Hi all,
I've just finished building a pair of 2.5 clones (exactly as per the instructions on the http://www.geocities.com/diyproac25/index.htm site), and yes it sounds wonderful.
However, I'm thoroughly perplexed by the port tuning. A simulation of the suggested 75mm diameter 140mm port in a 34l enclosure reveals a terrible response with a 4dB peak at around 55Hz. I figured there must be some explanation; perhaps the damping will increase the effective volume sufficiently, or perhaps its some weird over-the-top form of baffle step. For 34l a port of 450mm seems more appropriate for flat Butterworth response.
However, the bass sounds BAD. I haven't measured it yet, but it is clearly audible that something's wrong. There's a horrible peak around 60Hz (as expected) with rapid rolloff below (also expected). Furthermore the woofer cone excurts heavily on most material where other speakers' (also with Scan-Speak woofers) don't, a clear indication of a too high tuning frequency.
Obviously the results seem to be similar to the simulation, so now I ask: what gives? Many people have built this speaker and none complain about the bass.
Hoprfully someone can provide some insight.
I've just finished building a pair of 2.5 clones (exactly as per the instructions on the http://www.geocities.com/diyproac25/index.htm site), and yes it sounds wonderful.
However, I'm thoroughly perplexed by the port tuning. A simulation of the suggested 75mm diameter 140mm port in a 34l enclosure reveals a terrible response with a 4dB peak at around 55Hz. I figured there must be some explanation; perhaps the damping will increase the effective volume sufficiently, or perhaps its some weird over-the-top form of baffle step. For 34l a port of 450mm seems more appropriate for flat Butterworth response.
However, the bass sounds BAD. I haven't measured it yet, but it is clearly audible that something's wrong. There's a horrible peak around 60Hz (as expected) with rapid rolloff below (also expected). Furthermore the woofer cone excurts heavily on most material where other speakers' (also with Scan-Speak woofers) don't, a clear indication of a too high tuning frequency.
Obviously the results seem to be similar to the simulation, so now I ask: what gives? Many people have built this speaker and none complain about the bass.
Hoprfully someone can provide some insight.
mm id like to know ppls thoughts on this too, im in the middle of building mine, and the ports will be quite solidly fixed into place id hate to have to change there dimensions.
I cant imagine thered be an error on the dimensions page that hasnt allready been picked up, these clones , and the originals are quite tried and tested., this seems odd
I cant imagine thered be an error on the dimensions page that hasnt allready been picked up, these clones , and the originals are quite tried and tested., this seems odd
Disabled Account
Joined 2003
I've tried varioous computer simulations for 38Hz port tuning (as indicated by Stereophile) and graphs correspond with Stereophile's measured response which shows rising SPL from 100 to 50Hz.
It seems the Response 2.5/clone's bass tuning is on the high side intentionally. My own listening impressions seems to indicate a rather generous or ripe bass response, consistent with other ProAc speakers I've heard. eg. Studio 125, D25.
75mm x 140mm (Geocities) or 72 x 135mm (Troels) is more or less correct.
Double check you have the correct (33-34L) internal volume (remember that damping material takes up space) whereas stuffing material is not included.
IME too much 200gsm Dacron kills midbass response and too little gives too much midbass.
It seems the Response 2.5/clone's bass tuning is on the high side intentionally. My own listening impressions seems to indicate a rather generous or ripe bass response, consistent with other ProAc speakers I've heard. eg. Studio 125, D25.
75mm x 140mm (Geocities) or 72 x 135mm (Troels) is more or less correct.
Double check you have the correct (33-34L) internal volume (remember that damping material takes up space) whereas stuffing material is not included.
IME too much 200gsm Dacron kills midbass response and too little gives too much midbass.
Well I just did a frequency sweep with impedance measurement, and the tuning frequency seems to be around 52Hz - too high! I've come to know the speaker as rather fussy concerning placement, but nowhere the bass sounds right. I still have to experiment with damping though, maybe adding some will help.
I have made some minor changes to the design, but none should be the cause for such a big change. Firstly I changed the depth and height slightly to compensate for different thickness MDF available in my country - width is still the same. I used 22mm throughout. Secondly I left out the bitumen layer as I couldn't get some. This caused the total volume to be 38l, which should still be fine. Thirdly my damping is 270g/m2 Dacron instead of the 200g/cm2 specified, I compensated by using a little less and by plucking it somewhat looser. Lastly I used a precision dual flared port (still 75mm inside diameter) from Parts Express, but the instructions clearly states how to perform the conversion (centre tube 4" shorter than the desired [140mm] length).
