NUTS/Volts function generator in Sept.2006 has mistakes

[Timothy Rasch]

Hi ,the function generator on page 38 N/V Sept. in the sine circuit [LM386]
has many mistakes : this circuit should have a switched wien network for the course frequency and a dual variable resistor about 10k[part of the wein network] for fine frequency . The wien network goes from pin 5 in series to ground and the parallel resistor cap part of wien network high side goes to pin 3 [+] input ;between 50 uf cap [-] and pin 2 [-] input there should be a 390 ohm resistor and also on pin 2[-] a lamp 3v@15ma is connected and the other end to ground. The load resistor from 50uf [-] should be 100ohms to ground. Also connect a 10uf [+] to pin 1 [-] to 8 . Ground is pin 4 ,vcc is pin 6. The values for wien network for 1khz 47k,.01uf . I found this on application sheet from National Semiconductor.
Also there is a typo in the first paragraph in N/V PG.38 .It reads 486 .Should read LM386. Hope this helps on the mistakes.
Tim Rasch [email protected]

[Timothy Rasch]

Hi ,I read the National application sheet wrong for the LM386 in my old databook and went to National and downloaded their newest datasheet. The series resistor is 47k and series capacitor is .01uf;the parallel resistor is 4.7k and it's cap is .01uf in the wien network for a 1khz sine wave. Remember this circuit is a power oscillator and the signal is at a very high level [must be attenuated for use in preamps testing,Use a resistor attenuater network for this]
Tim Rasch [email protected]

[jollyrgr]

This article is filled with mistakes. Either the author has no clue, the translation to the magazine was really messed up, or a little of both. For instance take this line in the second column:

"A triangle waveform is obtained at pin 1 of the 4,700uF ceramic disc capacitor."

The author goes on to say he would prefer to use electrolytic but has not had trouble using CD capacitors. First, is there such a thing as a 4,700uF ceramic disc capacitor? Second, when did capacitors get "pin numbers" (especially ceramic discs)? The article text calls for capacitor C5 to be 4,700uf ceramic disc, the parts list 0.47uF electrolytic, the schematic a 4,700pF. Correct me if I'm wrong but doesn't 4700pF = 0.0047uF? So we have three values for C5. What gives?

Does someone have a source for a 4700uF ceramic disc?

[Newz2000]

There should be a peer-review process that should happen before articles get printed. I realize this increases the cost of the project, but the benefits would be enormous.

The peer review would work like this:
"Hey friend, here's my article, here's the parts you need, build this using nothing but what I've given you (you can use your own tools, but no extra parts or documentation) and let me know if you have any troubles."

Admittedly, this would be cost prohibitive for the projects that cost $200 or more, but at the same time,
a: How would your readers feel if they bought the parts to make a $200 project based on a magazine article and then it didn't work
b: Who builds $200 projects out of a magazine article?

[jollyrgr]

Another mistake I noted but maybe I'm missing something. Hopefully someone can expand on this.

In the article the author suggests using a doorbell transformer with a 12VAC output at 1.5A fused at 1A. He states the current of the 555 and LM386 is 1.5A. But the I max of a 555 is around 15mA. It can sink around 200mA max. See this: http://cache.national.com/ds/LM/LM555.pdf

The LM386-4 has a maximum power of 1000mW @ 16VDC. In simple terms this is around 63mA, right? See here: http://cache.national.com/ds/LM/LM386.pdf

Since I don't see any other active devices I don't think this could hardly exceed 300mA at best. Thus 1.5A is way overkill. But back to the transformer and power supply. I believe that most doorbell transformers are around 16VAC. But let's say for the sake of argument that there is a 12VAC step down transformer. I see a full wave bridge in the circuit. But I also see the DC voltage specified as 11.4VDC. Doesn't a full wave bridge take 12VAC and convert it to 12VAC*1.414 = 17VDC?

The author specified a Radio Shack part number for the Full-Wave bridge of 276-1181. This is a 6 AMP 200V rated device! A better choice (since the circuit is fused at 1A) would be 276-1152.

Let's assume that the bridge IS supplying 12VDC (not likely). The LED D2 and R1 (the related 2K resistor) see 12VDC. Assume that the LED is a standard red unit that drops a standard 1.5V. This leaves 10.5V for the 2K resistor. Taking the 10.5V/2000 Ohms gives 5.25mA. An LED MIGHT glow at that current level; the "usual" range is 20mA.

Maybe I'm not seeing this circuit correctly but it looks like it is a mess and would never work. Any novice trying to build it would have hard time.

[Timothy Rasch]

Hi, for a simple sinewave generator 1khz [fixed frequency power oscillator] for LM386 please go to[ LM386 datasheet in Google] and download the datasheets from National Semiconductor. The wien network values can be changed for higher or lower frequencies
fr=1/(2* 3.14159*R*C). This equation only applies when the resistors and capacitors are equal values . I don't know the actual equation when the resistors are a 10/1 ratio. The person building many wien networks must measure the frequency for each network and make a chart. Just make sure the ratios of resistors are maintained .IE in the application sheet 47k,4.7k a 10/1 ratio has to be maintained. Also for the lamp which is 3v@15ma,[200 ohms], maybe [2] 6v@75ma in series should work [160 ohms]. A person can develop a basic program for the wien networks if they wish for a complete audio range of frequencies and make a complete audio generator with switched wien networks usually in x2 steps or whatever they desire. Tim Rasch [email protected]

[philba]

The problems with this article are precisely why I did not renew my sub to N&V. Some of the articles are good but marred with significant editing errors (like omitting parts lists, for example) while others are just plain bad with dubious circuits. For every project, you can google the internet and find several that are better in some way. CCI is much more professionally put together though they seem to be more driven by sponsor products than I would like.

