Hello all, hows about helping a newbie

[justblair]

Hi all.

I have a set of Mission 753 (non-freedoms) and currently use a Musical Fidelity E100 amplifier with them. I like the sound, especially after doing the ever so simple tweeter mod on them, but the gainclone interests me. I just like to get my hands dirty. Some questions.

The setup I envisage (ultimately) will be 4 monoblocks driving the missions, each bi-amped. I'm thinking probably that I will go down the SMPS route, rather than building my own supply. (Decible Dungeon has been a fantastic resource so far) I may wish to go further in the futiure and bypass the speakes passive filters. But one thing at a time.

Now I am more or less ready to get going. I need some advice on the smps arrangement.

Decible Dungeon suggests using skynet 8080's, and I can get hold of them at reasonable prices. I was considering using 4. Either to produce +/-24v. or to produce 2x+/-12v. Im thinking also about using a linear regulator for each amp

As the missions are rated at 6ohm I'm fairly sure that the LM3886 chip is the one for me. The LM3875 seems from what read to prefer 8ohm.

Questions...

1. Am I right in saying that the LM3886 really needs at least 20v? If so I am looking at using the skynets at 24v.

2. Will the 24v arrangement supply enough VA for running 4 amps?

3. Is using linear regulators after the skynets just getting silly?

4. Other advice?

I know that someone will advise that I should be starting simple and working up to this, but as power supplies are the expensive bit, I'd rather start off high quality here than buy up (reasonably expensive torroids and diodes) only to scrap them. I will do my experimentation at the cheaper stuff on the amps.

Long first post I know, but advice would be good.

[Nuuk]

LM3875 (yes) with +/-24 volts from four Skynet 8080 modules. That will run your Mission 735s fine.

No linear regs.

Try that, then if you want a spearate PSU for each channel, add another four Skynet 8080s.

(I took pity on you with a name like that! )

[AndrewT]

Originally posted by justblair
I may wish to go further in the future and bypass the speakers' passive filters. But one thing at a time.

Hi,
go read the last 24hrs posts in DCX2496 thread.Behringer DCX2496 digital X-over
You may decide that passive bi-amping is as far as one may want to go.

I tried DCX active route and reverted back to passive, but I was not nearly as thorough as Pqr in post1054, 1056, 1057.

[justblair]

@NUUK Thanks for the advice, I was doubting the need for regs, due to the voltage drop more than anything. As for the LM3875, I had discounted that due to its reqirement for 8ohms.

Do I take it that 2ohms difference wont effect them badly in your opinion?

Does an inverting design make more sense in my case?

And thanks for the pity



@Andrew... Thanks for the heads up... I have not got far into my research regarding active crossovers, I will think very carefuly now before attempting such a thing!

Open to any other advice... Particularly a few opinions on the chip choice.

[Nuuk]

Yes, the LM3875 will be fine with 6 ohms - as I recall the 753 is not a difficult load.

I have a pair of IPL A2 kit speakers. They use pretty much the same drivers as the Mission 752 and I drove them from LM3875 amps with no problem.

As regards this active crossover business, I have said this a few times now but we know how easy it is for posts to get 'buried' on such a popular forum so I will say it again.

Going active is not just a matter of removing the passive crossovers and using active ones in their place. The removal of the passive crossovers can, and does change the speaker alignment. With a sealed box speaker, this usally isn't a problem but with a ported design, it can mean that you need to change the volume of the speaker cabinets, usually making them larger.

As an example, my A2 speakers with passive xovers had a cabinet volume of 27 litres. After moving to active xovers, I found the best size (by trial and error) was around 33 litres. Oh, and the port tuning needs altering too!

I can honestly say that I achieved very good reslts with my active A2s using nothing more than the simple discrete crossover that is described here

(you could try opamps instead) Of course, I may just have been lucky but for a two-way speaker, it is worth a go for what is involved!

End of sermon!

[justblair]

@NUUK thanks again, especially for such promptness in your reply. Funny enough I have DD open in another window ATM and was just reading on your experiances tuning the enclosures.

I dont see me going active in a hurry (if at all), I like when I start a project to know where it is going and plan for where it might lead. Once I have bored myself playing with the amp, I may try the active route... Following in your footsteps though I can see a preamp with a buffer being the next logical step.

