I've heard of people dropping a capacitor across Vdd & Vss on the board of education to prevent it resetting due to power drain from running servos or DC motors. Can anyone think of a down side to doing this?
1) Make sure the servo or DC motor power requirement is within the capability of the regulator (which doesn't have a heatsink ... look at the datasheet for the maximum current draw vs. temperature).
2) People often put a low value resistor or choke (inductance) between Vdd and the capacitor to limit the inflow current and to add a little extra filtering. The Propeller Protoboard documentation shows one example of this using a 1uH choke and a 1000uF electrolytic capacitor connected to the servo supply.
Mike Green said...
1) Make sure the servo or DC motor power requirement is within the capability of the regulator (which doesn't have a heatsink ... look at the datasheet for the maximum current draw vs. temperature).
2) People often put a low value resistor or choke (inductance) between Vdd and the capacitor to limit the inflow current and to add a little extra filtering. The Propeller Protoboard documentation shows one example of this using a 1uH choke and a 1000uF electrolytic capacitor connected to the servo supply.
The downside is extra cost and extra board area.
What I've got is a 1F 5v PowerStore Arogel "Super" capacitor (H01650-01E). So what you're recomending is to put a 1 microhenrie inductor in there as well? Or is there some sort of ratio I should try to maintain? I realy dont have much experiance with building an inductor circuit.
The "Super" capacitors are for very low currents, typically used in place of a battery backup for time-keeping chips. It's NOT at all appropriate for driving servo's or motors.
A 20 uF or 1000 uF electrolytic CAN provide for brief 'bursts' of current need, like on startup of a motor. However, the power supply being used MUST be large enough to provide for the 'steady state' -- the capacitor only reduces the very short term 'drop-out'.
allanlane5 said...
The "Super" capacitors are for very low currents, typically used in place of a battery backup for time-keeping chips. It's NOT at all appropriate for driving servo's or motors.
A 20 uF or 1000 uF electrolytic CAN provide for brief 'bursts' of current need, like on startup of a motor. However, the power supply being used MUST be large enough to provide for the 'steady state' -- the capacitor only reduces the very short term 'drop-out'.
The problem is not the motors or servos, but the basic stamp "board of education". Even when driving the servos that come with the Boe-bot kit the board is resetting due to low power as soon as the batteries lose even the slightest of charges. It is most noticable when the servos are both changing direction. For a while I thought it was a short in my sensors since the board would reset every time it detected somthing directly in fornt of the bot and tried to back up. As it turns out, this was just the act of both servos changing direction causing a spike in power draw.
When you say the problem is not the motors or servos, and place the blame on the BOE, this seems to me that you are misunderstanding
something. If the servos draw enough power where the voltage on the BOE drops below the 5V + regulator headroom, the BOE will
reset. This is primarily due to the internal resistance of the battery. Adding a choke and capacitor (suitably connected) reduces the
current spikes drawn by the motors, hopefully keeping the output battery voltage above the voltage needed by the BOE.
How are you driving your motors? I find that if I alternate pulses to the servos (that is, don't pulse both servos at the same time, but
offset the pulses by 10ms) that helps somewhat (because both servos won't have their peak current draw at the same time). Also,
try not to change direction too fast (stop too suddenly, turn too suddenly, accelerate too quickly) because this causes excessive
current draw too.
I wish there was a good place to get a 5-cell AA holder, so we could use NiCD instead (which have much less internal resistance than
Alkaline cells). Unfortunately if people started using one of those, someone would fill them with AA Alkalines giving 7.5 volts to the
servos, which is too much. But I think 5 NiCD AA's would do a better job at this than 4 Alkaline AA's.
rokicki said...
What kind of batteries are you using?
When you say the problem is not the motors or servos, and place the blame on the BOE, this seems to me that you are misunderstanding
something. If the servos draw enough power where the voltage on the BOE drops below the 5V + regulator headroom, the BOE will
reset. This is primarily due to the internal resistance of the battery. Adding a choke and capacitor (suitably connected) reduces the
current spikes drawn by the motors, hopefully keeping the output battery voltage above the voltage needed by the BOE.
How are you driving your motors? I find that if I alternate pulses to the servos (that is, don't pulse both servos at the same time, but
offset the pulses by 10ms) that helps somewhat (because both servos won't have their peak current draw at the same time). Also,
try not to change direction too fast (stop too suddenly, turn too suddenly, accelerate too quickly) because this causes excessive
current draw too.
I wish there was a good place to get a 5-cell AA holder, so we could use NiCD instead (which have much less internal resistance than
Alkaline cells). Unfortunately if people started using one of those, someone would fill them with AA Alkalines giving 7.5 volts to the
servos, which is too much. But I think 5 NiCD AA's would do a better job at this than 4 Alkaline AA's.
Just using basic AA's, but thinking of switching to a NiCd battery pack or a set of four C's. As for using five AA's producing 7.5v, last I checked the Board of Education had a voltage regulator on it and could take 6-9v in and put 5v out.