I hope the editorial staff learns from this but have little confidence.

[hole_puncher]

I think NUTS&VOLTS should extend everyones subscription by 3 months for free for not putting this article in the April fools issue. In all my years of reading electronics mags, dating back the the early days of Popular Electronics, I have never seen anything like this. Every line of the text is filled with mistakes and stupidity. The choice of the fuse was most hilarious as was the descriptions of how the circuits work. The title itself "Nine Volt Function Generator" - What is 9 volt about it?

The design is appalling:
A 0.1 uF supply filter capacitor????

Calling the Timing capacitor connection the triangle output?? Anything connected there would change the oscillation frequency/waveshape without buffering. The classic two op-amp square/triangle generator would have been a much better (sane) choice here.

The fuse and power switch on the secondary of the transformer?? Possible safety issues here.

I wont comment on the sine generator since as drawn that circuit wont do anything.

Hopefully the author reads this forum and is able to learn from it all but this should be a total embarrassment to nuts and volts.

[Timothy Rasch]

Hi,when is Nuts/Volts going to republish the corrected schematics and text of the entire article for the September issue? There might be other mistakes in the other articles too.I agree with the person before that said " Nuts/Volts should EXtend everyones subscription three months for FRee". I am the first person that started this topic of the function generator mistakes. Tim Rasch [email protected]

[Timothy Rasch]

Hi I recently built the sine generator as shown in LM386 data sheet from National Semiconductor [6 pages]and that circuit works great after I added 0.01uf ceramic cap[I recommend 0.1uf] to bypass VS at pin 6 to 4 [ground]. Data: 12v battery power supply,the caps in wien network measured 0.012uf ,845hz @ 4.8vp-p sinewave about .05% distortion like my Kenwood AG-203A commercial generator. I used a 12v@35ma lamp instead of 3v@15ma lamp. If anyone builds the LM386 circuit I don't recommend changing the two resistors in the wien network only the capacitors. I will further experiment with changing the cap values. Also the feedback resistor [390 ohms ] might have to be adjusted for each range[and switched in] to minimize distortion. The amplitude of sinewave I think will vary from high to lower as frequency is increased. For a fixed frequency low distortion sinewave it is a great simple circuit for anyone to build.Use an attenuator network of resistors for the 4.8vp-p sine to reduce it down for testing preamps to about 20 mvrms level. Build the circuit as small as possible to elliminate oscillation/feedback problems.
Tim Rasch [email protected]

[Timothy Rasch]

Hi ,I built the sine generator [using LM386]as in the preceeding letter. It works real well with very low distortion at 16.088khz ,4.8vp-p, 680pf 5% silver mica cap. It will not work good at 100khz [double oscillation with 100pf cap] . I am making chart of frequency referenced to the caps .[the caps have to be matched with the correct value desired for the frequency ]If you use decade steps in smaller capacitance the frequency increases in decade steps and the frequency lowers in decade steps when the capacitance is increased in decade steps. Contrary to what I thought about the amplitude changing with increasing frequency the amplitude and distortion stay the same. Also with lowering the supply voltage to 8v the 4.8vp-p output stays the same. This is the best circuit I ever constructed that kept the amplitude with varing supply voltage!!!! Other circuits would decrease amplitude . Everyone feel free to build this circuit and experiment with the caps in the wien network.Leave the 390 ohm resistor alone in the feedback circuit. Tim Rasch [email protected]

[bodgy]

I haven't subscribed to N&V for a little while now mainly because I subscribe to 4 other electronic magazines and one has only so mmuch time to read them .

So, as I haven't read the article in question this is a stab in the dark. About the capacitor's value - if the author like myself is from Europe, I wonder if the value is meant to be in nF, and the zeros in uF/pF editing/conversion have gone a little awry? Or perhaps the decimal point was a splodge somewhere - one reason why I tend to write 4n7F, 3k1 etc.

If the LED is a high intensity one, then 5-10mA probably would be sufficient


Colin

[philba]

I haven't subscribed to N&V for a little while now mainly because I subscribe to 4 other electronic magazines and one has only so mmuch time to read them .

So, as I haven't read the article in question this is a stab in the dark. About the capacitor's value - if the author like myself is from Europe, I wonder if the value is meant to be in nF, and the zeros in uF/pF editing/conversion have gone a little awry? Or perhaps the decimal point was a splodge somewhere - one reason why I tend to write 4n7F, 3k1 etc.

If the LED is a high intensity one, then 5-10mA probably would be sufficient


Colin

what? the N&V editors made a mistake? Not Possible.

[Timothy Rasch]

Hi, about the preceeding letter I don't get why you are talking about a"led having a 5-15ma range is sufficient". There is no led in this project unless you are refering to the 3v 15ma lamp in National Semiconductors LM386 datasheet , being replaced by Leds. I seriously think it won't work with leds . I will try an experiment to find out and let everyone know later Saturday. All that I know is Nationals' wien bridge oscillator circuit works good[ audio range] ,about .05% distortion comparing to Kenwood
AG-203A wien oscillator exactly. Tim Rasch [email protected]

[JPKNHTP]

-JPKNHTP