Direct driving Apogee Full Range speakers

Hi
I've owned several Apogee's off and on over the years and now have a pair o the Full Range. Even in their stock configuration they are impressive speakers. But i know they can do better with active 3-way XO and direct driving the ribbons. I think i have sufficient amps for the job: 2 pr of Krell KRS 200 for the bass and MR and Clayton S40 for the TW. the KRS should handle the bass directly no problem and i think(?) the Clayton can handle the TW directly as well. the big question, as always, is the MR. I am very concerned about damaging the ribbons and do not want to run the MR directly despite the KRS probably ability to handle it. My question is whether to use a series resistor on th MR (and maybe the TW?) to raise the resistance up to approx 1 ohm, OR use a Speltz Zeroformer to do the same? I view these options as being sonicaly superior to using a full on transformer. Yes? So i would greatly appreciate some feedback/guidance on these alternatives, or some other approach?

Lastly, i repeat my concern for ribbon safety - will either the series resistor or the Zerofromer provide any protection for the MR and TW ribbons? the very last think i want to do is re-ribbon these speakers. Would adding in-line fast-blo fuse be advisable?
thanks!
 
Putting a 1 or 2 ohm non-inductive resistor in series with both the Midrange and Tweeter ribbon will have very little negative sonic effect and should protect them from damage. Matching SPL levels can be done at the active Xover.


To protect the ribbons you want to avoid any amplifier clipping, plus all power-up and power-down DC or AC THUMPS. To protect the ribbons you want to keep low frequency BOUNCE off the midrange and tweeter ribbons. The original Apogee Xover included a series capacitor to protect the delicate tweeter ribbons. When the tweeter ribbons are directly connected to a large amplifier a small series resistor provides some damage insurance.

To protect the midrange and tweeter ribbons from stretching due to low frequency BOUNCE you want to use crossover slopes much steeper than the 6db/octave in the original design. You can get 18db/octave from a single amplifier stage. Discrete ClassA active crossovers, like the Krell KRX-1, provide a modest number of amplifiers for the crossover function such that 18db/octave is often used for 3-way speakers. The popular LR4 and LR8 steep slopes are typically constructed from IC opamps. Anything is possible with digital Xover.


Over the Full Range production lifetime Apogee made a few minor changes to the construction of the Bass Panel, midrange ribbon, and tweeter ribbon which affected their resistance, efficiency, and resonance.
MY SPEAKERS:
Bass Panel = 1.86 ohms
Midrange Ribbon = 0.14 ohms
Tweeter Ribbon = 0.88 ohms
 
thanks Line Source - very much appreciated. a couple quick follow up questions:

what current rating should i get on the resistors? Was thinking 5 amp.

What exactly is low frequency bounce?

i currently use a Pioneer D-23 crossover in 2-way mode at 6 db slope (replaces stock passive XO). i plan to use it in 3-way mode to tri-amp. It does have the ability to run 18 db slopes, but wont i need to do something about phase shift around the XO points if i go 18 db?

Thanks again. Love the FR!
 

george a

Member
2002-05-06 1:12 pm
oz
I have to chime in here.

Do NOT overlook any DC offset at the amp output - this will move a direct coupled speaker in one direction or another (in or out) possibly more so with a ribbon which may offer little resistance to the force.
This could stretch your tweet ribbons in particular .
I raised the issue (unsuccessfully) of "wobbly' low freq output driving my direct coupled ribbons by my diy pas a5.
i was using a series resistor att he time and ended up going back to transformer coupled to avoid dc offset in particular..

Cheers
 
thanks George. Ya, i will b measuring the dc offset today and see where i come out. a friend pof mine had FR's and was direct driving using tube amps with autoformers. every time he turned his amps on/off the MR and TW ribbons flapped around. any idea how much offset is acceptable?
 
HI guys
i measured my amps for dc offset and they ranged from 1.5 mv to 25mv. Would obviously want to use the amps with higher offset on the bass panels. i also meausered the ribbons:
bass - 2.2 ohms
MR - .5-.6 ohms
TW - 1.2 ohms
These measurements were including the pigtail wires.

