My measurement microphone and preamp/mixer are waiting for me when I get home today. I'm sure it's going to take a while for me to get reasonable measurements but does anyone have hints/suggestions for measuring a realitively long line source ribbon? In particular, I'm interested in the test baffle - should I go with the narrow "wings" of my test setup (~3" wide on each side) or should I go with a larger baffle to isolate edge effects? My starting point is going to be; keep the test stand I have, point the mike at the middle of the line at a one meter distance and see what happens. I've got the D'Apolito (sp?) book on measuring speakers and I will certainly read though it again, but I don't believe it has anything on line sources.
Thanks,
Denis
Thanks,
Denis
First Measurement !
Here is my first good (I hope) measurement of the full size ribbon. I'm using my new Berhinger mike and mixer/preamp and the mike is about 1.3 meters from the driver pointing at the center. I've added no special baffle so the test stand "wings" are about 4" on either side of the ribbon.
Caveats: this is my first time using Just MLS, I'm using the integrated audio on my Dell computer (i.e. no separate sound card), and I'm measuring in my HT Room and the walls are about 6 feet away. One last caveat - I still have not mastered the best way to calibrate the "sound card" by feeding both channels all the way through the mice amp and power amp. I'm just patching input to output and hitting the "calibrate" button.
It's a little ragged in places but over all is +- 3db from ~1k to 20k. There is a 1 octave rise below 1k with a peak of about 6db. This confirms my initial impressions that it had good output to about 500 hz. Not bad considering it's my first measurements. Especially when you considered that a month ago the driver was a collection of raw material at McMaster Carr!
The key test will be to see if I can repeat the measurements tomorrow (I've had some problems between yesterday and today. If I cross it at 1-1.5k with a third or fourth order filter I should be OK. I've fiddled a little bit with trying to smooth the response by putting 1/4 round adjacent to the magnets in the back and I've also tried some cardboard baffles extensions as well as with trying some 1/4" felt in a couple places. None did any real good and some made things a little worse. I've still got a bunch of things I want to try.
Here is my first good (I hope) measurement of the full size ribbon. I'm using my new Berhinger mike and mixer/preamp and the mike is about 1.3 meters from the driver pointing at the center. I've added no special baffle so the test stand "wings" are about 4" on either side of the ribbon.
Caveats: this is my first time using Just MLS, I'm using the integrated audio on my Dell computer (i.e. no separate sound card), and I'm measuring in my HT Room and the walls are about 6 feet away. One last caveat - I still have not mastered the best way to calibrate the "sound card" by feeding both channels all the way through the mice amp and power amp. I'm just patching input to output and hitting the "calibrate" button.
It's a little ragged in places but over all is +- 3db from ~1k to 20k. There is a 1 octave rise below 1k with a peak of about 6db. This confirms my initial impressions that it had good output to about 500 hz. Not bad considering it's my first measurements. Especially when you considered that a month ago the driver was a collection of raw material at McMaster Carr!
The key test will be to see if I can repeat the measurements tomorrow (I've had some problems between yesterday and today. If I cross it at 1-1.5k with a third or fourth order filter I should be OK. I've fiddled a little bit with trying to smooth the response by putting 1/4 round adjacent to the magnets in the back and I've also tried some cardboard baffles extensions as well as with trying some 1/4" felt in a couple places. None did any real good and some made things a little worse. I've still got a bunch of things I want to try.
Attachments
More measurements
A while back, somebody (Graeme?) asked me what the transformer response looked like. Today I measured it by connecting to the secondary of the transformer instead of the usual mike input, and keeping the JustMLS "probe" on the primary of the transformer. Then I used the subtract function to isolate the transfer function of the Xformer. I've attached the results and the response is pretty close to dead flat, within the limits of the sample rate and window size. That's about all I can measure until I get software that can measure distortion. I want to get a soundcard that can handle sample rates higher than 48k - although it is of only theoretical interest since I can't hear much above 15k.
A while back, somebody (Graeme?) asked me what the transformer response looked like. Today I measured it by connecting to the secondary of the transformer instead of the usual mike input, and keeping the JustMLS "probe" on the primary of the transformer. Then I used the subtract function to isolate the transfer function of the Xformer. I've attached the results and the response is pretty close to dead flat, within the limits of the sample rate and window size. That's about all I can measure until I get software that can measure distortion. I want to get a soundcard that can handle sample rates higher than 48k - although it is of only theoretical interest since I can't hear much above 15k.
