Hi Ross, Abraham here. Could we leverage the capabilities that the beam expander give the RF tube on a glass tube? I believe the answer may be yes, what sort of cutting advantage we may be able to get. I think a beam expander would be very useful in particular on large machines, where the beam quality degrades the larger the beam travel distance is. Maybe that myth I often read about a 100W RF machine been equivalent to a 150W glass is all due to that magic beam expander.
Hi Abraham Bean divergence is a real issue fir RF tubes as it is about 3 times greater than glass tubes. However, glass tubes are not immune because their divergence is about 2.5mm per meter. So if your 100 watt tube starts off with a 6mm beam as it exits the tube and it is a big machine (say 1m in Y and 2m in X, in the extreme path leng situations it could vary between 0.5m to 3,5m. That means the nominal beam of a glass tube will change from (6+1.25)= 7.25mm to 6+ (2.5X3.5) = 14,75mm over the table area. That implies much better cutting near to the tube. There is no logical reason why you cannot use a beam expander with a glass tube. it is just a matter of finding a way to mount it axially true to the beam just as it exits the tube. I have demonstrated the CUTTING advantages of a compact bem hitting the lens but it will not change the power. 100 watts is 100 watts As I am beginning to find out, there is no magic in RF technology. it's just a VERY expensive way to get instant tube switching which is great for high speed ENGRAVING. For CUTTING you do not need high speed switching or a high speed machine, just CNC motion control and lots of power. I still consider the glass tube technology as being the perfect jack of all trades. You can easily afford to buy a small 50/60watt machine for engraving and a bigger high power machine for cutting and have lots of change left. Best wishes Russ
That is a near impossible question to answer because I have no idea of your needs. Is this for a small business that you hope to start? is this just for personal hobby use? What is the biggest thing you might want to fit on the table? Do you plan to mainly do cutting ? Do you live in a 20th floor apartment? For £3000 you can buy a reasonable machine but whatever you buy for this budget will have some issues. It will not be a simple plug and play machine. You will have to be prepared to do some work and probably buy a new tube to make it into a reliable machine. If this is your first machine, them maybe spend as little as possible and stay small (300x600 or600x400mm) and use it as a learning vehicle. In a year or so you will have a much clearer idea of what you REALLY need and have experience of the machine and the technology to make a more informed purchasing decision. If you wish to attach your disguised email address to another comment I will reply and you can send me links to machines that interest you and I will review them for you. Best wishes Russ
Hi again Russ Sorry to bring up further boring stuff... ;) I am not yet convinced... :o) I planned to try to do a video to better explain my understanding of it since your last video, but some C related illness struck the family so it got pushed back.. (no worries about the family this far... keeping our fingers crossed... ) I’l see if i can throw something together in the weekend... But in the mean time.... I fully agree with your glass tube related findings regarding analogue vs pwm of a hv-power supply. With the additional note that the tube itself might also add some slowness to the system response... ie it takes some time to energise the gas cloud in the tube tube and establish a beam, and vice versa, it takes some time for the built up energy to dissipate from the tube by the lazing action. But the pwm filtering stuff is as you said implemented as a low-pass filter at the psu control input. In the rf tube case, the filtering stuff is made up of the complete system, where the main filtering effect is taking place in the gas related physics of exciting atoms and getting them to emit photons... Regarding the proof of the rf-tube pwm issue, I still think my point is valid, the result is in line with what i expected, because you run the tube at a pwm frequency that is below the filter effect i spoke about. The specified 100+100us corresponds to 5kHz, to make use of the averaging effect the pwm frequency needs to be well above that. They specify 25kHz as the recommended max freq so go for that. If you repeat the red card test with all the same settings but only change the pwm frequency from 5 to 25kHz, and then repeat the test at 50 vs 100% power level I think you will see another result... actual smooth analogue power control. (I hope I'm not upsetting you with my nagging about the pwm... I’m just trying to reveal some of the things i have seen a little further down into the rabbit hole of electronics and physics. They are of course my interpretation of what I’ve seen, but i hope that those can spread some light and usefulness in this much interesting journey into co2 laser technology. I sure have learned a lot from your videos.) Best regards
