Reasonable price for signal generator?

For a change it's nice to see some cons listed!. There is very little electronics in there, so maybe 98% of the build cost will be mechanical items, costed out in small quantity say about $11. There *should* be an assembly time labour cost of maybe 4$, giving total manufacture costs of $15 and a sell out price based on say a 1.5 mark up of say $22. But ... I wouldn't buy one at that price :( reason is you're going to have to offer far far more in the way of spec' than you are here. For example, rise and fall times could easily be 10nS 0-5Vpp 50ohms all the way to 10MHz using just (cheap) transistors and a 74HC logic gate. Another example, The battery life of 54Hours is unbelievable. You must assume that at least some part of the time a bulb or speaker just might be connected. The case volume you have does not allow for a battery offering a realistic life greater than an hour or so. Honestly, I'm not intentionally being negative, it's just that I've seen the disappointment in peoples faces resulting after much effort and time has been spent on these type of projects. Best learning curve though is personal experience :). A valid offer price can be reached simply by pricing up your parts and own labour cost. Add a suitable profit margin on and then offer the unit for sale. (this is a UK perspective, details may be different in US but the principle is similar) john

Err, no.

FORGET all about making "standard" pieces of test gear, buy the way nobody carries a squarewave generator in their toolkit, find yourself a "niche" device.

A few comments...

Price isn't the issue, it has to be useful to someone, it has to have a target market. Who needs a portable square wave only generator? The "TTL probe" kind can be handy for circuit work, and there are a few of those around already.

It needs a sine wave output.

It needs a low battery LED.

The rise/fall time is poor. I assume the TTL output is much better?

Dave :)

Re who needs.. I need one. There are endless projects where it comes in hand to test across frequency ranges, and others where just modulating a bias current of a photodiode is interesting or useful, for example.

It has a low battery LED.. the power LED gets noticibly dim once the regulator can no longer maintain 5 volts. It was a feature (low batt LED), but I decided the extra few dollars in parts doesn't reveal much more than the dimming power LED does.

I agree with earlier opinion that estimated power usage should account for different loads.

Ahh, it doesn't have a seperate TTL output.. but the output is TTL compatible level-wise and I guess edge-wise minimally.

I agree wholeheartedly with the fellow who says "FORGET about making standard test gear".. but it still makes sense as a strategy to develop 'rudimentary' products - for experiment at least.. and for bread, butter, and parts, not to mention a volunteer project and the student loan obligation. Possibly developing products and weaving services in later, like R&D, prototyping, engineering things. I've got at least one project that could fetch hundreds apiece no doubt, but there is R&D overhead (time) for developing the usb interface and software. I'm getting into programmable logic, CPLDs, just to broaden the scope of the tools at hand.

I'm surprised anyone would think this sort of device, a wave generator, is not so useful. The nice bench generators with all the useful features start over $500, with the cheapest one being at least $300, and then it's a hassle to use off the bench. For an experimentor / hobbyist, having some signal source for less than a hundred dollars could be a bonus, and the level they want to get in at technically. I'm not betting on it being a smart fledgling business move, just that I'd start there and crank out a few amplifier models as well.

When I made a pre-amp for capacitive-response musical keyboard, I could check the bandwidth went well above 20 kHz because I have this signal generator.

Generally, since variable level requires an output buffer in this case, op amps that can deliver current tend to be heavy quiescent current users, that or lack the slew rates necessary to preserve source spectrum. Also, a more expensive part doesn't really belong in something that is trying to be the Casio of signal generators, cheap but good.

Regarding the need for a sine wave, indeed, but replicating what the big boys are doing (in function) doesn't sound like a favorable enterprise... so I "niched" the approach to something that isn't really being offered, though it's time-intensive to start, I won't be too surprised if there are modest rewards. Thanks for your feedback. Geoff

Ok, that's good.

The 600ns rise/fall time makes it not very useful as a TTL generator unfortunately. I would not advertise this capability at all.

I'm not saying it's not useful. It's just that it is: a) Too big, way too big. and b) Lacks a sine wave feature

These are two very major drawbacks IMHO.

A sine/square wave generator can be made a lot smaller than this.

There is a sine/square battery powered kit available that costs about AU$20 (about US$15 + postage) and is smaller. Not as wide a range as yours, but it's got sine wave too.

There might be a market for it, but I would seriously look at making it smaller at the very least. Also, by making it a sine wave oscillator to start with you can simply square it up and that gives you square wave too for virtually no additional cost. So much extra advantage and market appeal for want of a circuit topology change.

As a designer of low cost kits (and I have published two low cost function generator designs), I would not have even considered a square wave only design of this size. That's just my 2 cents worth anyway.

BTW, got any photos of the unit?

Regards Dave :)

I wouldn't really want one, I would (and did) buy an old Wavetek 142, for about $100 plus $100 shipping. You might look at the front panel controls on that one, I believe that they are particularly well thought out. Having an extra output which is not 50 Ohm but can drive high currents (as you suggested) would be a nice extra feature for playing with AC motors etc.

Chris

Actually it's 3.5 inches by 1 inch by 5.5 inches in case size.. it's just that the banana jacks and knobs extend that to other dimension.

I've got some photos, and will mail you something from the first version which is nearly identical except for improvements in labelling.

Actually it's 3.5 inches by 1 inch by 5.5 inches in case size.. it's just that the banana jacks and knobs extend that to other dimension.

I've got some photos, and will mail you something from the first version which is nearly identical except for improvements in labelling.

Cool. Post them on the web so we can all see.

Dave :)

It does drive those currents, but the evaluation of impedance is certainly not exact, and is based on a nominal 40 ohm output impedance of the driver chip and 10 ohms in series with that (protects the amplifier from major overcurrent situations) .. the impedance goes up with frequency.. according to the spec sheet, but it does drive across the 1 to 3.5 volt range at a megahertz into a 50 ohm load... barely but indeed... at 1 MHz the slew limit squashes the range a bit.

I'll put the new label scheme on a new one and post a picture or that within 12 hours or so.

Thanks for your input . Geoff

It does drive those currents, but the evaluation of impedance is certainly not exact, and is based on a nominal 40 ohm output impedance of the driver chip and 10 ohms in series with that (protects the amplifier from major overcurrent situations) .. the impedance goes up with frequency.. according to the spec sheet, but it does drive across the 1 to 3.5 volt range at a megahertz into a 50 ohm load... barely but indeed... at 1 MHz the slew limit squashes the range a bit.

I'll put the new label scheme on a new one and post a picture or that within 12 hours or so.

Thanks for your input . Geoff

Images of the signal generator can be found at;

Thanks

Very nice!

Dave :)