Setting up a ESP active crossover with gainclone

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hi guys,

I'm after a little advice please,

I've just finished one channel of my active bi-amp gainclone setup.
I have built the ESP LR crossover in 2 way mode and am feeding it into 4 peter daniel premium LM3875's (dual 25V trafo 300VA into 6Ohm speakers)

Now, I have tested the gainclones as monoblocks without any active crossover and was really impressed by the sound.
I also fed the crossover into my 6ch surround amplfier (using separate speaker for L hi L low etc) which also impressed me (absolutly no noticeable coloration or hiss hum etc.)

Naturally when it came to testing the two together I expected to be blown away, but ermm wasnt!

Sounds good (cant really be more precise at the mo) but not quite what I think it can do .

I notice the ajustable level for each output on the LR crossover could someone point me in the direction as how to set it up both simple and also correctly (doing it 'by ear' doesnt seem to be great, I cant decide on a particular level)

I havent yet removed the existing crossovers (but have tested the channel with two separate speakers i.e low output fed into whole crossover on speaker so not to feed 300Hz into 2Khz Xover etc)

Or is it just me rushing it and need to wait until I finish the other chanel to be truly amazed?

Thanks

Dan
 
Disabled Account
Joined 2007
Dan_Steele said:

I havent yet removed the existing crossovers (but have tested the channel with two separate speakers i.e low output fed into whole crossover on speaker so not to feed 300Hz into 2Khz Xover etc)

Hi,
I'm doing a similar project:
http://www.diyaudio.com/forums/showthread.php?s=&threadid=104118

You probably will get better results with the crossover out. What are the speakers?
You mention the low frequency, but what did you do with the high? Letting these signals go through a passive crossover will show no impovement., about the same as bi-amping.
 
Speakers are Jamo E800 series, I have both the E800 bookshelf and E850 floorstanders. Cost around £1000 for the 4 about 3 years ago.

The set up I have for testing at the moment involves sitting the bookshelf ontop of the floorstander and leaving the bi-wire links in. The putting the low into the floor stander and high into the bookshelf.

I think I will have a go at removing the crossover today in one of the bookshelf's( I initally left them in as I figured they would serve some kind of protection for the speakers during testing)
 
Bit of an update:

Removed the crossover from the bookshelf..

Was a bit nervous at first, feeding 300Hz into a tweeter didnt seem right in my newbie(ish) books at first, also considering I havent implemented any delay in the amp outputs (maunaly connect after a few seconds) or speaker protection circuitry.

Still seemed a little under what i expected, but then I connected my heavily modded playstation (linear PSU, new output from dac etc etc)

Anyway, it now sings!!! Little bit of hum (seems dependant on where the phono input cable to the amps is located) but im sure I'll figure it out.

Would still be nice to have some kind of numerical setup on the crossover tho.

Regards

Dan
 
Ex-Moderator
Joined 2005
Hold on there Dan. You're feeding frequencies as low as 300Hz directly into a tweeter? You shouldn't be doing that!

On the ESP site the 3 way active crossover is meant to be used with a 3 way system - everything below 300Hz is sent to the woofer amp, 300Hz-3kHz is sent to the midrange amp, and 3kHz on up is sent to the tweeter amp. If you're doing a two way with a midrange and a tweeter you would send everything below 3Khz to the midrange and everything above 3kHz to the tweeter. Most tweeters are not meant to reproduce frequencies below 1-2kHz.
 
hmm yeah kinda what i thought! I will have a look tomoro to see if i can recalcalatlate the components. Is 3khz a generally accepted value? Can see me having problems getting info regarding the tweeter & optimum frequency response being a purchased speaker and all. It sounds so good tho (hopefuly indication that no damage has occured )

Dan
 
Ex-Moderator
Joined 2005
Most commercial two way designs with a ~1" tweeter and a 5.25-6" mid/woofer have a crossover frequency in the 1-3kHz range. You can use your computer as a source and generate test tones while only driving one driver at a time to try and determine by ear the approximate crossover frequency for your Jamo E800's.
 
Disabled Account
Joined 2007
BWRX said:
Hold on there Dan. You're feeding frequencies as low as 300Hz directly into a tweeter? You shouldn't be doing that!

On the ESP site...


Dan seems to be following what is found on the ESP site for two-way. A cross at 310Hz.

Quote:
"Figure 1A shows the general concept for a full stereo version, with two identical filter sections (but without balanced inputs). With the component values shown, these have a crossover frequency of 310Hz (refer to the article on Bi-Amping to see the reason for my choice of frequency). "

A mistake?
 
