What's difficult for friction is to measure that angle, the action is so short and so variable...
For the skating i was thinking to apply it only to pivoted ones, but surely the rig must be built with much different precision - tolerances ... More: the left and the right skipping force test must be applied to the same grooves: very difficult. if not impossible
Stylus drag: In fact my pendulum was used just for Stylus Drag and Skipping force measures ( on Dd thread "angling for 90° #1548 - #1569 -#1595. Drawings and photos) In that occasion, discussing method and results you talked about your previously made rig for carriage movement measuring.
c
For the skating i was thinking to apply it only to pivoted ones, but surely the rig must be built with much different precision - tolerances ... More: the left and the right skipping force test must be applied to the same grooves: very difficult. if not impossible
Stylus drag: In fact my pendulum was used just for Stylus Drag and Skipping force measures ( on Dd thread "angling for 90° #1548 - #1569 -#1595. Drawings and photos) In that occasion, discussing method and results you talked about your previously made rig for carriage movement measuring.
c
That would be the rig that inspired the suggestion I made in my previous post. Sorry, I should have credited you with the original thought. I remember you having trouble with getting consistent results which is why I suggested the refinement of damping the motion of the arm. This could take the form of a silicon fluid trough. Also instead of measuring the amount the arm swings measure the force required to stop it swinging.
Since my previous post I have had a couple of additional ideas. I believe that you used a normal record playing music in your stylus drag test. As the modulation of music constantly changes the drag will also change. You could instead use a test record with a single tone recorded at 0dB. This would eliminate variation in stylus drag and tell you what the drag is at a reference level. The same could be done for an unmodulated groove using the runout groove of any record. These two combined will tell you the range over which drag varies. If you have a record with a blank side with no grooves you can also measure the drag on this. It should be about the same as the unmodulated groove. Using the blank disc with your normal arm you can determine how much antiskate is required to keep the arm stationary. Take the ratio of the drag of the modulated disc to the blank disc. Now increase the skating force by about 80% of this ratio (60% if you mainly listen to classical) and your skating force should be set.
As an alternative, build a linear tracking tonearm and never worry about skating force again😁
Niffy
Since my previous post I have had a couple of additional ideas. I believe that you used a normal record playing music in your stylus drag test. As the modulation of music constantly changes the drag will also change. You could instead use a test record with a single tone recorded at 0dB. This would eliminate variation in stylus drag and tell you what the drag is at a reference level. The same could be done for an unmodulated groove using the runout groove of any record. These two combined will tell you the range over which drag varies. If you have a record with a blank side with no grooves you can also measure the drag on this. It should be about the same as the unmodulated groove. Using the blank disc with your normal arm you can determine how much antiskate is required to keep the arm stationary. Take the ratio of the drag of the modulated disc to the blank disc. Now increase the skating force by about 80% of this ratio (60% if you mainly listen to classical) and your skating force should be set.
As an alternative, build a linear tracking tonearm and never worry about skating force again😁
Niffy
Thanks Niffy, those tests were born from the observations on the Sirinx mk1 and mk2: on web the stylus drag was calculated (as physical result or with the stopping method) between 0.5 and 1.5 gr. and using such forces the arm was sliding well, while on the disc it did not move at all.
In fact, the pendulum test gave much lower values (0.2 - 0,4 gr, with 1,5 2,5 VTF.). A concert without peaks was chosen to have a practical measure, and it was amusing to see the arm lightly oscillating with the variation of the drag. Photographing with 1/2" exposure time, which included those small variations, gave me an average measurement.
Maybe the theoretical calculations did not take into account that part of the force is used to produce the electrical signal, not friction (if not, what would the cartridge made for?)
The tests you are suggesting are really interesting and accurate , sooner or later I will put my hands on.
As an alternative, build a linear tracking and never worry about skating force again.
It is what I'm doing with Lil Casey, not because I'm worrying about the skating but to see how to limit the opposite on linears, ( "generator misalignment" a perfect definition) with just a different geometry (always on the principle of "what if ....")
c
attachment - for those who do not like the Cartesian projections now the Lil Casey *.Step CAD file. To overcome my mistakes, or laugh at
In fact, the pendulum test gave much lower values (0.2 - 0,4 gr, with 1,5 2,5 VTF.). A concert without peaks was chosen to have a practical measure, and it was amusing to see the arm lightly oscillating with the variation of the drag. Photographing with 1/2" exposure time, which included those small variations, gave me an average measurement.
Maybe the theoretical calculations did not take into account that part of the force is used to produce the electrical signal, not friction (if not, what would the cartridge made for?)
The tests you are suggesting are really interesting and accurate , sooner or later I will put my hands on.
As an alternative, build a linear tracking and never worry about skating force again.