However, back to my question. Is there a recomended choke/capacitor circuit?
Okay, if you are using normal carbon batteries, that's your problem. There's no way carbon batteries can drive servos;
their internal resistance is *way* too high. Use Alkalines.
Surprisingly, C's may not actually do all that much better than AA's (the internal resistance of the larger batteries is
roughly the same as the smaller batteries). They will last longer, certainly, but the initial behavior will not be much
better, and you will be making your bot weigh a *lot* more (thus increasing the load on the servos and the current
demand), which may actually mean the C's will work more poorly than the AA's in total.
The BOE has a regulator on it, but the problem is if you run your servos directly from the battery pack and not from
the regulated 5V, and you don't want to put 7.5V into your servos.
In any case, if you are currently driving your servos from regulated 5V, I'd instead drive them directly from
battery V+ instead. (This gives you a bit more variability in their behavior, but it reduces the load on the
regulator and improves somewhat the isolation between the servos and the BOE.)
The standard choke/capacitor circuit has the choke going from the battery V+ to servo V+, and the capacitor
going from servo V+ to ground. As for values, well, the values listed above sound reasonable.
rokicki said...
Okay, if you are using normal carbon batteries, that's your problem. There's no way carbon batteries can drive servos;
their internal resistance is *way* too high. Use Alkalines.
I think of alkalines as "normal". Can you even find carbon batteries anymore?
Since you are having trouble, I'd recommend a few ceramic 0.1uF caps across Vdd and Vss.
Even though the BoE has a low dropout regulator, I'd recommend using NiMH or NiCd AAs and putting in 5 instead of the 4 that are standard. I made a pack for mine with tabbed AAs that still fits in the Boe-Bot chassis. You can also get a single cell holder and put it in series with the standard 4 cell pack.
If you go too crazy with bulk capacitance (500uF or more), you can cause the supply voltage to rise too slowly and some older BoEs won't reset properly when you turn them on. Holding the reset switch down after power on works there or you can use an external reset circuit.
I have to admit, now that I'm doing more with my BOEBot, I'm noticing the same thing (only for the Prop protoboard rather than
the Board of Education)---Alkaline AA's are not lasting very long at all, and when they start failing, frequent resets are the primary
sign. This last set I just removed was resetting when the battery voltage was 5.86V (unloaded); the current spikes from the motors
were sagging the 5V rail (and this was also, I guess, was showing up on the 3.3V rail). The sags were of pretty short duration
(a small fraction of a second) but still enough. Note that a 5.86V unloaded on four AA alkalines is not very discharged at all.
(These cells started at 6.4V.)
I'm going to figure out how to mount a five-cell pack (or a six-cell pack and I'll only use five of the slots) instead. I may end up
going to different supplies for the electronics and the servos, but I definitely want to use a regulated 5V on my servos for some of
the dead reckoning stuff I'm doing (this should strongly reduce the impact of input voltage on servo performance).
jfizer, are you having any better luck? Have you tried any solutions?
Harry, would you be willing to attach a picture of your "tabbed AA" solution? I'm having a hard time figuring out how to mount
five or six cells on my BOEBot.
There's a solution to the "four-battery problem" on the way from Parallax: the BOE-Boost. Attached is a preliminary copy of the user docs. The release date has not yet been determined.
I've used his Battery Boosters to eliminate the need for extra batteries for years.
The Battery Booster 12 was designed to eliminate the need for a 9v battery on Scott Edward's Mini SSC II, and a 5v version exists specifically to run a Stamp also.
Either one would work for most Stamp boards, depending on what pin you connect on the Stamp. The 5v would be more efficient running a Stamp as you can bipass the regulator.
They greatly extend the usable charge on NiCD and NiNH batteries. You can even run a Stamp on 2 cells, if you want.
Rokicki,
I use NiMHs with my Propeller Protoboard Boe-Bot. I use the standard 4 cell AA battery holder that mounts under the chassis between the servos and the back of the chassis. I mounted a single AA battery holder with pigtail leads on the top of the metal chassis along the rear edge (using glue) and wired it in series with the leads from the original battery holder. I use a Maha Universal Charger (MH-C777Plus-II) to charge it. The battery pack unplugs from the Protoboard and plugs into the charger with an adapter I made for it. The battery holder and power connectors all came from the local Radio Shack.
Comments
2) People often put a low value resistor or choke (inductance) between Vdd and the capacitor to limit the inflow current and to add a little extra filtering. The Propeller Protoboard documentation shows one example of this using a 1uH choke and a 1000uF electrolytic capacitor connected to the servo supply.
The downside is extra cost and extra board area.
What I've got is a 1F 5v PowerStore Arogel "Super" capacitor (H01650-01E). So what you're recomending is to put a 1 microhenrie inductor in there as well? Or is there some sort of ratio I should try to maintain? I realy dont have much experiance with building an inductor circuit.
