a delta1 "HAT TRICK" of sorts

delta1

Moderator: linux-works

a delta1 "HAT TRICK" of sorts

Postby linux-works » January 1st, 2011, 10:43 am

yes, a hat trick. 3 things in a row ;)

Image

Image

I'm doing an experiment to try out a few things.

this is an LCDuino-1 (on its back, to the far left) controlling THREE delta1 relay stacks. the idea, here, is that I'm doing an active crossover volume control and I have 3 outputs from the crossover: HP, BP, LP.

its passive (no buffah's yet, lol) but it is meant to be a test bed for the LCDuino firmware with more than 1 delta1 board. this is actually the first time I've had more than 1 d1 board on a test bed and had to re-arrange the programming code quite a bit to support it. also have to add support for 'balance' or fading between boards. its being done now.

see that large lcd display? its not quite a backpack style mount and it took some ribbon cable to get it to talk to the arduino but its still a 1:1 wiring connector. I plan to use more of that lcd 'real estate' for this particular application and so the wide lcd was a fun touch for this build. it was available locally at a surplus shop and I took a chance on it, not knowing if it would work or not. it seems to work ;) its 40x2 chars,lines.

are 3 relay boards very chattery? well, yeah, a little ;) but its a 'grouped burst' and they are all attempted to be done at once, as best as possible.

if things work well for a 3board version, the 2board d1 build (ie, a 'balanced' build) and of course the unbalanced single-ended 1board d1 build will all work.

currently, I have a new behringer dcx2496 (yummy!) to use as a source for all this. it accepts direct spdif input and runs thru 'math filters' in digital domain to create the HP,BP,LP analog outputs. those are the ones I take, directly (at pro audio +10db level) and feed into the rca-in of my new triple-delta1 box. each realy board is a full 8bit build and so you have 255 steps of half db each, giving 127.5db total range. even the +10db 'hot' signal from pro audio gear is not too much to overpower things once they go thru a 127db attenuator ;)

anyway, more stress testing for the LCDuino, the delta1 and the poor, poor i2c bus. it never gets any rest, now ;)
Sercona Audio (Control Systems for Analog and Digital)
Photos: http://www.flickr.com/photos/linux-works/
User avatar
linux-works
 
Posts: 2065
Likes: 1 post
Liked in: 40 posts
Joined: March 31st, 2010, 9:04 am
Location: Santa Clara, CA
Country: United States (us)

Re: a delta1 "HAT TRICK" of sorts

Postby DorkVader » January 1st, 2011, 11:28 am

The best part is, of course, the usage of the pretty ribbon cables.

Having two different manufactirers of resistors won't affert anything, will it? I imagine not. (I came up with an alpha20 project on Mouser, and had to go with Xicon, 'cause I couldn't find an 825-Ohm resistor from Dale)
User avatar
DorkVader
 
Posts: 493
Likes: 0 post
Liked in: 0 post
Joined: April 7th, 2010, 8:21 am
Location: Boston
Country: United States (us)

Re: a delta1 "HAT TRICK" of sorts

Postby linux-works » January 1st, 2011, 11:50 am

with that build I just used what I had 'on hand' (lol). I had a pair of the brown sausage (vishay) boards and was planningn on using that for 'balanced' testing. the blue xicon board is not only a diff vendor but also diff Z. I picked 16k, I think, as the impedance of that board when picking resistors from mouser. I did want each board to be consistent and its a little trial/error in using the online R/2R calc to find a set of R's and then trying to find ALL of them in stock from a vendor or type. you can re-run the calc and change the Z very slightly, for example, and have it regenerate all new R's for you to try. then see if you can get them all in your basket on some online order and if not, try AGAIN. its not easy sometimes.

the steps are all .5db and all boards have 8 bits (relays) and so the range is -127.5 -> 0, no matter what the input-Z is. the output-Z varies as you vary the atten and ideally you do want a buffah right after the delta1. if your amp is close by, its less mandatory to have the buffah.

if you use all identical amps, you probably do want to take care to make all 3 boards use the same set of R's. if I did build them all at once, I certainly would have done that.

from a software POV, as long as each board uses the same # of bits (relays) and each step is the same size (half db orr whole db or whatever) - that's all that matters. input or output z issues are not seen at the software level.
Sercona Audio (Control Systems for Analog and Digital)
Photos: http://www.flickr.com/photos/linux-works/
User avatar
linux-works
 
Posts: 2065
Likes: 1 post
Liked in: 40 posts
Joined: March 31st, 2010, 9:04 am
Location: Santa Clara, CA
Country: United States (us)

Re: a delta1 "HAT TRICK" of sorts

Postby suicidaleggroll » January 1st, 2011, 11:52 am

DorkVader wrote:Having two different manufactirers of resistors won't affert anything, will it? I imagine not.


Shouldn't
It's for a 3-way active build, so the board with the different resistors will just be for one of the bands (say, midrange). As long as it's consistent for the L and R channels I don't think there would be any issues.