Hi Klas Please take your time. Personal concerns assume a much a higher priority and be assured of my concern. I will never get frustrated or worried by rational and logical discussion about things I am willing to learn about, so please do not apologize. At present I have proved to my own satisfaction that that the raw PWM signals are not being filtered or interfered with but are driving the lasing process directly. I fully understand that the switching process for a glass tube may be virtually instant but then there is a significant delay while the flyback transformers to create the high voltage that in turn ionizes the nitrogen. The process of ionization, once triggered, is in itself almost instantaneous and starts creating photons immediately . The amplification of those photons to a usable power beam takes only a few reflections off the internal mirrors. Typically those mirrors will be about 1000 to 1200 mm apart. For a photon to travel such a distance takes about 3ns. Within an RF unit the distance is much shorter so lets say about 1ns travel time between mirrors. The excited CO2 electrons can only remain in their high energy state for about 1 millionth of a second before they randomly emit a photon while falling to their ground state. However that excited state is 1000 times longer than the rate passing photons that will STIMULATE their photon emission. With these times associated with the physics , provided you do not turn the ionization off, I can understand why the switching is almost instant. The whole point of RF technology is the pre ignition parameters that hold the nitrogen at the threshold of ionization so there is no noticeable switch on time. Even at 25khz where the cycle time is 40 microseconds, that is several magnitudes slower than the nanosecond physics. That is my very naive understanding of laser physics and it has been adequate to explain everything I have observed so far. There may well be practical physics phenomena that interfere with this oversimplified model and you are hinting at such conditions as the frequency increases.. But let me stop for a moment and ask how will it impact the use of this machine for cutting or engraving organic or mineral materials? For cutting the frequency is nullified because with 20 watts I will only be ever using 100% For photo engraving I will always be limited by the power available (as I have already found out) so that automatically causes me to run slowish (600 mm/s max) and setting low frequencies and generally 50% PWM to match power pulses to pixels. I have already proved that doubling the frequency may double the pulses per pixel but the net effect is 50% less burning and a very weak light result. Gray scale engraving is something I have yet to explore but I an very confident it will be slow(with 20 watts) but precise, provided I use multiple burns for 3D engraving. Photo engraving with greyscale is always going to be poor quality, not because of switching speed but because of the very limited tonal range available for the few organic materials that can be used. I have already tried this with the glass tube laser using the 20khz PWM signal and not switching the beam off . Very fuzzy results. My point really in all this is , for what reason will I ever need to use the high frequencies available? With 20 watts I will always be using low frequencies or CW. However it will be interesting to to see where my simplified model breaks down and for what reason.. Even if it will not affect my results, I am always happy to acquire new knowledge as it keeps my two remaining grey cells active. As I said, with your personal matters to attend to, there is no rush to educate me . I have been ignorant on this subject for ever, so another week or two will cause no harm. Sincere best wishes Russ
Hi. First thank you for the information, it is very helpful. Actually i was looking online for months in order to find any Hebrew translation of the machine's operation and functions. is there any chance that you will be able to help me with that? or alternatively provide me with any further ideas how to get something like this? Thanks!
The Ruida Controller can display in 17 different languages but I am afraid Hebrew is not one of them The onl option I can think of is to download a copy of the ruida RDC 6442 manual from my drop box link www.dropbox.com/s/b1aob1jb6bawuzd/RDC644XG_A%20Manual.pdf?dl=0 and work through it with Google translate. Best wishes Russ
this rabbit hole is starting to get very interesting even if I have no desire to go down the RF route. I am a glass tube guy into cutting, what really grabbed my attention was right at the end after your cutting test where you mentioned using a 7.5" lens coupled to a 2.5" lens. For some reason I seemed to have missed that RD Learning Labs video, can you tell me what number video it was on? On a seperate tack, a couiple of years ago I bought a top of the range desktop laser from HPC and assummed that a machine for nearly 4k pound would have been fully adjusted. I use it almost exclusively for prototyping small items and there fore concentrate all my work in one small area of the bed. For a long time I have been susoecting that something was amiss with the bed levelling so before moving the machine I decided to do a bed check as per your video from way back. What I found staggered me. The bed was totally skewed diagonally by over 5mm in one direction and nearly 7mm in the other diagonal, Just goes to show that you can not even trust a so called reputable UK company to do any real quality checks on their product, so why pay the premium for a UK sourced Red Sail Clone. The mis-alignment was so bad that the only way to correct it was to totally strip the entire bed lift mechanism, strip off the belts , remove the screw jacks and reset each by hand to a known reference. i also found under the belt adjusment that the base of the machine had been drilled multiple times arounf the locating bolts not what you would expect to find on a well set up machine, tere had certainly been a bit a of bodding going on, hopefully not in the UK, but who knows. Anyway, the bed is now perfectly levell and the machine cits well in all corners and in the middle. Two hours hard work turned an okay machine into almost a great machine. Okay enough waffle, keep the videos coming and help expand my knowledge of these facinating machines.