Attachments
After measuring the big ribbon I decided to measure my prototype "little ribbon" (re: page 1). I measured it for several reasons:
1. I was just interested in it
2. I was hoping to isolate the line source measuring affects of the big ribbon. I measured the big ribbon pretty close, much closer than I would listen. I wondered if some of the response lumpiness could be a function of cancellation effects from measuring it at a distance of ~ 30% of the line length.
3. I also wanted to compare the ribbon response with and without a baffle. My big ribbon measurements only had tiny "wings" and I made no serious attempt at a anything approaching a proper baffle.
I've attached the results of measuring the little ribbon with (top) and without (bottom) a baffle. One note on that baffle - It consisted of four pieces of cardboard taped together with masking tape. It is quite crude and I'll repeat this with a better baffle when I have the time and patience to make one.
You can see, the baffle made a significant difference below 1k. The response is still pretty rough in that area but I bet I can smooth it out with a better baffle. The response of the little ribbon, above 1k, is somewhat smoother than the big ribbon. My guess at this time is that the big ribbon response measurement was affected by line and or baffle effects. This is another area to dig into. I also plan to try out the little ribbon with the 7 micron foil that linesource kindly sent me. I tried it once before and confirmed an increase in efficiency. This time I can compare frequency response. It should improve, because of the lower mass and the mass may also affect ribbon resonances, which could also affect the frequency response and distortion (which I can't measure).
1. I was just interested in it
2. I was hoping to isolate the line source measuring affects of the big ribbon. I measured the big ribbon pretty close, much closer than I would listen. I wondered if some of the response lumpiness could be a function of cancellation effects from measuring it at a distance of ~ 30% of the line length.
3. I also wanted to compare the ribbon response with and without a baffle. My big ribbon measurements only had tiny "wings" and I made no serious attempt at a anything approaching a proper baffle.
I've attached the results of measuring the little ribbon with (top) and without (bottom) a baffle. One note on that baffle - It consisted of four pieces of cardboard taped together with masking tape. It is quite crude and I'll repeat this with a better baffle when I have the time and patience to make one.
You can see, the baffle made a significant difference below 1k. The response is still pretty rough in that area but I bet I can smooth it out with a better baffle. The response of the little ribbon, above 1k, is somewhat smoother than the big ribbon. My guess at this time is that the big ribbon response measurement was affected by line and or baffle effects. This is another area to dig into. I also plan to try out the little ribbon with the 7 micron foil that linesource kindly sent me. I tried it once before and confirmed an increase in efficiency. This time I can compare frequency response. It should improve, because of the lower mass and the mass may also affect ribbon resonances, which could also affect the frequency response and distortion (which I can't measure).
Attachments
Dhenry,
The Al foil I sent you should be 5.8 microns thick. I do not know how well this thickness is controlled over the different rolls I have. This is commercially pure aluminum and does not respond to any heat treatment. Cold working will cause hardening.
For dipoles, the baffle effect comes in at the quarter wavelength...
13,550 inches/sec velocity divided by 4*baffle width will give you a rough estimate of where the 6db/octave baffle effects will show up on your measurements.
The Al foil I sent you should be 5.8 microns thick. I do not know how well this thickness is controlled over the different rolls I have. This is commercially pure aluminum and does not respond to any heat treatment. Cold working will cause hardening.
For dipoles, the baffle effect comes in at the quarter wavelength...
13,550 inches/sec velocity divided by 4*baffle width will give you a rough estimate of where the 6db/octave baffle effects will show up on your measurements.
gl wrote earlier "One of the patents posted earlier on this thread stated that the gap width between the ribbon and the pole piece had a significant effect on distortion. Any opinions on this gap/distortion thing anyone? "
Below is the text from patent 4,580,014 that I think gl was referencing.
""One of the distortion factors that limits ribbon loudspeaker performance is air flow leakage through the clearance between the longitudinally extending edges of the ribbon and pole faces establishing the magnetic field. As the ribbon moves in response to electromagnetic forces in the course of its vibrations perpendicular to the plane defined by its surface, air leaks past the edges of the ribbon in response to the pressure difference set up by the forces on the ribbon. If the leakage flow were a linear function of the pressure difference across the edges of the ribbon, the effect of such leakage would merely be to linearly increase the amplitude of ribbon vibration. However, the airflow changes suddenly from laminar to turbulent flow at relatively low and ill-defined pressure differences. Both turbulent flow and the transition are highly nonlinear with respect to pressure, and result in nonlinear distortion of the ribbon vibration amplitude which is particularly strong at the low end of ribbon speaker’s frequency range. In practice, a clearance of about 1% of ribbon width or 5 mils for a ½ inch wide ribbon gives rise to approximately 10 to 15% third harmonic distortion whereas a clearance of about 0.1% yields 0.1 to 0.2% distortion. ""
The two most common ribbon designs are (1) magnets form ribbon gap; and (2) iron pole pieces form ribbon gap. Due to the brittle nature of magnets from high temperatures used in formation, most magnets have slightly irregular surfaces and will vary slightly in thickness across samples. Hence, machining iron pole pieces with adjustments to set the gap offers the tightest gap tolerance. Since iron saturates around 1.7T, using iron pole pieces to set the ribbon gap with todays super NdFeB magnets can limit the maximum magnetic field in the gap to a lower level than the topology where the magnets face each other to create the ribbon gap. If you look at Dhenry's ribbon pictures in this thread you can see how irregular the magnet edges are relative to machined steel.
There are design methods to reduce gap distortion when super tight tolerances cannot be achieved, but these are typically manufacturing secrets.
Below is the text from patent 4,580,014 that I think gl was referencing.
""One of the distortion factors that limits ribbon loudspeaker performance is air flow leakage through the clearance between the longitudinally extending edges of the ribbon and pole faces establishing the magnetic field. As the ribbon moves in response to electromagnetic forces in the course of its vibrations perpendicular to the plane defined by its surface, air leaks past the edges of the ribbon in response to the pressure difference set up by the forces on the ribbon. If the leakage flow were a linear function of the pressure difference across the edges of the ribbon, the effect of such leakage would merely be to linearly increase the amplitude of ribbon vibration. However, the airflow changes suddenly from laminar to turbulent flow at relatively low and ill-defined pressure differences. Both turbulent flow and the transition are highly nonlinear with respect to pressure, and result in nonlinear distortion of the ribbon vibration amplitude which is particularly strong at the low end of ribbon speaker’s frequency range. In practice, a clearance of about 1% of ribbon width or 5 mils for a ½ inch wide ribbon gives rise to approximately 10 to 15% third harmonic distortion whereas a clearance of about 0.1% yields 0.1 to 0.2% distortion. ""
The two most common ribbon designs are (1) magnets form ribbon gap; and (2) iron pole pieces form ribbon gap. Due to the brittle nature of magnets from high temperatures used in formation, most magnets have slightly irregular surfaces and will vary slightly in thickness across samples. Hence, machining iron pole pieces with adjustments to set the gap offers the tightest gap tolerance. Since iron saturates around 1.7T, using iron pole pieces to set the ribbon gap with todays super NdFeB magnets can limit the maximum magnetic field in the gap to a lower level than the topology where the magnets face each other to create the ribbon gap. If you look at Dhenry's ribbon pictures in this thread you can see how irregular the magnet edges are relative to machined steel.
There are design methods to reduce gap distortion when super tight tolerances cannot be achieved, but these are typically manufacturing secrets.
Actually, the magnets have a surface that is at least as smooth as the steel I'm using. The big difference is the small crack between the magnets on a given edge caused by the fact that the edges are not perfectly sharp 90 degree angles but instead have a tiny radius.
I would be very interested to know if and how any of you have been able to mount their ribbon so that it is 1/500 inches from the pole pieces without touching the pole pieces. Frankly I don't come all that close and I cant even get my corrugated ribbons to be straight within that tolerance, much less than the 1/5000 inches talked about in the patent. How close do the commercial ribbons come to the pole pieces?
Denis
I would be very interested to know if and how any of you have been able to mount their ribbon so that it is 1/500 inches from the pole pieces without touching the pole pieces. Frankly I don't come all that close and I cant even get my corrugated ribbons to be straight within that tolerance, much less than the 1/5000 inches talked about in the patent. How close do the commercial ribbons come to the pole pieces?
Denis
Just to add a couple of things:
1) I was not the one who enquired about the transformer response. I am building a ribbon with a corrugated length of 68" so I won't need a transformer.
