Notifications
Clear all

Linear power supply issue and question

6 Posts
3 Users
0 Likes
342 Views
Fritigern
(@fritigern)
Member
Joined: 2 years ago
Posts: 36
Topic starter  

I have tried to build the linear pawer supply as per the video in 2022.

OK, I didn't have a 2200uF 63V capacitor, I made do with a 2200uF 35V one. And used a 5V regulator (tested beforehand with reliable DC power supply and output tested as 5V).

All done on a breadboard, (which the legs of the big Capacitor made the holes unusable).

The DC Volts reported at the output of the rectifier were 11.2, but when I offed the cap it went up to 14.4. I presume the multimeter doesn't like irregular DC volts so it offers an average, and the capacitor filled in enough troughs to up the volts to 14.4.

The output from the Regulator was 5.19Volts, with or without the second Capacitor.

I guessed the issue may have been due to the max voltage of my capacitor being inadequate, though it was still nearly double the transformer output.

To test the theory I added another big capacitor in parallel- 100uF 100V to help the 2200uf 35V one, and this dropped the output gf the Regulator to 5.15V, so the fault is in the inadequate smoothing of the voltage in to the Regulator.

Question 1 Why did an inadequately smoothed input to the Regulator induce a Regulator output higher than normal output?

2 is there a formula to work out the specs of the Capacitor required to smooth the rectified voltage? The known inputs are AC Voltage and hertz.

Thank you

Fritigern

 


   
Quote
Ron
 Ron
(@zander)
Father of a miniature Wookie
Joined: 3 years ago
Posts: 6662
 

@fritigern Although I will be interested to see somebody with more smarts than me answer your question, I wonder if it matters all that much, 5V+-5% seems acceptable for hobby work, but maybe not good enough for medical though.

First computer 1959. Retired from my own computer company 2004.
Hardware - Expert in 1401, and 360, fairly knowledge in PC plus numerous MPU's and MCU's
Major Languages - Machine language, 360 Macro Assembler, Intel Assembler, PL/I and PL1, Pascal, Basic, C plus numerous job control and scripting languages.
Sure you can learn to be a programmer, it will take the same amount of time for me to learn to be a Doctor.


   
ReplyQuote
Fritigern
(@fritigern)
Member
Joined: 2 years ago
Posts: 36
Topic starter  

@zander, thank you for not responding "simply follow the instructions!". Which I would have done if I had the right component.

it does matter to me because I'm trying to learn this stuff, and to to that I need to be able to make mistakes (as above) and learn from them so that I can avoid them next time. In the above failure I didn't end up with 5v, I ended with a ratty voltage which my multimeter said was 5.19. i didn't need to check it with my oscilloscope because I knew from my test reading (9vDC in 5v out) and the change when I added more capacitor strength after the rectifier, that the regulator outputted a non stable voltage without the capacitor specified in the article. A formula would be useful so that all outputs could be achieved without having to check each with an oscilloscope.

Fritigern

 


   
ReplyQuote
(@davee)
Member
Joined: 3 years ago
Posts: 1607
 

Hi @fritigern,

 Please try to understand that I cannot see your rig, and although I can (and did) search for one of Bill's (@dronebot-workshop)'s excellent explanations, this still leaves more details undefined than defined. For example, the article covers lots of different options ... it is not like buying a product from (say) Ikea, in which a product code defines exactly what parts you are using. So to enable myself, and anyone else venturing an answer, please start with precise references ... e.g. web page address, and section heading, and provide details of the exact parts you have selected, circuit diagram sketch, etc. , as it greatly increases the chances that any answer will be relevant. Sorry, if this sounds a little unwelcoming, but it benefits both of us in the long run. And please don't think I am trying to put you off.

-----------

Now to try to address some of your points:

2200uF 63V capacitor, I made do with a 2200uF 35V one.

The voltage rating must exceed the maximum voltage the component will experience in operation, otherwise it is generally not important. (There are more subtle considerations if designing a mass production product, but you are unlikely to need to worry about them.)

So your example is fine, and will not affect operation of your circuit.

----------

used a 5V regulator (tested beforehand with reliable DC power supply and output tested as 5V)

It would be helpful to know which one. You don't state the input voltage, was the input AC or DC, and does that mean the output was 'about 5V' or 5.0000V ? I am asking because your later question refers to (presumably) the same device giving an  'about 5V' output. Also, what load was present?

Sorry, more questions than answers, but there are lots of reasons why measured results do not match expectations.

---------

We all use breadboards of various types and usually take all sorts of 'liberties' that the electrons really "don't like", and object by not working the way we would like.

