Post by thread:[EE]: Replacement for Intersil ICL8038

[Repost with tag] Hi, I plan to use ICL8038 signal generator chip from Intersil. Unfortunately, it's flagged "not recommended for new designs" on very first page of its data sheet: http://www.intersil.com/data/FN/FN2864.pdf Anyone know is there any close replacement for it? Not necessary pin to pin compatible since I am doing new design. Thanks & best regards, -- WH Tan

WH Tan wrote: > I plan to use ICL8038 signal generator chip from Intersil. > Unfortunately, it's flagged "not recommended for new designs" on very > first page of its data sheet: Hmm, I remember seeing it used on an italian magazine back in the '70s :) so definitely it is "old" ! > Anyone know is there any close replacement for it? Not necessary pin > to pin compatible since I am doing new design. Can't tell easily. I see some things here and there, and also got some samples (years ago, anyway). What does google suggest? -- Ciao, Dario

Check MAX038 On 10/4/07, Dario Greggio <spam_OUTadpm.toTakeThisOuTinwind.it> wrote: {Quote hidden}> > WH Tan wrote: > > > I plan to use ICL8038 signal generator chip from Intersil. > > Unfortunately, it's flagged "not recommended for new designs" on very > > first page of its data sheet: > > Hmm, I remember seeing it used on an italian magazine back in the '70s :) > so definitely it is "old" ! > > > > Anyone know is there any close replacement for it? Not necessary pin > > to pin compatible since I am doing new design. > > Can't tell easily. I see some things here and there, and also got some > samples (years ago, anyway). > What does google suggest? > > > -- > Ciao, Dario > -

On 10/4/07, João Seabra <.....jseabraKILLspam@spam@gmail.com> wrote: > Check MAX038 > > Alas, it's even deader: http://www.maxim-ic.com/quick_view2.cfm/qv_pk/1257 " This product was manufactured for Maxim by an outside wafer foundry using a process that is no longer available. It is not recommended for new designs. For further information, contact us. The data sheet remains available for existing users." -- Paul Anderson VE6HOP wackyvorlonKILLspamgmail.com http://www.oldschoolhacker.com "May the electromotive force be with you."

> > Check MAX038 > Thanks very much Joao. That is what I am looking for. If it helps - http://home.clear.net.nz/pages/joecolquitt/dual_osc.html Can throw in panel graphics too. Must have been put on CD, can't see them on this HDD Added some digital outputs to that as well. Intend, one day, to add a frequency meter. Have the schematic/code for a simple 50MHz PIC-based one http://www.siliconchip.com.au/cms/A_108353/article.html If you're starting from a triangle wave, this isn't too bad. Not sure what the upper F limit is but it's OK up to at least high audio. +/- 12V can be reduced to +/- 5V http://home.clear.net.nz/pages/joecolquitt/tri2sine.html

Paul Anderson wrote: > On 10/4/07, João Seabra <.....jseabraKILLspam.....gmail.com> wrote: > >>Check MAX038 >> > > Alas, it's even deader: > www.maxim-ic.com/quick_view2.cfm/qv_pk/1257 > > " This product was manufactured for Maxim by an outside wafer > foundry using a process that is no longer available. It is not > recommended for new designs. For further information, contact us. The > data sheet remains available for existing users." This was my fear too ... -- Ciao, Dario

I recommend you look at the PSOC chips from Cypress. They are cheap and may produce the waveforms you want. http://en.wikipedia.org/wiki/PSoC Contact me off list if you want more info. Cedric On Oct 4, 2007, at 12:23 PM, WH Tan wrote: [Repost with tag] Hi, I plan to use ICL8038 signal generator chip from Intersil. Unfortunately, it's flagged "not recommended for new designs" on very first page of its data sheet: http://www.intersil.com/data/FN/FN2864.pdf Anyone know is there any close replacement for it? Not necessary pin to pin compatible since I am doing new design. Thanks & best regards, -- WH Tan

Damn...Im getting old... :-) On 10/5/07, Paul Anderson <EraseMEwackyvorlonspam_OUTTakeThisOuTgmail.com> wrote: {Quote hidden}> > On 10/4/07, João Seabra <jseabraspam_OUTgmail.com> wrote: > > Check MAX038 > > > > > Alas, it's even deader: > http://www.maxim-ic.com/quick_view2.cfm/qv_pk/1257 > > " This product was manufactured for Maxim by an outside wafer > foundry using a process that is no longer available. It is not > recommended for new designs. For further information, contact us. The > data sheet remains available for existing users." > > > -- > Paul Anderson > VE6HOP > @spam@wackyvorlonKILLspamgmail.com > http://www.oldschoolhacker.com > "May the electromotive force be with you." > >