2) We need to add a third pole piece/ribbon interface to the list. The Magnepan interface has the ribbon bonded to and more or less touching the (supposedly) non-conductive ceramic magnet pole pieces.
Graeme
1) I was not the one who enquired about the transformer response. I am building a ribbon with a corrugated length of 68" so I won't need a transformer.
2) We need to add a third pole piece/ribbon interface to the list. The Magnepan interface has the ribbon bonded to and more or less touching the (supposedly) non-conductive ceramic magnet pole pieces.
Graeme
gl said:
Hi Graeme,
Could you explain this a little further? I'm not sure I understand what you mean by a "third pole piece/ribbon interface ".
WRT Magnepan: are you saying that the ribbon might actually touch the pole pieces, but depends on them being nonconductive to avoid shorting out? I would have thought that touching the poles would cause unacceptable rubbing noises. It would be very easy to put a thin, smooth nonconducting (and nonmagnetic) surface over my magnets if I did not have to worry about rubbing. I guess that's something else to try.
Regards,
Denis
I have seen the use of stips of plastic fridge magnet material on the surface of the magnets in the gap, can't remember where, would give a smooth surface and maybe some damping. Dont know what it would do to the field in the gap?
Just purchased some Pioneer PT-R29A Beryllium ribbons, will take photos and post if there is interest?
Regards
James
Just purchased some Pioneer PT-R29A Beryllium ribbons, will take photos and post if there is interest?
Regards
James
Hi Denis and Others,
Regarding the Magnepan ribbon: The edge of the ribbon is definitely bonded to face of the magnets. I understand that the bonding is done with dots of some kind of silicone or Loctite tak-pak type of material but I don't know for certain. I do remember reading that the magnets are non-conductive ( I believe in Jim Wineys patent ). The implication of this is that the ribbon does touch the pole pieces. The ribbon is crossed over at 1700Hz in the MG3.6 and 1000Hz or so higher in the 20.1 so the excursion is small. Then again maybe the pole piece surface does indeed have a smooth surface treatment.
I have heard that Magnepan sells kits to re-ribbon the tweeters for about $30. I would assume that there are peope on this web site who have first hand knowledge about this. I welcome their input.
What is the spacing between the ribbon and the pole pieces for some of the commericial ribbon units like the Ravens and the Founteks? Does anyone know?
Graeme
Regarding the Magnepan ribbon: The edge of the ribbon is definitely bonded to face of the magnets. I understand that the bonding is done with dots of some kind of silicone or Loctite tak-pak type of material but I don't know for certain. I do remember reading that the magnets are non-conductive ( I believe in Jim Wineys patent ). The implication of this is that the ribbon does touch the pole pieces. The ribbon is crossed over at 1700Hz in the MG3.6 and 1000Hz or so higher in the 20.1 so the excursion is small. Then again maybe the pole piece surface does indeed have a smooth surface treatment.
I have heard that Magnepan sells kits to re-ribbon the tweeters for about $30. I would assume that there are peope on this web site who have first hand knowledge about this. I welcome their input.
What is the spacing between the ribbon and the pole pieces for some of the commericial ribbon units like the Ravens and the Founteks? Does anyone know?
Graeme
I designed my wide bandwidth ribbons with a 0.05 inch gap on each side, which due to magnet irregularites drops to about 0.025" at the tightest point. I purchase rolls of pre-slit foil that has very good width control, probably 0.001"
A ribbon should never rub up against the magnets or pole pieces. This is very audible. For Magnepan ribbon assembly, shims are use to maintain a controlled gap between the ribbon and magnets while the adhesive is applied and drying. This adhesive is what maintains the gap over time. Below is a DIY description of Magnepan ribbon replacement.
-----------
There is not much double stick tape that could be placed on the tiny connector blocks...
The stretching of the ribbon is prescribed...! It could be that the run-in time is reduced by stretching the ribbon in a defined way.
In the meantime i talked with a repair person, and i got these informations:
- The length of the ribbon is defined by the edges of the thick black tape that holds the alu to the paper strip, ie. where the ribbon gets free of this tape.
- It is best to measure the prescribed stretch length of 62" (this is 2" longer than the whole magnet structure is) from slightly to the 'inside' (magnet side) of the contact springs screw hole, and mark the 62" length at the other side of the channel (the point is then beyond and outside of the magnet structure).
