Right Mark. Wooden floor resonance and vibration caused by the loudspeaker which is in contact with it, can change by small movement of the loudspeaker which will bring it closer or further from an underlying wooden supporting member. Nocking around the floor with a mallet, one can locate by the change in pitch where the beams are. Then there are methods and various materials for coupling loudspeaker over larger surface or decoupling it from the floor .
As for the room resonance, this is what I have experienced.
A loudspeaker repositioning by an inch or so will change the upper treble combing which may give the misleading impression of better bass (memories from the past
). Half a meter or so will alter combing at the treble/upper mids (stronger misleading impression of better bass plus real change in clarity/muddiness at treble/mid freq).
At bass region, for noticeable changes of dips or peaks in pressure/velocity from combination of all propagation modes in all three dimentions, speaker (or listener) need to move by more than half a meter. Check that out with multiple recordings with a microphone (viewing sweep results on the screen show much more dramatic than we perceive them acoustically)
For to get around the psychoacoustic trap (the misleading impression of altering one end of freq spectrum and noticing changes at the other end), drastically LP or HP filter the excitation signal and concentrate on one side of the spectrum each time.
George
Yes! this technique is useful for many kinds of listening tests.
For just one example: Awhile back I was using a software music synthesizer to experiment with different test tones. I was looking for something dense like pink noise, but less random. I constructed a multi-sine generator that could produce simultaneous sines at 1/48 octave intervals from 10 Hz to 20 KHz (528 total)*. I tweaked it so that the individual sines line up with the bins in REW's 1/48 oct. realtime spectrum analyzer, so any response changes show up very quickly with no averaging delays, but still with lots of frequency resolution.
I also was able to shut off portions of the spectrum in one octave chunks. Using a single octave of this sound, centered on my 2 KHz crossover frequency with tweeter polarity inverted, I was easily able to zero in on the deepest notch as I adjusted the tweet delay time back & forth in the DSP, from the listening position.
*The full-bandwidth output of this thing is one hellacious noise by the way. Here's a recording of the first 20 seconds or so after starting the program - it's a zipped mp3, with some audible compression artifacts of course - I suppose this would be a perceptual coder's worst nightmare, heh.
(mind your volume knob):
View attachment MultiSine_48x11_15sec.zip
Last edited:
Works to calm the kids down as well. We keep a couple of CDs in their room for just such a purpose.
I have run into a difficulty on my surveillance microphones. I use a THAT signal processing chip for a limiter in them. Seems there are very few analog chips made anymore for that. Even the NE570 is getting scarce! Enough folks have copied my design and placed orders for the chip that it is now backordered by 4 months. As THAT has its' own foundry that was surprising.
I did find 150 chips in the UK that only leaves me short 400 for my current order. I expect the final quantity to be over 10,000 units.
So a redesign using a discrete gain cell may be in order or would a complete custom analog IC be more reasonable?
I did find 150 chips in the UK that only leaves me short 400 for my current order. I expect the final quantity to be over 10,000 units.
So a redesign using a discrete gain cell may be in order or would a complete custom analog IC be more reasonable?
There used to be folks who put transistors and resistors on semi-custom chips. All you basically did was layout the top interconnect layer. Not sure if this is still offered.
On just my needs $500K is a reasonable expense!
Bill I didn't go with the 570 as the That chip had both a rectifier and VCA in one package and cost less. My PCB subcontractor is the one who delayed ordering the chips for reasons known only to them!
I also left a message for the THAT sales folks about wanting to buy a few reels. Emails and phone calls have no response.
I suspect my next move is to redesign using with some matched transistor chips.
Last edited:
Demian,
DSP seems to exceed my microphone power budget of 10 mA at 48 volts.
I found folks running the right processes and making similar parts. In quantities of 5,000 pieces each unit will cost under $3.00!!! Start up costs extremely reasonable.
I guess there is still room for obsolete parts.
DSP seems to exceed my microphone power budget of 10 mA at 48 volts.
I found folks running the right processes and making similar parts. In quantities of 5,000 pieces each unit will cost under $3.00!!! Start up costs extremely reasonable.
I guess there is still room for obsolete parts.
Yes, it would be tough. There might be a mobile codec that can do it all, but heavy learning curve I'm sure.
For example, something in the vein of:
WM8281 | Cirrus Logic
Demian,
Yes I use DC - DC converters now. Interestingly enough one fellow who was looking at one of my pieces was surprised I used an off the shelf unit rather than piece one together. To me the secret to using them is taming the resultant high frequency noise.
As you know for simple ones it can be a single chip, a few diodes and passives and for efficient designs a transformer. They actually make the transformers locally, but the last time I needed a sample they had to get it for me from one of their distributors as they keep no stock.
My current PCB supplier avoids 0102 size parts and I am unsure of BGA ability. Currently I have them do runs of 1,000 pieces for my stuff.
What has surprised me is the sales reps call on me as 1,000 piece orders are considered significant. Seems larger orders are all offshore folks.
This is a bit interesting as what I am paying about $20 for locally made is running $30 ish from China! It is almost 100% machine assembled locally and virtually none of the parts made in China.
But I will keep you in mind if I run into difficulties. Currently redisgning a product to increase the available parts suppliers.
Current products others are copying but they don't quite understand the tweeks. Also only one patent so far.
Next version should blow their minds. Lower cost, compatable with pro standards and the current crappy ones and performance better than what they understand to be possible!
Apparently as I think you know there is some advantage to 50+ years of experience in the audio field.
Yes I use DC - DC converters now. Interestingly enough one fellow who was looking at one of my pieces was surprised I used an off the shelf unit rather than piece one together. To me the secret to using them is taming the resultant high frequency noise.
As you know for simple ones it can be a single chip, a few diodes and passives and for efficient designs a transformer. They actually make the transformers locally, but the last time I needed a sample they had to get it for me from one of their distributors as they keep no stock.
My current PCB supplier avoids 0102 size parts and I am unsure of BGA ability. Currently I have them do runs of 1,000 pieces for my stuff.
What has surprised me is the sales reps call on me as 1,000 piece orders are considered significant. Seems larger orders are all offshore folks.
This is a bit interesting as what I am paying about $20 for locally made is running $30 ish from China! It is almost 100% machine assembled locally and virtually none of the parts made in China.
But I will keep you in mind if I run into difficulties. Currently redisgning a product to increase the available parts suppliers.
Current products others are copying but they don't quite understand the tweeks. Also only one patent so far.
Next version should blow their minds. Lower cost, compatable with pro standards and the current crappy ones and performance better than what they understand to be possible!
Apparently as I think you know there is some advantage to 50+ years of experience in the audio field.
Last edited:
The placement won't really be a problem for a BGA like that. The problem will be that you need laser drilled microvias to escape 0.5mm and lower pitch BGAs. No x-ray inspection might be less than ideal but maybe not a showstopper. I really don't mind BGAs at all, but I do mind the cost of prototyping HDI boards.