Silicon Chip ULD Mk3 amp - 0.0006% distortion

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
It had several non solved issues, dont know to what extent these
were solved in the current version..


There was an errata included with MK2 kits sold through local retailers Altronics and Jaycar after the fault was identified. 22K is actually a bit overkill. I’ve had no problems with 10K (bypassed with 10n) and I also substitute the bc639 with a 2SA2240 (lower cob and higher beta works best there). I did find that a small amount of VAS degeneration (22 ohms) was needed for unconditional stability.

There have been some smoked modules but mostly because people will fiddle with them 'live' and do not realise that the suppied generic 5A rail fuses will blow well after the output devices have internally melted down - like a Fukashima reactor. :D
 
Is this amp 6 ppm THD20 ? At what output ? 6ppm is not hard to achieve these days. But can it "hold" that @ 100v p-p 20Khz ? 4R , anyone ? . If it's not a triple , highly doubtful ....
It does look like a nice build. Sporting those T-traks and using a real Vbe + 2 of the internal diodes must make for near perfect thermal response. :) Still , it is what it is. :D
PS - I can now beat TMC ... without TMC ( 2 designs).
OS

Hello Outstripper

The real amp does 0.004% THD @20Khz at 100watts into 8R less at 4R.

this is achieved with double EF ouput and not a tripple. Please note the ouput consists of a total of 4 output BJT devices.

Regards
Arthur
 
I bought the mag yesterday & I can't help but wonder, why have they specified a 47uF bipolar input capacitor in the hp filter along with a 12k resistor? This would give a -3dB pt of .282Hz.... Surely this is a mistake and that they had meant this to be a 4.7uF cap giving a -3dB pt of 2.82Hz
 
"super sizing" electrolytic DC blocking caps has some justification - read
Self

basically if there's less V drop due to a bigger Cap's low Z then there's less "room" for distortion - Self shows AP distortion measurements in later editions of his amplifier book
 
why have they specified a 47uF bipolar input capacitor in the hp filter along with a 12k resistor? This would give a -3dB pt of .282Hz.... Surely this is a mistake and that they had meant this to be a 4.7uF cap giving a -3dB pt of 2.82Hz

"super sizing" electrolytic DC blocking caps has some justification - read
Self

basically if there's less V drop due to a bigger Cap's low Z then there's less "room" for distortion
a sensible 100ms filter can be achieved with low Rin and electrolytic or with medium Rin and large film or with high Rin and smaller film.
I favour film and adopt generally between 1uF and 4u7F allowing a range of Rin from a low of 20k to a high of 100k.

There seems to be no reason for adopting a large (or very large) electrolytic.
Well, there is a reason - electro is cheap !

At the NFB loop the DC blocker has to be electro to avoid the compromises that JLH adopted to allow a pair of film caps to be used.
If electro is adopted for the NFB DC blocker then it must be oversized such that the input filter determines the low frequency roll off.
If the NFB DC blocker is used to roll off the bass response then that guarantees that some AC voltage will be developed across the DC blocker.
 
Last edited:
July 2011 SC mag.

Hello from Tassie.:)


I just wanted to share some info with those of us building this amp. It's all good!


In a way I am giving Alltronics a bit of a plug because I am very happy with their service and the quality of their kits.


If you need a good transformer for the amp - albeit a little small (IMO), Alltronics have the 300VA + 7.5VA toroid on special for $60


Also; - Last week they had the ULD MK2 135Watter on special for $59 ea. I have performed the MK3 upgrade on that one and this has saved me $30 per board.


This is not a paid commercial, like those silly ads that come onto your tv at 3am in the morning, but a heartfelt plea for Aussies to support the locals. We need these people to stock the goodies or we will have to buy all our components from China. This is not a bad thing I reckon, but it takes ages for your gear to arrive.


If you are considering building the ULD in any form, and you live in Oz then keep an eye on the Alltronics site because they are a teriffic mob and I can assure you that their boards are of superior design and quality to the Chinese ones that I have brought time and again, only to find that I have been left up the creek without a paddle, when I find NO INSTRUCTIONS or even a SCHEMATIC!

Quote Me:- If Forestry Tasmania and the Tas Govt continue to trample swathes of scrub here, we will have no timber to build speakers and the Army will have to change it's uniform 'cause their camoflage will contrast with the naked red dirt of the wastelands.