- Place the shims rel. far into the channel, but still so that you can lift it with a needle nose plyer, similar, or pointed nails....
- Adjust the ribbon between the shims. Then place one end of the ribbon at the same point ('inside' of screwhole), fixed on top of the connector block by the ready double-stick tape.
- Stretch the other end carefully, between the shims to the marked point.
- Then go back with the free end and fix it to the other end block, placing the edge of the black tape just to the 'inside' of the connector leaf's screwhole.
- If there's a slight slack in the ribbon, (which you see in the tendency of twisting in the ribbon, specially at the ribbon support tools) you can without any real problem stretch the ribbon a bit more, max. up to 1 inch, until the twist disappears. The ribbon should not *lay* on the support tools in any way, it should only touch a bit.
- physical length differences of up to ca. 1 inch are not a problem at all in *practice*.
The problem with the syringes and the blue fluid is that the nozzle needs frequent cleaning, the dots should be about a pin-needle's head big (the metal ones, that are used to fix textiles before making eg. a seam).
That ribbon is really ingenously simple and efficient design-wise.
A ribbon should never rub up against the magnets or pole pieces. This is very audible. For Magnepan ribbon assembly, shims are use to maintain a controlled gap between the ribbon and magnets while the adhesive is applied and drying. This adhesive is what maintains the gap over time. Below is a DIY description of Magnepan ribbon replacement.
-----------
There is not much double stick tape that could be placed on the tiny connector blocks...
The stretching of the ribbon is prescribed...! It could be that the run-in time is reduced by stretching the ribbon in a defined way.
In the meantime i talked with a repair person, and i got these informations:
- The length of the ribbon is defined by the edges of the thick black tape that holds the alu to the paper strip, ie. where the ribbon gets free of this tape.
- It is best to measure the prescribed stretch length of 62" (this is 2" longer than the whole magnet structure is) from slightly to the 'inside' (magnet side) of the contact springs screw hole, and mark the 62" length at the other side of the channel (the point is then beyond and outside of the magnet structure).
- Place the shims rel. far into the channel, but still so that you can lift it with a needle nose plyer, similar, or pointed nails....
- Adjust the ribbon between the shims. Then place one end of the ribbon at the same point ('inside' of screwhole), fixed on top of the connector block by the ready double-stick tape.
- Stretch the other end carefully, between the shims to the marked point.
- Then go back with the free end and fix it to the other end block, placing the edge of the black tape just to the 'inside' of the connector leaf's screwhole.
- If there's a slight slack in the ribbon, (which you see in the tendency of twisting in the ribbon, specially at the ribbon support tools) you can without any real problem stretch the ribbon a bit more, max. up to 1 inch, until the twist disappears. The ribbon should not *lay* on the support tools in any way, it should only touch a bit.
- physical length differences of up to ca. 1 inch are not a problem at all in *practice*.
The problem with the syringes and the blue fluid is that the nozzle needs frequent cleaning, the dots should be about a pin-needle's head big (the metal ones, that are used to fix textiles before making eg. a seam).
That ribbon is really ingenously simple and efficient design-wise.
Member
Joined 2003
Hi gang,
Verhagen tested 30:1 gap slits (ribbon clearance) and 100:1 slits. He reports that higher ribbon clearance does increase both THD and IM distortion, though not as much as indicated in the patents. He seems more concerned about loss of low-frequency output with higher clearance.
Ferrite magnets I have measured are essentially non-conductive...of course not true for nickle coated NIB magnets. Practically, conductivity doesn't matter as dragging the edge of a ribbon generates terrible distortion (with pink noise it can sound like ringing or singing). So, as LineSource indicates, spacing is needed. It seems prudent to strive for near zero clearance...a temporary strip of paper in each slit can help with alignment.
I understand Arum Cantus ribbons use tape on the magnet faces, perhaps to prevent shorts and/or provide a slick surface in case of inadvertant contact. A thin Scotch transparent tape seems to serve well when I get reckless with low frequency
Paul
Verhagen tested 30:1 gap slits (ribbon clearance) and 100:1 slits. He reports that higher ribbon clearance does increase both THD and IM distortion, though not as much as indicated in the patents. He seems more concerned about loss of low-frequency output with higher clearance.
