Active Speaker

[Terry]

As for Sinclair amplifiers, this may be of interest ...

Looking at it again nearly nine years later, I've just spotted the typo!

[Mark Hennessy]

Ah yes - I recognise that tweeter! I have not got around to measuring them, but it's on my list.

On the original design that they come from, a pair of them was wired in parallel, and that combination was fed via a single 1uF capacitor. That would have resulted in a fairly high crossover frequency - impossible to calculate without seeing a plot of impedance vs frequency. Running a single one via 1u8 will result in a lower frequency, of course. But the frequency you need is going to depend on what the large drive unit is doing. In the absence of measuring gear, then trial and error is perfectly OK.

Are you not tempted to add a second Sinclair amp and try an active design? If nothing else, it would be a fun experience!

Yes, you can change the caps to alter the turnover frequencies. Of course, the Rotel design is a hi-fi amp, and so the turnover frequencies will be closer to the ends of the spectrum than might be the case for a non-hi-fi design. Increasing C514 is - as you've found - the way to bring down the turnover frequency to make it more appropriate for the context.

Noting the reversed treble control in the test hook-up, do bear in mind that it's usual to use logarithmic pots in passive Baxandall designs, so if that's what you have there, you might want to try reversing it to double-check, else the action (and response when at centre) will be different when you build the final circuit. I haven't actually looked inside the amp to verify if that's what Rotel used.

Hope this helps,

Mark

[Crackle]

I remember constantly having to change the output  transistors  (I think they were OC26) on my Z12 amps.

Hi Mark
I wonder if you could confirm, with the tweeter connected via a single capacitor and no chokes elsewhere in the speaker circuit, should the tweeter be connected in phase with the bass speaker.
Something in the back of my mind tells me the tweeter should be connected in opposite phase, or is that just in a xover with inductors and capacitors.

Are you not tempted to add a second Sinclair amp and try an active design? If nothing else, it would be a fun experience!

No I just want to keep it simple.

Thanks
Mike

[Crackle]

I have now more or less finished the control panel, I just have the mains socket, switch and fuse to wire up. Then I can put the back on and see what it sounds like.
My tests so far seem to indicate it goes a bit boomy with the back a loose fit.
If that's the case I will probably have to cut some large vents in the back panel. The original back would have been well vented with a cloth covering to stop the ingress of dust.
You can see what an original back would have looked like here.
http://www.kbmuseum.org.uk/kb_images/kb72/kb72.htm

I am not too sure about Baxendall tone controls needing to be logarithmic, I bought some new pots from Rapid, 2 x 22k log for the tone controls and 1 x 100k log for a volume control to help match the active speaker to the source of the audio.

Mike

[Mark Hennessy]

In general, if you want a flat frequency response when the tone controls are in the centre, and also wish to have symmetric lift and cut, then yes, the "passive Baxandall" requires logarithmic pots.

Confirmation: http://www.duncanamps.com/technical/tonestack.html - or: http://sound.whsites.net/articles/eq.htm

Interestingly, it's not correct to call the passive arrangement "passive Baxandall" - we should be calling it "James" - but unfortunately for EJ James, "passive Baxandall" is the common usage. Searching for "James tone control" does bring up a decent number of results, though. I will make an effort to use the right term in future.

Anyway, if you had bought linear pots, then the circuit would have provided considerable bass and treble boost when the controls were set to the centre, and there would have been much more cut available than boost. Buying log pots was the right thing to do here.

The active Baxandall circuit uses linear pots. These are often easier/cheaper to source, and for stereo applications the matching between the two sections is usually better with linear pots, so that's just a couple of reasons why manufacturers like the active scheme. No insertion loss either.

The input impedance will depend on the setting of the volume control and the tone controls. At very low volume settings, it will be close to 100k. It will fall to perhaps less than 10k as you turn up the volume to maximum. I haven't done any calculations, and it's probably not a huge problem, but just be aware of it. If a high input impedance is definitely required, then a buffer after the volume pot is a good idea. Even just an emitter follower might well do the job. We can cross that bridge if needed once you've used it for a while (all the best projects are never finished - they always evolve in light of experience).

Without knowing the TS parameters of that drive unit, it's hard to give a precise answer about the back panel. My best guess would be along the lines of "it was probably originally intended to fit in a valve radio with a well-ventilated rear panel, so if you put it in a sealed box, the Q would be too high, and it will "honk" at the bass end". I'd try a sealed box (with some stuffing) first of all, and compare it with and without the back, and perhaps slide it around to give different openings. In terms of LF, the original rear panel has such large openings that they would note have been "tuned" in any way - anything over perhaps 2 or 3 square inches probably falls into the same category, so there's plenty of latitude here.

[Crackle]

Hi Mark
Thanks for your reply and help.
The tone controls seem to do nothing from left to centre, then from centre to full right you get an increase of bass or treble. They were sold as log and marked "A".

With the back fitted it does make the speaker sound flat, it seems to emphasize the low mids, like a hollow box, which oddly enough is what it is.
It sounds very nice with the back removed. I will look for some foam to shove in there to see what effect that has.
But at the moment it looks as if;
A, it is the wrong type of bass speaker for the cabinet
B, the cabinet is made of the wrong wood, oak is too solid.
C, I may have to cut a new speaker baffle and fit 4 of those 4 inch speakers. They would be able to handle the full power of the Z30 so a change of PSU could be done.
D, I may just leave it as it is. and forget the project.

