Crossover Questions

[triceratops]

I've got two crossover questions that I'd like to ask for help on from some of the resident experts.

1. How does the older Klipschorn 500/5000 network (circa 1960) differ from the Type A?

2. The early 1960's Model H (Heresy) speakers were intended for center channel use and had significant roll off on the base end. One factory option was to request that the speakers be dialed in for mains use, which would presumably involve changes to the network. Later in the 1960's the cabinet was deepened to allow for a longer mid horn and the treble and mids were attenuated to bring up the bass. I have a matched pair of Model H speakers with consecutive numbers from 1963. The cabinets are the shallower original type (about 11" deep). These were obviously used as mains by the original owners, but have very modest bass response. The network is labelled 4R. Is that the right crossover to use these as mains? Any tweaks recommended?

Any help appreciated.

Best in horns,

Triceratops

[WMcD]

Page 1

[WMcD]

Page 2

[WMcD]

Page 3

[WMcD]

Sorry to do that piece meal. I zipped another article and could not upload it with a dial up modem. The bbs timed out.

In any event. This is a little known article by PWK. My guess is that what is described is the 500/5000 versus the A. But that is purely a guess.

I was not aware of any change in the horn used in the H. I'm not aware of much of early Heresy history.

My understanding is that the midrange eventually used in the K-Horn was 3 dB (107 dB) hotter than the bass or tweeter. So the autotransformer was used to knock it down 3 dB. Some would say it need even more. Of course these work out to 104 dB.

On the other hand, the bass unit in the Heresy was about 97 dB. To make it a "full range" with even output, both the tweeter and midrange would have to be attenuated via the autotransformer too. You see that in the E type network. I guess the tweeter is 7 dB down and the mid is 10 dB down (from full output) to bring them to 97 dB.

However. PWK was very concerned that a full range H would have distortion from the direct driver when it was wired as a full range. His solution was to use something like an A crossover where the mid and treble are running the same way (tweeter full out, mid down 3 dB) and the direct radiator was down from there at 97 dB.

In this way, when the H was used as a center, its midrange and tweeter were matched to the K-Horn level (efficency at 104 dB). The woofer in the H was not contributing too much output. On the other hand it was not putting out enough to be distorting. The H wired this way really didn't have much center channel bass. Most of the ear's need for a fill is in the treble.

As you point out, there was a market for a full range H. It made sense. As he said, 2/3rds of a K-Horn (almost) at less than half the price. Having two H models wouldn't make marketing sense and the guy didn't trust people enough to put in complicated switches.

I suspect that another justification for a full range was so that the H could be used as a center for two Cornwalls. The CW were running at 101 dB.

If the same logic was to be implemented with a center H for the CW, then there would be yet another adjustment necessary and another model H. (Granted the 104 dB mid and high of the center H could still be used, rather than a 101 dB H.) In a perfect world, this could have made for three model H (center for K-Horn, center for CW, and just by itself).

Now you know what I think.

Gil

[3dzapper]

Gil, thanks for another trip to the fountain of knowledge.

Rick

[JohnA]

Hey Rick,

Take out one of your 500-500 crossovers and sketch out the circuit. I and others would like to see it. There are several copies of the Type A on the Forums. I can send you a copy if need be.

Please do the same for your Model Hs and publish the driver complement.

[triceratops]

Gil and John,

Thanks for your comments and suggestions. Let's start with the Klipschorn. Attached is a drawing I adapted (crudely) from the article Gil posted. Note that C2 is two capacitors strapped together and wired in parallel. The top capacitor of the pair is labelled 4 ufd. What would be the value for a pair of these wired in parallel?

triceratops

[triceratops]

Gil,

Thanks for the article. Rather than the A network, I think the article describes the transition between the 500/5000 network and whatever came before it. I have a picture of this earlier network with three coils and no transformer, but the 500/5000 has two coils and one transformer like the "new" network in the article. The article is from late 1958 and my speakers are from 1960, so it all seems to fit chronology-wise. I know Tom Mobley's Khorns are just a few months removed from mine and I'm assuming they have 500/5000 networks too. I will try to post some pictures of these networks.

Best in horns,

triceratops

[triceratops]

Here is a partial photo of one of the 500-5000 networks. the transformer can be seen but coil L1 is out of frame. More to come.

Best,

triceratops

[JohnA]

The value of 2 caps in parallel add, so 2 fours make 8 uF.

Wow, your chronology may be correct. I always thought the 500-5000 was "first" crossover, at least for the production speakers.

[triceratops]

Hi John,

Well I don't know about the nomenclature. Maybe both styles described in the article Gil posted were called 500/5000 networks. They may have changed the design but not the name. In any case, the drawing I posted is wrong because if the cap values are additive in parallel, then C2 will be more than 4 ufd. Is it possible that caps of unequal value could be wired in parallel? If so, maybe the cap underneath is a 1 ufd to provide the 5 ufd total shown in the drawing that Gil posted.

Attached is someone else's partial photo of a transformerless early crossover like the one referred to in the article. I will try to do a more complete photo of my later crossover with transformer. I will also make a separate post on the insides of the 1963 Model H's.

Could you post a photo and diagram of a Model A network? How close is it to what I posted in the 500/5000 drawing (taking the incorrect C2 value into account)? Thanks for all your help.

Best in horns,

triceratops

[WMcD]

I kinda jumped the gun.

It was nice to have an early morning question and I had that article on the computer just ready for an upload.

The "new" crossover is close to being an A in topology. However, we have to look at the midrange. It is a passband design. The inductor in series with the driver makes it so. The A does not have it.

On the other hand, the "old" seems to me to be a passband too. My read is that it blocks lows at 12 dB per octave, and highs at 6 dB.

The article is of great interest in that that it starts talking about the merits of autotransformers. Later, PWK goes into the merits of amplifiers. He touches upon feedback and the interplay of good final stanges and poor input stages. Or vice versa. Dare I say, local versus universal feedback?

I've never looked into this amp issue with any seriousness. However, it is uncanny that the article, which didn't make it to the Klipsch Papers, touches upon a subject which is so dear to the amplifier crowd today. Again, the man was on to something from the first and some decades later it is better recognized.

Best,

Gil

[JohnA]