StromTrooper banner

1 - 20 of 78 Posts

·
Registered
Joined
·
441 Posts
Discussion Starter #1 (Edited)
Well, about a month ago I was kind of commissioned to make a Yosh Box to adjust a bunch of San Diego area member bikes. Obviously, the idea is to make an idle adjustment only so I went to the task of trying to read the schematic from this post at VSRI and make it work for my application: Home Made Yosh Box - mum told you not to do it!

Key Features:
  • Idle Circuit Only
  • Only for V Twins (no cyl 3 or 4)
  • Uses as cheap of components as possible (eliminates rotary switch)
  • Uses small board; 25 x 15 holes
Beyond that, if someone can tell me if it looks right when comparing it to the schematic on page 2 here, it would be greatly appreciated. http://www.vstrom.info/forum/forum_files/Manuals/Box_Instructions.pdf

Thanks a ton! Oh, the image below sucks, but gives the idea. See the attachment for a better quality version.

 

Attachments

·
Registered
Joined
·
441 Posts
Discussion Starter #2
Well, I guess I will just build it myself. Most folks must just dump the cash on the Power Commander.

Call me, Frugal McDougal.
 

·
Registered
Joined
·
7 Posts
Interesting idea. Idle only Yosh Box. I like it. Seems to look ok to me. Not sure what type of power LED you intend to use, just make sure you have the proper resistance for your current limiting resistor with that LED. It is easy to visualize when the unused components are removed from the schematic. As long as the voltages are within spec once you power up and test the voltages per the instructions.
 

·
Registered
Joined
·
441 Posts
Discussion Starter #4 (Edited)
Interesting idea. Idle only Yosh Box. I like it. Seems to look ok to me. Not sure what type of power LED you intend to use, just make sure you have the proper resistance for your current limiting resistor with that LED. It is easy to visualize when the unused components are removed from the schematic. As long as the voltages are within spec once you power up and test the voltages per the instructions.

Thanks a ton for responding. I have been holding off a bit waiting for someone to confirm that my mods to this schematic are not totally insane.

Great idea whiting out the regular schematic. Doh! :hurray:

I will be heading to fry's electronics tomorrow and picking up the parts. The German Schematic located here show the led the way I did it. Not sure which type of LED to pick up but it will be small and red.

Oh, and I even added another part to the schematic to put test points on the top of the box for pins 4 & 5 so you can see exactly how much voltage you are sending to the ECU with the Rich/Lean knob. Never seen it done, but have seen it talked about. That way you can really pin point how much you are richening the mixture. It is a bit touchy as I understand.
 

·
Registered
Joined
·
30 Posts
330ohm may be a little low for the power LED. Thats' 36mA of current by my calculations. In general its better to design for 15mA , so 800ish ohms would be better(best to design with the specs for a particular LED, but 15mA is a good guess for arbitrary led). Then again, its not exactly a high precision part of the circuit.

You could probably do without the 0.1uF cap, if you wanted to save components. I'd have to do some research into the exact specification for that circuit, but caps are usually +/- 20%... so that 22uF is going to be +/- 4.4uF... no real gain in adding 0.1uF.

From your diagram, I cant tell if you intend to mount the switches and pot or not. I highly recommend it. I've worked with rigs with floating switches held up by wires, it doesn't go well :headbang:
 

·
Registered
Joined
·
7 Posts
Thanks a ton for responding. I have been holding off a bit waiting for someone to confirm that my mods to this schematic are not totally insane.

Great idea whiting out the regular schematic. Doh! :hurray:

I will be heading to fry's electronics tomorrow and picking up the parts. The German Schematic located here show the led the way I did it. Not sure which type of LED to pick up but it will be small and red.