Hello Russ, first and foremost, thank you for enlightening me with your videos about lasers. They have been a good point of reference, since I just acquired a 60w Co2 Laser machine. I've been three or four nights watching all the videos of your blue china machine and the new tiger in Spanish subtitles. I have been amazed at how you gut to the maximum all the concepts that influence both cutting and engraving. Another thing that I saw in your videos a couple of days ago was to control the air in the nozzle with the example that you put by closing and directing the air jet ... Since then my recordings are much cleaner, which I thank you for. Heart, since the sellers as you say in one of your videos only tell you how it works, but they have no idea how to get the most out of the machine. I am a professional photographer and I only want my machine to engrave all my packaging. I see that getting a clean and perfect engraving is not easy. I would appreciate some advice on how to get the most out of my engraving machine. Thank you very much and greetings from Arahal Sevilla Spain
Russ, is there a contact email for contacting? I have an old epilog 25W Profile. I would like to discuss if you think the Ruida board would run it. Right now it has to have a LPT port and windows XP computer to run it. I would like to get it running again but use USB
Hi If you would care to add your disguised email to another comment I will reply and carry on a discussion off line. I will also delete your comment Best wishes Russ
Hi Russ, thanks for enlightening me with your series on lasers they have been a good point of reference. as I have just acquired a old dead machine and have completely rebuilt it as a lock-down project now working, looking forward to seeing what it can do.? just a briefly return of info PWM pulse width modulation is the time between the pulses change, what you are explaining with the sketch changing the height of the wave is AM aptitude modulation hope it helps.? I have just signed up to RDworks PM me if you need too. Great series.
Hi Ron From what I have seen so far there is no change of amplitude with PWM. It is a 0 and 5v square wave . As a mechanical engineer dipping my toe into an electronics ocean, there was a time when I thought there may be some amplitude variation as well to add another dimension to the PWM signal but thinking about it logically tells me that would destroy the whole principle ( Width Modulation). So I am sorry if there was some confusion in my video. Best wishes Russ
So, if i lower the nozzle from 8mm that i am currently, to 4mm, i should be able to run much much faster than 27mm/sec that i normally run... I think now i see the point of fiber lasers that the nozzle almost touches the work...
Hi Polis There will be an advantage for all materials but some materials like wood will see more gain than MDF ( lots of plastic plus sawdust) or acrylic.. It depends on your nozzle design (length of thread) that will determine what washer thickness youncannuse. Best wishes Russ
Yet another interesting video, Russ. Thanks for sharing. Although I am bit puzzled where is the video you mention at 39:02. I only remember that you have uploaded the taster for that group of tests.
Hi Nico You are correct. It was a quick exploratory taster and I have not yet found time to study the promising possibilities in depth. I have been cornered with so many other things. The good news (in a way) is that the fiber laser is no longer with me. It was a loan machine that has now had to be recalled because it was being used to fulfil an urgent sales order. That happened just right for me to work on designing and preparing parts for the Russ Spec machine to send to Cloudray. Just as I was clearing that task the RF unit arrived from Cloudray and that has forced me to work on the Tangerine tiger and get even more distracted with manufacturing all the parts for sale. I am just starting to get a bit clearer now but I need to push on with running through all the basic things that the RF machine is supposed to be able to do. I have not forgotten the 4" lens work because I am convinced that diverging a beam onto the lens will exploit a weakness in lens theory and that is that the closer you get to the lens axis the less you can focus and now associate that with the intensity distribution within our beam where the highest intensity is at the center of the beam. Bring these two factors together and it is already looking like there may be some significant CUTTING benefit to be had . Best wishes Russ
Hi Ross, Abraham here. Could we leverage the capabilities that the beam expander give the RF tube on a glass tube? I believe the answer may be yes, what sort of cutting advantage we may be able to get. I think a beam expander would be very useful in particular on large machines, where the beam quality degrades the larger the beam travel distance is. Maybe that myth I often read about a 100W RF machine been equivalent to a 150W glass is all due to that magic beam expander.
Hi Abraham
Bean divergence is a real issue fir RF tubes as it is about 3 times greater than glass tubes. However, glass tubes are not immune because their divergence is about 2.5mm per meter. So if your 100 watt tube starts off with a 6mm beam as it exits the tube and it is a big machine (say 1m in Y and 2m in X, in the extreme path leng situations it could vary between 0.5m to 3,5m. That means the nominal beam of a glass tube will change from (6+1.25)= 7.25mm to 6+ (2.5X3.5) = 14,75mm over the table area. That implies much better cutting near to the tube.
There is no logical reason why you cannot use a beam expander with a glass tube. it is just a matter of finding a way to mount it axially true to the beam just as it exits the tube. I have demonstrated the CUTTING advantages of a compact bem hitting the lens but it will not change the power. 100 watts is 100 watts As I am beginning to find out, there is no magic in RF technology. it's just a VERY expensive way to get instant tube switching which is great for high speed ENGRAVING. For CUTTING you do not need high speed switching or a high speed machine, just CNC motion control and lots of power. I still consider the glass tube technology as being the perfect jack of all trades. You can easily afford to buy a small 50/60watt machine for engraving and a bigger high power machine for cutting and have lots of change left.
Best wishes
Russ
If you had ~3,000£ to spend on a laser for cutting which would you buy?
That is a near impossible question to answer because I have no idea of your needs. Is this for a small business that you hope to start? is this just for personal hobby use? What is the biggest thing you might want to fit on the table? Do you plan to mainly do cutting ? Do you live in a 20th floor apartment? For £3000 you can buy a reasonable machine but whatever you buy for this budget will have some issues. It will not be a simple plug and play machine. You will have to be prepared to do some work and probably buy a new tube to make it into a reliable machine. If this is your first machine, them maybe spend as little as possible and stay small (300x600 or600x400mm) and use it as a learning vehicle. In a year or so you will have a much clearer idea of what you REALLY need and have experience of the machine and the technology to make a more informed purchasing decision.
If you wish to attach your disguised email address to another comment I will reply and you can send me links to machines that interest you and I will review them for you.
Best wishes
Russ
Hi again Russ
Sorry to bring up further boring stuff... ;)
I am not yet convinced... :o)
I planned to try to do a video to better explain my understanding of it since your last video, but some C related illness struck the family so it got pushed back.. (no worries about the family this far... keeping our fingers crossed... )
I’l see if i can throw something together in the weekend...
But in the mean time.... I fully agree with your glass tube related findings regarding analogue vs pwm of a hv-power supply. With the additional note that the tube itself might also add some slowness to the system response... ie it takes some time to energise the gas cloud in the tube tube and establish a beam, and vice versa, it takes some time for the built up energy to dissipate from the tube by the lazing action. But the pwm filtering stuff is as you said implemented as a low-pass filter at the psu control input. In the rf tube case, the filtering stuff is made up of the complete system, where the main filtering effect is taking place in the gas related physics of exciting atoms and getting them to emit photons...
Regarding the proof of the rf-tube pwm issue, I still think my point is valid, the result is in line with what i expected, because you run the tube at a pwm frequency that is below the filter effect i spoke about. The specified 100+100us corresponds to 5kHz, to make use of the averaging effect the pwm frequency needs to be well above that. They specify 25kHz as the recommended max freq so go for that. If you repeat the red card test with all the same settings but only change the pwm frequency from 5 to 25kHz, and then repeat the test at 50 vs 100% power level I think you will see another result... actual smooth analogue power control.
(I hope I'm not upsetting you with my nagging about the pwm... I’m just trying to reveal some of the things i have seen a little further down into the rabbit hole of electronics and physics. They are of course my interpretation of what I’ve seen, but i hope that those can spread some light and usefulness in this much interesting journey into co2 laser technology. I sure have learned a lot from your videos.)
Best regards
Hi Klas
Please take your time. Personal concerns assume a much a higher priority and be assured of my concern.
I will never get frustrated or worried by rational and logical discussion about things I am willing to learn about, so please do not apologize.
At present I have proved to my own satisfaction that that the raw PWM signals are not being filtered or interfered with but are driving the lasing process directly. I fully understand that the switching process for a glass tube may be virtually instant but then there is a significant delay while the flyback transformers to create the high voltage that in turn ionizes the nitrogen. The process of ionization, once triggered, is in itself almost instantaneous and starts creating photons immediately . The amplification of those photons to a usable power beam takes only a few reflections off the internal mirrors. Typically those mirrors will be about 1000 to 1200 mm apart. For a photon to travel such a distance takes about 3ns. Within an RF unit the distance is much shorter so lets say about 1ns travel time between mirrors. The excited CO2 electrons can only remain in their high energy state for about 1 millionth of a second before they randomly emit a photon while falling to their ground state. However that excited state is 1000 times longer than the rate passing photons that will STIMULATE their photon emission. With these times associated with the physics , provided you do not turn the ionization off, I can understand why the switching is almost instant. The whole point of RF technology is the pre ignition parameters that hold the nitrogen at the threshold of ionization so there is no noticeable switch on time. Even at 25khz where the cycle time is 40 microseconds, that is several magnitudes slower than the nanosecond physics.
That is my very naive understanding of laser physics and it has been adequate to explain everything I have observed so far. There may well be practical physics phenomena that interfere with this oversimplified model and you are hinting at such conditions as the frequency increases.. But let me stop for a moment and ask how will it impact the use of this machine for cutting or engraving organic or mineral materials? For cutting the frequency is nullified because with 20 watts I will only be ever using 100% For photo engraving I will always be limited by the power available (as I have already found out) so that automatically causes me to run slowish (600 mm/s max) and setting low frequencies and generally 50% PWM to match power pulses to pixels. I have already proved that doubling the frequency may double the pulses per pixel but the net effect is 50% less burning and a very weak light result. Gray scale engraving is something I have yet to explore but I an very confident it will be slow(with 20 watts) but precise, provided I use multiple burns for 3D engraving. Photo engraving with greyscale is always going to be poor quality, not because of switching speed but because of the very limited tonal range available for the few organic materials that can be used. I have already tried this with the glass tube laser using the 20khz PWM signal and not switching the beam off . Very fuzzy results.
My point really in all this is , for what reason will I ever need to use the high frequencies available? With 20 watts I will always be using low frequencies or CW. However it will be interesting to to see where my simplified model breaks down and for what reason.. Even if it will not affect my results, I am always happy to acquire new knowledge as it keeps my two remaining grey cells active.
As I said, with your personal matters to attend to, there is no rush to educate me . I have been ignorant on this subject for ever, so another week or two will cause no harm.
Sincere best wishes
Russ
Hi. First thank you for the information, it is very helpful. Actually i was looking online for months in order to find any Hebrew translation of the machine's operation and functions. is there any chance that you will be able to help me with that? or alternatively provide me with any further ideas how to get something like this? Thanks!
The Ruida Controller can display in 17 different languages but I am afraid Hebrew is not one of them The onl option I can think of is to download a copy of the ruida RDC 6442 manual from my drop box link www.dropbox.com/s/b1aob1jb6bawuzd/RDC644XG_A%20Manual.pdf?dl=0 and work through it with Google translate.
Best wishes
Russ
this rabbit hole is starting to get very interesting even if I have no desire to go down the RF route. I am a glass tube guy into cutting, what really grabbed my attention was right at the end after your cutting test where you mentioned using a 7.5" lens coupled to a 2.5" lens. For some reason I seemed to have missed that RD Learning Labs video, can you tell me what number video it was on? On a seperate tack, a couiple of years ago I bought a top of the range desktop laser from HPC and assummed that a machine for nearly 4k pound would have been fully adjusted. I use it almost exclusively for prototyping small items and there fore concentrate all my work in one small area of the bed. For a long time I have been susoecting that something was amiss with the bed levelling so before moving the machine I decided to do a bed check as per your video from way back. What I found staggered me. The bed was totally skewed diagonally by over 5mm in one direction and nearly 7mm in the other diagonal, Just goes to show that you can not even trust a so called reputable UK company to do any real quality checks on their product, so why pay the premium for a UK sourced Red Sail Clone. The mis-alignment was so bad that the only way to correct it was to totally strip the entire bed lift mechanism, strip off the belts , remove the screw jacks and reset each by hand to a known reference. i also found under the belt adjusment that the base of the machine had been drilled multiple times arounf the locating bolts not what you would expect to find on a well set up machine, tere had certainly been a bit a of bodding going on, hopefully not in the UK, but who knows. Anyway, the bed is now perfectly levell and the machine cits well in all corners and in the middle. Two hours hard work turned an okay machine into almost a great machine. Okay enough waffle, keep the videos coming and help expand my knowledge of these facinating machines.
Hello Russ, first and foremost, thank you for enlightening me with your videos about lasers. They have been a good point of reference, since I just acquired a 60w Co2 Laser machine. I've been three or four nights watching all the videos of your blue china machine and the new tiger in Spanish subtitles. I have been amazed at how you gut to the maximum all the concepts that influence both cutting and engraving. Another thing that I saw in your videos a couple of days ago was to control the air in the nozzle with the example that you put by closing and directing the air jet ... Since then my recordings are much cleaner, which I thank you for. Heart, since the sellers as you say in one of your videos only tell you how it works, but they have no idea how to get the most out of the machine. I am a professional photographer and I only want my machine to engrave all my packaging. I see that getting a clean and perfect engraving is not easy. I would appreciate some advice on how to get the most out of my engraving machine. Thank you very much and greetings from Arahal Sevilla Spain
Hi Manuel
If you like to add your disguised email address to another comment I will contact you and delete your comment
Best wishes
Russ
Russ, is there a contact email for contacting? I have an old epilog 25W Profile. I would like to discuss if you think the Ruida board would run it. Right now it has to have a LPT port and windows XP computer to run it. I would like to get it running again but use USB
Hi
If you would care to add your disguised email to another comment I will reply and carry on a discussion off line. I will also delete your comment
Best wishes
Russ
Hi Russ, thanks for enlightening me with your series on lasers they have been a good point of reference. as I have just acquired a old dead machine and have completely rebuilt it as a lock-down project now working, looking forward to seeing what it can do.? just a briefly return of info PWM pulse width modulation is the time between the pulses change, what you are explaining with the sketch changing the height of the wave is AM aptitude modulation hope it helps.? I have just signed up to RDworks PM me if you need too. Great series.
Hi Ron
From what I have seen so far there is no change of amplitude with PWM. It is a 0 and 5v square wave . As a mechanical engineer dipping my toe into an electronics ocean, there was a time when I thought there may be some amplitude variation as well to add another dimension to the PWM signal but thinking about it logically tells me that would destroy the whole principle ( Width Modulation). So I am sorry if there was some confusion in my video.
Best wishes
Russ
So, if i lower the nozzle from 8mm that i am currently, to 4mm, i should be able to run much much faster than 27mm/sec that i normally run... I think now i see the point of fiber lasers that the nozzle almost touches the work...
Hi Polis
There will be an advantage for all materials but some materials like wood will see more gain than MDF ( lots of plastic plus sawdust) or acrylic.. It depends on your nozzle design (length of thread) that will determine what washer thickness youncannuse.
Best wishes
Russ
Yet another interesting video, Russ. Thanks for sharing. Although I am bit puzzled where is the video you mention at 39:02. I only remember that you have uploaded the taster for that group of tests.
Hi Nico
You are correct. It was a quick exploratory taster and I have not yet found time to study the promising possibilities in depth. I have been cornered with so many other things. The good news (in a way) is that the fiber laser is no longer with me. It was a loan machine that has now had to be recalled because it was being used to fulfil an urgent sales order. That happened just right for me to work on designing and preparing parts for the Russ Spec machine to send to Cloudray. Just as I was clearing that task the RF unit arrived from Cloudray and that has forced me to work on the Tangerine tiger and get even more distracted with manufacturing all the parts for sale. I am just starting to get a bit clearer now but I need to push on with running through all the basic things that the RF machine is supposed to be able to do. I have not forgotten the 4" lens work because I am convinced that diverging a beam onto the lens will exploit a weakness in lens theory and that is that the closer you get to the lens axis the less you can focus and now associate that with the intensity distribution within our beam where the highest intensity is at the center of the beam. Bring these two factors together and it is already looking like there may be some significant CUTTING benefit to be had .
Best wishes
Russ
Love your work but, as an electronics engineer, your video gave me a headache.
Round and round in too many circles.