And then finally - experiment with crossover filter modifications that could potentially improve the overall performance
Update on my JBL L100 Rebuild.
- JBL L100 New crossover design with modern components Made for these drivers only: 123A bass units, LE5-2 Midrange drivers, LE20, LE25 or LE26 tweeters.
- JBL Century L100 Restoration. 4.2015 - By Greg Roberts Page Four. After the finish cured for a few days, I began working on the crossovers. I brought one of the speakers to my listening room at home, and set.
- As mentioned, the L-100 has a few weak spots, some of which decreased over the years with new versions of the L-100, other models and other drivers. Commonly claimed weak spots of the L-100 are: The JBL.
The original crossovers are not really crossovers at all. They consist of two capacitors, one for the mid and one for the tweeter, and two potentiometers to control the output of the mid and the tweeter. The midrange capacitor is a 8uFd and the tweeter a 13uFd, which at first glance is the opposite of what you would expect. Normally the mid would have a larger value hi-pass cap. The mid has no low-pass on the upper end (bandpass), and the woofer has no filter on it at all. Not a conventional crossover to say the least.
I'm not going to publish too many of the tests here, nor will I describe the testing procedure in detail. Instead I will boil it down and give you the end results. This whole procedure took more than twenty hours, over a four day period.
One thing to note - the components used in these speakers sound a lot better than they look in tests. I wasn't surprised by this, as it is consistent with my experiences where a good test does not necessarily correlate to a good sound, and vice versa.
Here's a simple test showing the frequency response on the woofer with no filter. I should note here that the frequency response of my testing in this room will not be accurate in the low bass. To accurately test bass response, it takes a much more sophisticated setup than I have available to me. Nonetheless, this is still useful information, and something that I can work with.
Here's a test showing the midrange frequency response with a 8uFd new capacitor, no attenuation.
Here's the tweeter with a 13uFd new capacitor, no attenuation.
Here are the three components together.
I was surprised to find that there was a LOT of overlap between the mid and the tweeter. In fact, using the original crossover design, the tweeter has more output at 900Hz than the midrange. Another interesting note - the midrange and tweeter tested exactly the same with the new caps as they did with the old caps. However, the speaker sounded better with the new caps and L-pad resistors (same design as the old crossover) than it did with the old crossover.
I began calculating and trying different filters on each of the components in an attempt to smooth out the response, and to optimize the bandwidth of each. After many hours of work, I had some very good looking test results. Smoother response, a better low-pass on the woofer, a better hi-pass on the tweeter - some very nice looking slopes that cross-over one another exactly how you'd draw it up. I had a test crossover built with various components sitting on top of the test speaker, alligator clips holding everything together.
And how did it sound with music? Worse than the old original crossovers!
Just too many crossover parts I thought, but actually, after some investigation, I figured out that it was really the midrange that I had screwed up by pushing it too low, causing an increase in distortion. I'm giving you the short version here. There were hours and hours of trial and error, testing and listening - all coming back to the fact that the JBL engineers had it pretty well sorted out to begin with.
So it was day four of working on the crossovers, and I had a huge amount of data from all my trials, and well over 100 tests. What to do with all of it? I kept working on it through the day and finally ended up with two small changes to the crossover design that made improvements both in tests and in listening. The first change rolls off the objectionable 900Hz peak in the tweeter, leaving it sounding more like a tweeter and less like another midrange. The second change smooths out the upper midrange peak in the midrange driver, flattening the response and taming the sharp, edginess that I detected in the upper midrange.
Here's the new crossover
The new crossover is certainly an improvement over the old, and all it took was a couple of small changes and some new parts. Gone are the tiny, thread-sized wires on the original potentiometers that carried the entire mid and tweeter signals, gone are the terrible connections inside the pots between two corroded plates scraping together, gone are the old caps.
The new crossover gives us new caps, nice connections throughout, nice wiring, L-pad resistors for attenuation rather than the pots, and a couple of small design changes that help smooth out the response.
To mount the crossovers in the cabinet, I chose to drill one hole in the back center, where a bolt would be inserted through the back and through the crossover board to hold it in place. I also drilled out the binding post holes to accept the new binding posts.
New wiring
Resistor L-pad modules to set different levels for the mid and tweet.
Here they are with the components installed and new foam rings for the tweeters
And with the Cane-style grill cloth
8/14 More finished photos to come
13 hours labor to put cloth on frames, build crossovers, install components, install new binding posts, install crossovers, install foam rings, make new wiring. Materials for grill cloth, crossovers, binding posts, wiring, foam rings $158.