It is what I'm doing with Lil Casey, not because I'm worrying about the skating but to see how to limit the opposite on linears, ( "generator misalignment" a perfect definition) with just a different geometry (always on the principle of "what if ....")
c
attachment - for those who do not like the Cartesian projections now the Lil Casey *.Step CAD file. To overcome my mistakes, or laugh at
My money is on this one, a mere 30k, for the arm only that is, active, laser guided
http://www.dereneville.de/Documents/News/2017_01_Hifi_Tunes_DE.pdf
Btw, mine will be more accurate and much, much cheaper, think 1-1,5k
Hahaha
http://www.dereneville.de/Documents/News/2017_01_Hifi_Tunes_DE.pdf
Btw, mine will be more accurate and much, much cheaper, think 1-1,5k
Hahaha
Last edited:
Hi walterwalter,
All mechanical bearings will have some level of static friction (stiction) so it is still present in my arm. I only measured static friction as this is always greater than dynamic (moving) friction. If you are going to have a problem with friction it will be with static friction.
All bearings exhibit some degree of variation in the level of friction as they rotate. For this reason I took many separate measurements of the bearings, rotating the wheels slightly between each. I measured for movement both to the left and right. In total I took about 300 separate measurements for each type of bearing tested.
I then plotted all of these in a graph that showed the coefficient of friction against the probability of that friction occuring. A lower friction would show towards the left of the graph. A more consistent bearing that showed less variation would produce a narrower taller peak. The area under each plot is the same and adds up to a probability of one.
The graph at the bottom of this post shows the friction of 4 of the bearings I tested.
They are for my current bearings that are jewelled bearings with tungsten carbide wheels running on tungsten carbide rails.
Tungsten carbide wheels on tungsten carbide rails with my home made steel pin bearings.
Stainless Steel wheels on tungsten carbide rails with my home made pin bearings.
Finally there is boca hybrid ball race bearings on borosilicate glass rails. The boca hybrids are very high quality ball race bearings, about as good as this type of bearing gets.
The all jewelled and tungsten carbide bearings are clearly the best. The average coefficient of friction for this bearing is 0.00174 which with my carriage equates to a force of 0.931mN (0.095g). The mode, which is the most common and is where the peek on the graph is, has a coefficient of friction of 0.00157. With my carriage this is 0.843mN (0.086g). With my 22um/mN cartridge and a 0.931mN side force would cause a tracking error of about 0.23°, this only occurs momentarily just before the carriage starts moving. (less than the average skating error of a 9"pivoted arm, this happens all the time)
It is interesting to note that the ballrace bearings have a double peek in the graph. These bearings have 6 or 7 balls, can't remember of the top of my head. The lower peek seemed to be where there was a single ball at the bottom of the race and the higher when there where two. These bearings are designed for high speed use where this variation would not be a problem. The problem is only apparent with very slow stop-start motion. This effect was much more pronounced with lower quality bearings, not shown in this graph.
View attachment 729499
Niffy
Really precious infos, Niffy. I can not download the chart but I've seen some for the ball race bearings that resemble the behavior of my tests: an initial friction drop and then a non-linear slowdown.
The coefficients found for non recirculating are about 0,002, that with my cart of 14,5 grams give 0,290 mN. I have found half of it, which is incredibly low: so I think that with your jewells that are surely better the carriage weight must have much less influence than calculated with coefficients.
That measure may be of course a reading or calc mistake, but even that the 4 balls are not in contact each other, while normally there are many more and crawl between them, or on the cage.
Anodization has a notable hardness (8-9 Mohs) but the surface is not super slick. Chrome steel and delrin balls tested roll quite the same way, but the delrin seems to sound better, sharper. (damping of a so light carriage-head shell unit?)
carlo
The friend that 3d-printed my PLT arm has several types of recirculating at hand, round and flat: every ball reaching the curve to go back, you feel the friction increase-decrease. Not to talk of the horrible chattering: noisy like a flock of sheep
The coefficients found for non recirculating are about 0,002, that with my cart of 14,5 grams give 0,290 mN. I have found half of it, which is incredibly low: so I think that with your jewells that are surely better the carriage weight must have much less influence than calculated with coefficients.
That measure may be of course a reading or calc mistake, but even that the 4 balls are not in contact each other, while normally there are many more and crawl between them, or on the cage.
Anodization has a notable hardness (8-9 Mohs) but the surface is not super slick. Chrome steel and delrin balls tested roll quite the same way, but the delrin seems to sound better, sharper. (damping of a so light carriage-head shell unit?)
carlo
The friend that 3d-printed my PLT arm has several types of recirculating at hand, round and flat: every ball reaching the curve to go back, you feel the friction increase-decrease. Not to talk of the horrible chattering: noisy like a flock of sheep
Hello carlthess40,
There are several of us "here" who have made proper working arms. I am one of them. Please look here:
A Revolutionary Pivoting Tangential Tone Arm
Sincerely,
Ralf
S#!t Ralf,
That's one beautifully crafted arm. I seem to have missed your thread somehow. I'll definitely be studying your design in detail when I have time to give it the attention it looks like it deserves.
