Misadventures in DIY Audio Electronics

[Amphiprion]

Today I literally boiled the insulation off magnet wire, blew up some MOSFET's in a half-bridge DIY switching amplifier of my own poor design (boo - expensive FET's), burned my thumb on a 25W wirewound resistor that just happened to be dissipating over 100W due to a circuit failure, killed a half bridge FET driver (only one! yay at $4ea), and destroyed a transformer (the one with the toasted magnet wire).

But progress was made. Oh yes, progress was made.

Anyone else have some fun lately? On the plus side, my scope, multimeter, power supplies save the transformer, and soldering station all survived. Less than $100 destroyed in components and transformer and no injuries other than a burned thumb.

[Letifo]

:jawdrop:

[Jed]

Well at least you have all your fingers. And thumbs :T

[JoshK]

How did it all start?

I haven't destroyed too much yet. I burned up a softstart by shorting the output. That was $25, but the parts are all replaceable. I think the resistor is the biggest victim.

I 'play' with tube amp building. I have to have a lot of respect for the project I'm working on given the voltages involved.

[cjd]

Dang. Worst I've done is hook up the 75v to the 25v side, (and reverse) popping a cap and... um. well, ok, so I never did get anything on that board working again and gave up on it. Man, caps are loud when they pop tho.

C

[Amphiprion]

Yea, last weeked I popped a pair of electrolytics - 1000uF 16V Panny FM's. I hooked my 13A power supply up backwards. Very nice dual 'pop' and even some fibrous residue on my desk from where the pressure scores on the top just barely split.

I wasn't working with anything more than 50V this weekend even including switching transients, but there was some serious current on tap required for this experiment. A couple of sparks from probing around with the scope/multimeter but nothing fantastic.

Another week, another board spin.

[Saurav]

Nothing as exciting as yours. A few days ago I made my power supply get very hot when I put one of the opamps on the board backwards, and I guess that shorted out the power supply rails, which of course I didn't discover for several minutes because instead of running basic PS DC voltage checks first, I went straight to testing the filter's transfer function. Other than that I've been fairly lucky for the most part. Touch wood... or solder, or whatever's appropriate here.

[Johnloudb]

Hey, if you're not blowing stuff up ... you're not trying IMO. I think that makes sense. :scratchhead:

I've blown up some electrolytic caps, and ruined a transformer, so far. Can't wait to build my first amp.

[Hdale85]

Destroyed one side of a 2 channel amp board, 1 transformers, and almost one PSU board for my DAC.

[Amphiprion]

Ouch. Been able to fix the amp board? If I've designed a project myself I usually find it easy to fix, but debugging a board someone else designed is never fun. A lot like programming, actually.

[Hdale85]

No I probably could have but it was a cheap chinese kit I bought online that had no instructions. Basically the chips should have been isolated and I didn't know. Was my fault but oh well was one of the first amps I did.

[cjd]

Whenever I first power up a supply, I wear eye protection, ear protection, and have a blast shield over the parts. If all is quiet, I check voltages, etc. If that all checks out OK, I move on.

C

[chasw98]

Nothing quite as spectacular as your results, but I did get a working prototype going this weekend.



It is a very low noise, low distortion 3 way balanced electronic crossover.

Chuck

[Saurav]

What boards are you using? Or did you design them yourself? And where is the enclosure from?

[Glen B]

I've received shocks from 350-400V B+ rails, burns from molten solder, blown up a few electrolytic caps and nearly melted down a few transformers during my fledgling tube amp building days as a teen during the 1970's. These days I get soldering iron burns every once in a great while but no longer blow stuff up. I always double-check my work first, then check again before applying power.

[Glen B]

Nothing quite as spectacular as your results, but I did get a working prototype going this weekend.



It is a very low noise, low distortion 3 way balanced electronic crossover.

Chuck

Nice, clean layout. That chassis looks like a Par-Metal.

[chasw98]

What boards are you using? Or did you design them yourself? And where is the enclosure from?

Marchand XM9 and LD28 boards with some different parts. It is a Par-Metal chassis. www.par-metal.com

[Saurav]

Thanks.

[cjd]

Anyone doing balanced multi-channel setups really owe it to themselves to check out the TwistedPear Audio Joshua tree relay based attenuator. Stackable, so you can control *n channels (as long as the power is there in the control board), and one board can either be two single-ended or one balanced line. Easy to tweak for perfect level matching.

[Hdale85]

Yes Twisted Pear makes some awesome stuff Thats where my DAC came from.

[cjd]

You can probably blame me for Twisted Pear going as modular as they have. I still have all my original circuits somewhere, and some day I really will get around to working on them more. I was really pushing the modular thing though, got a bit of grief for it on DIYAudio but hey. Good stuff.

Par-metal is good to have around too. I really should get some stuff from them.

Too many projects!