So i am thinking of padding the MR with 1.5 ohm resistor and the TW with 1 ohm. That way amps will all see approx 2 ohms. I do not know whether the dc offset will be an issue when using the series resistors or not. Does the power rating of the resistor have any impact on both sonics and level of 'protection'. I.e would a 5w resistor be better/worse than 20w? My listening levels are such that peaks are in the low 80's db. If these speakers are as inefficient as has oft been reported then i should think 5w might be too low. Would appreciate your feedback.

Thanks!
 
Padding the ribbons up to a 2 ohm load is a good idea. You can always experiment with 1 ohm loads after safe operation has been proven over many hours.

I would purchase 25W or 50W non-inductive resistors from Caddock or Dale. Resistors this size often require some heat sinking for full power, so read the specs on free-air vs. chassis mounted power rating.
==========


Starting with just the resistors on the amp outputs would allow you to test the amplifier chain for poor servo DC, or DC/AC turn on/off transients under load.


I use the term ribbon BOUNCE for the large physical midrange ribbon movement when a deep bass transient is run through a shallow 6db slope Xover. The tweeter ribbon also shows some bounce off midbass transients. Transient ribbon bounce will quickly stretch out the ribbon pleats, requiring taking off the speaker grill_sock and re-tensioning the ribbon at the bottom clamp point. This will change the ribbon resonant frequency. If possible, I would start with 18db or 24db slope Xovers. The ribbons are resistive loads, so you will get very good transient response from these higher order slopes.

When your amp can effortly push 2,000 watts into 1 ohm
Be Afraid!
Be Very Afraid!

P.S.
Your measured resistance values are higher than "typically reported" on this forum. You may want to log onto the Apogee forum and ask if the Apogee factory started to use thinner Al foil on late model production.
bass - 2.2 ohms
MR - .5-.6 ohms
TW - 1.2 ohms
 
thank you Line Source.
i have used both pr of KRS on my Scintillas (1 ohm version) and never noticed any transients from turning on/off. But the Scintillas do have capacitors in the passive XO so that may have muted any transient.
My ribbons are original as far as i can tell and already have a fair amount of movement in them, so re-tentioning is probably in the not too distant future. They need new socks anyway. Are there any specs and/or instructions for proper tensioning? When you consider that Graz has several different versions and they are all probably different from OEM it gets a little confusing as to what the proper tension would be , not to mention different impedances.
Anyway, i will order up some power resistors and see if i can get these guys singing wihtout screaming in pain. As a test i was thinking i could leave one leg of each speaker cable unattached to MR and TW ribbons, then power everything up and very briefly touch the unattached lead to the MR ribbon pigtail and see if there is any movement at all. If not then i may be in the clear. Do the same for TW ribbon If it moves then its back to the drawing board
 
Are there any specs and/or instructions for proper tensioning?
As a test i was thinking i could leave one leg of each speaker cable unattached to MR and TW ribbons, then power everything up and very briefly touch the unattached lead to the MR ribbon pigtail and see if there is any movement at all. If not then i may be in the clear. Do the same for TW ribbon If it moves then its back to the drawing board

I would never "hot" connect an open amp onto a speaker. Perhaps consider putting an 8 ohm dummy load resistor across the amp output before you switch in the ribbon circuit for test.

