ARP Odyssey Oscillators

[etaoin]

Both Bridechamber and Magic Smoke sell 2k tempco's.

Ray Wilson also sells them, but that is the expensive PT146 version.

[LetterBeacon]

Cool, thanks for that. Bridechamber also stock the CA3086 that this circuit requires.

[e-grad]

http://www.emulatorarchive.com/

stocks 1K87 TC, though PRICY.

The site is well worth a look for anyone interested in building ARP-cicuits. The cicuit for the Bananalogue Arp filter-clone was done by this guy.

[LetterBeacon]

I emailed him a week or so ago, but he's out of stock.

How does the Bananalogue Arp filter-clone on that site compare with YuSynth's 4072 clone, by the way?

[Tim Servo]

[urbanscallywag]

Tim,

You might sign onto the forum at AVRfreaks.net and look for someone there. There are many capable people who could probably help you out (I mean working for you, not just helping you with AVR).

[LetterBeacon]

[Broadwave]

[ericcoleridge]

Andy, or anyone,

Is there a simple way to get a triangle wave from this VCO? I'm thinking, especially in LFO mode, a Tri-wave would be very useful. Perhaps I should just make a new post asking about Saw to Tri wave shapers-- but thought I'd ask here first.

edit: Just found this post with a few Saw/Tri Symetry Waveshapers:

http://electro-music.com/forum/viewtopic.php?highlight=triangle+waveshaper&t=27934

[numbertalk]

in wanting to understand things a little better, what exactly is the ppm rating for tempco resistors? For example, I find these 1.87K thermistors at Mouser:

http://www.mouser.com/Search/ProductDetail.aspx?qs=J16uGQxykBg%252bidq1SFDwWw%3d%3d

but notice their ppm rating is 100, which is much lower than the ones used in synth modules I've seen so far (3500ppm, right)? So what exactly does this rating mean and why aren't the ones at Mouser sufficient?

[Tim Servo]

There is a very nice SMT tempco from Digi-Key that has the proper specs:
Digikey P1.8KCDCT-ND (page 1814 in the current catalog)

Note that these are surface mount, and are 1.8K, rather than 1.87K. Should work, but it's probably worth a test run first. I believe the tempco resistors you were looking at from Mouser are the NTC (negative temp coefficient) type, which are primarily used for surge protection.

Dave Brown did a really nice job placing a couple of these resistors (two 1.0K units in series) in epoxy on top of the 2164 chip in a Mankato
http://modularsynthesis.com/magicsmoke/mankato/mse.htm


Tim (should work, maybe) Servo

[frijitz]

[numbertalk]

Thanks Ian! Going to read your site now - looks like a lot of math so going to plan to take a close look. Offhand not sure if the 100ppm/K tempco being more accurate and stable is a good thing or a bad thing for a synth builder. I'm sure reading the page you link to will clear that up for me.

[bubblechamber]

the DigiKey tempcos work really well and i believe there are a couple commercial VCOs that use them. the easiest way i've seen these used is to solder the tempco under a LM395 or SSM chip then solder the chip in with out a socket. it's a bit of a pain, but at $.55 each it's worth it.

[Dan Lavin]

Numbertalk,

OK, I guess it’s that time of year for my yearly lecture on what people on this forum refer to as “tempcos”. If you walk up to any EE and start talking about “tempcos,” you will either get a blank stare or a return question of “Tempco of what?!?” Tempco is short for temperature coefficient or how much something changes in relationship with temperature. All electronic components, (as far as I know), have some type of relationship with temperature and therefore all components have “tempcos”. Some have more change with temperature change and some less.

The component that is used mostly to temperature compensate electronic music circuits are technically called “linear positive temperature coefficient thermistors.” People in the industry, (as I have been for more years that I care to admit), would call them linear PTC thermistors. They are linear because their resistance changes linearly with temperature and they have a positive temperature coefficient because their resistance increases with temperature. There are switching type PTC thermistors, which have a zener-diode-like resistance increase with temperature, and there are logarithmic NTC (negative temperature coefficient) thermistors, both of which are actually more common than the linear PTC Thermistors we use in our E-M circuits.

Linear PTC thermistors are primarily speced with 2 pieces of information: the resistance at 25deg C (77degF) and the rate of change. Rate of change for NTC’s are expressed in %/degC (or K…makes no difference) because their rates of change are quite dramatic: 3-5%/degC. Linear PTCs have rates of change less than 1%/degC, so they are expressed in terms of ppm or parts per million. 1% = 10,000ppm, 0.1% = 1000 ppm, so 3350ppm = 0.335%/degC (or K, whichever you prefer).

Here’s how a 1K ohm/+3300ppm works: At 25C, the device is, of course, 1000 ohms. At 26C, the device would be 1000 ohms + (1000 * 0.0033) = 1003.3 ohms. At 24C, the resistance would be 1000 ohms – (1000*0.0033) = 996.7 ohms.

I blame the whole calling linear PTCs “tempcos” on the early data books from SSM. SSM was the first company marketing E-M specific chips and thus widely read by us early E-M enthusiasts. They labeled the thermistors as “1K +3300ppm tempco” for instance. I think people assumed “tempco” was the name of the device rather than an appreviation for temperature coefficient.

Sorry if this is more information than you needed and I hope I haven’t put too many people to sleep. Now imagine doing this for decades and you’ll understand why I am the way I am!

[LetterBeacon]

Very informative - thanks a lot!

[numbertalk]

Yes, thanks for all that info! With the link from Ian that's so informative, even if some of it is above my head at the moment (mostly the math on Ian's site )

[frijitz]

[Dan Lavin]

[frijitz]

Dan --

I agree that "thermistor" includes more than just the common ceramic types. But if you make the term *too* general, then *any* conductor could be included.

I believe the term "tempco" as used in the EM community is shorthand for "temperature compensation" or "temperature compensating resistor". Easy to get this confused with "tempco" as "temperature coefficient". But "temperature compensating resistor" is pretty standard descriptive nomenclature -- e.g., see the old Analog Devices Nonlinear Handbook.

To get back to more practical matters, I think wirewound temperature compensating resistors are always going to be around. I just noticed that two of the suppliers that I have used in the past now have SMD wirewound temperature compensating resistors:

http://www.krlbantry.com/products/Precision/sect-311.htm
(see 3d bullet)

http://www.precisionresistor.com/pt-sp-series-positive-tcr-compensators-c-22.html
(see bottom of table)

It might be worth getting a quote on these. If they are as inexpensive as the Panasonic units Digi carries, it might be worth it to get some big batches made up for the particular temperature compensating resistors we need for different applications.



Ian

[numbertalk]

Another "tempco" related question - aren't the bodies of these usually physically fixed to a chip on the board using a thermal compound? I don't see any ICs within reach of where those resistors go on this board (and Andy I don't see them this way in the photos of your boards), Will they still work as well soldered in just like a normal resistor this way?

[LetterBeacon]

I think the tempco is supposed to be attached to the pair of transistors next to it on the PCB. Although Urbanscallywag said this on page two:

[numbertalk]

[LetterBeacon]

[numbertalk]