Assuming you mean the 'micro pegboard' style breadboards in which you push a wire into a hole, then these are really only meant for low current, low frequency circuits .. the main smoothing capacitor of power supply is likely to be a higher current flow situation, so in addition to any mechanical concerns caused by pushing a thick wire into a small hole, I would also have electrical concerns.

 So please bear in mind that with power supplies, there can be substantial currents flowing, so that any small resistances (and inductances + capacitances) can substantially affect the characteristics of the circuit. At least some of what you observing may be the result of your mechanical/electrical approach.

-------------

The DC Volts reported at the output of the rectifier were 11.2, but when I offed the cap it went up to 14.4. I presume the multimeter doesn't like irregular DC volts so it offers an average, and the capacitor filled in enough troughs to up the volts to 14.4.

To understand what this meant, it would be more helpful if I knew:

What is the AC voltage from the transformer?

What is the rectifier configuration ... does it use 1, 2 or 4 diodes? (Note that some rectifiers can be 2 or 4 diodes in a single package.... e.g. a 4 diode bridge rectifier will have 4 connection wires.)

Plus, unfortunately it also depends upon the meter characteristics ... which for low cost meters is a little indeterminate, especially when fed 'unusual' waveforms.

--

A 'perfect' or 'ideal' diode would pass the entire voltage without loss in one 'forward' direction, and nothing in the other 'reverse' directoin.

A real diode will drop a small voltage in the 'forward' direction ... its reverse direction characteristics may not be 'perfect' but providing you do not exceed its maximum rating, you can usually assume they are 'perfect' in this power supply scenario. The forward voltage drop depends on several factors, but a figure of about 0.6V per diode is reasonable starting place for general purpose rectifier accompanying a mains transformer. Note this is PER diode ... the common 4 diode full wave rectifier bridge circuit effectively has two diodes in series, so the total drop. will be about 1.2V.

I don't know the AC output voltage from the transformer, but lets pretend it is 10.0V AC RMS and a perfect sine wave.

RMS (Root Mean Square) is the means of defining the DC equivalent of a repetitive waveform .. that is, 10V AC RMS passing through a 5 Ohm resistor, will dissipate 10(V) * 10(V) / 5 (Ohm) = 20 Watts, the same as 10 VDC across the same resistor.

When the waveform is a perfect sinewave, then the peak voltage will be

   10 (V AC RMS) * Square root(2) = 10 * (approximately) 1.414 = 14.14V

If this was rectified by a perfect (4 diode) full wave bridge, and capacitor, with no load, then the capacitor would charge up to the maximum (peak) voltage of 14.14V

If the diodes are 'real diodes', and there is a small load, but the capacitance is sufficient to smooth virtually all of the ripple (that is it would maintain very close to the peak voltage)

Then the voltage would be about 14.14 V - 1.2 V (diode drop) = 12.9V

This voltage is dependent on several factors, so a variation of at least (say)  +/- 0.5V is to be expected from this oversimplified analysis.

--

With the capacitor disconnected, then a 'general purpose and affordable' DC meter will struggle to make an accurate reading, but clearly it will be considerably less than the peak value with the capacitor, which is consistent with your finding.

-----------------------

The output from the Regulator was 5.19Volts, with or without the second Capacitor.

The precise interpretation of what is happening here is vague, partly due to not being able to see the circuit, etc.

I am guessing 'second capacitor' is a relatively small capacitor, maybe 100 microFarad?, connected to the regulator output.

I am also guessing, that you are suggesting a regulator that previously output precisely 5.0V is outputing 5.19 V in this configuration.

There is more than one possible explanation, but my first guess is that the regulator likes to have at least a small capacitance on both input and output to avoid it oscillating, and this need is in some way not met. The complication is that the capacitance must be connected by very short wires to the regulator, and that large electrolytic capacitors are not very good at doing the job. Unfortunately, this is one of those problems that is difficult to intentionally set up in a lab, but can occur when you least expect it.

As I alluded to earlier, power supply development on a breadboard is prone to difficulties, and 'odd' phenomenon  problems are probably more of a norm, than unusual.

If you had an oscilloscope, then this problem would show as a high frequency oscillation superimposed on the nominally averaged 5V output.

------------------------

I guessed the issue may have been due to the max voltage of my capacitor being inadequate, though it was still nearly double the transformer output.

Sorry, but I don't support that guess .. a capacitor rating of above 30V is more than enough for circuit producing a peak voltage of around 15V

----------------

To test the theory I added another big capacitor in parallel- 100uF 100V to help the 2200uf 35V one, and this dropped the output gf the Regulator to 5.15V, so the fault is in the inadequate smoothing of the voltage in to the Regulator.

