My DIY Amp project

[Hdale85]

I know this isn't the normal DIY sort of thing on these boards but I've been a member here for a very long time and thought I would share.

I just picked up some PCB's for a Pass Labs Aleph amp clone. They can do any of the Aleph designs up to the Aleph 5. I chose to do the Aleph 5 design which will give me 60 watts at 8ohms full Class A power.

Here are the PCB's



Ok I also ordered the heatsinks. I'm using these heatsinks off ebay going to get 6 of these that are 6" long and turn them verticle 3 per side for the sides of the chassis. I'm building the chassis my self as well. This is what they look like.

Image not available

This is what I've ordered so far there will be more to come for sure and I will keep this thread fully updated with everything as I'm building. Next paycheck I'll be ordering components for the PCB's.

[Gir]

How will you be building the enclosure? Sheet metal, plastic, wood?

Good luck on the project!

[Hdale85]

Aluminum, I'm ordering pieces from a company that will cut it to order. And the front and rear panels I will get form another company that actually has a panel designer. So I should be able to build it with very few tools.

[JonMarsh]

Sounds great, Dougie! I used to design audio electronics for a living (pro sound), and I've built my own share of power amps from scratch, it will be fun seeing your Pass project. Unfortunately there's not as wide a group interested in that- a little more challenging on the electrical side. I've been working on a new channel board design for my Aragon's, but a distinct lack of free time relative to the work involved has put that project to a stand still the last couple of years. I need to get my current speaker projects done, then I'll probably return to electronics myself.

Do you have access to a scope for bench testing before you hook it up to live hardware?

[TacoD]

I have build a Pass Labs Aleph-X clone, very nice sounding. In the near future I try to build the Krell KSA50 kit I bought from Jed . Jon, if you finish your design some day I am interested. I only have the skill to build the electronics, I can only design loudspeakers .

[littlesaint]

If your going to use aluminum, I recommend having it anodized. It makes for a great finish, and you can usually pick any color. You can have it laser etched for a great professional look.

[littlesaint]

Sounds great, Dougie! I used to design audio electronics for a living (pro sound)...

How long ago?

I have a friend that used to work for Soldano. He builds his own amps now. He's had a few auditioned professionally. He helped me build my SET monoblocks.

[JonMarsh]

Late 70's. Did some consulting in the 90's, too. My personal interests are more for non loop feedback stuff, like Ayre Acoustics. Strongly believe in the benefits of very stout power supplies, so stuff I do tends not to be so "commercially" feasible (not for "making money". ) Biggest personal complaint against many commercial designs is not enough storage capacitance and not enough wideband bypass. Shortly after the Dyna 400 came out I did a version with 100K uF supply, plus 100 uF of PPF bypass caps, and doubling up the number of outputs (the Dyna was designed by James Bongiorno, and used a stacked set of lower voltage output transistors - like Ampzilla). That amp was effortless at levels that had a Luxman 6000 or Stax DA-300 sounding rather strained - mostly because the transistors were working from a very low impedance wideband power supply, and the feedback loop wasn't trying to process/reject PSSR issues. In retrospect I wondered if Dyna was looking over my shoulder, when a few years later they came out with the ST-416 with auxilary capacitor bank. The only guy I knew of at the time doing similar things was Frank Van Alstine, same guy who's still around running Val Alstine Sound. I did get one of his preamps later on, a good unit, but it was stolen in an apartment break-in.

ThomasW has heard Frank's current stuff at the RMAF, says it's quite good, but I haven't had a chance to hear it- I'm looking forward to this October and attending RMAF.

~Jon

[Johnloudb]

Amp

Dougie,
Looks great. I'm looking to build an amp eventually. I'm working on a modified Bride of Son of Zen preamp right now.
Where do get these clones?