2007/10/5, Jinx wrote: > If it helps - > > http://home.clear.net.nz/pages/joecolquitt/dual_osc.html > > Can throw in panel graphics too. Must have been put on CD, > can't see them on this HDD > > Added some digital outputs to that as well. Intend, one day, > to add a frequency meter. Have the schematic/code for a > simple 50MHz PIC-based one > > http://www.siliconchip.com.au/cms/A_108353/article.html > > If you're starting from a triangle wave, this isn't too bad. Not > sure what the upper F limit is but it's OK up to at least high > audio. +/- 12V can be reduced to +/- 5V > > http://home.clear.net.nz/pages/joecolquitt/tri2sine.html Thanks. That's a bonus, if I later decide to ignore the not-recommended 'advice' :) -- WH Tan

2007/10/5, Cedric Chang wrote: > I recommend you look at the PSOC chips from Cypress. > They are cheap and may produce the waveforms you want. > http://en.wikipedia.org/wiki/PSoC > Contact me off list if you want more info. Cedric, Firstly, let me have some research - to find out what is a PSOC... probably will contact you later. Thanks & best regards, -- WH Tan

2007/10/5, Alan B. Pearce wrote: > I think Maxim do, but you probably do not want to know that ... ;)) I think I might be able to grab some ICL8038 if I want. But since this is a new design I'll try to find other solution. I'll probably take it as a last bet. Yes, apparently the MAX038 is not the answer either. Best regards, -- WH Tan

On Fri, 2007-10-05 at 19:27 +0800, WH Tan wrote: > 2007/10/5, Cedric Chang wrote: > > I recommend you look at the PSOC chips from Cypress. > > They are cheap and may produce the waveforms you want. > > http://en.wikipedia.org/wiki/PSoC > > Contact me off list if you want more info. > > > Cedric, > > Firstly, let me have some research - to find out what is a PSOC... > probably will contact you later. It's basically an MCU with analog and digital generic building blocks included. These blocks can be dynamically configured to allow the creation of custom peripherals. In the case of the analog blocks you can create op amps, A/D converters, even PLLs perhaps. The digital blocks can be configured to form counters, UARTs and other things like that. They are neat, but I've been hit with the same issue I have with Atmel: they obsolete parts too quickly, to the point that the newest software doesn't support the parts I have anymore. Annoying, and it's the reason I've never gone back to them. TTYL

On 10/5/07, WH Tan <KILLspamwhsiung.myKILLspamgmail.com> wrote: > > > I think I might be able to grab some ICL8038 if I want. But since > this is a new design I'll try to find other solution. I'll probably > take it as a last bet. > > It occurs to me that you could use a DSP chip, like a dsPIC or something similar to implement such a device in software. -- Paul Anderson VE6HOP RemoveMEwackyvorlonTakeThisOuTgmail.com http://www.oldschoolhacker.com "May the electromotive force be with you."

On Fri, 2007-10-05 at 12:47 -0600, Paul Anderson wrote: > On 10/5/07, WH Tan <spamBeGonewhsiung.myspamBeGonegmail.com> wrote: > > > > > > I think I might be able to grab some ICL8038 if I want. But since > > this is a new design I'll try to find other solution. I'll probably > > take it as a last bet. > > > > > It occurs to me that you could use a DSP chip, like a dsPIC or > something similar to implement such a device in software. Something like that wouldn't even need DSP, just a fast MCU with PWM would be enough (just a lookup table, and a physical output filter). Hitting the upper freq's available with the 8038 may be a problem though. Distortion may also be an issue depending on what you'll be using those waveforms for. TTYL

2007/10/6, Herbert Graf wrote: > On Fri, 2007-10-05 at 12:47 -0600, Paul Anderson wrote: > > It occurs to me that you could use a DSP chip, like a dsPIC or > > something similar to implement such a device in software. > > Something like that wouldn't even need DSP, just a fast MCU with PWM > would be enough (just a lookup table, and a physical output filter). > Hitting the upper freq's available with the 8038 may be a problem > though. Distortion may also be an issue depending on what you'll be > using those waveforms for. Hi Paul and Herbert, After more than a week of searching, I am convinced that most of the discrete analogue chips are gone. PIC or even dsPIC is an option now. I am not worry too much about the bandwidth because my application is targeting audio frequency range. I didn't consider them in first place because I would like to get 'true' sine-wave rather than approximated sine-wave. Probably I should switch my attention to find out if I can get an audio codec to do the same things. Best regards, -- WH Tan

