Mrboo's Diy Preamp

[markiemrboo]

Yep yep, I am at it again I am going to try and update this one as I go along, but I am quite useless when it comes to writing things up so... uhm.. yeah, might take some time before it's all complete and in a case!

 

First off, I didn't really design any of it (well, I actually could have designed the fan controller that can be seen in the picture and it has been slightly altered by myself! I'll probably make a separate post for the fan controller though). I used other peoples schematics, and drew my own PCB's again. I thank the people who drew the schematics very much indeed for making them freely available to me to use!

 

 

For the actual preamp part, I followed http://sound.westhost.com/project88.htm. I have NE5532's and OPA2134's to try there, mainly just because I am curious as to whether I will be able to hear a difference between the two!

 

The power supply for the preamp is identical to my power amp power supply, only on a revised (and still not bug free: explained later) layout. This was designed by 'carlosfm' over at diyaudio.com. http://myweb.tiscali.co.uk/nuukspot/decdun...nub.reg.psu.png

 

It also features a 'delayed startup', which I got from Decibel Dungeon (nuuk from diyaudio.com owns the site) http://myweb.tiscali.co.uk/nuukspot/decdun/gc/555delay2.gif This shorts the output of the preamp on power on and off, for a certain time period, to avoid any funny electrical noises getting through to the power amp and thus the speakers.

 

 

Following are some pictures of current progress. Unfortunately I *still* haven't got my camera back from RMA, so it's webcam only again Same methods as the power amp, print out on my inkjet, photocopy at my local staples, iron on to PCB and then etch. I could post the Eagle PCB files, but I don't want anyone trying to make the PCB's before I have managed to test them myself to make sure they don't explode!!

 

I've got the power supply complete so far. I ran in to two issues with my layout this time. I underestimated the size of the resistors, so they don't quite lay flat to the board (I dont think this is too big of a deal, hopefully) and... I *seriously* underestimated the size of an output capacitor. It was way too small and had to be mounted on the underside of the board. Not really a bad thing to be honest, as people always say "as close to the pins as possible" when it comes to these capacitors anyway... and this one's directly on the pins of the LM338's

 

Here you can see a piece of PCB artwork on paper (this is the fan controller), and then the copper PCB's just after ironing the paper artwork on to them.

 

Here are the boards after etching. As you can see, this time around I have tried to keep a lot of copper on the board to reduce etching times and make the etchant last longer (as it has to remove far, far less copper).

 

Here's the power supply PCB's with the components soldered on, a top view. I've actually separated the filtering capacitors from the regulator board this time ( modular? or just stupid? ). This would allow me to upgrade things easier if I want to, and also allows me to have two regulator boards coming off one filtering board. It also makes it easier for me to have multiple filtering boards on one transformer (though I haven't drilled the holes on this board, because I won't be needing to). Basically, the filtering board has two sets of inputs and two sets of outputs. The regulator board also has two sets of outputs, so I could power two preamps off of it if I wanted to. I'll actually be using this to provide power to the delayed start up.

 

Filtering is currently 680 x 3 = ~2040uF per rail. I just happened to have them laying about. I'm hoping this will be enough, but if not there's room on the board for up to 20,000uF per rail

 

Here's the underside of the boards. You should be able to see the two caps on the bottom soldered directly to the pins of the LM338 regulator. Whoops... You might also noticed that there's alot of silverness (solder) and no copper. Why have I done this? Well, I bought some lacquer spray, to stop the copper from oxidis(z?)ing from Maplin (GRR), and basically I hate it. Didn't seem to work well at all, so I just 'tin' the whole thing instead.

 

Here's the toroidal transformer for the preamp. It is *so* tiny and cute compared to the one in my power amp! Dual secondary 15VAC, 30VA (1A per secondary, plenty enough). Picture taken with the PCB's in view, and in my hand, as an attempt to show how stupidly small it is

 

 

I intend to put the finished product in a 1U version of the case which my power amp is in, from Maplin. It could very easily fit in something a heck of a lot smaller but meh...

