stevie

The 70Hz resonance has been bugging me for a few days now, because it shouldn't be there and it's causing a noticeable dip in the frequency response at the same frequency. At first I thought it was an air leak and decided to seal all the battens on the back panel. It didn't make any difference. So I tried another driver and the resonance was still there. That ruled out the driver as the cause of the problem. It had to be the cabinet. To cut a long story short, I eventually discovered that the impedance blip (or resonance) was caused by vibration of the baffle. The solution - and I'm sure Phil will include this in his drawings - is to fit a brace across the front baffle above the driver.

Here's the frequency response of the finished system with crossover.

The tuning with the standard port is in the upper 50s. The cab does works tuned to that frequency but power handling of the lowest frequencies is compromised. So I'd like to get it down to 50Hz - which means we need to lengthen the port. There are a couple of ways of doing it. The first is to buy a second port (at £1.27 or what ever not really a problem), cut it to the extra length needed, and use plastic glue (the stuff you use to glue plastic drainpipe) to glue the piece of port to the main one. Support with gaffa tape. The second way is to cut some cardboard (Cornflakes packet, for example) and tape it to the existing port to extend the length. Once I've tuned the system, I'll advise on the length of port needed.

Yes, that's about the size of it.

I've got chocolate block crossover assembled and will post the info tomorrow. It'd be interesting to see what you''ve come up with, @Downunderwonder

LspCAD and LAUD are my current tools of choice too, although I have an Arta setup ready to go as soon as I have time to learn it. I'm not sure it will add much to what I have now but at least I'll be able to use a laptop instead of an MS-dos PC.

Dunno about that @nilebodgers, It looks like a CD horn to me. B&C say this: " Hyperbolic cosine geometry combines best of exponential and CD characteristics." The reason we can get away without CD correction is that the crossover frequency is relatively high (3.5kHz) and I'm not bothered about getting a flat response above 6kHz. Still, the main thing is that we have a simple circuit that works really well and can be assembled in minutes.

Here's a diagram of the crossover. It doesn't get any simpler than this. I've assembled the circuit using a (chocolate block) terminal strip. So - no soldering needed. I'll post a frequency response of the finished system and of the assembled crossover later today. The Pulse 10 is wired directly to the input and this circuit connects to the HF unit.

You live in Germany, @Si600, which is the centre of the universe for speaker kit building IMO. Check out the next issue of Hobby Hifi, available at your friendly kiosk in about two weeks' time. It should contain a complete design for an 'updated' version of the LS3/5a using modern components. I assume you speak German. Falcoln and Wilmslow are stuck in the sixties, but IPL have some very nice transmission line designs, and Ivan knows his stuff.

Although I'm not keen on their hotter P-Bass pickups, I really rate the Duncan SPB-1, which is an accurate replica of a vintage P-Bass pickup and IMO has plenty of top end.

Fane used to use Bostik all-purpose clear glue to glue their cones. ATC used Araldite. The Bostik is easier and cheaper.

The large polepieces that work well on the Duncan Strat and Tele pickups have never really done it for me on the Precision Bass. They're louder than stock pickups but you lose the unmistakable character of a normal output P-Bass pickup like the SPB-1. The same applies to the Duncan Hot P-Bass pickup, incidentally, which sacrifices too much top end and tends to be muddy and characterless.

Wasn't this amp voted the worst amp in the 'What's the worst bass amp you've ever used?' thread?

For comparison, here's an Eminence Basslite.

I doubt very much whether they'd do that, but it wouldn't account for the differences. Apart from the 1dB difference in sensitivity, which could be down to my mic placement and is neither here nor there, the main differences are in the size of the peaks and dip between 2 and 3kHz and the fact that my curve shows a dip at 6kHz. Other than that, they are remarkably close. Peaks and dips can be minimised by using smoothing, a technique that's commonly used to make frequency response curves look a bit better. It looks like Celestion has made some effort to tame the peak at 2.5kH and have turned it from one large peak into a less worrying twin peaks and a dip. Bear in mind that this is an inexpensive chassis and that virtually all bass guitar cabs have resonances at around this frequency, most of them much worse than what we're seeing here.

