[luvdunhill]

Quote:

Originally Posted by NeilR /img/forum/go_quote.gif As soon as you get your scope you will want a function generator so you have a signal to run through your amp. I ebay'd a 4MHZ Wavetek 182A for about $100. That is a fairly common model. It is a decent basic generator and the square waves look ok up to about 100KHZ or so, which is all we probably care about. The rise is not perfect (a touch of overshoot) but it is good enough for what I need. If anyone is interested I could probably shoot some images and post it. A 4MHZ is probably a good bandwidth for these things. Lesser bandwidths will have more limited decent square wave range and higher bandwidth models get a bit more expensive. Then I ebay'd a lot of twelve 6 foot BNC-BNC cables for $20 or so. I cut them up and make cables as needed. BNC connectors are expensive; I thought that was a decent deal and I have a lifetime supply of pre-crimped connectors. You'll want one BNC-BNC to connect to your scope's trigger; that simplifies the triggering and eliminates constant playing with the trigger level to keep the image stabilized. Another set or sets to connect to your various amp input jacks (3.5mm and/or RCA) and maybe a fallback set with alligator clips or grabber clips.

Niel:

This was a very informative post! Thanks! I am looking for a Wavetek function/sweep generator and was wondering if anyone could expound on whether or not to go under 4 Mhz? I'm looking at the Wavetek 180 and 188 and have compared them as follows:

Frequency range: 0.004Hz to 4MHz (188)
Frequency range: 0.010Hz to 2MHz (180)

Amplitude: 1V to 10V p-p into 50O (188)
Amplitude: 30mV to 10V p-p into 50O (180)

Wave Forms: sine, square, triangle and TTL pulse. (188)
Wave Forms: sine, square, triangle, ramp, DC and TTL pulse. (180)

Sweep rate: 30ms to 1 min. (nominal) (188)
Sweep rate: 30ms to 20 sec. (nominal) (180)

I'm not too sure what I should be looking for. I tried the software function generation, which is decent, but I couldn't get a square wave worth anything out of my motherboard's sound card

Plus, these guys are cheap and I can wait around until I find a good deal on one! I noticed that amb uses a Wavetek 188, but I was wondering if the 180, which is sometimes cheaper would work just as well!

Thanks!!

 

[tangent]

Square wave testing is one of the best reasons to have a high-quality function generator. It's probably the easiest way to test for instability in an otherwise marginal amp design, for a number of reasons. First, it forces the amplifier to work at high frequency, where instability risks are the greatest. Second, square wave testing pushes a wide spectrum of frequencies through the amplifier at once, which is a more difficult test than simply using a single high frequency sine wave. Finally, the output is easy to interpret: is it still a square wave after going through the amplifier or not? (This requires an oscilloscope capable of higher frequency than the generator, but you should have that before you get a function generator anyway.)

100 kHz is the lowest frequency I'd accept for this sort of test. Since the generator you're looking at has half the bandwidth of the 188, it can probably only put out 50 kHz square waves with equal quality to the 188.

Instead of throwing good money at the problem before you can afford the right tools, I would recommend making your own generators for now. For about the same cost as an inexpensive DIY headphone amp, you can make a good quality fixed-frequency square wave generator. What you're paying for in a commercial generator are for pots, packaging, and profit.

 

[luvdunhill]

tangent:

thank you for your advice! I have found a few kits out there:

MAX038-based, 20 Mhz complete kit ($85)
http://store.qkits.com/moreinfo.cfm/QK101

Exar 2206-based 1Hz to 100kHz, various wave forms ($25/kit)
http://www.jameco.com/Jameco/Products/ProdDS/20685.PDF

LM/NE555-based Square Wave (1Hz 10Hz 100Hz 1KHz 10KHz & 100KHz) ($12/pcb)
http://store.qkits.com/moreinfo.cfm/QK111

Another rather simple solution using a NE566 chip (up to 1 MHz):
http://linearparts.com/documents/ne566.pdf

LM348-based, various wave forms (6-7kHz) ($10/pcb)
http://store.qkits.com/moreinfo.cfm/QK23

Intersil 8038
(no kits being offered, but decent chip, and discussed a lot)

I'm assuming this is what you were referring to? I could also just make a simple generator out of an opamp, but the ones above seem like they might be good investments.

Finally, do you think the Wavetek 188 is a decent unit for audio? It seems like you may be implying that it's okay, but maybe not the best choice.

Thank you so much!

 

[amb]

Quote:

Originally Posted by luvdunhill /img/forum/go_quote.gif Finally, do you think the Wavetek 188 is a decent unit for audio? It seems like you may be implying that it's okay, but maybe not the best choice!

