you think of the datasheet formula, it's for a static condition, and where you can expect x% change with y% change in load as stated in one of the "electrical characteristics" fields (10%, or something like that I think).

Don't remember any "typical application/application hint" schematic for wide change in load, maybe datasheet has one though, you never know.

Good luck!

Hi,

Thank you for the insight although I am indeed using the 3V3 for the microcontroller and ADC reference (After sufficient filltering and serpartion as analog rail.)

Although, now on I will consider using a zener and test results. TL431 as shown better result for EFT compliance.

- - - Updated - - -

I would want to clearify that the ripple on 3V3 is a concern only at low auxiliary supply i.e 180VAC-190VAC after this the Vcc rail is very clean. 190VAC onwards and their is no operational problem whatsoever.

As for 230VAC supply product 20% tolerance is a requisite, it is mandatory to make the product work 180VAC onward.

 

[d123]

I think it's that you can't use a regulator there in that circuit, perhaps?

Even so, this application note from Micrel could be useful for future reference, if you fancy searching for it: an-16 "Improving Adjustable Regulator Accuracy" - it places a voltage reference IC (they use a 1.225V one, the 4040 I think, maybe the 4041), ( and also by memory...) on feedback pin and output to the regulator to improve stability and accuracy.

- - - Updated - - -

and in case the other AN is of any use, here you are:

View attachment an-16 Improving Adjustable Regulator Accuracy Micrel.pdf
View attachment an-9 Design Considerations for 5V to 3.3V Pass Regulators Micrel.pdf

 

[KlausST]

Hi,

I often use a 3V3 power supply voltage regualtor (low dropout)

and a 3.0V voltage reference for the ADC. Like MAX6003.


Klaus