[Johnloudb]

DreadNaught

Late 70's. Did some consulting in the 90's, too. My personal interests are more for non loop feedback stuff, like Ayre Acoustics. Strongly believe in the benefits of very stout power supplies, so stuff I do tends not to be so "commercially" feasible (not for "making money". ) Biggest personal complaint against many commercial designs is not enough storage capacitance and not enough wideband bypass. Shortly after the Dyna 400 came out I did a version with 100K uF supply, plus 100 uF of PPF bypass caps, and doubling up the number of outputs (the Dyna was designed by James Bongiorno, and used a stacked set of lower voltage output transistors - like Ampzilla). That amp was effortless at levels that had a Luxman 6000 or Stax DA-300 sounding rather strained - mostly because the transistors were working from a very low impedance wideband power supply, and the feedback loop wasn't trying to process/reject PSSR issues. In retrospect I wondered if Dyna was looking over my shoulder, when a few years later they came out with the ST-416 with auxilary capacitor bank. The only guy I knew of at the time doing similar things was Frank Van Alstine, same guy who's still around running Val Alstine Sound. I did get one of his preamps later on, a good unit, but it was stolen in an apartment break-in.

ThomasW has heard Frank's current stuff at the RMAF, says it's quite good, but I haven't had a chance to hear it- I'm looking forward to this October and attending RMAF.

~Jon

Very Interesting!
Electronics are more my area of experience than speaker building. I agree, about filter capacitance. 12 years ago, I added larger electrolytic caps and bypass caps to my Dad's Dunlap-Clark DreadNaught 500 (1974). Wow, Its great bass was even better, along with very musical, detailed high frequencies. I'm going to resurrect it with a new circuit, some day.

Image not available

[Jed]

I'd like to have a pair of Aleph 2s someday. One time I heard the Pass stuff and some Ariel 10Ts. It was quite memorable, even if it was a long time ago. How much do you save DIY versus picking up a pair of used Pass amps?

[Jed]

Next paycheck I'll be ordering components for the PCB's.

If you don't mind, could you put links to where you got the stuff too? Just want to price stuff out etc.

[JonMarsh]

I'd like to have a pair of Aleph 2s someday. One time I heard the Pass stuff and some Ariel 10Ts. It was quite memorable, even if it was a long time ago. How much do you save DIY versus picking up a pair of used Pass amps?

"Saving" and "DIY" rarely mix, in my opinion. UNLESS you have access to test equipment and tools you don't have to mix in to the cost of the project. So, if you have the low distortion oscillators, distortion measuring euqipment, scopes and loads sitting around that I already have, DIY "might" save you some money.

Then, there's the matter of sweat equity.

So, there's a reason I have some good used amps sitting around. If you're looking at Pass gear, I'd go out and listen closely to an X series amp and see what you can dig up at Audiogon.

If you look around, you'll probably notice it's pretty tough to come across a used V5 or V6 Ayre on Audiogon. There's a reason for that, and it's not because lots of them haven't been sold new.

I would love to have a couple of pairs of Ayre MX-Rs, but no can afford. In comparison, a complete channel rebuild of an Aragon, while not providing the classy packaging, may meet my needs with a lot less investemnt. Fortunately I know from discussions with Chas about the MX-9 circuitry at CES 2006 that we were on similar tracks. He's done a lot more work, since, before the producct final release. There's an Ayre Dealer in Berkely I may have to visit someday soon, which has also just opened a showroom in SF. They also carry Avalon. This is their "blurb" on the Ayre MX-R.

Simplicity, confidence, and understated elegance. Ayre has surpassed even their own set of exacting standards by creating one of the most unique and innovative amplifiers ever to have graced the hallowed halls of high-end audio. The MX-R monoblocks serve as the new flagship pieces of the Ayre armada, providing the enthusiast with one of the finest high-powered solid-state mono-blocks of our time. Packing 300 watts into 8 ohms on a fully balanced circuit (that’s right, this means no RCA-ins on the back), smooth, effortless and majestic couldn’t be used to describe anything else, not any more. Fine machining on the units would lead one to believe they were not built by human hands (though they truly are handmade) but instead crafted by advanced beings from a place far more enlightened than our planet. You’ll have a gentle landing when you realize these amps were not in fact derived from some form of alien technology. And that they even have a price at all might also surprise you. In the realm of amplifiers occupying the 15-20,000 dollar price-point, the Ayre MX-Rs provide a ridiculously stunning presentation that won't be easily outshined in years to come.

That's the price for a pair of monoblock MX-R.

Of course, as dear has time has become, maybe I'm not looking at the cost/time tradeoffs correctly, either.

I just really like messing around with this stuff myself, though. At times the elctronics have been a much bigger love, as in the latter 70's and early 90's, and in the mid to late 90's, when I was doing some outside stuff.