Oh, and I even added another part to the schematic to put test points on the top of the box for pins 4 & 5 so you can see exactly how much voltage you are sending to the ECU with the Rich/Lean knob. Never seen it done, but have seen it talked about. That way you can really pin point how much you are richening the mixture. It is a bit touchy as I understand.
I like the idea of being able to read and log actual voltages when settings are applied. :thumbsup:
 

·
Registered
Joined
·
441 Posts
Discussion Starter #7
330ohm may be a little low for the power LED. Thats' 36mA of current by my calculations. In general its better to design for 15mA , so 800ish ohms would be better(best to design with the specs for a particular LED, but 15mA is a good guess for arbitrary led). Then again, its not exactly a high precision part of the circuit.

You could probably do without the 0.1uF cap, if you wanted to save components. I'd have to do some research into the exact specification for that circuit, but caps are usually +/- 20%... so that 22uF is going to be +/- 4.4uF... no real gain in adding 0.1uF.

From your diagram, I cant tell if you intend to mount the switches and pot or not. I highly recommend it. I've worked with rigs with floating switches held up by wires, it doesn't go well :headbang:
Is that based on 5 volts? If so, I will grab a 1k just in case. I think I will keep the 0.1uF cap because it is there to begin with. Like you, not sure what the purpose of the small value cap is, but I can't imagine it would impact things. Maybe it is a secondary filter or something.

I was planning on floating the switches and stuff mainly because the board I am using is REALLY small. To combat the issue, I was going to run long enough wires that the switches and pot can wire up to the top part of the case and mount the board to the bottom. I am used to working on guitar amps with tons of floating leads. Maybe you can elaborate as to why I shouldn't do that? I am totally open to any ideas.
 

·
Registered
Joined
·
441 Posts
Discussion Starter #8
I like the idea of being able to read and log actual voltages when settings are applied. :thumbsup:
Yeah, I can't imagine how that would hurt things. My bike is over rich right now so it is killing my fuel mileage. But I hope to find a sweet spot by actually being precise in adjustment. A silly knob is not accurate enough! :thumbdown:
 

·
Registered
Joined
·
522 Posts
I guess I must be kinda stupid. I bought a used Yosh Box for $200 and did my bike now it runs properly. I'll do a few friends for free and a few others for $$. Then I'll have ZERO dollars invested in mine. Why try and re-invent the wheel.
 

·
Registered
Joined
·
30 Posts
Is that based on 5 volts? If so, I will grab a 1k just in case. I think I will keep the 0.1uF cap because it is there to begin with. Like you, not sure what the purpose of the small value cap is, but I can't imagine it would impact things. Maybe it is a secondary filter or something.

I was planning on floating the switches and stuff mainly because the board I am using is REALLY small. To combat the issue, I was going to run long enough wires that the switches and pot can wire up to the top part of the case and mount the board to the bottom. I am used to working on guitar amps with tons of floating leads. Maybe you can elaborate as to why I shouldn't do that? I am totally open to any ideas.
I assumed 12V for that LED measurement. R = V/I, I = 15mA V = 12V R=800ohm. That way, if I'm wrong and its a 5V line, at worst your LED is just dim. You should be able to measure the voltage for that line with a meter.

If you can mount the switches to the top of the case, that will work just fine. I find that, without mounting of some kind, you end up stressing solder joints by moving wires back and forth and then spend an hour trying to figure out why the mixture just isn't adjusting because a joint broke (or worse, you used solid core wiring and broke the wire inside its insulation... there's an un-fun story behind that one!)

The only thing I can think of the 0.1uF doing is being an ultra-high frequency filter (on the order of GHz probably). Radial caps have some inductance to them, and the parts list shows the 22uF being a radial. If someone (like the original designer of the circuit) thinks that is needed, be sure to mount the cap with as short of leads as possible. That will minimize inductance, but I have a sneaky suspicion this is a case of engineer overkill where the guy who wrote the schematic just always filters his power lines like that.