Edited January 13, 2007 by markiemrboo

[romeo55]

Wow man... just wow.... heh, I can't believe your printed the PCB yourself, I tried that method as well and pretty much screwed everything up Heh, ended up calling a company to print it for me

 

Let us know how it goes! :thumbs-up:

[Andrewr05]

Looks pretty darn good dude!!

It would be neat to etch your name or a cool symbol into your boards too...

That way if you for some reason ever sold your work people would know who made it.

Plus it would at to the "GEEK FACTOR"

Edited January 14, 2007 by Andrewr05

[hardnrg]

you should put it in a tin of baked beans (minus the beans/sauce of course )

 

baked bean amp! yeah! lol... much better than boring snoring rackmount cases

 

nice pcb layout

[Makaveli]

Wow! Looking good! Keep up the good work.

[markiemrboo]

Wow man... just wow.... heh, I can't believe your printed the PCB yourself, I tried that method as well and pretty much screwed everything up Heh, ended up calling a company to print it for me

 

Let us know how it goes! :thumbs-up:

 

I made the PCB's for the power amp too A company would be much more professional as you usually get a silkscreen and stuff. I have to look at the PCB layout on screen to know where components go, as it's blank on the component side. I think I might be able to get a dodgy silkscreen by just... ironing on a laser print out on the top aswell. I might try this... one day. Anyway, was it expensive? Who did you use? What was it you built?

 

 

Looks pretty darn good dude!!

It would be neat to etch your name or a cool symbol into your boards too...

That way if you for some reason ever sold your work people would know who made it.

Plus it would at to the "GEEK FACTOR"

 

I tried a bit of writing on the softstart for my power amp. It didn't come out very well really though. I think the writing would have to be fairly large = a bit of a waste of precious PCB space It would be nice to have my initials or something on them though.

 

you should put it in a tin of baked beans (minus the beans/sauce of course )

 

baked bean amp! yeah! lol... much better than boring snoring rackmount cases

 

nice pcb layout

 

lol minus the tin plating or whatever it is would probably be quite nice too Unless you enjoy short circuits?! (who doesn't?!) The tin would have to be rather large though. Can you get industrial size tins of baked beans?

 

Wow! Looking good! Keep up the good work.

 

Thanks!

 

What annoys me is that I can't really start doing stuff till like 3pm..... when my brother wakes up :angry2: Urgh! Still. I hope to have the preamp board done today. I can't finish the delayed turnon thing as Farnell didn't have the relay in stock, so that's back ordered or whatever. I'll still be able to test it without that though! The many joys of modular...

 

My monitor is also playing up, but hopefully it'll continue to just about work (backlight is giving up I think) for a few more days / until I can get a replacement

 

Brother is awake... might be time to get to work

Edited January 14, 2007 by markiemrboo

[markiemrboo]

The preamp PCB has been drilled and all components mounted (bar the opamps, which are socketed). Also drilled the timer PCB and mounted all but the relay, as I don't have that yet. I also made a couple of bridge rectifiers on prototype board, using 1N4007 diodes, should be more than enough... they're rated 1A. I'm going to use the 35A beast for the disconnecting network, as this could potentially carry larger currents (i.e a mains fault?).

 

I have tested the timer with a 9v battery, and it seems to work OK. There's a delay of about a second (set via a pot). I think the pot may be slightly too large in value though, as it's incredibly sensitive and to get < 5 seconds it has to be twisted around nearly all the way to the right. Still, it works

 

I have also tested the regulator with a 9v battery, and this seems to work.

 

Unfortunately, I do seem to have run in to a stupidly strange and mysterious problem. The transformer seems to be OK (I get a sine wave from the secondaries on my 'scope). The bridge rectifier seems to be OK, I think (I get a rectified half sine wave). However, when I hook up the output of the bridge to the filter cap PCB I get nothing! No expected DC output from the PCB at all Fuses aren't blowing, nothing's getting hot, caps aren't exploding etc. The caps seem to hold a charge, and a connecting a battery to the cap PCB I get DC output.... aaaaand the caps then slowly discharge when I take the battery away. I've checked the polarity of wiring over and over and over, but it's all correct. I have absolutely no idea what it could be! It's left me really quite stumped.