I'm going to experiment with some more stuffing tomorrow, and afterwards take out the port and try the simulated length.
I have made some minor changes to the design, but none should be the cause for such a big change. Firstly I changed the depth and height slightly to compensate for different thickness MDF available in my country - width is still the same. I used 22mm throughout. Secondly I left out the bitumen layer as I couldn't get some. This caused the total volume to be 38l, which should still be fine. Thirdly my damping is 270g/m2 Dacron instead of the 200g/cm2 specified, I compensated by using a little less and by plucking it somewhat looser. Lastly I used a precision dual flared port (still 75mm inside diameter) from Parts Express, but the instructions clearly states how to perform the conversion (centre tube 4" shorter than the desired [140mm] length).
I'm going to experiment with some more stuffing tomorrow, and afterwards take out the port and try the simulated length.
Well, you've pretty much confirmed that the problem is the port. Maybe the rule of thumb you used doesn't work for such a short port.
I would really recommend using bitumen or an alternate damping for the walls. This isn't the major problem in your case, but you are definitely going to have increased panel resonances.
I would really recommend using bitumen or an alternate damping for the walls. This isn't the major problem in your case, but you are definitely going to have increased panel resonances.
Tiroth, I think you're correct about the equation for the flared port not being applicable to short lengths... the actual center tube distance is only about 30mm! Pity as it was a rather nice-looking port with low chuffing. Fortunately esthetics aren't important; they were built to be used as guinea pigs for digital crossover development at my University's DSP lab. (We're already looking for a nice multiway speaker with diffcult drivers and a nightmare crossover to test our new filters with, any suggesions? It has to be one for which an optimal passive one is available for comparison of course. The ProAc is the 1st one, but next we'd like to try something like a 3way with a Seas Excel midrange and ribbon/ringradiator tweeter).
What have you guys used for crossover components? Leftover budet expenditure allowed me to go all-out with Hovlands on the tweeter, Auricaps on the woofer, Goertz coils all round and Mills resisors. Especially the Hovlands made a huge difference IMO and although 4 in total are needed were needed for the 9500 crossover, are well worth it.
Regarding leaving the the bitumen, I've decided that it would be easier, cheaper and perhaps more effective to increase the bracing. I've used pieces of broomstick wedged randomly and it rendered the walls pretty dead. As I'm using the 9500 tweeter with Troels's crossover, it really isn't a clone anymore, so no need to try and reproduce the original.
What have you guys used for crossover components? Leftover budet expenditure allowed me to go all-out with Hovlands on the tweeter, Auricaps on the woofer, Goertz coils all round and Mills resisors. Especially the Hovlands made a huge difference IMO and although 4 in total are needed were needed for the 9500 crossover, are well worth it.
Regarding leaving the the bitumen, I've decided that it would be easier, cheaper and perhaps more effective to increase the bracing. I've used pieces of broomstick wedged randomly and it rendered the walls pretty dead. As I'm using the 9500 tweeter with Troels's crossover, it really isn't a clone anymore, so no need to try and reproduce the original.
pwatts, if u cant find a suitable port, what I have done is use 75mm PVC pipe and mounted it so the end of the port is in about 8mm from the face of the rear baffle , then filled in the rest with Body filler, then used a dremel to carve out a nice flare , cutting into the MDF at the outer edges, its worked surprisingly well and was much easier than i anticipated.
for the inside of the port ive dremel'd the edge into a small flare leading to a sharp edge right round, a flange would ahve been better but I didnt have anything handy without making way more mess cutting mdf up.
for the inside of the port ive dremel'd the edge into a small flare leading to a sharp edge right round, a flange would ahve been better but I didnt have anything handy without making way more mess cutting mdf up.
Different Port
I used a 2-1/4" port x 5" long. It is in a slightly smaller cabinet than the ProAc 2.5s and is 11-1/4" below the center of the midbass driver. I can say the bass is smooth and plays to 30-32 Hz with authority. I sold my Peerless XLS 10" subs. Did not need them as the ProAcs are more accurate.
I used a 2-1/4" port x 5" long. It is in a slightly smaller cabinet than the ProAc 2.5s and is 11-1/4" below the center of the midbass driver. I can say the bass is smooth and plays to 30-32 Hz with authority. I sold my Peerless XLS 10" subs. Did not need them as the ProAcs are more accurate.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.