Ferrite magnets I have measured are essentially non-conductive...of course not true for nickle coated NIB magnets. Practically, conductivity doesn't matter as dragging the edge of a ribbon generates terrible distortion (with pink noise it can sound like ringing or singing). So, as LineSource indicates, spacing is needed. It seems prudent to strive for near zero clearance...a temporary strip of paper in each slit can help with alignment.
I understand Arum Cantus ribbons use tape on the magnet faces, perhaps to prevent shorts and/or provide a slick surface in case of inadvertant contact. A thin Scotch transparent tape seems to serve well when I get reckless with low frequency
Paul
Thanks for all the responses. I hadn't thought about putting a temporary spacer while installing the ribbon. I will certainly go that way. The patent seemed to require almost unattainable (at least for me) clearances equal to ~ a sheet of paper just to get 10% distortion and 1/10th that to get really low distortion. I can get the kind of clearances that linesource was designing to, which is a little less than a 1/16th of an inch - for those of us who think in terms of grammar school rulers. The patent did not specify at what frequency and at what level of ribbon excursion to get the stated distortion. I suspect that if the quoted distortion numbers are true in general, they are for low frequencies (for a ribbon) and high excursions.
Regarding insulated magnets. When I made my first prototype I started out by putting regular scotch tape on the magnet surfaces. It did not work out well. Some part of the edge would start to lift and it would get worse from there. Before long, the top and bottom edges starting curling and contacting the ribbon. With my big ribbon, I just gave the whole driver; steel and magnets, a coat of clear spray paint. I did this mostly to stop the steel from rusting but it will also provide some insulation for the magnets. You really don't need much insulation because you are talking about very low voltages across the ribbon. In my case it is probably no more than 3-4 volts. I've bought some coil dope and may try putting a thin coat on the ribbon surfaces if insulation ever becomes an issue.
One last story about uninsulated magnets; I was doing some testing to destruction of my prototype ribbon to see how much power it could take before melting. I'm sure I saturated the transformer so I can't tell you the power, but when the ribbon finally melted it flopped around in the gap touching against the magnets. It kind of looked like arc welding as it produced some pretty impressive sparks. The current in the ribbon can easily hit dozens of amps - impressive when you consider how fragile it is.
Denis
Regarding insulated magnets. When I made my first prototype I started out by putting regular scotch tape on the magnet surfaces. It did not work out well. Some part of the edge would start to lift and it would get worse from there. Before long, the top and bottom edges starting curling and contacting the ribbon. With my big ribbon, I just gave the whole driver; steel and magnets, a coat of clear spray paint. I did this mostly to stop the steel from rusting but it will also provide some insulation for the magnets. You really don't need much insulation because you are talking about very low voltages across the ribbon. In my case it is probably no more than 3-4 volts. I've bought some coil dope and may try putting a thin coat on the ribbon surfaces if insulation ever becomes an issue.
One last story about uninsulated magnets; I was doing some testing to destruction of my prototype ribbon to see how much power it could take before melting. I'm sure I saturated the transformer so I can't tell you the power, but when the ribbon finally melted it flopped around in the gap touching against the magnets. It kind of looked like arc welding as it produced some pretty impressive sparks. The current in the ribbon can easily hit dozens of amps - impressive when you consider how fragile it is.
Denis
LineSource,
Thank you for the info on the Magnepan tweeter. Especially the part about the shims used to space the ribbon away from the pole pieces. Is there any additional info on the spacing of the attachment dots? I would imagine that they would need to be irregularly spaced to smooth out resonances.
Denis,
For tape I would suggest using some form of tough, thin plastic film, like teflon plumbers tape, glued to the pole pieces with the appropriate adhesive. Most adhesive tapes are OK for quick experiments but IMHO they aren't really reliable long term.
Paul W,
I really do need to buy Verhagens book. Thank you for reminding me of this. I generally don't take the stuff in patents as fact so I'm not at all surprised that Verhagens gap/distortion numbers were more modest.
Regards,
Graeme
Thank you for the info on the Magnepan tweeter. Especially the part about the shims used to space the ribbon away from the pole pieces. Is there any additional info on the spacing of the attachment dots? I would imagine that they would need to be irregularly spaced to smooth out resonances.