I had a DCX2496 for a while. I was planning on building a set of 3-way active towers using Scan Speak Revelators with the DCX as the active crossover, but ended up abandoning it. Now I just have a pair of Zaph Audio ZRT 2.5s running passive with dual mono amps and a DEQ2496 for room correction :)

I really like those Behringer 2496 processors, my DEQ has been a godsend, and the DCX I'm sure would have been fantastic as well if I ever got around to completing the towers (and buying all that damn amplification!)
suicidaleggroll
 
Posts: 136
Likes: 0 post
Liked in: 0 post
Joined: October 7th, 2010, 11:24 am
Location: Colorado
Country: United States (us)

Re: a delta1 "HAT TRICK" of sorts

Postby linux-works » January 1st, 2011, 12:19 pm

I got the DEQ along with the DCX. one directly feeds the other via a short xlr-xlr patch cable. even bought a measurement mic to toy around with.

for amps, I figure the chip-amps can handle the tweeter and a midrange that does not go too low. that will be 4 channels; and the other 2 will come from either a powered sub or maybe a class-d amp powering some 6.5" drivers in an enclosure for midbass. I want to mess around with that later on - but for now I'm just using some bookshelf 2way spkrs as the HP test units and a powered sub as the bottom end. by using the various settings in the DCX, I can get overlaps and boost some weak areas in the bookshelves, for example.

eventually I do want to have 4 channels from chip-amps and 2 from some class-D thing for a discrete 3 way tower system.

the DCX has 6 analog outs and so the hat-trick box can do the vol control on those and let the DCX run its analog outs at full max (no bit shaving, that way).
Sercona Audio (Control Systems for Analog and Digital)
Photos: http://www.flickr.com/photos/linux-works/
User avatar
linux-works
 
Posts: 2065
Likes: 1 post
Liked in: 40 posts
Joined: March 31st, 2010, 9:04 am
Location: Santa Clara, CA
Country: United States (us)

Re: a delta1 "HAT TRICK" of sorts

Postby suicidaleggroll » January 1st, 2011, 12:56 pm

Why a class-d?
I haven't looked into them too much recently, but a few years ago I know manufacturers were having issues making them full range without a lot of noise and distortion. The few that were full range (or at least capable of a few kHz) were not cheap. The only real advantage to them was efficiency, so if efficiency isn't that important to you a class A/B would probably have better specs, sound better, and be cheaper.

I have a pair of Emotiva UPA-1 monoblocks on my towers, and I'm really liking them.
suicidaleggroll
 
Posts: 136
Likes: 0 post
Liked in: 0 post
Joined: October 7th, 2010, 11:24 am
Location: Colorado
Country: United States (us)

Re: a delta1 "HAT TRICK" of sorts

Postby linux-works » January 1st, 2011, 1:00 pm

I was thinking if it was subwoofer range, class D might be reasonable enough. it would make a small enough box that would be close to what the chipamp box would be like, in size.

one reason I did the vol control as its own box is that I have freedom to pick ANY set of amps. the thought crossed my mind to build the d1 with a gainclone follower in the same box but that would tie my hands too much and also make the box a lot bigger and uglier inside.
Sercona Audio (Control Systems for Analog and Digital)
Photos: http://www.flickr.com/photos/linux-works/
User avatar
linux-works
 
Posts: 2065
Likes: 1 post
Liked in: 40 posts
Joined: March 31st, 2010, 9:04 am
Location: Santa Clara, CA
Country: United States (us)

how should LCDuino work with this? what's your opinion?

Postby linux-works » February 26th, 2011, 7:50 am

my application is to have 3 analog stereo pairs out of a 3 way crossover (does not really matter that the crossover is digital inside) and be able to drive 3 amps in tri-amp config to the various speaker drivers, 1:1 on each amp.

I was initially thinking that the hat-trick 3 way vol control would be a balance/fader box as well, but now I'm rethinking this.

perhaps what is really needed is 2 separate things: one is the combined ganged group volume control that is just like your classic multi gang pot and when you turn the knob they all move the same amount and completely in unison. the 2nd thing is a new box and it would act as a final 'trimmer' so that each amp and spkr can be balanced or faded on its own.

the nature of these 2 things are different in a few ways. first, the group volume control will click up and down as the user varies the volume. if its loud or chattery, that's less good (even though vol change events are not something you sit and listen to; you change to a level and leave it there, usually for quite a while). if the group vol control keeps all the values the same, the noise is not as bad as if it were sending a different value to each of the relay boards. keeping the same value across all boards in this box has a benefit, here.

the 2nd box usually is setup once (based on microphone testing in the room and stuff like that). you get the balance between the amps and stuff setup and then you leave it. this box and its relay attens would NOT run up and down as you change volumes. BUT, it has a higher demand on precision, I think, since its whole purpose in life is very fine control over balance. in this situation, I'd use those 25db 0.1db/step boards and they would be 'glued' to an amp and be part of that amp's config in this room. you could even have an IR/display on them and be able to vary its volume, entirely on its own, remotely ;) but it would not be the main vol control and once its set, it gets hidden away or stashed away. it would not have to take up front stereo shelf space at all.

so, the more I think about it, the more I see 2 boxes and 2 roles here, not a combination. what do you guys think? how would you use an lcduino and delta1 boards?

single ended (1 delta1) is simple and obvious. balanced with 2 delta1's is the next step up and it relates to this discussion in terms of need (or non-need) to have to vary the atten differently on each of the 2 delta1's in a balanced config. amb and I are thinking that if you are in balanced config, there is no need to have a diff value sent to one d1 board vs the other.

but on a 3 delta1 config in tri-amp or even a 2 delta1 config in bi-amp mode, maybe its needed to have d1/d2 board to board balance or offset variation? remember, you can only carve down to the native step size of the board, usually .5db. is having .5db offset (or multiples) worthwhile? (in the .1db stepper, yes, it is; but I wonder how useful .5db is; seems too large a step to be really useful in any channel fader app).
Sercona Audio (Control Systems for Analog and Digital)
Photos: http://www.flickr.com/photos/linux-works/
User avatar
linux-works
 
Posts: 2065
Likes: 1 post
Liked in: 40 posts
Joined: March 31st, 2010, 9:04 am
Location: Santa Clara, CA
Country: United States (us)


Return to δ1 relay-based R-2R stereo attenuator

Who is online

Users browsing this forum: No registered users and 0 guests

Quick Links

AMB audio DIY resources
AMB audio shop