I've just come across Celestion's own published measurement of the Pulse 10 chassis. It's interesting to compare it with the one I took, as it looks like it's been smoothed slightly. I also measured 1dB more sensitivity.

I decided to do a quick crossover hookup yesterday evening and managed to sort it out in an hour. You'll need three, cheap components, which is just what we wanted. The cab sounds surprisingly good too. More info and measurements to follow.

Sorry for the confusion with the graphs. It's not exactly easy to delete and replace images (although I finally managed to figure it out).

Here are a few more measurements. The first is the Pulse 10 at one metre. Unfortunately, I didn't calibrate the measurement, but from 200Hz to 1,7kHz it's 96dB. Between those frequencies it's really smooth - better than much more expensive drivers in fact. You can see the effect of the cone breakup between 2 and 3kHz. Not brilliant, but not as bad as your average Eminence. The compression driver (second graph) works well with the horn - thank goodness. Celestion say it can be used down to 2.5kHz, which is really good for such an inexpensive HF unit, but our crossover frequency here will be dictated by the bass unit rolloff, which is higher than that. Crossing over at 2.5kHz would also require a more complex crossover than I'm aiming for here.

Phil's working on the cabinet assembly instructions at the moment and will post them as soon as he's sorted. Once I've retuned the cab, located the cause of the 60Hz resonance and measured the bottom end response, I should be able to sign off the tweeterless cab design. Then you can build to your heart's content. The two-way version will obviously take longer.

I've been a bit busy with work recently, which is slowing the project up, but I managed to fit the drivers to Phil's cab and measure the in-cab impedance of the bass driver this morning. The tuning frequency with our 100mm diameter x 120mm port is around 58Hz. We need to get it lower and I've already planned for that. Those of you familiar with WinISD might like to enter the parameters of the Celestion Pulse 10 to see what's happening. The impedance curve is a useful troubleshooting tool, which tells you a lot more than just the load the driver is placing on the amp. Notice the bump just above 2kHz, which was there when I measured the impedance in free air. That's the main breakup point of the cone. There's likely to be some irregularity in the response here, depending on how well the manufacturer has managed to damp the resonance - if at all. There's also a faint bump at about 350Hz. That's caused by the vertical standing wave inside the cab. It looks better than I'm used to. The most interesting - and unusual - bump is at 60Hz. I have a good idea what that is but will have to troubleshoot it. Any suggestions?

That question opens up a big can of worms, @martyy. Yes, at a pinch, you could use it as a PA speaker. However, cabs used for bass guitar normally take into account the extra bass response you get from using the cab on the ground. So it will sound bass-light when used on a pole in the air and you'd need to eq it accordingly. The design goals of this project are not the same as a PA speaker, where the midrange is the focus of your attention. Remember also that this is a budget project. The Basschat Mk III cab was designed with no compromises in the midrange (expensive compression driver, low crossover frequency, good system power response) and works really well as a floor monitor or a PA speaker (in the latter case with the necessary LF eq). In fact, before lockdown, a group of us compared it with an active RCF cab and agreed that its reproduction of voice was superior. I'll be testing the finished design using all kinds of material, including full range audio - and I'll report back.

Sounds like an interesting project. You could probably keep the cab the same size, as the few litres that the amp would take up won't be a problem. You'd just have to alter the port length slightly.

You'd have to watch for vibration and air leaks. Otherwise, it might be worth a try.

One of those grilles would work fine. It's a cheap solution and protects the cone, but does look a bit home made. The round grilles made from punched metal, i.e. with small, round holes, look a lot better (if you can find them) but are more expensive. There's enough space on the baffle to fit a picture frame for a cloth grille. Then there are the metal grilles from Ebay mentioned above. I had real problems getting the etch primer to stick on the aluminium one - I tried two different brands. Although heavier, steel grilles are easier to finish with the type of spray paint that goes straight onto metal (Hammerite, Platikote and the German discounters' specials when they're available).