How much will this device be used? If you think it will be used several hundred times by finicky users, tuning to within the same tiny range on the pot each time, I'd consider adding a fine-tuning knob (connect just the wiper and one end of the smaller pot between the large pot and pin 3). Its kinda overkill on-par with the 0.1uF cap, but if we're in the business of over-killing, I had a sub-woofer that I kept adjusting the gain on, back and forth, a few degrees each way (I'm an audiophile, cut me some slack!). Eventually I ate a divot in the resistive material or something, and ended up with some pretty strange resistance readings right where i kept trying to set it. A fine control will help spread that wear out. Alternatively, when a user complains that they can't get a particular voltage across the test pins, simply offer to replace the pot.
 

·
Registered
Joined
·
441 Posts
Discussion Starter #11
I guess I must be kinda stupid. I bought a used Yosh Box for $200 and did my bike now it runs properly. I'll do a few friends for free and a few others for $$. Then I'll have ZERO dollars invested in mine. Why try and re-invent the wheel.
Well, I haven't been able to find a Yosh Box and for me, the initial cash outlay of even $200 bucks is steep. You aren't stupid, you are just in a better financial position than me! :thumbup:
 

·
Registered
Joined
·
441 Posts
Discussion Starter #12
I assumed 12V for that LED measurement. R = V/I, I = 15mA V = 12V R=800ohm. That way, if I'm wrong and its a 5V line, at worst your LED is just dim. You should be able to measure the voltage for that line with a meter.
Roger that. I recall reading in a lot of my research that the power coming into the box is 5V. Actually, the testing procedures in the original write up say to get a 5V power supply. Or, just hook it up to the bike but only pins 3 & 4 (I think) to get the power for voltages.

If you can mount the switches to the top of the case, that will work just fine. I find that, without mounting of some kind, you end up stressing solder joints by moving wires back and forth and then spend an hour trying to figure out why the mixture just isn't adjusting because a joint broke (or worse, you used solid core wiring and broke the wire inside its insulation... there's an un-fun story behind that one!)
Yeah, I definitely plan to mount the switches up nice and sturdy. I will be using guitar stomp box materials since those are built to be very tough. These things get kicked around all over stages like crazy. If I had the time, money and inclination, I would do a true PCB with a custom etch and solid mount everything like the Boss pedals are done. Super tough!

The only thing I can think of the 0.1uF doing is being an ultra-high frequency filter (on the order of GHz probably). Radial caps have some inductance to them, and the parts list shows the 22uF being a radial. If someone (like the original designer of the circuit) thinks that is needed, be sure to mount the cap with as short of leads as possible. That will minimize inductance, but I have a sneaky suspicion this is a case of engineer overkill where the guy who wrote the schematic just always filters his power lines like that.
GREAT info there! You are quite the EE guy, aren't you? I am just a DIY tinkerer and dabble in this stuff for fun. I recalled that on guitar amps, especially tube amps, they use big ass capacitors to filter the power supply before hitting the tubes. So I figured this must be the same thing. I too thought that a cap that small must be over kill, but it must be there for a reason. I will make sure to cut that thing down small before soldering it. I might even move it over one set of eyelets to make sure.

How much will this device be used? If you think it will be used several hundred times by finicky users, tuning to within the same tiny range on the pot each time, I'd consider adding a fine-tuning knob (connect just the wiper and one end of the smaller pot between the large pot and pin 3). Its kinda overkill on-par with the 0.1uF cap, but if we're in the business of over-killing, I had a sub-woofer that I kept adjusting the gain on, back and forth, a few degrees each way (I'm an audiophile, cut me some slack!). Eventually I ate a divot in the resistive material or something, and ended up with some pretty strange resistance readings right where i kept trying to set it. A fine control will help spread that wear out. Alternatively, when a user complains that they can't get a particular voltage across the test pins, simply offer to replace the pot.
I don't think it will be used hundreds of times at all, but you do have a very cool idea. It can be modded in at a later time too if finding out that voltages are difficult to move a small amount. Which might be the case since some folks complained a lot that the knob was very sensitive. Mitigating a large change in voltage seems like a great idea.