I dont want to try and reinvent the wheel right away. I am in the early learning centre of the diy electronics high street right now. I recently taught myself the basics on PCB manufacture (photo resist) and this is an opportunity to use and refine my newly aquired skills.

DD by the way is a fantastic resource, you should be proud. My decision to take up this project was hugely influenced by your site. Your writing style and the structure of the site really makes a project like this seem both attainable and desirable. Hope I dont sound too sooky, but I like to give credit where its due!

Blair

[Nuuk]

Thanks Blair, there are quite a few hours in the old Dungeon so I am very pleased when I hear that it has been of help and inspiration to others.

I think getting started is the hardest part of this hobby, and I intentionally wrote DD to help people over the first hurdle!

And you are doing the right thing in concentrating on one part of your system at a time.

[justblair]

Ok...

I have a few more questions. Here is my shopping list so far...

4 x Skynet 8080 (on route to me right now)
4 x Lm3075 (on route to me now)
8 x 1000uf Nichicon fine gold caps (in my grubby paws)
4 x Maxim ds1802 Digital volume controls (samples requested)
4 x QT110 touch sensors

My plan is to run the signal (Pc source) through the Maxim volume controls, to a buffer and then to the 4 lm3875 circuits probably IGC configeration.

The DS1802 I was planning to use in the circuit described by Pedja Rogic

DS1082 circuit

I was thinking along the lines of having two of these running, one per opamp (ie per channel) daisychaining them to get unversal control. In the future this would also give me the option of using remote control

Next off I will use the simple discrete buffer described at Decibel dungeon

Simple buffer here

Then into the IGC circuit at the same site.

Power comes from the SKynets, set up to run 24v, The buffers regulated down to +/-15v.

I will regulate the 5v supply from the Skynets down to +/-2.5v for running the DS1802 chips and the QT110's

Questions.

1. Are there any advantages in running 4 op amp stages ie. one per power amp?

2. What is the imput impedence of the discrete buffer?

3. Am I going to run out of power from the Skynets?

4. Can you see any other nasties ahead for me?

[justblair]

Up and running now with 4 SMPS and two IGC amps, having so much fun on this thanks for the help so far.

I thought I would share my early experiances on using SMPS. I had the option to try out differnet combinations of the Skynet 8080. I used of course Decibel Dungeons instructions to wire up.

One unmodded SMPS feeding two amps - Not brilliant at all, although initial impressions were that the amps were bright and engaging, clipping came very quickly. Later experiments showed also that detail was missing throughout the frequency range.

Two unmodded SMPS feeding two amps - Better. The brightness was still there, and clipping was less evident. However the detai was poor in the bass range. NUUK described this as flabby and I tended to agree, My flatmate however pointed out that the bass sounded compressed. As soon as he said it, it struck me that he was right.

Two modded SMPS in series feeding two amps - Things got a whole lot better in the Bass range having +/- 6amps available. Though the bass was flabby still, the compression had reduced dramatically. The sound was still very bright though smooth and I thought it was engaging and very musical, especially with Jazz and female vocals. That was untill I tried 24v from 4 SMPS.

This is where my experiance differs from NUUK. The 24v option was dramtically better on my setup. The controll over the bass was suddenly happening, But not just that, the top end improved as well. Although previously I thought that the 12v sound had been detailed it now seemed unsubtle. What I had mistaken for smoothness before was obviously a blurred sound. NUUK mentioned on his site that he would be pushed to choose between 12v and 24v, In my setup the choice is far clearer. Regardless of the type of music being played the 24v setup is several times better. The 12v is upfront and in your face, switch to 24v the soundstage goes back the way, deep behind the speakers. The imaging is so much better as well.

I have also tried a zobel network with 10r and .1uf cap. I am not sure if I am supposed to hear a difference. I think I do, but I have listened now to so much I think that I might be imagining it.

Plan is next to introduce a buffer to the mix to see if it improves things further. Its hard to imagine it will, but the parts are all sitting here in plastic baggies so might as well give it a go.

Thanks again all!

Blair

[AndrewT]

Hi,
the Zobel should not be heard.
It should help stop oscillation.
If oscillation did occur you might hear the remanants down in the audio band.

It sounds like the chipamps are sensitive to PSU and particularly low supply voltage (=lots of clipping on transients) and lack of current. Are you allowed to add smoothing capacitance between the reg and chipamp?
My camp and the other disagree on how much capacitance is needed to support good bass.

Have you tried them on an unregulated supply yet?