Ex-Moderator
Joined 2005
It's not a mistake. Have a look at his article on bi-amping and check out section 1.0. The first figure tells all. When he mentions a two-way crossover at 310Hz he intends it to be used to separate low frequencies for a woofer amp and mid/high frequencies for a mid/tweeter amp. His diagram clearly shows a passive crossover to split the frequencies for the midrange and tweeter. He explains why he chose to do that further down the article. If you're going to bi-amp a midrange and tweeter you will need to choose a higher crossover frequency. You run a high risk of frying a tweeter if you cross it over at 300Hz!
 
Disabled Account
Joined 2007
Yes, I read through the article again. Thing is, he should be clearer on the project page. A two way active filter sounds like it is for a two way speaker.
At the end of the bi-amping article he does provide a handy calculator for the crosover frequency component values.
A nominal 2500Hz should be good for most 2 ways.
 
Thanks MJL21193, reassuring to know its wasnt just me who read it like that.

Right I have got some 1k resistors on order now, coupled with the 47nf capacitors already installed (this way I only have to repalce half the components) gives a calcualted Xover frequency of 2394Hz, which if I had come up with in the first place would of made me a little more confident!

I guess then that some of the informaiton between 300Hz and around 2KHz was missing then when I was doing initial listening tests (only way of explaining how the tweeter didnt jump out of the cabinet?) which means I have even more listening pleasure to come!

On a side note, about the hum, I have found if I give the monoblock enclosure a tap, the noise builds up quite considerably and then dies down again a few seconds later, as simple as a dry joint?

And I thought I was almost finished!

Beleive it or not started this project as a little interim whilst I was waiting for the granite in my tube amp to be machined!
 
Dan_Steele said:
.........I havent yet removed the existing crossovers (but have tested the channel with two separate speakers i.e low output fed into whole crossover on speaker so not to feed 300Hz into 2Khz Xover etc)
you were lucky that you kept the passive crossover in the feed to the tweeter.
Your active rolled off the lower signal at ~300Hz and fed it to a crossover set at ~3kHz. The bass/mid was playing all the signals below 300Hz at correct level.
The frequencies above ~300Hz were sent to the crossover and it filtered out the information below ~3kHz, with the result that the treble driver played the signals above 3kHz at the correct level.

You were only missing the information between 300Hz and 3kHz. This is the band where all intelligible speech is and much of our musical info is in this band as well.

Yes, the project would seem a little flat played like this but, it saved your tweeter.


It might be worth retaining a big cap in the tweeter feed to protect it from faults coming from the amp.
Aim for the cap to roll-off about two octaves below the active filter to preserve much of the phase just below the passband.

You may still find that a simple two way filter sends the wrong information to the two drivers. The passive crossover may also equalise the signals sent to the drivers and you might find that the sound is not quite right, at best, or pretty damn terrible if the drivers are heavily equalised by the passive crossover. Be prepared for further research and much experimentation
 
Hi andrew,

you were lucky that you kept the passive crossover in the feed to the tweeter.

Well thats what I did first of all, but then removed it alotgether and fed the 300hz into the tweeter with nothing in between (yes it seems stupid now!) - But nothing bad 'seems' to have happnened, the tweet still sounded as crisp as ever - i will put the pass xover back in to verify this ~ I think (and hope!)I may have been lucky!

With regards to implementing tweeter protection with the cap, can you point me in the direction of what this is refered to as so I can do some reading on the topic to better understand how to implement it - calculating values etc. It would cetainly help me sleep easier, even tho I connect the tweet after the amp is powered on, I do flinch slightly just incase something untoward has happened since the last time it was powered up.

Nordic, motoboating always makes me giggle - do a search for its definition of urbandictionary.com !
 
Hi,
the series cap is effectively a high pass filter, but it's purpose is to filter out DC from the amp.
Look up single pole passive filter.
The turn-over frequency (=roll-off = F-3db) is set to about 800Hz if the tweeter is doing duty from 3kHz upwards.
25uF is about right for an 8ohm driver. 22uF or 27uF could also be used. It's not a passband filter (that's in the active crossover) so absolute accuracy is not required.
 
Thanks andrew for your help,

I have a few 22uf solens left over from the tube amp (630V met. polyprop) will insert one of those and listen out for any degrading in the sound.

If its a 4 Ohm driver (will measure tonight) what are the likely if any effects of the 1800Hz roll off as opposed to the suggested 800Hz
 
I think 1800Hz is too close to the 3000Hz of the active filter.
The two filters start to add together and the carefully engineered phase and attenuation built into the active crossover have been defeated.

However, there is only one way to KNOW. Try it.

Alternatively, for a 4ohm driver just parallel a second protection capacitor i.e use 50uF.

Have you found the formula yet?
F=1/2/Pi/R/C
F= frequency in Hertz, R=impedance in ohms (at the frequency being considered), C=capacitance in Farads.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.