Ribbon Tensioning: I once read a spec where "V" bent plastic coated solid copper wire, as used in home power wiring, was slipped onto the ribbon foil between pleats just below the bottom clamp as a weight to set the tension. I do not remember what AWG was used ... probably 12. Gratz would be an expert in ribbon tensioning techniques.
 

tyu

Member
2009-07-18 8:16 pm
Thanks for this post....
I have Had A pr of Duettas an from the 90s 5 pr of the stages..
The pr of Stages i have now were made in the 93...the MT have always moved as the Duettas did all over the place...an yes we think it All has to do with the 6db crossover letting low fre in the the MT ribbons....
You would think looking on paper at the passive crossover this is the case..
I have been looking at this in my Apogees speakers an have found a way to stop About 80% of the movement....
I have found that if the Crossover coiles are not gluded down... Like all are ..thats with one side of the coils pressed down to some type of board....or what ever
I have pull the coils lose in the MT Stock crossover an then put 1/2 rubber spacers in only two pr coil... an glued the rubber to the coils an then the board..
This stops this from moving a Lot...
Now I too thought it was the Base driver that was adding a LOt to the movement in the MT...if it is it the other 20%..
An there is know more Twesting of the MT ribbon with this setup...

Now when the movement stops the MT the sweetness of the passive crossovers relly come in...Jason liket the passives crossover sound better than any other...
I got in to this becaues My Magnepans true ribbons Move more than any ribbon i have ever owened...way to much to give the best sound ..an i found that geting the coils lose from the boards stop this 50-60%...an the Maggy ribbons are nothing the wind can kill one..
Just one ribbon lovers finding....Good luck
 

john65b

Member
Paid Member
2005-01-09 2:32 am
Chicago
On my Duetta Sigs, I notice the tweeters were flopping all over the place too...I was using a DCX2496 active crossover, so I steepened the slope to 18db from 6 and moved the crossover away from 500hz. Worked great.

But then I noticed the foam support about a foot down from top was dust (top 12" of ribbon canted back about 1 inch for vertical dispersion)...I put in a new foam support and took quite a bit of the slack out. Put the crossover points back as original and back to stock. The ribbons now are quite a bit more stable and don't wobble anywhere near as bad as before...
 
Can you please briefly elaborate on the 8 ohm dummy resistor idea?
Not sure i understand how this would work.

I was thinking about:
1) putting just a dummy load(no ribbon) of say 8-50 ohms across the output terminals to place some load on the power supply, close the feedback loop, and allow the DC servo to stabilize. You could measure the DC offset now if you are interested. You could test for turn on/off THUMPS now if you are interested. If you have an O-scope you could look at turn on/off transients .... just for fun.

2) with this dummy resistor still connected to the output, connect-in the ribbon circuit(1.5 ohm resistor plus ribbon = 2 ohm load) in parallel with the dummy resistor load and test if the DC servo continues to provide low offset with this much lower resistance speaker load. You could again test for turn on/off THUMPS by watching the ribbon.

3) When you are confident that the low resistance ribbons are safe from funky amp effects you can remove the 8-50 ohm dummy test resistor for normal operation.

Some feedback circuits, and some DC servo circuits perform differently when a low resistance load is attached. Also, Krell amps typically have several opamps that monitor output current, voltage, and power in order to perform dynamic management.

When your amp can effortly push 2,000 watts into 1 ohm
Be Afraid!
Be Very Afraid!
 
Last edited:
thanks again LS
Sorry for so many simple questions, im not an electronics expert and dont want to make any potentially expensive mistakes.
Do you think using a lower value on the resistor power rating would serve to act as a fuse against current spikes? Maybe 5w, or less?
i dont see any way around dc offset. hopefully the tweeters can live with 5-10 mv.
 
PERSONALLY;

I would purchase 50 watt (1-ohm or 1.5 ohm) (Caddock or Dale) non-inductive resistors and study the datasheets on heatsink requirements. Caddock have superior sonics, but require more heatsinking. (I constructed non-inductive resistors by gluing household 1um Al foil to plexiglass and cutting a serpentine). You want linear R over wide power.

I would put JUST the (1-ohm or 1.5-ohm) resistor across the Krell output terminals and test over, and over, and over again if the KRELL DC servo, or Krell ON/OFF THUMP, or Krell special bias circuits, cause NASTY currents that could destroy the ribbons. If both Krell amps are solid, I would connect a (1-ohm or 1.5-ohm) resistor in series with each M-T ribbon to the proven Krells. 5mv DC Servo offset is fine. (You are a lucky Dude). The active Xover gain will set the final SPL.