 

EDIT: What a donkey. Forgot to attach pictures!

 

Preamp:

 

Timer:

 

 

The resistor spacing did pose as a little bit of a problem for the preamp. My own fault really for buying 0.75W resistors without thinking about the size I think 0.4/0.5W metal films would have fit just fine. Still, I made them fit just about

 

 

 

 

 

EDIT: Figured out the power problem I am using 'faston' / spade connectors for pretty much all the wiring this time around, and basically... it was just loose connections on the transformer secondaries. You see, I don't have a crimp tool.... so.... I was using .... a hammer

 

Maplin have a crimp tool. Not sure if it'll work well on the female crimp connectors I have but I think I am gonna give it a go ( or one from Rapid if there's one cheap enough... as I need to order stuff )

Edited January 15, 2007 by markiemrboo

[markiemrboo]

A small update. Fun filled stories from the land of mrboo

 

 

First of all, I will have a replacement monitor tomorrow I actually managed to get through to LG this time (I tried about 5 months ago and no one could get through apparently). Pretty easy process, and they're gonna come to my house with a replacement and swap it there and then. Pretty cool

 

As for the preamp, well, as mentioned I got DC output eventually. I've redone the wires using my trusty hammer, with longer exposed wire at the end and a lot of force when hammering. All making good contact now. Output from the caps is ~23v DC. AC to DC from a transformer is apparently DC = AC * 1.414, so with a 15V transformer this would mean 15 * 1.414 = ~21v. Close enough

 

The regulator works fine, I'm getting ~+-13v from that. Dropping about 10v, so according to the LM338 spec sheet I think I should be seeing a maximum continuous current limit of 8A. Power supply can't even deliver that, so I am in the clear there really. I assume it's continuous anyway, as below that figure there is a peak figure which is 12A. Slightly off the 15v I wanted, but nothing to worry about.

 

I connected the actual main preamp board today too, which is where all the fun lies Popped in an NE5532 on the first stage (buffer), turned it on, thought I heard a sizzle. Turned it off instantly. While shuffling about checking things I heard a similar noise which was just the pins scratching on the surface. So I turned it on again and got no sizzle, I figured it really was just the scratchy underside-of-the-board pins. Checked the output pins and saw 12-13v (DC offset, I should be seeing mV's here). Couldn't see anything wrong anywhere, so I took out that opamp (socketed) and popped it in to the second stage (gain) instead. At the moment, signal wise, it's not connected to the first stage. I still should have been seeing mV's on the output, but still saw 12-13v! Stumped, I stuck my other 5532 in to the first stage (a stupidly stupid idea given the DC offset, but shh).

 

Wanna guess what happened next?

 

 

 

 

 

 

 

**BANG**

 

 

 

 

 

 

 

The other opamp that I just put in exploded! It made a sound like a firework. The noise from such a small thing was absolutely insane! Half of the plastic casing flew off Scared the **** out of me quite frankly. So my next question was, of course, why the bloody hell did that happen.

 

A quick glimpse at the datasheet, comparing with the layout of my PCB explained it all very quickly. It was a simple case of....... accidently connecting the power supply the wrong way around. I had connected the -ve rail (-13v) to the +ve pin of the opamp and the +ve rail (+13v) to the GND / -ve pin of the opamp. Did I blow up the original opamp? Yes, I actually DID hear a sizzle, and it wasn't the pins scratching like I thought. It just seemed to fail as a short rather than explode.

 

At this point I had one more cheapo opamp laying about, an LM358N. So, I tried this in the first stage, this time with the power supply plugged in the correct way, and I saw mV of DC offset at the output pins. That's more like it! In the second stage it reads 0.1mV offset. My 'scope seems to show no AC noise at all on the output of the first stage, with the input shorted,, even with a X1 probe. Sort of unbelievable really. I shall, of course, be re-measuring when it's all properly done. Possible some time later if I dare to plug in the more expensive 2134's....