> audio frequency range. I didn't consider them in first place because > I would like to get 'true' sine-wave rather than approximated sine- > wave I'd suggest you try the triangle-to-sine converter then (which does work rather well). You'll notice that the ICL8038 has a converter based on the same principle Or an external DAC. You can expect reasonable results using logic gates and resistors, but either that will require some fiddling or could be that you'd get a DAC comparitively cheaply

WH Tan wrote: > Hi Paul and Herbert, > > After more than a week of searching, I am convinced that most of the > discrete analogue chips are gone. > > PIC or even dsPIC is an option now. I am not worry too much about the > bandwidth because my application is targeting audio frequency range. > I didn't consider them in first place because I would like to get > 'true' sine-wave rather than approximated sine-wave. Probably I > should switch my attention to find out if I can get an audio codec to > do the same things. Look up AN-31 and AN-263 pdf's from National Semi. Then google stuff like "bob pease sine generator" and sine generator ap note "jim williams " site:linear.com and so on. Analog devices also has published valuable stuff in this realm. Other methods include turning square waves (easy to produce) into sines by filtering. In particular, you can us the MF10 filter chips to do this. The output may or may not be adequate for your purposes.

2007/10/7, Marcel Duchamp wrote: > Look up AN-31 and AN-263 pdf's from National Semi. Then google stuff > like "bob pease sine generator" and > > sine generator ap note "jim williams " site:linear.com > > and so on. Analog devices also has published valuable stuff in this realm. > > Other methods include turning square waves (easy to produce) into sines > by filtering. In particular, you can us the MF10 filter chips to do > this. The output may or may not be adequate for your purposes. > -

WH Tan wrote: >>Look up AN-31 and AN-263 pdf's from National Semi. Then google stuff >>like "bob pease sine generator" and >>sine generator ap note "jim williams " site:linear.com I've been following this thread. Now, just a simple/trivial hint: but, what about ROM-based tables with waveforms (sine/triangle/square) and a software loop to output them to a DAC? Subsampling/oversampling should generate a wide range of frequencies. Frequency will not be *very* high, but shold be good for audio range and a bit more. -- Ciao, Dario

On 10/9/07, Dario Greggio <TakeThisOuTadpm.toEraseMEspam_OUTinwind.it> wrote: > I've been following this thread. > Now, just a simple/trivial hint: but, what about ROM-based tables with > waveforms (sine/triangle/square) and a software loop to output them to a > DAC? > Subsampling/oversampling should generate a wide range of frequencies. > > Frequency will not be *very* high, but shold be good for audio range and > a bit more. Good idea. Actually many SPWM/SVPWM (for motor drive application) examples are using ROM (or now Flash) based data tables.

2007/10/9, Dario Greggio wrote: > Now, just a simple/trivial hint: but, what about ROM-based tables with > waveforms (sine/triangle/square) and a software loop to output them to a > DAC? > Subsampling/oversampling should generate a wide range of frequencies. > > Frequency will not be *very* high, but shold be good for audio range and > a bit more. Firstly, thanks again to all who respond to this thread. I appreciate all the valuable inputs. I will carry out some experiment soon... My apologies for not telling much about my requirements. Initially, I thought it would be pretty easy to get the replacement for ICL8038... I should have made up my mind when I failed at Google. But instead I thought probably Google might miss something :-) How come those great chips can be abandoned by their respective manufacturer ? :) Basically I am doing some kind of 'instrumentation' application... I want to generate audio range sine wave, which will be fed to a DUT (device under test). Initially I have put all those requirement together: 1) Output freq can be configured by software (not a requirement but would nice to have, isn't it?) 2) Amplitude can be configured by software 3) Distortion should be less then 2% ROM-table driven technique should be good to satisfy my requirements... still searching for a suitable DAC/ probably Audio Codec. Or I also thinking of several fixed-freq circuits using the suggestions by Jinx or Marcel. Those circuits will be hardware fixed to some interested frequency (say 400Hz, 1KHz and 10kHz) and software will choose which one to use. I will let you know once I complete the experiment. Thanks again & best regards -- WH Tan

WH Tan wrote: > 1) Output freq can be configured by software (not a requirement but > would nice to have, isn't it?) > 2) Amplitude can be configured by software > 3) Distortion should be less then 2% > > > ROM-table driven technique should be good to satisfy my > requirements... still searching for a suitable DAC/ probably Audio > Codec. > > Or I also thinking of several fixed-freq circuits using the > suggestions by Jinx or Marcel. Those circuits will be hardware fixed > to some interested frequency (say 400Hz, 1KHz and 10kHz) and software > will choose which one to use. Most of the analog sine generator techniques can easily have adjustable frequency and amplitude. Use digital potentiometers driven by your PIC. Then you can have good quality sine wave and be in control as well.