C

[Hdale85]

I think the modular thing is good...gives you many different configurations depending on each persons application.

[Amphiprion]

I frequently suffer from the 'plug it in and see what blows up' syndrome. I do populate power supply components and test them first, but beyond that just turning it on and blowing things up will help me find and diagnose things quicker. Young and impatient, this Hayenga is I always buy extra components of everything since you get the discount at multiples of 10 at Digikey and an extra board or two since any PCB after the first has such a low marginal cost.

[chrismercurio]

Mark,

Be safe.

I firmly believe in having solder jockey work checked out by someone competent and letting THEM bring it up with a variac AFTER QC'ing my work. Power amplifiers often have voltages that can kill and at the very least harm. I admit to being electrocuted and burned on more than one occasion, so my opinion comes from one who has lived to tell.

Chris

[Amphiprion]

I am perhaps not so cavalier as I might like others to believe But I do have a penchant for true experimenting. Anyone note the lack of details I post about my projects, other than what I blow up? One of these days I'll finish something successfully and be able to post about it.

Right now I am working with low frequency (<100kHz) discrete PWM designs, with dead time and shoot through looking like they will be the next two bugaboos I have to deal with. I'm probably going to go full bridge as well. I listen to KMFA out of Austin on my computer because local stations, well... let's just say I'm not EMI compliant.

Hey, if you're not blowing stuff up ... you're not trying IMO.

I'm totally stealing this for my tagline.

[Hank]

Well, Kid, when I was about 17, I bought an H.H. Scott LK-48B integrated tube amp kit and built it. Very complex, all point-to-point wiring. I made only one mistake, but it made for quite a show - wired an electro backward and when I did the smoke test, one of the output tubes had a spectacular internal blue vesuvius display. I've still got the amp and plan to re-cap it some day. When I got out of the Army, I took the DeVry Institute electronics correspondence course, which included a Heathkit breadboard, oscilloscope and their legendary 25" TV. After I built the 'scope, I was taking voltage measurements at test points to calibrate it and accidentally encountered the strange muscular contraction effect of putting myself between 800 volts and ground. :E

[Amphiprion]

Hey Bogey,

Does anyone have any poker games on the horizon? I haven't heard anything from Tubes or any of the usual suspects. I might have something in two or three weeks by then to show you (under an NDA of course). Right now I'm waiting on replacement parts, but it's really a very portable system.

I'll warn you though - it contains radioactive elements

[Johnloudb]

I'm totally stealing this for my tagline.

Ummm, Mark, that tag line ("If you're not blowing stuff up, you're not trying hard enough") is my intellectual property. I'll need royalties ... I'm thinking 10 cents per post.

[Hdale85]

That counts towards old posts right? Because now its readable on those as well. So that means he's up to 55.20 already.

[Johnloudb]

Good point. Yeah, I think your right about that. Thanks Dougie :T

I'll accept Paypal.

[Hank]

Does anyone have any poker games on the horizon?

Kid, I want to have another fajita grilling/Texas Hold-'em poker to-do soon, maybe Friday the 29th - bring your radioactive concoction. After this many years of sniffing lead-bearing solder fumes, a little radioactivity can't hurt. I'll put out a feeler e-mail for poker. I do know that Big Dog is back in town after an extended California vacation.
-Bogey

[Johnloudb]

Mark, you're certainly welcome to use that tag line, no charge . Actually, I'm flattered you decided use that. Good idea.

[Amphiprion]

Thanks John, I was afraid I was going to have to get my lawyer involved

Hank, that date works for me. I've contacted a fellow in Austin about auditioning his Orions sometime; if it works out would you mind if I extended the invite to him? I've never met him before but hey, he has a set of Orions

[Hdale85]

Just to get on Mark's side of this :B John your not the first person I've heard use that term anyways :P

[Amphiprion]

I like it because it has the same feel to it as "If it ain't broke, fix it" but is more akin to my treatment of semiconductors. I think I've smelled every semiconductor technology out there on fire.

FWIW, the absolute WORST was a transient voltage suppresion zener diode I had 'protecting' a MOSFET. It would handle a 1.5kW 10us transient pulse, but it was nice and small (DO-201 case). I can't remember what happened, but it got so hot the plastic casing cracked in half and everyone in the 20x20x40' work bay complained about the smell for the next four hours. It was awful.

Oh, and new transformers get here today! We will see how long they last. I found a very basic, stupid error in my circuit which is why I toasted it last time. It should be an easy fix this time, hopefully, and shouldn't require another board rev. If it works, I get to go buy a variac and start cranking the supply voltage from 24VDC up to ~170VDC (rectified wall voltage). That'll be when it gets scary.

[Hank]

The Kid wrote:

would you mind if I extended the invite to him?

Nope, not if he's bad at poker and bets heavily :W

[ThomasW]

Gentleman,

Email, PM or IM are a good way to arrange social events...