As you mentioned earlier, I have one of these. It's an oldie but a goodie for audio use. Surely it's not the best, but good enough to have nice clean 100KHz square waves with much faster slew than just about any audio amp I'm likely to test. Even though it's called a "4MHz function generator" and it could indeed produce signals to that frequency, the square wave shape deteriorates as you go past several hundred KHz. Nevertheless, a square wave's edge is what's most interesting, and a 1KHz square wave is as good a test as a 1MHz square wave as long as you have your scope's X-axis time scale zoomed in far enough.

I bought the 188 used for a song (it was DOA but was an easy fix -- replaced blown output transistors), and it has since given reliable service for several years. I have no complaints or desire at this point to replace it.

Whether it's a good buy for you depends on the price. If it's cheap, and in good working order, then you can't go wrong.

 

[tangent]

Quote:

Originally Posted by luvdunhill /img/forum/go_quote.gif MAX038-based, 20 Mhz complete kit

These generator-in-a-chip deals are probably going to be equal in quality, at best, to an inexpensive commercial generator. (Most of those generators are probably built around these chips!)

I recommend avoiding that sort of thing because I had a bad experience with a quality-brand generator running ~$400 new. The sine waves had visible distortion -- 1.8%! -- and the square waves rang like bells. It was almost completely useless for audio.

On reflection, this isn't too surprising. I opened it up and found that the parts I saw probably totaled $100 or so; very unimpressive, and completely DIYable if one cared to. The normal markup over the parts cost for this sort of product is 3:1 to 5:1, so $100 in parts is about right.

What I recommend instead when DIYing are fixed-frequency circuits. It's a lot easier to tune a fixed-frequency fixed-function circuit than something that tries to be all things to all people.

Since my bad experience, I've decided that there are only three reasonable courses:

1. Keep it cheap. Use a good PC sound card with generator software, build single purpose-built DIY circuits, etc. I can't imagine spending more than about $100 going down this path, and that's if you get nice software and replace your sound card.

2. DIY audio-specific test sets, or buy used ones. For DIY, there have been a couple of series on DIY audio test gear in audioXpress and its predecessors. In the used market, there's the venerable HP 339, anything from Audio Precision or Sound Technology, etc. Now you're up in the several hundred dollar range. That might sound expensive, but I'll pit a $300 dedicated audio test set up against a $300 20 MHz do-everything function generator any day, if audio testing is the only goal. For this kind of money, you can have wide frequency range or accuracy, but not both.

3. Go big: new generators by Tek, Agilent, Rohde & Schwarz, or Stanford Research. Now you're up to around $1000 in the used market, and at least $1,600 new.

I partition the situation like this because I see big dead zones in the gaps. $50-200 is particularly worrisome. Here you find the half-assed DIY projects, and the off-brand commercial crap. Above several hundred dollars but not yet well over a thousand, I'd be worried but not immediately scared off. Shop carefully in that range.

Quote:

LM/NE555-based Square Wave

That's more in line with what I had in mind, but instead of the kit you link to, I'd just get a 555 from Radio Shack and build something from a schematic found online, or the datasheet.

There are also popular square wave circuits based on standard logic ICs. The 4049 CMOS inverter comes to mind for some reason. Also, Forrest Mims.

Quote:

Finally, do you think the Wavetek 188 is a decent unit for audio? It seems like you may be implying that it's okay, but maybe not the best choice

All I'm saying is, it barely meets my lowest standard for utility, on paper. How it performs on the bench is of course much more important, but since I've never used it, I cannot say more.

 

[adhoc]

Quote:

Originally Posted by n_maher /img/forum/go_quote.gif Ditto that request, I've been thinking about a scope for a while and with a couple of upcoming projects it may be required.

Same here. Snubbing done correctly.

Damn - I was considering a PC o'scope.

 

[flecom]

I would really like to know what you guys think of some kit I picked up from someone who was leaving the amateur radio hobby... I got it all fairly inexpesnive for what it is but I was wondering how good it is now...

first was an oscilliscope, a Philips (Fluke?) PM3320... wasnt able to find a whole lot of information about this guy online... just that it is a 200mhz 2 channel digital storage scope...

pic here...

http://www.valuetronics.com/Details....Philips_PM3320

one of the other items was an HP 3325A function generator

http://www.amplifier.cd/Test_Equipme...rs/HP3325A.htm

besides that, i also picked up an HP 8562A 1KHz-22GHz spectrum analyzer... tho not sure how useful this would be for audio? and i wish i wish i wish i could get a tracking generator for this thing... but they didnt make one for the 22ghz model

http://www.atecorp.com/Equipment/HP/8562A.htm

I also picked up a very nice Rohde & Schwarz signal generator that goes up to several GHz but I cant remember the model... it needs a little work in the power supply section becuase its a little noisy... some of the tant caps are bad i think... and finally a Lambda LK350-FM giant heavy as heck adjustable voltage / current limiting power supply...

so overall i bought myself a pretty nice bench i think... but i would love to hear what you guys think about the stuff as far as how releveant it is in audio... (i bought all this stuff for RF work... and audio over RF etc... and then later realized duh, i can use it for audio too no?)