> easy to get the replacement for ICL8038 My very first function generator was made from a design in the Maplin catalogue ca 1979. It's the classic op-amp + thermistor. The spec claims 0.01% sine wave distortion at 1kHz. I still have it, but the the RA53 was appropriated for another project, so she no go no more It's a very simple circuit. I'll post it shortly

--- WH Tan <RemoveMEwhsiung.myTakeThisOuTgmail.com> wrote: > Basically I am doing some kind of 'instrumentation' > application... I > want to generate audio range sine wave, which will > be fed to a DUT > (device under test). Initially I have put all those > requirement > together: > > 1) Output freq can be configured by software (not a > requirement but > would nice to have, isn't it?) > 2) Amplitude can be configured by software > 3) Distortion should be less then 2% You can do DDS(direct digital synthesis) with a PIC or similar micro but you will need to add a D/A converter. There are also dedicated DDS chips but they tend to be pricey since they usually even the simplest of them are for RF. As I recall some that I played with from Analog Devices were in the $10-15 range. Cheerful regards, Bob

>> easy to get the replacement for ICL8038 > > My very first function generator was made from a design in the > Maplin catalogue ca 1979. It's the classic op-amp + thermistor. > The spec claims 0.01% sine wave distortion at 1kHz. I still have > it, but the the RA53 was appropriated for another project, so > she no go no more > > It's a very simple circuit. I'll post it shortly > I'll betcha it's a "twin-T" oscillator! I think one of those was HP's first product. They used a light bulb to adjust the gain so it WOULD oscillate but WOULD NOT clip. The twin T is, as I recall, just an amplifier and a voltage divider. The voltage divider has a series RC, then a parallel RC to ground. The amplifier input is connected to the top of the parallel RC. The amplifier output is connected to the top of the series RC. At f=1/(2*pi*R*C), the "voltage divider" has a "gain" of 1/3 at zero degrees. If the amplifier has a gain of 3 at zero degrees, you have an oscillator! Harold -- FCC Rules Updated Daily at http://www.hallikainen.com - Advertising opportunities available!

On 10/9/07, Bob Blick <bbblickEraseME.....sbcglobal.net> wrote: > As I recall some that I played > with from Analog Devices were in the $10-15 range. > > There is the DDS-60 kit: http://www.amqrp.org/kits/dds60/index.html The chip they use can be sampled for free. -- Paul Anderson VE6HOP EraseMEwackyvorlongmail.com http://www.oldschoolhacker.com "May the electromotive force be with you."

--- WH Tan <RemoveMEwhsiung.myEraseMEEraseMEgmail.com> wrote: > > Frequency will not be *very* high, but shold be > good for audio range and > > a bit more. Actually there are some cheap DDS chips from Analog, $3.95 gets you a 10 bit model with 25MHz max clock. You should be able to get a pretty accurate output through the audio range. Check out the AD9833. Serial I/O in a 10 pin package, 2.3 to 5.5 volts, voltage output of sine/triangle/square wave. Cheerful regards, Bob

Bob Blick wrote: > --- WH Tan <RemoveMEwhsiung.myspam_OUTKILLspamgmail.com> wrote: > >>>Frequency will not be *very* high, but shold be >>good for audio range and >>>a bit more. > > > Actually there are some cheap DDS chips from Analog, > $3.95 gets you a 10 bit model with 25MHz max clock. > You should be able to get a pretty accurate output > through the audio range. Check out the AD9833. Serial > I/O in a 10 pin package, 2.3 to 5.5 volts, voltage > output of sine/triangle/square wave. Yeah, I know those from AD, and actually I found it strange that nobody were advicing about them. So I guessed that OP wanted something in the audio range, such as the original 8038. But I seem to remember, yes, that those DDS can do audio as well. -- Ciao, Dario