Add 1 hour for packing the speakers and $10 in packing materials
Project cost totals
20 hours crossover design time - no charge
Labor is 41:50 - adjust to 39 hours @ $60 per hour = $2,340
Materials $343
Total = $2,683
Plus shipping cost TBD
You Are On Page Four
Back To Page Three
Back To Page Two
JBL Century L100 Restoration
Jbl L100 Crossover
4.2015 - By Greg Roberts
Page Two
The first thing I did was to remove the drivers and store them away. Here's a picture of one of the beautifully made woofers.
Tweeters and mid drivers
Crossovers, if you can call them crossovers. Two caps, one for the mid and one for the tweeter. Two potentiometers for controlling the output level of the tweeter and mid. No components on the woofers - they run full range. I think there's room for improvement here.
Crossover and plate from the front of the speaker
The thin alumimium plates came off easily with a heat gun and putty knife
One hour to remove components.
Next up - cabinet repairs. The cabinets are sanded with a random orbit sander using 80 and 120 grit discs. This 'cleans up' the surface of the cabinets and gives me a chance to inspect and find all the trouble spots. I remove any loose veneer or crumbling MDF with a putty knife, exposing holes and raw spots that will need to be filled and sanded. I use epoxy filler like this for the holes
I use a lower viscosity epoxy glue for loose pieces that I want to glue in place before filling. On the backs of these speakers, there's a recessed area that leaves a lip, or reveal all around the back. Damage to this inside lip area can be very time consuming to fix, so I prefer to keep that lip intact and repair with glue, rather than digging it out and filling with epoxy. This saves a lot of time and money. In this picture you'll see where I've used the epoxy glue in a piece of the lip that had cracked and was loose, clamping it in place to set up before filling with the heavier epoxy filler.
I also use the lower viscosity epoxy to fill in loose seams at the corners of the speakers. Note in this picture how the joinery at the seams is not simply a 45 bevel, but there is a tongue and groove part as well.
If I can get the putty knife more than 1/4' into the seam, I use the lighter epoxy and push it down into the seam to re-glue it
Here's a bad spot on the inside of the lip on the back
Another long crack on the lip - lighter epoxy, clamp to hold, then fill with epoxy filler
The repairs look scary sometimes, but in the end, these will sand up nicely and provide a flat, solid area for the veneer to hold to
Here they are, initial sanding done, initial repairs done - 2 hours 15 minutes for this work
I waited a few days before sanding the epoxy, to let it cure really good. Normally you can sand this stuff in 24 hours, but since I put such a thick layer on, and because I had other projects to work on anyway, I figured I'd give it some extra time.
These images show the first coat repairs sanded. Note the voids and rounded corners which will need an additional coat of epoxy to bring perfectly flat and sharp.
On the back edges, I tried to make repairs to the broken areas that would NOT interfere with the inside revealed edge of the back. Once that inside edge is damaged, it takes a lot more time to repair. Here's that really bad back edge damage that was repaired without getting into the inside reveal. Had I broken that piece out, this five minute repair would have turned into a half hour repair.
This next series of photos shows that one back edge reveal damage that needed repair. Very light filler to start, so the sanding is easier. Using a hand sanding block with 80 grit paper. Sanding very carefully up to the edge of the tape. Another light coat of filler. I'm hoping just one more coat after this will do it.
I got to talk to my customer on the phone today, and we made some decisions about this project. He really liked how the Sapele/cane grill cloth looked on the JBL restoration that I did many years ago. In fact he watched that restoration back when I did it, and has kept track of my work ever since, just waiting for the right time to have his speakers done. We've decided on Rosewood veneer with cane cloth for his speakers. I will be adding trim to the front of these cabinets and in-setting the grills as I did with the older restoration.
By the way, my customer mentioned that he was reading the comments on the Lansing Heritage Forum thread about this project, and he told me that this was definitely a labor of love for him. He's had these speakers for many years, through college and everywhere in between, and so he's very happy to finally be able to bring these speakers back to new, or even better than new condition. He'll be enjoying these for many years to come. He knows he could buy L100 speakers in decent condition and for a lot less money than this restoration is going to cost, but they wouldn't be 'his' speakers, and they wouldn't have Rosewood veneer, cane grills, wood trim on the fronts, upgraded crossovers, and be in perfect condition. He feels the money spent for these things will be money well spent. This is exactly why I'm in this business, and I'm happy to be offering this level of service to my customers.
So, since we're adding trim and using different grills, and we're not going to be using the potentiometers on the crossovers anymore, many of the holes in the fronts of the speakers need to be filled in.
One hour for sanding and second coat epoxy
Corners are sharpening up nicely
Small third coat of epoxy is on
I cut the trim pieces for the fronts out of MDF.
One hour for sanding, third coat epoxy and cutting trim for fronts
Jbl L100 Crossover Kit
I know this project so far is a bit tedious, but it will start to get more exciting very soon as we get into the veneering process. So just hold on, more to come!
You Are On Page Two
Jbl L100 Crossover Diagram Pdf
Go To Page Three