Hi carlthess40,
That's quite a challenging project. Looks like a very complex design. The main advantage of a linear tracking arm is not the reduction in lateral tracking error. The main advantage is that you can have a much shorter armtube. This massively increases rigidity and reduces colouration. The arm you showed looks like it has an armtube no shorter than a conventional arm so the main advantage has been lost.
Even if the arm has zero tracking error there will still be lateral tracking error on any eccentric record and virtually all records are eccentric to some degree. The average tracking error due to eccentricity on an average record is about 0.15-0.2°. If you really want to achieve zero tracking errors then you have to employ some method of centering your records. If you have perfectly centred records then you can't beat a passive air bearing arm. A servo arm requires a very small error in order to know how much to move and in which direction. With an air bearing on a perfectly centered record you wouldn't even get this small error. You also wouldn't have the complex mechanism which is highly likely to have a negative effect on system resonance.
I don't mean to be disparaging but I would be remiss in not pointing out that a servo linear arm doesn't really solve the problem it sets out to solve. Still a fascinating project and I wish you success.
Niffy
That's one beautifully crafted arm. I seem to have missed your thread somehow. I'll definitely be studying your design in detail when I have time to give it the attention it looks like it deserves.
Hi carlthess40,
That's quite a challenging project. Looks like a very complex design. The main advantage of a linear tracking arm is not the reduction in lateral tracking error. The main advantage is that you can have a much shorter armtube. This massively increases rigidity and reduces colouration. The arm you showed looks like it has an armtube no shorter than a conventional arm so the main advantage has been lost.
Even if the arm has zero tracking error there will still be lateral tracking error on any eccentric record and virtually all records are eccentric to some degree. The average tracking error due to eccentricity on an average record is about 0.15-0.2°. If you really want to achieve zero tracking errors then you have to employ some method of centering your records. If you have perfectly centred records then you can't beat a passive air bearing arm. A servo arm requires a very small error in order to know how much to move and in which direction. With an air bearing on a perfectly centered record you wouldn't even get this small error. You also wouldn't have the complex mechanism which is highly likely to have a negative effect on system resonance.
I don't mean to be disparaging but I would be remiss in not pointing out that a servo linear arm doesn't really solve the problem it sets out to solve. Still a fascinating project and I wish you success.
Niffy
I no longer can build things like that any longer. I have severe carpal tunnel syndrome in both hands and arthritis and I’ve already had two back surgeries and need one more
I am a highly accomplished machinist with the wood and metal and also a very accomplished welder. I’ve been a master mechanic for 45 years and with all of the years of doing this type of work has taken a toll on my body
I was hoping to find someone who has one or who can make one to a certain level that I could finish or take over a project
I do agree with you 100% that you do need to have the length of the arm as short as possible, the longer the arm, the less stiffness you have and the arm and start to transfer some vibrations through the arm itself from the needle. It’s basic engineering and physics
I just wish my hands would work a 1/10 of a%
As what my Head has gone through it with ideas for making turntables and arms
I have a bunch of 1 1/2 bulletproof glass that I am wanting to make a turntable out of It
I just need to get out to my brothers and use his shop and start milling and running the lathe to make some parts
I am a highly accomplished machinist with the wood and metal and also a very accomplished welder. I’ve been a master mechanic for 45 years and with all of the years of doing this type of work has taken a toll on my body
I was hoping to find someone who has one or who can make one to a certain level that I could finish or take over a project
I do agree with you 100% that you do need to have the length of the arm as short as possible, the longer the arm, the less stiffness you have and the arm and start to transfer some vibrations through the arm itself from the needle. It’s basic engineering and physics
I just wish my hands would work a 1/10 of a%
As what my Head has gone through it with ideas for making turntables and arms
I have a bunch of 1 1/2 bulletproof glass that I am wanting to make a turntable out of It
I just need to get out to my brothers and use his shop and start milling and running the lathe to make some parts
With my 22um/mN cartridge and a 0.931mN side force would cause a tracking error of about 0.23
Niffy, could you explain me please how to do this calculation? does the length of the arm have to do with it? I would like to check if the measurements made on the carriage match somehow to the 0.13 mm bending found on the video frames.
Carlthess 40 -Since Ralf told you about his PLT, magnificent but challenging, may i point out my poor, proletarian 3D Toy? (Angling for 90° - #1387-88 and others). Completely different of course, but it works honestly, tirelessly, and you have just to 3D print the files, make the ballpen tip bearings and assemble few pieces.