All I've done a few years ago is work in simuluation with SIMetrix, and refine device models, and do once test PCB layout.

I've made a vow to return to this once the M12ta and Isiris are done.

[JonMarsh]

DreadNaught

Very Interesting!
Electronics are more my area of experience than speaker building. I agree, about filter capacitance. 12 years ago, I added larger electrolytic caps and bypass caps to my Dad's Dunlap-Clark DreadNaught 500 (1974). Wow, Its great bass was even better, along with very musical, detailed high frequencies. I'm going to resurrect it with a new circuit, some day.

Image not available

I remember Dunlap-Clarke! Talk about a blast from the past! :B

~Jon

[Jed]

Good points Jon.

Maybe I'll look into upgrading my Forte Model 4. Already increased capacitance and replaced the rectifiers. Nelson Pass left before the Forte Model 4 was produced, apparently the Model 4 uses IGBT output devices, which Nelson does not think is appropriate for audio. Apparently they aren't very stable, although I haven't had any problems. Are there better alternatives that are plug and play for these output devices?

I'd like to learn more about electronics as well, but this is definitely not in my background-- which is art by the way. Looks like we've got a lot of artists on this board-- CJD, Thomas etc.

[audioslave]

Pieces and Parts

I am about to start on a Aleph 2. I've been looking into these for a while. Here are a few links for info and parts:

trannys:

http://www.victoriamagnetics.com/

http://www.plitron.com/

http://www.avellindberg.com/

PCBs:

http://www.kk-pcb.com/

Heatsinks:

http://www.mmmetals.com/index.html

http://www.r-theta.com/

http://www.conradheatsinks.com/welcome.htm

Caps, fets...etc

Digikey, mouser for new
Trading Post on DIYAUDIO for matched parts
APEX Jr for used
http://www.ampslab.com/trans_irfp9240.htm output devices

Should be easy right?

mike

[relder]

Should be easy right?

mike

Getting all the parts is the hardest part.

[Hdale85]

I got the pcb's from chipamp.com they were designed by BrianGT.

http://www.diyamps.com/aleph/

Components will be coming from Mouser and Digikey most likely. Maybe a few from Partsconnexion. Still waiting for a BOM to start from and after I build my own BOM I will post here. Nice to see so many interested in this though!


Oh and no I don't have a scope. Maybe I should look for a used one on ebay or something? How much would a decent used scope go for Jon? I shouldn't really need one if I double check everything and test the parts before I put them in.

As for cost I'm guesstimating that this project will cost me something like 450-650. Shouldn't be more then that anyways. PSU parts are about 100 bucks, the parts for the boards about 150ish maybe 200 at most so thats 300. The chassis is going to be about 100-150 bucks depending. Then just a tranny and some connectors and what not. This is for a 2 channel stereo amp. I plan on building 2 more of something for my theater/audio setup. I might do the Aleph-X for the other channels and use the Aleph 5 for the far rears.

[Jed]

Those are really nice looking boards, and if you can build this amp for less than $650, thats a pretty good incentive to DIY versus buying a used one for $1300-1500 IMO. Might take a year of work though.

[Johnloudb]

Good points Jon.

Maybe I'll look into upgrading my Forte Model 4. Already increased capacitance and replaced the rectifiers. Nelson Pass left before the Forte Model 4 was produced, apparently the Model 4 uses IGBT output devices, which Nelson does not think is appropriate for audio. Apparently they aren't very stable, although I haven't had any problems. Are there better alternatives that are plug and play for these output devices?

I'd like to learn more about electronics as well, but this is definitely not in my background-- which is art by the way. Looks like we've got a lot of artists on this board-- CJD, Thomas etc.

You could also put a small resistance 0.16 - 0.22 ohms (10+ watts) after the rectifiers - before the filter caps (both pos. & neg. rails). To filter more noise. But, without knowing your amplifiers schematic. Won't hurt anything, can just see how it sounds. Right Jon? RF/EMI is bad!!

I wouldn't change transistors, affects the bias current. IGBTs are probably well suited to your amp.

Those Alpha PCBs do look nice!

[Hdale85]

I don't think this is going to take me a year. I'm hoping 3 months or so at most.

[JonMarsh]

EMI/RFI is bad. Fast rectifiers are good, snubber networks on fast rectifiers even better. Even a small common mode ring filter isn't a bad idea (high current, not that high an inductance.