 

[amb]

Jeez flecom, I'd say those are quite overkill for audio...
Nice stuff though.

 

[psilosome]

Quote:

Originally Posted by tangent /img/forum/go_quote.gif That's more in line with what I had in mind, but instead of the kit you link to, I'd just get a 555 from Radio Shack and build something from a schematic found online, or the datasheet.

I strongly recommend a CMOS-based 555 - it will give a much cleaner square wave especially at higher frequency. Here's a photo of a 83 kHz square wave using the CMOS chip (TS555IN) versus the regular (BJT?) chip in the same circuit. The regular chip only went to 75 kHz. This was using a 9V battery for power so maybe the problems wouldn't be as bad with a beefy power supply.

 

[luvdunhill]

Quote:

Originally Posted by psilosome /img/forum/go_quote.gif I strongly recommend a CMOS-based 555 - it will give a much cleaner square wave especially at higher frequency. Here's a photo of a 83 kHz square wave using the CMOS chip (TS555IN) versus the regular (BJT?) chip in the same circuit. The regular chip only went to 75 kHz. This was using a 9V battery for power so maybe the problems wouldn't be as bad with a beefy power supply.

psilosome:

thank you so much! this does raise another point, in that Tangent was recommending a minimum of 100kHz, and both of these solutions seem to fall short (at 9V at least). I've also looked at the LM566C, voltage controlled oscillator, but my guess is it won't perform that much better. It's a start though. Am I right to assime that the TS555IN will probably perform as well as the 4049 CMOS inverter Tangent recommends?

tangent:

thanks again for setting me on the right track! I'm going to digest this information and maybe try and hunt down the AudioXpress articles.. for now I think I'll invest in the above solution and maybe a new soundcard for testing with RMAA

 

[psilosome]

Quote:

Originally Posted by luvdunhill /img/forum/go_quote.gif psilosome: thank you so much! this does raise another point, in that Tangent was recommending a minimum of 100kHz, and both of these solutions seem to fall short (at 9V at least). I've also looked at the LM566C, voltage controlled oscillator, but my guess is it won't perform that much better. It's a start though. Am I right to assime that the TS555IN will probably perform as well as the 4049 CMOS inverter Tangent recommends?

I'm really not very qualified here, but I will tell you that I did not push the 555 to its limits. I didn't even de-flux the protoboard. The circuit I used (with the circuit essentially right out of the datasheet) called for 100 kHz with those component values. Also, I doubt that going to a higher voltage would help much in getting higher frequencies - a power supply better able to deliver instantaneous current might reduce the glitches in the waveform from the regular 555.

 

[luvdunhill]

hm, here are a "few"

more circuits:

http://www.epanorama.net/links/oscillator.html

I guess I'm shooting for around 100 kHz... Maybe just pick a few of these circuits and try them out? I really don't have a good itea on how to compare their performance, unless I could compare rise/fall rates? I suppose I need something that is useful given the slew rates of the op-amps I'll be testing... The OPA637 has a slew rate of 135 V/us, I suppose that's a decent benchmark.

Anyways, I have more than enough to keep me busy here!!

 

[luvdunhill]

okay! Here are the circuits I've decided to get parts for and experiment with. If any of the circuits on the link about look like they'd perform better than these, I'd appreciate the heads up, otherwise I'll just try these out:

CMOS 4047 IC-based -- Logic - Monostable Multivibrators CMOS LP Mono/Astable
* simply add a resistor and capacitor!

CD4007UB IC-based -- Logic - Buffers, Drivers, and Transceivers Dual Comp Pair Plus
* a few more resistors, but then I will just need to vary one resisitor and one capacitor

LMC7215 -- IC COMP MICRPWR R-R CMOS 8-SOIC
* low power, and another design to mess around with... just add a few resistors, and vary a resistor and capacitor

CMOS 4011 NAND-gate -- simply two NAND gates, two diodes, two resisors and a capacitor. Simple circuit!

TS555IN -- Straight from the datasheet... CMOS '555 timer circuit.

Thanks all for your help and I'll let you know how things go!