 

 

 

Fuzzy, but you should still get the idea:

 

 

 

 

EDIT: Another small update (which doesn't deserve a new post). Found the other LM358N and popped that in to the second stage for testing, soldered the volume pot wires to the board (and pot of course). Volume control seems to work, in that the DC offset from the first stage doesn't get through to the second stage when the pot is all the way to the left, and when all the way to the right I see DC offset at the second stage The DC offset from the second stage is nothing to worry about by the way, as there are output caps which remove this.

 

All that should be left is sorting out grounding, soldering RCA connectors, getting the case and chopping holes, and mounting all the stuff inside the case At the moment I don't have a "source selector", so it's just.... "in" (from computer to pre) and "out" (from pre to power), but this is easy to add afterwards should I get a source other than my PC

 

If I manage to get the RCA sorted out I might be brave and hook it up to my power amp later on today! At which point I will attempt to get pictures of it all connected and running on my bed

Edited January 16, 2007 by markiemrboo

[hardnrg]

wrong way norris!

 

looking forward to the upcoming pictures of it all connected

 

Heinz Baked Beans 840g

 

[markiemrboo]

wrong way norris!

 

looking forward to the upcoming pictures of it all connected

 

Heinz Baked Beans 840g

 

 

840g of baked beans. Who knew!

 

 

I done one set of RCA's, so I have tested it today with various signals. Firstly with the LM358's, because, well, they're crap for audio and I don't care if they explode too I used the RCA's for input and scoped the output of both channels. They both seemed to be identical in every way, which is nice

 

Simple 60Hz on both channels (it works!!!)

 

60Hz one channel and 120Hz the other (it still works!!!)

 

I then tried 15KHz, which came as quite a shock... I believe what you see here is "crossover distortion". I'm tempted to attribute this to the quality of the opamp, but I don't know enough to actually know for sure whether it is or not It should be a smooth wave anyway

 

Here's the same 15KHz signal...... but after I swapped in the OPA2134's. Much better now!

 

60Hz and 120Hz, but with the OPA2134's and a slightly different horizontal scale from previous. Whoops.

 

Finally, here's a fuzzy picture of the amazing mess that is... preamp... If I could, I would have made it one huge board. But I can't, so it's wire heaven

 

 

I have ordered a case, which might hopefully be here tomorrow. I ended up having to go with a 2U again though, mainly just because of the height of the caps. Ah well! Might have it finished tomorrow or I might be lazy and have it finished the day after (most likely to be honest!)

[jammin]

Looking good.

I'm guessing this has been a pretty good learning experience.

 

It's always nice blowing things up as well (as long as they don't cost much).

 

On a slight side note but somewhat related. The integrated amp I'm using while at uni has a bit of a buzz to it.

Anything you can do to reduce that, which you know of? I did some quick searches but didn't come up with much.

Edited January 16, 2007 by jammin

[markiemrboo]

Looking good.

I'm guessing this has been a pretty good learning experience.

 

Rather

 

It's always nice blowing things up as well (as long as they don't cost much).

The NE5532's are only like... ~30p or something!

 

On a slight side note but somewhat related. The integrated amp I'm using while at uni has a bit of a buzz to it.

Anything you can do to reduce that, which you know of? I did some quick searches but didn't come up with much.

 

A buzzing... sort of clicky / scratchy or more of just a 'smooth' low frequency hum? Either could be a "ground loop" AFAIK (I have had both from not 'properly' grounding in my power amp). Without messing about inside (where you'll probably have to cut traces and stuff if it's a commercial amp), I believe you can buy things called "ground loop isolators", not that I have used one myself.

 

You might also check that the input sockets (assuming RCA), that the outer barrel metal (ground contact of the RCA plug) isn't touching the metal casing (which should be attached to safety earth, and if it is I think this would create a loop).

 

It might even be inteferance from the transformer. Is it a toroid (round donut type thing)? or like a big square lump? or...? The big square ones do tend to radiate quite a lot apparently, where as toroids it's largely concentrated down the middle hole part and doesn't spread out around it so much.

 

Apparently dimmer switches also sometimes wreak havoc by creating such noise, though I don't know the details as to how there

 

Are you using unshielded input cables? If they're running close to mains cable with unshielded signal cables, I discovered this can also create noise

Edited January 16, 2007 by markiemrboo