Biggest issue in building a known good design is being sure to follow the ground wiring recommendations so you don't have any power supply ripple current being induced in the signal path.

It all depends on how much free time is available- the metal work takes time, stuffing and soldering boards isn't to bad, particularly not for the Aleph designs which don't have many components.

I NEVER hook up a power amp untested to other equipment - a miswired connection can cause a full PS DC offset to appear on the speaker or even at the input.

Of course, I have my own load setups, a dedicated high power test supply with variac for bringing up channel modules (can probably find a picture of that), a Crown IMA (Intermodulation Distortion analyzer (that one show's my age), a Kronhite low distortion oscillator, and an HP 8903B Distortion analyzer. My scope I'm not super happy about because my older Tek rack mainframe finally died on me, and I picked up a good deal on a used HP 500 MHz digital, but I don't really trust it that much yet.

[cjd]

I have an amp project I've been working on for almost two years :P

Long as you're in it for the "fun"...

C

[Hdale85]

I'm in it for the fun. I won't be doing much metal work my self other then assembly. The top and bottom and front and rear panels will all be precut. The heatsinks are the sides and I'll jut put them together. I'll be drawing up some pics sometime here. I have to install Sketchup on this pc.

[Jed]

How many of those heatsinks from ebay do you need? $20 per foot seems like a good deal.

[Hdale85]

He cuts them for free and i'm going to use 3 6" pieces per side turned vertical so the fins will go vertical and they will be side by side so the depth will be about 16 and 1/8". I'm going to attach them with a copper rod that will span them and the output fets will be mounted to the copper this is to help spread the heat evenly over the sinks.

So 3 of the 12" sinks basically should be 70 bucks with shipping.

And by the way the whole chassis will be annodized black. But thats further down the road .

[Hdale85]

Jon you think something like this would be good?

http://cgi.ebay.com/TEKTRONIX-TAS250-dual-Channel-Oscilloscope-DC-50-MHz_W0QQitemZ300133897953QQihZ020QQcategoryZ104247QQssPageNameZWDVWQQrdZ1QQcmdZViewItem

Whats a decent used scope I can get for 200 or less? 150 would be better.

[ahaik]

Looks great Dougie,
I may be following in your footsteps soon, I too had dropped the F4 option because I don't have a sutible preamp/speakers combo for it.
At my job we have a full blown high tech machine shop and a friend that works there told me he can felp me with the parts and build for the chasis.
I will be following your thread closely.

[Johnloudb]

Scope

Jon you think something like this would be good?

http://cgi.ebay.com/TEKTRONIX-TAS250-dual-Channel-Oscilloscope-DC-50-MHz_W0QQitemZ300133897953QQihZ020QQcategoryZ104247QQssPageNameZWDVWQQrdZ1QQcmdZViewItem

Whats a decent used scope I can get for 200 or less? 150 would be better.

I wouldn't buy an Oscilloscope with less than 100Mhz bandwidth. If you just want to test your gear to see if it performs ok, True RTA (free version - level 1) has a built in O-scope with a pathetic 48kHz bandwidth
(http://www.trueaudio.com/). But it's good enough to check your amp with a 1kHz - 20kHz sine-wave. It has a built in wave function generator too!

O-scopes are nice, but ... if your going to buy, you should look for +100Mhz bandwidth, dual channel. I have an HP and have used Tektronix. There's many good ones.

[JonMarsh]

I'd usually recommend a bandwidth of 100 MHz or more, especially with MOSFET amps, because they can oscillate at pretty high frequencies - 100 MHz and above.

Also, try to stay with analog- low bandwidth digitals have real problems with seeing a waveform accurate- due to aliasing artifacts. In fact, the only digitals I really recommend are ones with what they call digital phosphor which samples and displays high frequency envelopes when they might otherwise be aliased out due to sampling artifacts.

My last analog scope cost me $2,000 - in 1978. Look for used gear on places like Test Equity- you'll pay a bit more, but the stuff will be in better shape than what you might often find on GeeBay. (as in, "Gee I wish I hadn't bought that...")

I realize I usually go over the top, and buy stuff for the long haul. If you have any serious interest in doing this, more than once, look for a new or used Agilent DSO3062 or similar series- bandwith models from 60 MHz to 300 MHz, starting around a grand new. 1 GSA digitizing rate.



