Query/Feedback to Colin @ TalkingElectronics

[Mike - K8LH]

Colin55,

Since your Talking Electronics site is pretty much "one way", where you get to make all the comments and criticism without any method for readers to make queries or provide feedback, I hope you don't mind an open-ended question placed here.

As I was browsing through your site today I noticed your critique of Jose Pino's ANP (almost no parts) Clock on the Spot The Mistakes (page 11) page (near the bottom of the page). The original Jose Pino page can be found here; Jose Pino's Amost No Parts Clock

Your complaints were (1) no segment current limiting resistors, and (2) no digit driver transistors. I believe the first issue is subjective. I myself omit current limiting resistors in favor of software current limiting. Your other complaint is confusing to me since Jose is only lighting one segment at a time on this Charlieplexed display and so there's no need for digit driver transistors. Jose confirmed this when I asked him about it a few years ago, which brings me to my question.

Did you ever consider asking Jose about his circuit before criticizing it unfairly based on a faulty assumption?

Regards, Mike

 

[colin55]

[FONT=&quot]

Your complaints were (1) no segment current limiting resistors,

I really thought you would have more understanding of electronics to say that no current limiting resistors are needed on the output of a microcontroller.
I have already gone over this point on my site and if you look through it with a “fine-toothed comb” you will see the explanation.
Google has every paragraph on my site indexed and I get 4GB to 9GB of traffic each day and you are the first to query the point. Just Google for the answer.
The point is this.
The output FETs are only permitted to deliver 25mA, but if you short-circuit the output they will deliver over 50-100mA.
Putting a LED directly across the output is nearly shorting the output and more than 50mA will flow.
Normally the wattage dissipated is 0.025A x approx 0.2v = 0.005Watts.
When you put a LED on the output, the wattage increases to: 0.05 x 3 = 0.15Watts. This is an increase of 3,000% - shared between two outputs.

By not using driver transistors, you are using a "run of 28" or 4% illumination time. You cannot possibly expect to get any sort of worthwhile brightness at this current.

[/FONT]

 

[Mike - K8LH]

Hi Colin,

I apologize for not presenting my question more clearly. While I mentioned both of your complaints for Jose's ANP clock, my question had to do with your second complaint about the lack of digit driver transistors. It seems you assumed that Jose was driving the clock one digit at a time when he's actually driving it one segment at a time and so there really is no need for digit driver transistors.

When I came across Jose's clock several years ago, I was curious too, so I asked him about his circuit via email instead of assuming that he might be doing it one way or the other. It really didn't take much time or effort on my part and after he responded, I was delighted to see how much brightness he was actually getting with only about 1-ma "average" current per segment.

So, would you indulge me if I repeat the question, please?

Did you ever consider asking Jose about his circuit before criticizing it unfairly based on a faulty assumption?

Also, now that you know your criticism about the lack of digit drivers isn't valid, I'm curious if you will modify your critique of his circuit on your web page?

Thank you for your time and kind consideration.

Cheerful regards, Mike McLaren

 

[colin55]

I emailed Jose about some other fault on this site and got no reply. So I doubt if you ever got through to him.
I never assumed anything. I realised it was individually scanned as he mentioned that on his website.
It's just an absurd way to scan a display.

I notice you didn't answer any of the statements I made in my reply, but just kept prattling on about unimportant details.

 

[upand_at_them]

Now I'm curious...What makes Charlieplexing an absurd way to scan a display? I've done it, and it works. And it uses a minimum of parts.

 

[be80be]

When you don't no what your doing you add more parts. But some people have a better understanding then the rest and came up with simple way of doing the same thing.

Like Charlieplexing leds you control the pin current in the software with the on and off time of the Led and if you do it right your leds will be nice brite.

 

[ghostman11]

[FONT=&quot]
and I get 4GB to 9GB of traffic each day and you are the first to query the point


[/FONT]

thats seems a awfuly high number for 1,800 visits and 4000 page views a day!! you must have a very very heavy amount of download traffic? i would have assumed a higher ranking with those figures. obviously i am not questioning you i am just curious at the figures
regards jason
i exspect no one else has pointed it out because users are disproportionately male, and they are disproportionately less affluent, moderately educated users browsing from school and home. BTW not being rude just giving a possiable reason why it hasnt been pointed out before! the comment on users was a cut and paste from alexia

 

[Mike - K8LH]

I emailed Jose about some other fault on this site and got no reply. So I doubt if you ever got through to him.

After you posted this disturbing comment yesterday I sent Jose an email asking if he remembered our exchange. Here's an excerpt from his reply;

excerpt from Jose Pino email of 8/28/11 9:35 PM

Mr. McLaren,

Thank you for bringing this to my attention.

Surely I remember you. Even if I don't pay attention to several emails like the one mentioned here:

I emailed Jose about some other fault on this site and got no reply.

Unfortunately for some people, I only answer e-mails that ask the right questions; So I really don't remember colin55's emails.

If you can refrain from any further personal attacks, can we continue, please?

At this point may I ask if you're even willing to fix a mistake in your "Spot the Mistakes" column, when one is pointed out? For example, if I were to point out a simple, obvious, and glaring mistake in one of your column posts, would you even consider fixing it? If we can agree that mistakes should be corrected then maybe we can work toward a solution on the ANP Clock article.

 

[colin55]

I don't know where you got 1,800 visits and 4,000 pages per day from.