Since we are dealing with a base line of 2.5v and a swing up to 5v and down to 0v, is there a resistor I can throw in to slow the effect of a knob twist or is that effectively what the 2nd pot will be doing (since a pot is a variable resistor in the first place!)?
 

·
Registered
Joined
·
30 Posts
Yeah, I picked up one of those EE degrees. Don't use it, mind you, but its fun to have!

As for the pots, there isn't much one can do to slow down the voltage swing across a pot (other than buying a worm-drive pot that does multiple turns to start with). In the end, the voltage on pin 5 is solely dependent on the difference between the resistance from ground to wiper and the resistance from power to wiper. The idea of having 2 pots is to have one that swings the full 10kohm, and one that may swing only 1kohm, using the smaller one to change the resistance on one side. Here's how I'd connect it:
+5V (if indeed it is +5V) connects to one leg of the 1kohm worm-gear pot
the wiper of the 1kohm worm-gear pot connects to one leg of the 10kohm pot
the other leg of the 10kohm pot connects to ground
the wiper for the 10kohm pot connects to pin5, as you have it

It doesn't really matter if the 1kohm pot is on the ground side or the +5V side, so I arbitrarily put it on the high side. One leg of the 1kohm pot is not connected because its not needed. The one disadvantage of that setup is that its hard to tune all the way to +5V or 0V because you have to turn both pots to the extremes to do that, but from what I read, anyone tuning their idle circuit to the extreme probably needs their engine looked at anyways!
 

·
Registered
Joined
·
441 Posts
Discussion Starter #14
Yeah, I picked up one of those EE degrees. Don't use it, mind you, but its fun to have!

As for the pots, there isn't much one can do to slow down the voltage swing across a pot (other than buying a worm-drive pot that does multiple turns to start with). In the end, the voltage on pin 5 is solely dependent on the difference between the resistance from ground to wiper and the resistance from power to wiper. The idea of having 2 pots is to have one that swings the full 10kohm, and one that may swing only 1kohm, using the smaller one to change the resistance on one side. Here's how I'd connect it:
+5V (if indeed it is +5V) connects to one leg of the 1kohm worm-gear pot
the wiper of the 1kohm worm-gear pot connects to one leg of the 10kohm pot
the other leg of the 10kohm pot connects to ground
the wiper for the 10kohm pot connects to pin5, as you have it

It doesn't really matter if the 1kohm pot is on the ground side or the +5V side, so I arbitrarily put it on the high side. One leg of the 1kohm pot is not connected because its not needed. The one disadvantage of that setup is that its hard to tune all the way to +5V or 0V because you have to turn both pots to the extremes to do that, but from what I read, anyone tuning their idle circuit to the extreme probably needs their engine looked at anyways!
Haha! So true. Most folks need to hover between 1% to 2% on idle from what I gather. But getting a finer adjustment is the trick. I have heard of those worm-gear pots. I figure I could do that in replacement of the 10K pot anyway, right? At least that is what was indicated somewhere on the net.
 

·
Registered
Joined
·
30 Posts
Haha! So true. Most folks need to hover between 1% to 2% on idle from what I gather. But getting a finer adjustment is the trick. I have heard of those worm-gear pots. I figure I could do that in replacement of the 10K pot anyway, right? At least that is what was indicated somewhere on the net.
Yeah, you could definitely replace it instead of the series trick I was doing. Most worm pots I've seen are 10 turn, so you'd have 10x the control of the original design. Putting them in series lets you get 100x or better instead (by using a smaller resistance worm pot), but does require more hardware. 10x may be all you need to fine tune the result, so your way will work just fine.

Either way, your idea of putting test points to measure the voltage was spot on: you'll need those if you start using worm pots (lest you have to start counting turns to figure out where your voltage is!)

