Strange noise in DC-DC converter

[pikulin03]

My task was to build very simple buck DC-DC converter with input voltage in the range 15-30V, output - 12V.
The key idea was to build the control circuitry not using control IC chips, but by means of usual OP AMPS and comaprators in order to be free to modify any of its blocks.

The operation frequency was chosen very low to avoid high frequency effects, as the converter has to be used for some aducational purposes.
Amplifier used is the NE5534
Comparator used is the LM311
The ramp generator was build using the simple 555 circuit.
All control circuits are powered from the imput voltage through the 7812CT.

When I first tested the converter- ir seamd OK- the output voltage was about 12V, as Vin was changed from 15 to 30V. But than I looked at the waveforms and was surprised.
ALL waveforms in the converter contained the high-frequency harmonics and rather high spikes.
Including the RAMP voltage, the power voltage of IC's, output error voltage and even input voltage from the ordinary lab DC power supply.
The mentioned spikes cause comparator switch at very high frequences.

These spikes are shown on the pictures I've attached.
I've added bypass capacitors 0.1u between the VCC and groung of every IC, changed the capacitors at 7812CT to higher values. Tried several lab. power supplies- however the spikes are still present!

Maybe somebody has an idea- where do they come from?

I would appreciate any help!

 

[goldsmith]

Hello my friend
Those spikes , are usual . and they created by your inductor , and you can improve , them with using a faster diode , and use an snubber network , although they are not important at your aim.
Best Wishes
Goldsmith

 

[pikulin03]

Thank You for the answer. However I've tried to change the diode to fast FR153 and used snubber- nothing helps, but in my case these are not just spykes, but noise signal, not allowing the comparator to operate normally, causing multiple switching after one general switching cycle.

 

[mister_rf]

Read here:
**broken link removed**
video also:
**broken link removed**

 

[dick_freebird]

The spikes may just be the ground eddy currents across
assembly L,R; they also might be B-field pickup in the probe
ground loop.

 

[goldsmith]

Again Hi
Why you did not use compensation network for your comparator? it is very important thing .
Best Wishes
Goldsmith

---------- Post added at 21:06 ---------- Previous post was at 20:58 ----------

BTW , i'm sure , that your triangle wave , is not good for this aim , because , it has the DC voltage about 1/3 vcc . why you used that way to triangle wave creation? probably , it is your problem.
Respect



---------- Post added at 21:09 ---------- Previous post was at 21:09 ----------

[/COLOR]And for 12 volt , why you didn't use a simple 7812 series , regulator ( why you used , an LM317?)

 

[pikulin03]

Read here:
**broken link removed**
video also:
**broken link removed**

Thank you for the video- it was really interesting and very aducating! Hope it helps me!

---------- Post added at 09:20 ---------- Previous post was at 09:14 ----------

Again Hi
Why you did not use compensation network for your comparator? it is very important thing .
Best Wishes
Goldsmith

---------- Post added at 21:06 ---------- Previous post was at 20:58 ----------

BTW , i'm sure , that your triangle wave , is not good for this aim , because , it has the DC voltage about 1/3 vcc . why you used that way to triangle wave creation? probably , it is your problem.
Respect



---------- Post added at 21:09 ---------- Previous post was at 21:09 ----------

[/COLOR]And for 12 volt , why you didn't use a simple 7812 series , regulator ( why you used , an LM317?)

1. I didn't use the compesation network as the aim of this circuit is to study instabilities in DC-DC converters and theitr effect on overall parameters of the circuit. But not the instabilites of the kind the spikes are.
2.I Thought it would be the easiest way to create the ramp signal and why does the 1/3 vcc matter in this case? In any case- I'll try to study the problem using simple lab. wave generator.
3. I really used the 7812 series- You misunderstood smth.

THANKS!

---------- Post added at 10:08 ---------- Previous post was at 09:20 ----------

Read here:
**broken link removed**
video also:
**broken link removed**

Unfotunately this techique did not solve the situation- those high frequency noises even did not become smaller a bit!

 

[mister_rf]

Poor PCB layout techniques and improper ground referencing schemes for the control circuitry can introduce unexpected frequency noise on the outputs.
To reduce ripple ceramics capacitors must be placed close to the regulator input pins (they have the extremely low ESR that is needed to reduce the ripple voltage amplitude) but this implies multiple capacitors connected in parallel, also add an LC filter at the input control circuits supply pins and very important: signal ground re-wiring.

Read here:

https://www.ti.com/lit/an/slta055/slta055.pdf


Can you attach photos of the PCB circuit you are using?

 

[goldsmith]

My friend
At first step , the compensation network can help to your comparator , to do it's duty , correctly.
And at 2nd step , you can use a relaxation oscillator with not gate such as 40106 ( it should be schmitt trigger not gate ).
and then you can take an integral from that , with a simple BJT and capacitor .
And each easy way , is not ok. don't try to find easiest , way , try to find best way.
Good Luck
Goldsmith