Day Number of visits Pages Hits Bandwidth
15 Aug 2011 4,380 25,257 407,230 9.78 GB
16 Aug 2011 4,288 23,805 399,928 7.07 GB
17 Aug 2011 4,284 11,347 383,627 5.97 GB
18 Aug 2011 4,277 22,300 398,949 6.55 GB
19 Aug 2011 4,066 16,821 375,278 6.47 GB
20 Aug 2011 3,402 9,444 314,269 5.90 GB
21 Aug 2011 3,629 15,503 362,407 6.84 GB
22 Aug 2011 4,736 13,762 452,318 8.50 GB
23 Aug 2011 4,798 25,736 515,841 10.39 GB
24 Aug 2011 4,451 17,527 432,338 8.68 GB
25 Aug 2011 4,281 12,581 393,860 10.12 GB
26 Aug 2011 4,194 14,664 384,754 8.71 GB
27 Aug 2011 3,582 10,529 363,511 16.01 GB
28 Aug 2011 3,791 17,239 392,639 12.14 GB

 

[ghostman11]

i got it from here https://www.statshow.com/http://www.talkingelectronics.com/ was just curious to the amount of bandwidth used for what seemed like low numbers of visitors, so i assumed must be down to a heavy amount of users downloading the pdf's etc from the site. i did try alexa but they dont do figures for low ranking sites (>100,000) well not that i could see. anyway i think mike has a more important point thats needs answering rather than my curiosity over bandwidth.

 

[phalanx]

I believe the first issue is subjective. I myself omit current limiting resistors in favor of software current limiting.

Like Charlieplexing leds you control the pin current in the software with the on and off time of the Led...

Software cannot limit the instantaneous current in an LED. This isn't a subjective matter, you absolutely need current limiting resistors to prevent both the LED and microcontroller from experiencing current loads outside of what they are designed to tolerate.

 

[colin55]

Software cannot limit the instantaneous current in an LED. This isn't a subjective matter, you absolutely need current limiting resistors to prevent both the LED and microcontroller from experiencing current loads outside of what they are designed to tolerate.

Thank goodness someone understands electronics.

 

[be80be]

Phalanx If you turn off the power there is no power? If you switch the power on and off fast it can only get what you let it have. And it's that simple.

PWM can be used to adjust the total amount of power delivered to a load without losses

 

[colin55]

https://www.electro-tech-online.com/members/be80be.html Learn a little bit about electronics.

 

[phalanx]

Phalanx If you turn off the power there is no power?

You are correct about this point...

If you switch the power on and off fast it can only get what you let it have. And it's that simple.

This reasoning is completely wrong...

The short period of "ON" time is still completely uncontrolled unless you have a current limiting resistor in series with the I/O port and LED. Uncontrolled current damages the transistors in the I/O port and the junction in the LED.

 

[be80be]

I guess this boils down to when you don't Know what your doing you add a lot of parts LOL

I can set the power out a pin to a easy 10.2 mA all day long and it lights a led nice

Collins you need to grow up and be a Man boys post like this

be80be - Learn a little bit about electronics.

It's thinking like yours that keeps me in business

And I guess this is all wrong to https://en.wikipedia.org/wiki/Pulse-width_modulation I wonder if they think the same as You.

O and one more thing instantaneous current can only come into the picture when you have a fixed supply when you change the Duty your lower the power

witch make away to dim led's and slow a motor control current.

 

[colin55]

https://www.electro-tech-online.com/members/be80be.html listen to what phalanx is saying. You obviously don't understand what he is saying AT ALL.
You are completely mixing up the concept of PWM and what PWM is actually providing in the way of CURRENT PULSES during each cycle.

 

[phalanx]

I guess this boils down to when you don't Know what your doing you add a lot of parts LOL

I can set the power out a pin to a easy 10.2 mA all day long and it lights a led nice

Unless you have access to a more expensive multi-meter like a Fluke 189 or better, you aren't seeing the whole picture. The better meters are capable of measuring the overshoot as opposed to the average value you are talking about.

If we modeled this as an ideal system, the current through the diode would be infinite since there is no series resistance to control it. The diode would vaporize. In the real world there are resistances in the I/O pin, wires, PCB traces, etc., that will limit the current to some maximum value which is still outside the range of what the I/O pin and diode are capable of handling. During the "ON" portion of your PWM waveform, current through the LED will be at this maximum value which is actively contributing to the early demise of your controller and LED.

A less capable multimeter will report the average current value through the diode based on your PWM dutycycle. Over the half second it took to measure an average value of 10.2mA, the diode was slammed thousands of times with peak currents of 100mA or greater.

And I guess this is all wrong to https://en.wikipedia.org/wiki/Pulse-width_modulation I wonder if they think the same as You.

All the info in that link is correct. Your understanding of it is incorrect.

O and one more thing instantaneous current can only come into the picture when you have a fixed supply when you change the Duty your lower the power witch make away to dim led's and slow a motor control current.

Your microcontroller is running off of a fixed supply therefore any time your PWM is in the "ON" portion, you are subject to the instantaneous currents through the system.

 

[be80be]

Collins you can change the duty cycle and lower the current output. Are you trying to say that PWM can't be use to make a controllable current source it you think that your both wrong. If you pulse a output pin at say 5KHz and a duty of 5% that pin will only be able to output 10.2 mA. The led will look nice a bright.

 

[phalanx]

Collins you can change the duty cycle and lower the current output.

Average current yes. Instantaneous current NO!

Are you trying to say that PWM can't be use to make a controllable current source it you think that your both wrong.

He isn't saying that because if you know what you are doing, PWM can be used in a more complex system to control both average and instantaneous current in a device. Your example of open loop PWM control of an LED is not a controllable current source since there is no way to limit instantaneous current to acceptable values.


I don't know what you do for work but if it involves telling people that it's ok to design systems this way, you are doing them a great disservice.