Marcel Duchamp wrote: >>1) Output freq can be configured by software (not a requirement but >>would nice to have, isn't it?) >>2) Amplitude can be configured by software >>3) Distortion should be less then 2% >> >>ROM-table driven technique should be good to satisfy my >>requirements... still searching for a suitable DAC/ probably Audio >>Codec. I used a I2C DAC and a SPI DAC, and they both can reach up to audio limits - SPI much more of course. > Most of the analog sine generator techniques can easily have adjustable > frequency and amplitude. Use digital potentiometers driven by your PIC. > Then you can have good quality sine wave and be in control as well. In theory, this could be done creating tables in RAM with the desired amplitude, but, even at 16bits (i.e. slow - or a dsPIC is needed), this will not be as "smooth" as a analog trimmer. -- Ciao, Dario

2007/10/10, Bob Blick wrote: > Actually there are some cheap DDS chips from Analog, > $3.95 gets you a 10 bit model with 25MHz max clock. > You should be able to get a pretty accurate output > through the audio range. Check out the AD9833. Serial > I/O in a 10 pin package, 2.3 to 5.5 volts, voltage > output of sine/triangle/square wave. Thanks Bob. This one sounds really interesting... I am still reading its data sheet. Unfortunately can't do any test on this one right now as I have to wait for the part. Best regards, -- WH Tan

2007/10/10, Jinx wrote: > > I'll betcha it's a "twin-T" oscillator ! > > Note - > > Designed to run off one 9V battery > Original components shown (eg BC108) Sometime analogue design scares me ;) It's time for me to do some refreshment on those kind of analogue thing. Had been trying some simple design with op-amp... At the moment really don't have enough luxury to build this one :) Certainly will keep it and try to build one later. Thanks. -- WH Tan

> Sometime analogue design scares me ;) I know what you mean. But if you break that circuit down, ignore the switches etc, it's a basic op-amp configuration that you can find many examples of on the web, although many probably don't have a thermistor as the feedback element As I mentioned, it's sitting idle because I pinched the RA53, and now find those are bit expensive from my sources. But recently I've been stripping down some quite large switch-mode supplies and find they each have at least a couple of disk thermistors, of around 80R at ambient. Might see if I can adapt the circuit to use one. Not sure exactly what the thermistors are doing there. They are always in pairs, close to other. It'll be for compensation of some kind, but whether that's for frequency or current, not sure

Jinx Re the thermistors in the SMPS. They may be there as protection. We use PTCs to protect against short circuits on 48V systems instead of fuses. They are cheaper than polyswitches, the initial fault current is lower and the voltage drop under normal conditions is not of concern (and they reset themselves). RP On 19/10/2007, Jinx <RemoveMEjoecolquittTakeThisOuTspamclear.net.nz> wrote: {Quote hidden}> > > Sometime analogue design scares me ;) > > I know what you mean. But if you break that circuit down, ignore > the switches etc, it's a basic op-amp configuration that you can > find many examples of on the web, although many probably don't > have a thermistor as the feedback element > > As I mentioned, it's sitting idle because I pinched the RA53, and > now find those are bit expensive from my sources. But recently > I've been stripping down some quite large switch-mode supplies > and find they each have at least a couple of disk thermistors, of > around 80R at ambient. Might see if I can adapt the circuit to use > one. Not sure exactly what the thermistors are doing there. They > are always in pairs, close to other. It'll be for compensation of > some kind, but whether that's for frequency or current, not sure > > > -

> They may be there as protection. We use PTCs to protect > against short circuits on 48V systems instead of fuses. They > are cheaper than polyswitches, the initial fault current is lower > and the voltage drop under normal conditions is not of concern > (and they reset themselves). You could be right. These are 48V 60A Nokia supplies, and not a fuse in sight. Some pairs are down at the coal-face amongst the transformers and FETs (possibly operating a or bypass crowbar ? although there is a seriously chunky relay too), some are up in the voltage indicator (LM3914s/bar LEDs) area, and look to be connected to comparators

On Fri, Oct 05, 2007 at 02:23:34AM +0800, WH Tan wrote: > [Repost with tag] > > Hi, > > I plan to use ICL8038 signal generator chip from Intersil. > Unfortunately, it's flagged "not recommended for new designs" on very > first page of its data sheet: > > http://www.intersil.com/data/FN/FN2864.pdf > > Anyone know is there any close replacement for it? Not necessary pin > to pin compatible since I am doing new design. > > > Thanks & best regards, > > -- > WH Tan I'' a little late replying but you might check on the XR2206. Ray Warren