[Mark Hennessy]

Simple way to check if log or lin:

1. Rotate pot to mechanical centre.
2. Measure resistance between wiper and the anticlockwise end of the track.

If the resistance measurement is about half of the value, then it's lin. If it's about 10% of the value, then it's log. If it's about 90% of the value, then it's anti-log (or - quite likely in my case - you mis-identifed the anticlockwise end )

Most often a pot marked "A" is log (A for audio), but sometimes "A" means linear - it just depends on the manufacturer. So I measure them and label them accordingly.

There is a chance that Rotel designed the circuit to work with lin pots - as I said earlier, I haven't looked inside - if so, simply copy one of the many other circuits out there - they are much the same apart from a few component value changes. Testing and experimentation is to be expected in any project - I never attempt to copy something with the expectation of it working perfectly for my needs first time.

Regarding the speaker/box, I really doubt the cabinet material is making a significant contribution when you look at all the other factors at play here. There's an excellent BBC paper available on the R&D website where they compared wood types for cabinets if you fancy a bit of bedtime reading: http://downloads.bbc.co.uk/rd/pubs/reports/1977-03.pdf

Far more important is the TS parameters of the drive unit, and the size of the box, and whether it's sealed, or whether there's a tuned opening like a port (which could be just a simple hole rather than a length of pipe). I won't go into all that as you probably know about it all, but suffice to say, it's not difficult to measure the relevant TS parameters of the drive unit, and with the help of something like WinISD, model the performances of that driver in that cabinet. Just depends on how keen you're feeling.

If you're thinking about subbing in those 4" bass drivers, I'd just fit a single one. It might seem wrong given that you'd be able to fit several in there, but multiple drivers won't allow you to go any lower in the bass end. Yes, they would handle more power, but a single one is easily good for 20 or 30 watts. Also, if you put several drive units together, you get an effect called "lobing", which means the frequency response in the midrange-treble region will vary enormously as you move your head around relative to the cabinet. It's horrible.

At the risk of rambling, I ran into exactly this when I started helping out at the local arts centre. On each side, they have 2 cabinets (Martin Audio AQ10), each containing a 10" woofer and a horn tweeter. They are side-by-side (touching), and driven in parallel. As you walk along the rows of seats, you hear a completely different sound as a result. So I unplugged one on each side, and that effect has gone away completely. And one per side is more than loud enough, so the other cabs are just sitting there as "hot spares". Because I know a lot of people would have a hard time understanding/believing me, I took a whole series of measurements at each seat in a row, and have a written report standing by for whoever is first to notice that I've unplugged some of the speakers... I'd love to meet the saleman who recommended this setup!

Anyway, from what I've seen of those 4" drivers, they will definitely have a smoother response through the mid and will extend further at the treble end, and I wouldn't be surprised if they go lower in the bass - knowing that old radio speakers have very stiff suspension and a high Q as they are designed to be used in open cabinets - even quite large vintage speakers can resonate at more than 100Hz. The box is probably a bit on the large side for those 4" units, but that really won't make a huge difference as you only have a couple of watts to drive it. Seal the back and bung a bit of damping in there to stop reflections causing a midrange peak (which you've already heard). Be aware that a lot of foam has practically no effect acoustically - you need the right stuff. Alternatively, get some wadding from some old speakers - the white BAF is OK, but many alternatives exist.

In conclusion, the scope for experimentation is endless, and I would encourage anyone to play around for as long as they're having fun and learning stuff. If it's becoming boring or feels like a chore, then cut your losses and go with what you've achieved so far, but always consider re-visiting the design once you've listened to it over a few weeks or months. I really love playing around with speakers. I've got a design on the go here which has been brewing for months - every now and again I'll get the measurement mic out and try a new crossover idea, then listen to it while I'm working on other things for a few weeks, then repeat. Passive crossover design is just about one of the hardest things I've done - the circuits might look simple, but they are extremely complex in their operation - and I enjoy the challenge. Active speakers - where there's no interaction between the crossover filters and the complex impedance of the drive units - is too easy!

Enjoy

[Crackle]

Hi Mark
I removed one of the tone control pots just to check, and measuring from what I would know as the hot end of the track is 18k and from the earthy end of the track which I believe is the same as your anticlockwise end is 3k. So it would appear to be log law.
       

I have not as yet tried any foam in the cavity to see if it improves the sound when the back is on.

Mike

[Crackle]

I have cut a piece of 4 inch thick foam to the size of the inside of the cabinet.
It makes a subtle difference with just the foam, but there is still a worsening of the fidelity when the back is placed over it.
Even with a 2 inch slot at the top it still sounds flat and uninteresting.
I will wait till my son returns home from work, his musically trained ears are better at mine at distinguishing what is wrong and missing with audio.
But I reckon that an open back is the way to go with this 10 inch driver.
The problem is probably something to do with the large thin paper diaphragm and it not coping well with the added air resistance caused by closing the back.

Mike

[Mark Hennessy]

A honk in the mid-bass (100-200Hz) will be because of the high Q of that driver, as explained earlier. An open back is likely to be the best bet for that particular speaker - I doubt its TS parameters will be ideal for either sealed or ported applications.

Mid-range peaks are likely down to reflection within the box - the damping material will help a lot with that.