Murphy's Law and Theremins - Electronics Guru Help Needed

Hmmm. When I earlier checked the voltage from pin 12 of the LM13700 to ground, it must have been during a stable period when no hiccups were present.

Because just now, it is hicupping badly. And when I tuned "L11" according to the manual, I can never get it stable to zero volts. The voltage reading from pin 12 is fluctuating randomly between -0.9XV to 1.1XV. edit: in fact the range of fluctuation is more - between 1 to 3V.

Before assuming L11 is bad, what other components should I check? I see a couple of caps around there plus some 3904 transistors, how do I check these?

To check if L11 is defective or not, just see if you are able to change the VCA voltage continuously from -12V up to >0V with it.

Go and buy a non conductive ice spray at your local electronic parts store. Use it and alternatively a hair-dryer in order to detect the temperature sensitive component. You will see that applying ice-spray on the defective component will increase these hick-ups. Re-heating it will reduce them. Candidates are: Q8, C17, C18, C13, Q6, Q7, D1

Edit: Sorry for my bad English, I found that it must be called freeze spray. The US version of a current European product may be found here (http://www.crcindustries.com/ei/content/prod_detail.aspx?PN=14086&S=Y)

Thanks Thierry,

You are too modest, your English is excellent! I will try and get hold of that ice/freeze spray. I had an idea what it was already, just not sure if the local stores have it.

Hope that the problem just turns out to be an errant cap or transistor that would be easy to get and replace locally.

No freeze spray to be found locally, so on a whim I fashioned a nozzle on wifey's hair dryer and thought I could heat up the sections of the circuit instead (being a cold day today).

When it was hiccuping, I warmed up different board sections and I think I have localized the trouble - thank you Thierry!

Turns out all things must start from the very beginning as is intended. I now really think the problem emanates from the Power Supply circuit (Don's initial hunch!).

When heat is applied to the Power Supply section of the board, the hiccups go away!

The external power adapter delivers 14VAC 200mA to two half-wave rectifier diodes and two voltage regulators on the board. Checked with a multimeter, the "79L12" regulator outputs a stable -11.92V.

The other, a "Zetex 78L12" reads +11.84V but drops and fluctuates badly everytime the hiccups occur, and only when the hiccups occur. Reapplying heat generally around it (can't be precisely directed), stops the voltage drops.

I would hate to think that it's the 78L12 which is failing, it's probably not available locally.

I am hoping that it could be a bad solder somewhere here, or one of the other components in this circuit section - two large 2200uF caps, two .47uF caps and two 0.1uF caps plus the two IN4001 diodes.

How do I test these components? How do I pinpoint the exact problem now?

Hmmm, statistics and lifelong experience would take me to resolder first the power supply area. After that the 78L12 would be my first candidate for replacement. You may check before if its input voltage is more or less stable, independent of the hick-ups. If this is the case, you may be almost sure that the diodes and the big capacitor work fine.

Voltage readings at the big caps were a relatively stable +18V and -18V respectively, slight fluctuations. So the diodes and big caps are ok.

Input voltage of 78L12 is relatively stable hovering around +18V, even during hiccups. The output voltage is +11.84V and fluctuates/drops drastically during (and causing?) hiccups.

So, I took the board off and flipped it over, touched up solder joints on the regulator and the two adjacent capacitors, crossed my fingers and plugged in the ac adapter. Output reading was at first stable, then after a couple of minutes it started fluctuating badly.

So, I'm about to remove the 78L12 positive regulator. The part reads "Zetex ZSR 1200". A quick check at Digikey shows it as non-stock. It is the 12V 200mA TO-92 version, right?

Are there any equivalents? Similar ones I can find are rated 12V 100mA. I have a nasty feeling I won't be able to find this in a local shop. Gah!

senkun, where are you located? Even here in a small french town (70000 people) I was already able to find a 78L12 in order to fix a B3 theremin... the price was 0.62 euros.

The 78L12 is produced by several manufacturers (Fairchild, Texas Instruments and so on). They are all compatible. If you really cannot get one, you may try out a standard 7812 (1A max output current) which is perhaps better available.

I'm stuck out in the boondocks here in Taiwan. And to think Taiwanese companies OEM most of the notebooks and gadgets today, but exact replacement parts just can't be found easily in the local shops.

A search at Mouser shows all the 78L12s at 100mA only, and the original 200mA Zetex part ZSR1200CL is listed as obsolete.

Wait! Is the original part 12V 100mA or 200mA?

Edit: Ahh it is 200mA
http://www.zetex.com/3.0/product_portfolio.asp?pno=ZSR1200CL

Will a regular 7812 fit? The form factor is different.

Edit2: The manual states the instrument draws around 30mA in operation, but incoming from adapter is 200mA. Would it be safe to assume that using a replacement 78L12 with max output 100mA instead of the original 200mA be ok?

I would give the 100mA version a try. If this doesn't work, you'll have enough experience to replace it once more... ;-)

The 7812 is the 1A version. You cannot kill anything with it. So you could try it too (take care of the different pin layout). And if it doesn't work... see above... :-)))

Haha! It's not that I'm worried of blowing up my theremin (although the wife would jump for joy if that happened), I just like to make sure of everything beforehand and then do it perfectly first time round. Soldering fumes I like, but solder braids and pumps don't agree with me. :)