Cheers - Phil
 
Thanks jcx. I might just try this with my next design and measure the difference between large and small input caps. I've recently recieved his latest edition of the power amplifier design handbook and I'll make sure to look this up

Hi there Ollie' ;) My name is Phil and I just wanted to ask if that book is a good reference? I have another book; "Self on Audio Second Edition" it is a teriffic source of info. Sometimes I feel that you have to read this info two or three times before you actually adopt some of his methods and incorporate them into your designs.

His name keeps popping up on these sites because I think he explains things pretty well. When I look at Silicon Chip's ULD 135 Watter, in schematic form I see the speaker protection module is a reasonable copy of self's DC protection circuit, but Silicon chip has added and tweeked it to suit our requirements, so I'm not complaining.

I would have built a Blameless from Self's site (the site referenced in the aforementioned book) knowing that It would be spot on and just what the doctor ordered, but I'm afraid they are a little out of my price range.

Cheers
 
Hello from Tassie.:)

In a way I am giving Alltronics a bit of a plug because I am very happy with their service and the quality of their kits.


Hi Phil,

Just checked the Altronics website but cant see the MK3 on there??
Can you provide a link.. I know it's in the product catlogue, but when I was in there last not even the sales staff knew about the uld-mk3's availability..

Ta
CM
 
Amp Modules

Hi Murph' its nice to meet you. I purchased the ULD Mk2 kit #K5151, this is essentially the Mk3 minus the VBE multiplier.

I didn't want to pay $100 each for the MK3 so bought the MK2 and upgraded it to MK3 by changing the NFB cap to 1000uF, winding 5 more turns cw on the output inductor and adding the VBE Multiplier onto a little board that mounts onto the heatsink, that board is only 20mm x 40mm and a small piece of veroboard will do the trick. There are a few other changes also, but you end up with the MK3 - the extra components are probably in your junk box.

If you would like to read about the upgrade, go to the library and get a copy of Silicon Chip from September last year and they explain the few small changes that have to be made.

I have just taken a look at the Alltronics site and the boards have just gone back up in price to $80 so I guess this was all an exercise in vanity! None the less, the trannies are good value. Oh s*^t, they removed that one too.

I am very sorry Murph' but they were on special and very good value.

Nothing good lasts forever!!!

Cheers mate.
 
I always get local aussie made custom trannies for any amp projects anyway..

One day I'd like to be wealthy enough! I'd get them to make it with their special wire.

One day I hope they find a better mounting system for them toroids because the bolt through the middle is not a real good idea (IMO - for what its worth), because the mounting bolt seems to compress the encu to a certain degree if you want the thing to stay put.

perhaps next time I will place a short length of nylon dowell in the centre of the toroid - say 32mm diameter and say 50mm long, then drill the dowell as well as the chassis. This method would give just enough grip and the dowell will stop the compression and any lateral movement within the chassis. It probably isn't an issue, but I often think about all that juice that's running around in there and dont want to experience any more sparks in my life.

On that topic it was interesting in the Silicon Chip article when they said to be careful not to touch the fuse-holders on the 56V rails as you might receive a lethal shock! I must just be lucky because my last amp had the same configuration and I was poking around in there with both hands! - some blokes are just lucky I guess.

Cheers matey, and give us a hoy when you get the big beastie up and going. I know they are a good design and the previous home made one sounded absolutely perfect for my taste, however my board layout wasnt right or the fact that I had 2x200VA lumps inside the enclosure caused a terrible hum that just annoyed me even though it was only audible inbetween tracks. The bass response is awesome and brought out qualities in my speakers that previously were inaudible.

That amp had a loudness & balance preamp as per the NE5532 application note 142 but the preamp only had to act as a buffer for the signal and the loudness concept was lost on an amp such as this. Loudness is not required (I feel) because the damn thing sounds good at low volume too.

Adios amigo
 
use VA rating from one to two times the total maximum power output of the amplifier/s.

I find that 1.5Times well.

if your ULD is predicted to give out 100W into 8r0 then a pair of them can be powered by a 200VA to 400VA transformer.

If you decide you want to use 4ohms speakers then the predicted power jumps up to just less than 200W into 4r0 and the transformer requirement changes to 400VA to 800VA for a stereo pair of 200W+200W into 4r0+4r0.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.