[Johnloudb]

I like it because it has the same feel to it as "If it ain't broke, fix it" but is more akin to my treatment of semiconductors. I think I've smelled every semiconductor technology out there on fire.

FWIW, the absolute WORST was a transient voltage suppresion zener diode I had 'protecting' a MOSFET. It would handle a 1.5kW 10us transient pulse, but it was nice and small (DO-201 case). I can't remember what happened, but it got so hot the plastic casing cracked in half and everyone in the 20x20x40' work bay complained about the smell for the next four hours. It was awful.

Oh, and new transformers get here today! We will see how long they last. I found a very basic, stupid error in my circuit which is why I toasted it last time. It should be an easy fix this time, hopefully, and shouldn't require another board rev. If it works, I get to go buy a variac and start cranking the supply voltage from 24VDC up to ~170VDC (rectified wall voltage). That'll be when it gets scary.

I cracked the case of a regulator once by wiring it backwards. Fortunately my projects have usually emitted some smoke before anything gets destroyed.

That sounds like a monster class D amp you're building. If you don't mind me asking, what's your plans for it? Is it going to be a full range amp, or a subwoofer amp?

[Amphiprion]

Fine, I'll go ahead and spill the beans. My avatar should be a dead giveaway as to what I'm working on. If that's not enough, it's something like this.....(this not mine, just something off of youtube):

http://www.youtube.com/watch?v=rasp88nbsRw

I'm doing something very different though with the operation of the PWM amp and the high voltage transformer I'm using (NOT a flyback) compared to what I've seen out there. The radioactive joke earlier comes from the use of thoriated tungsten welding electrodes (ground to a fine point which will never melt) that are totally harmless. The transformer I blew last weekend was my HV transformer. UPS should have the new ones here this evening. I'm also totally regretting my choice of half bridge chip - I so should have gotten the IR2304. 0.5us of dead time is terrible to look at on the scope.

I'm really hoping to find a magnetics manufacturer that will do a prototype or two for me of a new HV transformer. That's where I'm limited right now. Eventually I would like to be driving it with rectified or even doubled and rectified AC (169 or 338VDC) in a full H bridge config. Right now I'm just in testing mode though, and I'm still able to get sparks even with just 24VDC in a half bridge mode.

[Johnloudb]

Wow, that's really cool. :T

I remember the Plasmatronics loudspeaker, that had a plasma tweeter. They also used helium. This is obviously different. Well, I don't expect you to give away all your secrets. One question though. A problem with the Plasmatronics was the release of Ozone, or so I've heard. Those Sharper Image high voltage air purifiers also release small amounts of Ozone. Is that a problem with type of design?

[Amphiprion]

Note to self: When working with non-isolated power supply and scope, make sure what you think is ground is ground before you go clipping on your probe ground.

Followup note to self: GFCI is a wonderful, wonderful thing.

Followup to the followup note to self: Tektronix makes a heck of a rugged scope. Although it does need to be run the the auto-cal again for channel 1.

[Amphiprion]

Wow, we posted at about the same time!

Yes, I am definitely making O3, as well as any other magnificent compounds that are formed when the ionized gases decide to recombine. The guy on youtube says it's no problem in a two electrode setup, but I'm getting distinct ozone emissions. Makes the lab smell better in any case. Helium was used because you could get more SPL out of it, argon can also be used and is much cheaper but doesn't have that advantage from what I've read. Right now I'm atmospheric.

[Johnloudb]

Hey, open a window already!! Don't kill your lungs. xx)

[Amphiprion]

Hmm. I wonder if I took up smoking if the ozone would cancel it out? It probably doesn't work like that.

On the other hand, the 170VDC test (actually 168.9VDC according to the meter) went off without a hitch. I took photographic evidence too:


Images not available


The best I could get with the 24VDC supply was about 5kV based on arc breakdown distance, and it was non thermal (it was noisy zappy blue sparcs that couldn't be maintained). That's an easy 10KV right there of thermal plasma, totally silent. This is just the "zero signal' test. I'll be running test tones through it this weekend to see if I can get it to make some sound, and I made the input compatible with your standard stereo output so I can play some music through it if all goes well.

After that, I have to start cutting traces on the board to turn the full bridge + 200VDC electrolytics into a voltage doubler and run the half bridge amp section at ~340VDC but that takes some serious hacking of the board. I still need to find a better transformer too.

I really was going to hold off and keep this secret, but it's just too cool to not share. Also please note my use of thoriated tungsten electrodes, and proper HV insulated wire unlike in the youtube vid Unfortunately you can't solder to tungsten, so I have to use clips for that. I think you can 'braze' a connection between copper and tungsten, but I've got to read up on that.

[Amphiprion]

No replies? NO REPLIES?

I'm hurt! Literally. It shocked me.

(no just kidding )

Today is the "can it make sound" experiment. I should have results by this evening.