Denis,
For tape I would suggest using some form of tough, thin plastic film, like teflon plumbers tape, glued to the pole pieces with the appropriate adhesive. Most adhesive tapes are OK for quick experiments but IMHO they aren't really reliable long term.
Paul W,
I really do need to buy Verhagens book. Thank you for reminding me of this. I generally don't take the stuff in patents as fact so I'm not at all surprised that Verhagens gap/distortion numbers were more modest.
Regards,
Graeme
SHIMS 101.
get a metric ruler with 1mm scale.
search web for table of paper thickness like
http://www.paper-paper.com/weight.html
1mm = 10 sheets of laser printer paper(20lb. bond = 3.8 - 4 point stock = 0.004 inch thick ~ 0.1mm)
40 lb. Bond paper is 7.3 points = 0.0073" = 0.185mm
67 lb. Bond paper is 10.5 points = 0.01" = 0.25mm
0.1mm ~ 1 sheet of 20lb bond paper
14 E-6 mm/mm /*F aluminum coefficient of thermal expansion
200 *F heating on 25.4mm (1") wide aluminum ribbon = 0.072 mm
0.2 - 0.3mm tweeter ribbon gap seems like reasonable target
0.4 - 0.6mm midrange ribbon gap seems like reasonable target
get a metric ruler with 1mm scale.
search web for table of paper thickness like
http://www.paper-paper.com/weight.html
1mm = 10 sheets of laser printer paper(20lb. bond = 3.8 - 4 point stock = 0.004 inch thick ~ 0.1mm)
40 lb. Bond paper is 7.3 points = 0.0073" = 0.185mm
67 lb. Bond paper is 10.5 points = 0.01" = 0.25mm
0.1mm ~ 1 sheet of 20lb bond paper
14 E-6 mm/mm /*F aluminum coefficient of thermal expansion
200 *F heating on 25.4mm (1") wide aluminum ribbon = 0.072 mm
0.2 - 0.3mm tweeter ribbon gap seems like reasonable target
0.4 - 0.6mm midrange ribbon gap seems like reasonable target
Dear dhenryp
This looks wonderful to me. My first ever "ribbon experience" is with some 12 year old Carvers with the aluminium/mylar ribbons, and I really liked the easiness of the sound. Yours look like "true" ribbons which I have never heard well done (heard some in a Dali something or other, but the speaker tore my ears out.)
Are you getting around to putting together a "How to make you own Ribbon Tweeter/Midrange Driver Page", because without relatively explicit instructions, I would not be able to emulate your feat ... and I am actually thinking about making a set of speakers something like the Carvers and some "home made" drivers would be the bees knees. All quite exciting really.
Regards,
George
This looks wonderful to me. My first ever "ribbon experience" is with some 12 year old Carvers with the aluminium/mylar ribbons, and I really liked the easiness of the sound. Yours look like "true" ribbons which I have never heard well done (heard some in a Dali something or other, but the speaker tore my ears out.)
Are you getting around to putting together a "How to make you own Ribbon Tweeter/Midrange Driver Page", because without relatively explicit instructions, I would not be able to emulate your feat ... and I am actually thinking about making a set of speakers something like the Carvers and some "home made" drivers would be the bees knees. All quite exciting really.
Regards,
George
"bee knees" - that's one I've not heard in a long time! Thank You!
I am working on a general DIY Audio web page and the ribbon experience will be part of it. I did not take detailed step by step construction pictures of the first ribbon but I will when I build the second one. The page will probably be up in a month or two. I'll be sure to mention it in this thread and to get it added to the big list on the "Loudspeakers" forum.
I bought a couple sheets of masonite last night to use as test baffles so I hope to have better measurements some time this weekend. Lately, I've been spending a lot of time working on the design for the line array cabinets. I also bought (what I hope will be) beautiful veneer off e-bay. This will be my first foray into raw veneer.
Denis
I am working on a general DIY Audio web page and the ribbon experience will be part of it. I did not take detailed step by step construction pictures of the first ribbon but I will when I build the second one. The page will probably be up in a month or two. I'll be sure to mention it in this thread and to get it added to the big list on the "Loudspeakers" forum.
I bought a couple sheets of masonite last night to use as test baffles so I hope to have better measurements some time this weekend. Lately, I've been spending a lot of time working on the design for the line array cabinets. I also bought (what I hope will be) beautiful veneer off e-bay. This will be my first foray into raw veneer.
Denis
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.