Btw, when you build it and start measuring voltages, its very likely that the computer has a 8-bit ADC to convert your voltage to a number internally. They're cheap, common, and often even built into the chip (sometimes you'll see 10 and 12 bit ADCs, but they're often not even run in that mode). Measuring a 5V signal with 8 bits (256 values) means each increment on the chip is about 0.02V. You don't have to worry about accuracy on your voltmeter reading any further than that. You can use this to figure out whether or not you need to do the series pots or just a single worm pot.
 

·
Registered
Joined
·
441 Posts
Discussion Starter #16
Yeah, you could definitely replace it instead of the series trick I was doing. Most worm pots I've seen are 10 turn, so you'd have 10x the control of the original design. Putting them in series lets you get 100x or better instead (by using a smaller resistance worm pot), but does require more hardware. 10x may be all you need to fine tune the result, so your way will work just fine.

Either way, your idea of putting test points to measure the voltage was spot on: you'll need those if you start using worm pots (lest you have to start counting turns to figure out where your voltage is!)

Btw, when you build it and start measuring voltages, its very likely that the computer has a 8-bit ADC to convert your voltage to a number internally. They're cheap, common, and often even built into the chip (sometimes you'll see 10 and 12 bit ADCs, but they're often not even run in that mode). Measuring a 5V signal with 8 bits (256 values) means each increment on the chip is about 0.02V. You don't have to worry about accuracy on your voltmeter reading any further than that. You can use this to figure out whether or not you need to do the series pots or just a single worm pot.
You are such the man, it's not even funny. If I were crazy, I would have an LCD read out showing the voltage as I turn the pot.

Baby steps! Off to Fry's Electronics I go!

Thanks again!
 

·
Registered
Joined
·
441 Posts
Discussion Starter #17 (Edited)
Well, $49 later, I have all the parts I need. I already had a perfboard though. But I did pay retail prices since I didn't want to be patient and I still need the Suzuki plug. I would expect things to be significantly less if bought on line from Mouser or something.

Oh, and I did get the 10 turn pot and the testing lead contact points for a voltmeter. That was a $17 upgrade.
 

·
Registered
Joined
·
30 Posts
Well, $49 later, I have all the parts I need. I already had a perfboard though. But I did pay retail prices since I didn't want to be patient and I still need the Suzuki plug.
Yeah, it always frustrates me to see something like

10uF capacitor:
Radio shack: $1.27
Digikey (quantity 1): $0.26
Digikey (quantity 1000): $0.03 * 1000

.. and then of course, because of time, I buy the Radio shack one, cursing their markups :furious: I guess that's the markup you have to have when the 3 cent part may sit in a bin in an expensive store for 2 years before someone buys it.
 

·
Registered
Joined
·
441 Posts
Discussion Starter #19
Yeah, it always frustrates me to see something like

10uF capacitor:
Radio shack: $1.27
Digikey (quantity 1): $0.26
Digikey (quantity 1000): $0.03 * 1000

.. and then of course, because of time, I buy the Radio shack one, cursing their markups :furious: I guess that's the markup you have to have when the 3 cent part may sit in a bin in an expensive store for 2 years before someone buys it.
Exactly! But, in the end it won't matter.

Below is the filled out board from the front.


View from the back (sloppy, I know, but it should work).


And the box I chose:

  • Top Left: On/Off
  • Top Middle: Power Led
  • Top Right: Write Switch (Momentary)
  • Bottom Left: Cylinder Select
  • Bottom Middle: 10K Linear 10-Turn Pot
  • Top, Top of Box: Test Points for Voltmeter
 

·
Registered
Joined
·
30 Posts
Looks awesome!

Have you checked the voltages before you plug it into the bike? You should be able to test all of them with a battery across pins 3 and 4. Pins 1 and 2 might be a little tough to test the 'switch open' case, because they're floating, but they're also the easy ones.
 
1 - 20 of 78 Posts
Top