For something more budget oriented working with a PC directly, try PicoScope.

PicoScope







But there good ones cost more than the Agilent.

[MarcE]

It has a built in wave function generator too!

hm, just don't try the square wave feature ... unless you have always wondered what ringing looks like 8)

[chasw98]

Tektronix 2235 and 465 scopes can be had for relatively cheap dollars. They were workhorses in their day. Look for them in good condition as Jon says. No broken knobs, intermittent trace, etc. If possible, find out if they have been calibrated recently.

Chuck

[Johnloudb]

hm, just don't try the square wave feature ... unless you have always wondered what ringing looks like 8)

I'll bet it can handle a 100Hz square wave ... And it's free.

[Hdale85]

Yeah I'm sure I would use it a lot in the long run (for starters I'll be starting school before too long majoring in EE) But I really don't have the money right now to drop a grand on a scope.

[Hdale85]

This TrueRTA program how good is it? I could definitely spend the 100 bucks easy on the full blown version. Do I need a good sound card for it? Does it like use the inputs and outputs? Could probably pick up an M-Audio sound card if need be as I was planning on upgrading anyways.

[Hdale85]

http://cgi.ebay.com/Tektronix-2235-100MHz-Oscilloscope-NEAR-MINT_W0QQitemZ290143045217QQihZ019QQcategoryZ104247QQssPageNameZWDVWQQrdZ1QQcmdZViewItem

Might be able to swing this. Although shipping for 50 bucks and they are in the same state as me seems very steep.

[Hdale85]

Was playing in SketchUp a bit and so far this is what I got.



I need to find a black annodized brushed aluminum paint stamp. But you can see how I plan on doing the heatsinks and what not in 3 pieces and everything. The fin count is not right and they are too big but hey best I could get. Is there a way to change the units in Sketchup? Or something make it so you can use smaller snap spaces.

[JonMarsh]

Haven't used sketch up- there should be a snap setting, but how low it goes may be a limitation, if it uses real world coordinates when doing things like building, depending on how the internal coordinate system was programmed. That's why God invented floating point...

Looks good, though!

~jon

[cjd]

That's why God invented floating point...

I used to get in trouble back in grade-school for ignoring the decimal and then dropping it in place after I had done the arithmetic. Funny, now we have a fancy name for it (which probably existed already and wasn't in the vocabulary of my teachers...) But then, I never did math "properly" - I always had to work backward from the answer.

C

[Hdale85]

I've got to draw some more details in there. Like I want to put a piece of copper that is about 1" wide that will go the length of the heatsinks and the fets will be mounted to those copper strips. This should help spread the heat evenly over the heatsinks. I'll have to draw that in. Maybe add the connectors on the back and draw in the pcb's and stuff (although with very little detail) to figure out how I'm going to lay everything out. To get my sinks with that many fins I had to draw them in feet instead of inches. So yeah have to look for the snap setting right now its very large.

[cjd]

So, unless you can machine the surface of the copper and the heat-sinks to match perfectly, I would recommend against the copper bar idea. Copper is better, yes - but aluminum is still very good. In order for copper to be an advantage you need to be sure the copper/aluminum connection is perfect, or you may end up being able to dissipate less heat than mounting straight to the aluminum.

C

[chasw98]

http://cgi.ebay.com/Tektronix-2235-1...QQcmdZViewItem

Might be able to swing this. Although shipping for 50 bucks and they are in the same state as me seems very steep.

Dougie:
The shipping is usually high because these are very large units and weigh quite a bit. The one in that ad doesn't look too healthy. Look at the screen around the grid lines.

If you don't need the scope right now, just sit back and watch that auction place until the right one comes along. You should be able to snag a decent working 2235 for ~$200 shipped.

Look at this one. http://cgi.ebay.com/Tektronix-TEK-22...QQcmdZViewItem

Chuck

[Hdale85]

That ones missing all kinds of knobs and stuff.

[Dennis H]

Yeah I'm sure I would use it a lot in the long run (for starters I'll be starting school before too long majoring in EE) But I really don't have the money right now to drop a grand on a scope.

Heck, save your money then. The school should have all kinds of scopes and other test gear you can use.

[Hdale85]

Yeah but won't be in those classes for a while I think that TrueRTA program should be enough for my little amp projects for now.