#2557 "Scope" mean oscilloscope or macro scope? some experience with the second
carlo
Niffy, could you explain me please how to do this calculation? does the length of the arm have to do with it? I would like to check if the measurements made on the carriage match somehow to the 0.13 mm bending found on the video frames.
Carlthess 40 -Since Ralf told you about his PLT, magnificent but challenging, may i point out my poor, proletarian 3D Toy? (Angling for 90° - #1387-88 and others). Completely different of course, but it works honestly, tirelessly, and you have just to 3D print the files, make the ballpen tip bearings and assemble few pieces.
#2557 "Scope" mean oscilloscope or macro scope? some experience with the second
carlo
Sorry Carlthess40, wrong (inverted numbers) links. From #1837-#1838 you may download the print files. All can be said by me and others on that arm starts at #1761.. Look also to the Rabbit 3D, more evoluted design.
the Rabbit - Bunny belted PLT story starts long before.
ciao carlo
sorry for offtopic
the Rabbit - Bunny belted PLT story starts long before.
ciao carlo
sorry for offtopic
Hi Carlo,
Compliance is a measurement used for springs and things that behave like springs. It is a measure of how much something will move for a given applied force. It is the reciprocal of stiffness which is a measure of how much force is required to move a given distance.
My cartridge has a compliance of 22um/mN. This mean that if I apply 1mN of side force to the stylus it will be deflected relative to the cartridge by 22um.
My bearings have an average friction of 0.931mN. This means that my carriage has to be pushed with a force of 0.931mN before it starts to move. This force is applied to the carriage by the stylus. For the carriage to move a force of 0.931mN must be applied to the stylus. This force will cause the stylus to move by
0.931mN x 22um/mN = 20.48um
The cantilever in my cartridge is about 5mm long (5000um). I had to measure this myself as its not something that manufacturers tend to stick in their spec sheets. It is difficult to accurately determine exactly where the pivot point in the cartridge is. I measured the length at between 5-6mm. I decided to use to shortest length as this would give the worst results. I prefer to under quote specifications rather than over quote as I aim to be as honest with myself as possible and avoid falling into the trap of confirmation bias.
From the length of the cantilever and the distance the stylus has moved the angle can be calculated using the sin rule.
Sin^-1(5000um/20.48um)= 0.23°
Arm length has no direct bearing on this equation. Arm length will probably have an effect on the amount of force required to overcome bearing friction. My test rig applied the side force at the same location as the stylus so this is already taken into account.
Most manufacturers quote dynamic compliance. This is the compliance at a set frequency, normally 315hz though sometimes 100hz. The compliance of most cartridges decreases with increased frequency. This means that at very low frequencies the compliance will be higher than that specified. Again manufacturers aren't very forthcoming with details of how much. It cannot be very different at very low frequency or the standard equations for calculating arm resonance wouldn't work. As it can't be very different I am reasonably happy to work assuming that compliance does not vary.
Niffy
Compliance is a measurement used for springs and things that behave like springs. It is a measure of how much something will move for a given applied force. It is the reciprocal of stiffness which is a measure of how much force is required to move a given distance.
My cartridge has a compliance of 22um/mN. This mean that if I apply 1mN of side force to the stylus it will be deflected relative to the cartridge by 22um.
My bearings have an average friction of 0.931mN. This means that my carriage has to be pushed with a force of 0.931mN before it starts to move. This force is applied to the carriage by the stylus. For the carriage to move a force of 0.931mN must be applied to the stylus. This force will cause the stylus to move by
0.931mN x 22um/mN = 20.48um
The cantilever in my cartridge is about 5mm long (5000um). I had to measure this myself as its not something that manufacturers tend to stick in their spec sheets. It is difficult to accurately determine exactly where the pivot point in the cartridge is. I measured the length at between 5-6mm. I decided to use to shortest length as this would give the worst results. I prefer to under quote specifications rather than over quote as I aim to be as honest with myself as possible and avoid falling into the trap of confirmation bias.
From the length of the cantilever and the distance the stylus has moved the angle can be calculated using the sin rule.
Sin^-1(5000um/20.48um)= 0.23°
Arm length has no direct bearing on this equation. Arm length will probably have an effect on the amount of force required to overcome bearing friction. My test rig applied the side force at the same location as the stylus so this is already taken into account.
Most manufacturers quote dynamic compliance. This is the compliance at a set frequency, normally 315hz though sometimes 100hz. The compliance of most cartridges decreases with increased frequency. This means that at very low frequencies the compliance will be higher than that specified. Again manufacturers aren't very forthcoming with details of how much. It cannot be very different at very low frequency or the standard equations for calculating arm resonance wouldn't work. As it can't be very different I am reasonably happy to work assuming that compliance does not vary.
Niffy
Last edited: