Mongoose DX project
- Thread starter oofnik
- Start date
TylerDurden
100 GW
Check out Miles' thread: http://endless-sphere.com/forums/viewtopic.php?f=1&t=10063oofnik said:
recumpence
1 GW
oofnik said:
That is the exact impeller I made for my AXI.
It would take about and hour.
Matt
Updated version, now with more air volume and angled vanes:
I don't know how much (if at all) the 30 deg. angle helps, but most impeller designs incorporate angled vanes so it must be a good idea.
I'm looking for quotes, emailing machine shops that are listed on CL, talking to friends at other engineering schools with less red tape than mine (pretty much any other engineering school in the country) ... arg. It looks like my friend at Virginia Tech will be able to make it for me at little to no cost though. He's involved with their robotics department (RoMeLa) - awesome stuff. I hope he can follow through!
I don't know how much (if at all) the 30 deg. angle helps, but most impeller designs incorporate angled vanes so it must be a good idea.
I'm looking for quotes, emailing machine shops that are listed on CL, talking to friends at other engineering schools with less red tape than mine (pretty much any other engineering school in the country) ... arg. It looks like my friend at Virginia Tech will be able to make it for me at little to no cost though. He's involved with their robotics department (RoMeLa) - awesome stuff. I hope he can follow through!
More update - I just spent the last several hours learning ExpressPCB / ExpressSCH retracing Fechter's throttle circuit. Not fun. 
But I got my own layout done that's a little more compact; I've omitted the current sensor, the caps, and the regulator circuit as I already have that stuff in my electronics box. Once I get this board etched and soldered up I will have a drop-in replacement for my servo tester with current limiting functionality.
That is, if I'm lucky enough that it works. Me and electronics don't get along so well most of the time. :?
But I got my own layout done that's a little more compact; I've omitted the current sensor, the caps, and the regulator circuit as I already have that stuff in my electronics box. Once I get this board etched and soldered up I will have a drop-in replacement for my servo tester with current limiting functionality.That is, if I'm lucky enough that it works. Me and electronics don't get along so well most of the time. :?
recumpence
1 GW
Oofnik,
What is your downtube diameter? As I mentioned in my PM to you, I have a 1.5 inch diameter prototype clamp set here. If these will work for you, I will ship them to you at no charge. You can keep them. I would like your input on them before I produce them. That is, of course, if they will work for you.
Matt
What is your downtube diameter? As I mentioned in my PM to you, I have a 1.5 inch diameter prototype clamp set here. If these will work for you, I will ship them to you at no charge. You can keep them. I would like your input on them before I produce them. That is, of course, if they will work for you.
Matt
I'm glad I gave you the impression that I know what I'm doing.swbluto said:
I'm actually just getting done with my second year in undergraduate mechanical engineering. Electrical stuff has always interested me, but I decided that I was more comfortable with mechanical. I've built several little electronics projects that have taught me a lot of the basics but I wouldn't call it in-depth knowledge..For transferring the board layout I plan on using a Sharpie. Yup. Print out the silk screen, overlay it on the copper, drill the holes, then just sharpie the connections front and back then etch.
Matt - awfully kind of you to offer! Unfortunately my downtube is of a strange shape, kind of a semicircle with an oblong top. I posted a drawing of the profile back on page 6. The round part diameter is 1-3/8". The top tube is round but with the same diameter. Perhaps I could make some wood spacers to put in between the clamps and the frame though? Hmm..
recumpence
1 GW
The clamps go down to that size without any shims. However, if you want shims, you can get 1.385 to 1.5 shims.
Matt
Matt
Oofnik,
In what area did you have problems with the magnets coming loose on your 5330 motor? I took my motor apart and noticed the magnets are set into machined pockets on one side of the casing. I assume they moved (slid over to one side) on the other side? If that's the case I plan on putting some dabs of epoxy between the magnets only on the side without the pockets.
My motor actually has what looks like a lot of epoxy under the magnets already. More than I expected anyway.
In what area did you have problems with the magnets coming loose on your 5330 motor? I took my motor apart and noticed the magnets are set into machined pockets on one side of the casing. I assume they moved (slid over to one side) on the other side? If that's the case I plan on putting some dabs of epoxy between the magnets only on the side without the pockets.
My motor actually has what looks like a lot of epoxy under the magnets already. More than I expected anyway.
A few of my magnets tended to slip outward from the bell due to the great force from the stator trying to pull the bell inward.
Maybe they started reinforcing the magnet glue on later revisions.. I definitely didn't see any epoxy on my magnets. Hm.
If it looks solid to you I'd say try to run it as is. If they come unglued you can always re-glue them. If you use CA glue just don't be an idiot like me - push them back in to place securely before applying the glue, NOT after! :? The stuff sets within seconds.
Maybe they started reinforcing the magnet glue on later revisions.. I definitely didn't see any epoxy on my magnets. Hm.
If it looks solid to you I'd say try to run it as is. If they come unglued you can always re-glue them. If you use CA glue just don't be an idiot like me - push them back in to place securely before applying the glue, NOT after! :? The stuff sets within seconds.
I finally got around to securing the battery pack better using tie downs instead of bungee cords.
What prompted me was my ride today where one of the bungees snapped and wound itself around my derailleur several times, becoming devoured in the process. But everything's okay, no bent/damaged parts.
The batteries don't jump around at all any more with the tie downs.
Also, I finally took an official, accurate efficiency measurement. According to the amp / volt data from the logger plus the distance from my bike computer, I used 9122 mAh to travel 8.30 miles. Numerically integrating the current and voltage in a spreadsheet yielded almost exactly 300 Wh of energy consumed. So that's... oh jeez.. 36 Wh per mile.
Wow.
Now that's with full throttle 80% of the time, minimal pedaling, mostly acceleration. Believe me, my brakes weren't much happier than my motor. I was just blasting it. So I guess that's not so so bad for blasting full throttle everywhere on a 3kW setup, right? Oh yeah and it was really windy today too.
On another note, my batteries are getting super hot.. hot enough where it's uncomfortable to touch them. I know Headway's data sheet lists up to 85 C operating temp, but I don't like that. So I need to either stop blasting (NO!!), build my A123 booster pack, or allow heat to escape. I'm going to probably combine option 2 and 3. I'm also going to make all efforts to keep the battery compartment sealed, so I may just replace the top panel of the box with an aluminum plate. Right now, it's plastic, MDF and felt, which all insulate very well. The aluminum should allow a majority of the heat to escape.
edit: I attached the logger data so you guys can see how I arrived at the Wh measurement if you're curious. Feel free to plot it and look at the pretty graph too.
What prompted me was my ride today where one of the bungees snapped and wound itself around my derailleur several times, becoming devoured in the process.
But everything's okay, no bent/damaged parts.The batteries don't jump around at all any more with the tie downs.
Also, I finally took an official, accurate efficiency measurement. According to the amp / volt data from the logger plus the distance from my bike computer, I used 9122 mAh to travel 8.30 miles. Numerically integrating the current and voltage in a spreadsheet yielded almost exactly 300 Wh of energy consumed. So that's... oh jeez.. 36 Wh per mile.
Wow.Now that's with full throttle 80% of the time, minimal pedaling, mostly acceleration. Believe me, my brakes weren't much happier than my motor. I was just blasting it. So I guess that's not so so bad for blasting full throttle everywhere on a 3kW setup, right? Oh yeah and it was really windy today too.
On another note, my batteries are getting super hot.. hot enough where it's uncomfortable to touch them. I know Headway's data sheet lists up to 85 C operating temp, but I don't like that. So I need to either stop blasting (NO!!), build my A123 booster pack, or allow heat to escape. I'm going to probably combine option 2 and 3. I'm also going to make all efforts to keep the battery compartment sealed, so I may just replace the top panel of the box with an aluminum plate. Right now, it's plastic, MDF and felt, which all insulate very well. The aluminum should allow a majority of the heat to escape.
edit: I attached the logger data so you guys can see how I arrived at the Wh measurement if you're curious. Feel free to plot it and look at the pretty graph too.
oofnik said:
I was going over this again because your response is fascinating. When you say silkscreen, do you mean actual white ink that's commonly used to denote and identify parts on a board? If so, tell me you how you do it, please, pretty please with sugar on top.
You know, now that you mention it, I probably could print out the layout on one of those iron-on t-shirt transfer sheets, and then use a copper board instead of a t-shirt.
*15 seconds of googling*
There are so many ways to do this. Google "iron on pcb" or "inkjet pcb". Fascinating. I'll have to look in to exactly how I want to do this.
So I think I found my favorite method: http://www.riccibitti.com/pcb/pcb.htm. Magazine paper, laser printer, clothes iron, etching chemical. That's it.
*15 seconds of googling*
There are so many ways to do this. Google "iron on pcb" or "inkjet pcb". Fascinating. I'll have to look in to exactly how I want to do this.
So I think I found my favorite method: http://www.riccibitti.com/pcb/pcb.htm. Magazine paper, laser printer, clothes iron, etching chemical. That's it.
Uhhhh... I thought silkscreen was typically white ink that makes parts on a pcb board identifiable(i.e., C1, R2, L43, etc.). I might be wrong, though.
Anyways, yeah, I'm looking into actually doing "print on" transfers. I just ordered a laser jet printer that appears like it's suitable for PCB layout(For some reason, Brother's toners don't work and the newer HP P-series uses a toner that doesn't work). I plan to try magazine covers, avery mailing labels, officeMax's Glossy Photo paper, Press n' peel and an iron until finding something that repeatably works with 10 mil traces. Apparently there's a "photo etching" method for transferring a layout, but the per-unit costs and start-up cost seems significantly higher, despite having a reputation for greater repeatability and consistency. But, if it appears that the iron-on method doesn't work that well, I'll probably move to the UV realm.
Anyways, I've been interested in some kind of homebrew silk screen method. Some green solder masking would also be nice, but I think I'll leave that to the professionals.
Anyways, yeah, I'm looking into actually doing "print on" transfers. I just ordered a laser jet printer that appears like it's suitable for PCB layout(For some reason, Brother's toners don't work and the newer HP P-series uses a toner that doesn't work). I plan to try magazine covers, avery mailing labels, officeMax's Glossy Photo paper, Press n' peel and an iron until finding something that repeatably works with 10 mil traces. Apparently there's a "photo etching" method for transferring a layout, but the per-unit costs and start-up cost seems significantly higher, despite having a reputation for greater repeatability and consistency. But, if it appears that the iron-on method doesn't work that well, I'll probably move to the UV realm.
Anyways, I've been interested in some kind of homebrew silk screen method. Some green solder masking would also be nice, but I think I'll leave that to the professionals.
gwhy!
100 kW
oofnik said:
I have used this method and it works very well, it tends not to be as clean as uv exposure through a transparency but its very good for prototyping and small pcb runs as long as the layout isnt to tightly populated what i mean by this is taking tracks through ic pins ( can be done but check carefully after etching ) :wink:
edit: just read the link... I must try the magazine stuff
I have only used the the ironon pcb paper that you can buy. thanks for the link...
recumpence
1 GW
oofnik said:I finally got around to securing the battery pack better using tie downs instead of bungee cords.
What prompted me was my ride today where one of the bungees snapped and wound itself around my derailleur several times, becoming devoured in the process.
The batteries don't jump around at all any more with the tie downs.
Also, I finally took an official, accurate efficiency measurement. According to the amp / volt data from the logger plus the distance from my bike computer, I used 9122 mAh to travel 8.30 miles. Numerically integrating the current and voltage in a spreadsheet yielded almost exactly 300 Wh of energy consumed. So that's... oh jeez.. 36 Wh per mile.
Now that's with full throttle 80% of the time, minimal pedaling, mostly acceleration. Believe me, my brakes weren't much happier than my motor. I was just blasting it. So I guess that's not so so bad for blasting full throttle everywhere on a 3kW setup, right? Oh yeah and it was really windy today too.
On another note, my batteries are getting super hot.. hot enough where it's uncomfortable to touch them. I know Headway's data sheet lists up to 85 C operating temp, but I don't like that. So I need to either stop blasting (NO!!), build my A123 booster pack, or allow heat to escape. I'm going to probably combine option 2 and 3. I'm also going to make all efforts to keep the battery compartment sealed, so I may just replace the top panel of the box with an aluminum plate. Right now, it's plastic, MDF and felt, which all insulate very well. The aluminum should allow a majority of the heat to escape.
edit: I attached the logger data so you guys can see how I arrived at the Wh measurement if you're curious. Feel free to plot it and look at the pretty graph too.
Your efficiency sounds about right. My recumbent is about the same at full throttle with alot of accelleration. Normaly cruising at 20mph will cut your WH per mile in half at least. My guess is you should see about 18 WH per mile at 18 to 20mph without any hard accelleration. That would be typical for an RC setup.
Those hard launches really eat up the WH.
Matt
gwhy!
100 kW
swbluto said:
Yes clearances/spacing , dont go to thin with the track width if you really dont have to and try to avoid having one copper track to close to another, as I said it is possible to route a track through 2 ic pads without touching but it is a 'fine line'
no pun intended it depends on how clean the edges are on the track after its been transferred to the copper and then you have to be careful not to over etch the copper, the only real way of finding out what is the min you can get way with this method is to experiment and get a feel for it. Any tracks that are very close to another track always check with a meter that there isnt any shorts before you start populating the board. The real problem I have found using this method is pad pilot holes not being transferred to the copper which can make drilling the holes a bit tricky, you are better off using a bigger pad hole than standard when drawing the pcb. Trial and error is the best way with the materials you are using to find out what you can get away with. Hope this helps. I have made double sided pcbs using the iron on pcb paper that have come out rather well but getting the alignment is a bit tricky but can be done.
RWP
1 kW
What a great project!
I am using the same Towerpro motor and just yesterday three of the magnets came loose and hit the mount flange. Tried to get all the magnets out but none of the others seemed to want to come loose on the bench. I mixed a batch of JB Weld and slathered it into the notches between the magnets and a bit on the magnet ends with a coffee stir. I hope this will keep the magnets in place. I will let you know.
I am so impressed with all the fine work here at ES!!! And the community is over the top.
Roy
I am using the same Towerpro motor and just yesterday three of the magnets came loose and hit the mount flange. Tried to get all the magnets out but none of the others seemed to want to come loose on the bench. I mixed a batch of JB Weld and slathered it into the notches between the magnets and a bit on the magnet ends with a coffee stir. I hope this will keep the magnets in place. I will let you know.
I am so impressed with all the fine work here at ES!!! And the community is over the top.
Roy
I just glued mine (before using it) with a little extra epoxy. It really looks to me like most places recommend epoxy for gluing magnets. I'm interested to find out which glue ends up working. I'm almost 100% sure the factory used epoxy on the Towerpro motors, just not quite enough.
Here's a pdf file I've been looking at that talks about using epoxy (I don't think it says what kind though) mixed with microbubbles to hold the magnets:
http://www.torcman.de/index_e.htm Click on the file that says "Assembly Procedure TM280, 350, 430".
Here's a pdf file I've been looking at that talks about using epoxy (I don't think it says what kind though) mixed with microbubbles to hold the magnets:
http://www.torcman.de/index_e.htm Click on the file that says "Assembly Procedure TM280, 350, 430".