I would use 18 or 24db Xover slopes at 200Hz and 6Khz to minimize ribbon bass BOUNCE, and also to control tweeter beaming. The ribbons are resistive and will provide great transients, even with steep slopes. Shallow 6db/octave Xovers can destroy the ribbons. Sorry, textbook 6db Xover transient perfection are ribbon death.

We have Ficus trees behind our Apogees for diffusion(and home CO2 to O2 conversion).

Purchase a few 24bit, 96Khz recordings of that are your family favorites. Recording quality is KEY to audio LOVE.
 
thanks again LS for all of your help !!

Greatly appreciated on my end. i am hoping the extra work to direct drive the FR's will pay off. Having a set of fully rebuilt Scintilla's, with big upgrades to the caps, gives me a good reference. in stock form the FR's beat my Scintilla in just about every area important to me. the Scintilla crossovers impart a very nice tone which the FR still lack somewhat but thats about it. Big line sources like FR, IRS V and Dali Megaline have a way of 'breathing' life into the music and have a relaxed quality to the music that is addictive. Some say the image is too big. i find it very life-like.
the adventure continues.....
 
well i tested directly hooking up amp to both MR and TW using series resistors of 1,2 and 8 ohms. there is definitely ribbon movement when amps come on line and it is in proportion to the amount of offset. as you can imagine the ribbon moves in one direction or the other, depending on the phase of the offset and then just stays there until amp is powered off. The MR movement is anywhere from next to nothing, up to about 3/16" with the amp that has 20 mv offset. the TW moves more and its mainly the rear segment that moves, but it moves as much as 1/2 inch and then hits the 'stops'.
At this point i am debating if this is a good idea or not. It will be hard to get all amps down to zero offset. So how much bias can the ribbons tolerate over time?? clearly it will put some stress on them, but i don't know what levels would be considered harmful in the long run.
i guess plan B would be to just use transformers and still tri-amp? Does anyone know if the Speltz Zeroformer will still transfer the DC? Otherwise i would need someone to build custom transformers. I dont know what specs to use however.

Plan C is to stick with the original OEM transformer box and upgrade the transformers AND caps. that seems both expensive and sonically less appealing than plan B.
:headbash:
feedback/comments gratefully accepted.
 
A coupling cap will prevent steady-state DC from deflecting the ribbons fore/aft, but not turn on/off transients.....especially if it's a large value capacitor.

An autoformer is not an inductive coupling like a transformer, so DC offset (although possibly reduced) will still be passed to the transducers.

I recommend to proceed very carefully whatever path you take. Direct-driving of fragile ribbons like Apogee's is potentially hazardous if power amp (or some other unexpected) failure occurs.

Cheers,

Dave.
 
The servo circuit on my Krell KSA250 amp is constructed with discrete components.(circuit and pictures posted on diyAudio solid state) The main servo DiffAmp inputs are JFETS and there is a trim pot that is connected between the two JFET source pins which sets the DC offset.

I do not know what circuit is used in your amp, but a quick look(Picture) at the main PCB might have the servo trim part marked on the PCB. The DC offset could be adjusted and measured with just the dummy 1.5-2 ohm resistor across the output terminals. Due to thermal drift, any final DC offset measurements would be performed on a warm amp.

As I noted, I use the standard midrange transformer and standard tweeter capacitors on my Full Range, plus an 18db/octave active woof-MT crossover for robust, safe operation.


If you cannot get the DC offset reduced on your amps to ~5mv, you might consider using 3-way electronic Xovers, with 3-way amplification, with the standard mid-transformer and tweet-capacitors. robust, safe operation

After some experience with 3-way Xover and amps, you will have knowledge to decide if a new mid-transformer or difference tweeter-capacitor values are worth exploring.


My DIY direct drive amps used low offset opamps and had ~1mv offset. I would expect ~5mv offset on all of your amps is possible, but might still risk ribbon damage from on/of transients.