Nice try, but again I think your theory is again flawed. Adding a 100uF capacitor in parallel with 2200uF only increases it to 2300uF. (In practice the wide capacitance value tolerance of electrolytics means adding 100uf to 2200uF is meaningless in smoothing capacitor circuit.)

I think a more plausible explanation is that the change in voltage is due to the general change of intrinsic wiring inductance, and possibly small reduction in the effective internal resistance of the capacitor(s) on the input to the regulator, contributing to a slight change in the oscillation.

-----------

You don't state which regulator you used, but regulator manufacturers often recommend one or two small (in uF) capacitors on both input and output, connected with very short leads. As a bare minimum, I would consider adding 0.1uF ceramics, and probably small electrolytics, say 10uF, in parallel .. but the manufacturer's data sheet should be your first reference, all I am offering is a broad guideline.

------------

I think I have indicated answers to your two specific questions in the main text of the sermon.

I hope you find this useful and illuminating. Sorry it is rather detailed, and in some places probably tough going, but I have tried to be informative. Please feel free to query any bits I have got wrong or are difficult to understand.

Best wishes, Dave


   
ReplyQuote
Fritigern
(@fritigern)
Member
Joined: 2 years ago
Posts: 36
Topic starter  

@davee, thank you for a fabulous and informative reply. For what it's worth I'll try to answer the questions in the text, but I'm not expecting you to respond to them. Yes, the regulator had earlier produced a measured 5v, but when I tried to use a small transformer, rectifier, capacitors and a regulator, all supposedly within the parameters to work, this voltage increased to 5.19. Transformer output 18v (250mA), rectifier, one of the 4 diode bridge rectifiers with four legs, regulator L805 from STI with 2V dropout. The "second capacitor" was the 10u 50V Tantalum capacitor specified by Bill on the output of the regulator. I don't have a tantalum one that size so i used electrolytic film one and replaced it with a 10pF ceramic. The only current was the quiescent current of the rectifier and regulator. My vague description of the tested regulator output as "about 5V" is because I didn't take notes, the output was either 5 or 5.01 volts (I was satisfied that a. it worked and b. it was within the parameters advertised. I never imagined that the parameters might be dependent on its input conditions) the input was 9volts dc from an adjustable bench supply. I didn't take readings with loads applied. So even with a 5.19 volt reading I should have used the oscilloscope to check the level of variation of the output. And thank you for the information about the maximum voltage of my capacitor being adequate. Maybe the specs sheet was for a higher voltage transformer.

Again, thank you very much for taking the time, I really appreciate it.

fritigern


   
ReplyQuote
(@davee)
Member
Joined: 3 years ago
Posts: 1607
 

Hi @fritigern,

  I hope my answer was useful ... you might find it useful to revisit it when you have played with your power supply a bit more .. it is a long sermon to digest in one go!

-----------

Fixed voltage regulators are usually fairly accurate, even though their spec may allow a wider tolerance, so it would be unusual, but it is feasible a 5V one might be as high as (say) 5.2V, even when working correctly, which is why I was querying your earlier experience to see if it gave any guidance about the 'normal' behaviour of that particular device. My guess though, is that one is oscillating.

----------------

I just downloaded the sheet for the regulator you specify:

https://www.farnell.com/datasheets/2307057.pdf

I note it says output voltage 4.9V to 5.1V for a 'benign' test and 4.8V to 5.2V under a wider range of conditions.

Also an 'application circuit' Figure 4, shows input capacitor of 0.33uF and output capacitor of 0.1uF ... I would expect these to be ceramic, and connected with wires as short as reasonably possible ... remember wire lengths of just 1 cm can have sufficient inductance to undo all the good effects of the capacitor!

------------------- 

You say you replace a 10uF with a 10pF .. perhaps that was a typo, but 10pf is one-millionth of 10uF ... 10pF is extremely low and would not be 'noticed' by a voltage regulator.

If you use your oscilloscope to look for oscillation, depending on both what the regulator was doing, and the bandwith of your 'scope, you might see a 'nice clear' oscillation, you might see a kind of 'mush', or it might even look stable with the oscillation being largely confined to the regulator chip itself, because sometimes the oscillation is at very high frequencies that are difficult to resolve, but it is worth a look to see what you can learn.

-----------------

I am very confused if you say your transformer AC output is supposed to be 18V .. if that is rectified with a 4 diode bridge rectifier, with capacitor, then on no-load I would expect roughly 24 V dc:

     Using the formula I described previously ... (18 * 1.41) - 1.2 =  24.2 V

Certainly not 11-14V that you were describing previously ..... Are you sure that transformer is playing nicely? I would check its output with a meter.

Happy 'gremlin' hunting! Best wishes, Dave


   
ReplyQuote