silk purse from sows ear contest?

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I was curious if anyone would be interested in making a communal thread / contest to make the best measuring speakers possible, for some absurd low driver price limit. I thought a contest like nature might be that extra bit of accountability for people like myself to actually get something done in a timely manner. How about setting up a initial set of rules / judgements of:
- use USD or EU, convert your currency. (if you are in US or EU, you cannot convert)
- $30 limit USD (or 37 EU) prices per speaker (drivers, active/passive crossover components, DSP, purchased radius ports)
- $ free no limit wood, pvc ports, screws, glue, wire, stuffing, terminals (goal is to take cheap components and design / build good enclosures)
- DSP not allowed unless it is part of the budget (SGTL5000 type chips make this feasible actually).
- amp not part of contest UNLESS you go for DSP, then your entry takes RCA input instead of speaker level input
- prize = bragging rights within this thread.
- deadline = 9/1/18 2.5 mo should be plenty of time.
- judges = anyone who submits a design with photos / simulations and a 1W/M REW response curve. You may vote a single time for someone else's project.
- 2 categories, simulated as well as IRL.
- do NOT include shipping costs, Taxes, VATS, Customs, anything else.
- do NOT include incidentals like wire, terminals, glue, caulk, screws etc.
- use best guesstimate for winding your own coils
- NO External dimension or volume limits on enclosures.
- use 1/24 smoothing in REW plots.
- out of budget entries welcome for 'close but no cigar, still ineligible but thanks for helping us learn!'
- You may enter multiple designs even reusing exact same components.
- NO driver minimum, CAN be single full range driver in a 4x8 sheet of plywood if you can make that sound good and convince peers to vote for it!


possible judging guidelines, but remember this is peer with 1 vote, not something scientific:
- Mathematical formula for judging 20-20k, add a point for each db higher or lower than reference at each standard graph axis label in REW: (20, 30, 40, 50, 60, 70, 80, 90, 100, 200, 300, 400, 500, 600, 700, 800, 900, 1000, 2000, 3000, 4000, 5000, 6000, 7000, 8000, 9000, 10000, 20000) meaning the lowest score aka flattest response wins.
- Another formula that does not focus so much on the lower frequencies, maybe octaves? 20,40,80,160,320,640,1k,2k,4k, 8k, 16k ??

Extra imaginary internet points for:

- dumpster diving!
- creative reuse of things ( unwind a motor into an inductor for example)!
- inventing creative cheats!
- money left over
- not using buyout drivers
- high peak SPL
- HD plots that are actually good
- painting it red or gold
- using OSB instead of decent wood
 
Last edited:
Yeah, I would expect some people would use REW and the laptop mic.. Not really expecting a great scientific comparison from this, but a cheap way to get some feedback / provide peer feedback, and focus on the engineering under our control vs throwing the $ at the engineering that we cannot really control. Maybe it would encourage some mini meetups for testing if logistics aligned.
 
Just another Moderator
Joined 2003
Paid Member
This would not be possible for me. I doubt I could find any drivers for the $20, I could do bipolar electros for caps, and coils I could wind myself (how would I work out the cost, my magnet wire I bought ages ago). If I had to buy coils then again would be very difficult to meet the cost.

The shipping alone would be more than the budget, if I were to buy something from OS where the prices are more reasonable :)

You could have two categories. One for actual builds, and one for simulated builds :) I'd be able to participate in the second, provided I could find some cheap drivers that had FR and impedance curves published.

edit: perhaps shipping could be excluded from the cost, that would make the playing field more even.

Tony.
 
Last edited:
Given that half of the judging frequencies are below 700Hz, the winner is probably not going to be the person who builds the flattest speaker but the person who knows how to measure below 700Hz without significant room/environmental effects.

That's a good point that the measurement conditions would probably have to be standardized. Also, the set of measurements would have to be standardized as well. For example, should this include off-axis measurements, or on-axis only? If off-axis is included, do you do both vertical and horizontal?
 
Well I know $20 per speaker can be done easily. The Dayton B652 uses known cheap drivers and a single cap on the tweeter for a crossover.

Poly buyout woofer $5.88
GRS 1tm-4 $2.90

$9 in drivers leaves $11 in creative improvements over the B652.

Full disclosure, this is exactly the system I had in mind to improve apon, mostly trying to figure out a way to low pass the woofer, and use a silk dome tweeter such as:
peerless BC25SC06 $10 only leaves $4
or
dayton ND25FA $13 only leaves $1

I would be incredibly impressed (but not too surprised from the minds on this forum) if someone did come up with something "excellent", and I suspect the feedback and collaborative suggestions will help us all. I look forward to doing my first measurements of a speaker, showing some pics, and learning how to do better.
 
Last edited:
That's a good point that the measurement conditions would probably have to be standardized. Also, the set of measurements would have to be standardized as well. For example, should this include off-axis measurements, or on-axis only? If off-axis is included, do you do both vertical and horizontal?

Excellent points. Initially I had only thought for on axis. I think it would be interesting for the rules to allow some creative gamification and "cheating" though. For example, if the purpose is to optimize a 1W/1M sweep, who cares if it hits Xmax at 2w ? Or if it is only for on axis, who cares if it is intollerable 30deg off. As long as the 'cheat' is explained so the rest of us can 'appreciate' it!
 
It's most likely impossible to purchase anything at all here (maybe except binding posts and glue?), and keep under 20$ limit.

Good idea, but I wish the limit was a little bit higher, at the moment it seems that these restrictions make the competition only for residents of the US.

I would be more in favor of a "scavenge only" contest.

Edit:
Did not see the "DO NOT include shipping cost", but this is still pretty vital. It could be as much as 50$ or even more in shipping alone to get 20$ worth of whatever few drivers are needed. The amps and such is not much of a concern really.
 
Last edited:
The cheapie Dayton B652 is interesting. Dayton Audio B652 6-1/2" 2-Way Bookshelf Speaker Pair

A well behaved 6" polycone and an 8" paper cone can get to much the same place, as the Vifa PL17WJ-00-08 FR below shows. A question of mechanical cone damping. It always needed a bass coil at least though.

I scavenged these 8" Sony E44 with a similar cheapie balanced-drive coneish mylar tweeter for £20. Simple 3.3uF capacitor to tweeter. Nothing else.

626011d1500178190-classic-monitor-designs-sony_e44-jpg


They sounded like dreck, TBH. I did a better crossover and they were then remarkably good.

These days I am working on a WLM La Scala clone, which went for 2,500 Euros originally, with a 4th order tweeter filter and available drivers in the same cabinet. Surprisingly, I am finding the modern Visaton bass isn't actually so musical or enjoyable as the original Sony retro bass cheapie.

687319d1529236762-6-8-woofer-system7-wlm-la-scala-clone-jpg


But I cannot do this for $20. :(
 

Attachments

  • Vifa PL17WJ-00-08 6 inch Polycone.PNG
    Vifa PL17WJ-00-08 6 inch Polycone.PNG
    23.9 KB · Views: 419
Last edited:
... You could *almost* pair up a Peerless TC6F 2" with a GRS 8PR-8 8", but the price is 88 cents over budget. And there's not much info on the GRS 8".
Couple the TC6F with the 17.10$ 6,5" Peerless SDS-160F25PR01-08, and it's in a whole 'nother league. But still fairly cheap.

Could try just placing some pipe/cup on the back of the TC6F, use 1st order 400hz-1khz scavenged parts for xo.
 
Last edited:
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.