FYI: Ascii code 230 is µ
Post Edited (jfizer) : 6/12/2007 11:16:53 PM GMT
A 20 uF or 1000 uF electrolytic CAN provide for brief 'bursts' of current need, like on startup of a motor. However, the power supply being used MUST be large enough to provide for the 'steady state' -- the capacitor only reduces the very short term 'drop-out'.
The problem is not the motors or servos, but the basic stamp "board of education". Even when driving the servos that come with the Boe-bot kit the board is resetting due to low power as soon as the batteries lose even the slightest of charges. It is most noticable when the servos are both changing direction. For a while I thought it was a short in my sensors since the board would reset every time it detected somthing directly in fornt of the bot and tried to back up. As it turns out, this was just the act of both servos changing direction causing a spike in power draw.
When you say the problem is not the motors or servos, and place the blame on the BOE, this seems to me that you are misunderstanding
something. If the servos draw enough power where the voltage on the BOE drops below the 5V + regulator headroom, the BOE will
reset. This is primarily due to the internal resistance of the battery. Adding a choke and capacitor (suitably connected) reduces the
current spikes drawn by the motors, hopefully keeping the output battery voltage above the voltage needed by the BOE.
How are you driving your motors? I find that if I alternate pulses to the servos (that is, don't pulse both servos at the same time, but
offset the pulses by 10ms) that helps somewhat (because both servos won't have their peak current draw at the same time). Also,
try not to change direction too fast (stop too suddenly, turn too suddenly, accelerate too quickly) because this causes excessive
current draw too.
I wish there was a good place to get a 5-cell AA holder, so we could use NiCD instead (which have much less internal resistance than
Alkaline cells). Unfortunately if people started using one of those, someone would fill them with AA Alkalines giving 7.5 volts to the
servos, which is too much.
Just using basic AA's, but thinking of switching to a NiCd battery pack or a set of four C's. As for using five AA's producing 7.5v, last I checked the Board of Education had a voltage regulator on it and could take 6-9v in and put 5v out.
However, back to my question. Is there a recomended choke/capacitor circuit?
their internal resistance is *way* too high. Use Alkalines.
Surprisingly, C's may not actually do all that much better than AA's (the internal resistance of the larger batteries is
roughly the same as the smaller batteries). They will last longer, certainly, but the initial behavior will not be much
better, and you will be making your bot weigh a *lot* more (thus increasing the load on the servos and the current
demand), which may actually mean the C's will work more poorly than the AA's in total.
The BOE has a regulator on it, but the problem is if you run your servos directly from the battery pack and not from
the regulated 5V, and you don't want to put 7.5V into your servos.
In any case, if you are currently driving your servos from regulated 5V, I'd instead drive them directly from
battery V+ instead. (This gives you a bit more variability in their behavior, but it reduces the load on the
regulator and improves somewhat the isolation between the servos and the BOE.)
The standard choke/capacitor circuit has the choke going from the battery V+ to servo V+, and the capacitor
going from servo V+ to ground. As for values, well, the values listed above sound reasonable.
I think of alkalines as "normal". Can you even find carbon batteries anymore?
http://www.bluebelldesign.com/Power1.htm
Since you are having trouble, I'd recommend a few ceramic 0.1uF caps across Vdd and Vss.
Even though the BoE has a low dropout regulator, I'd recommend using NiMH or NiCd AAs and putting in 5 instead of the 4 that are standard. I made a pack for mine with tabbed AAs that still fits in the Boe-Bot chassis. You can also get a single cell holder and put it in series with the standard 4 cell pack.
If you go too crazy with bulk capacitance (500uF or more), you can cause the supply voltage to rise too slowly and some older BoEs won't reset properly when you turn them on. Holding the reset switch down after power on works there or you can use an external reset circuit.
Harry
Blue Bell Design Inc.
the Board of Education)---Alkaline AA's are not lasting very long at all, and when they start failing, frequent resets are the primary
sign. This last set I just removed was resetting when the battery voltage was 5.86V (unloaded); the current spikes from the motors
were sagging the 5V rail (and this was also, I guess, was showing up on the 3.3V rail). The sags were of pretty short duration
(a small fraction of a second) but still enough. Note that a 5.86V unloaded on four AA alkalines is not very discharged at all.
(These cells started at 6.4V.)
I'm going to figure out how to mount a five-cell pack (or a six-cell pack and I'll only use five of the slots) instead. I may end up
going to different supplies for the electronics and the servos, but I definitely want to use a regulated 5V on my servos for some of
the dead reckoning stuff I'm doing (this should strongly reduce the impact of input voltage on servo performance).
jfizer, are you having any better luck? Have you tried any solutions?
Harry, would you be willing to attach a picture of your "tabbed AA" solution? I'm having a hard time figuring out how to mount
five or six cells on my BOEBot.
Post Edited (rokicki) : 6/21/2007 5:31:33 PM GMT
-Phil
www.medonis.com/
I've used his Battery Boosters to eliminate the need for extra batteries for years.
The Battery Booster 12 was designed to eliminate the need for a 9v battery on Scott Edward's Mini SSC II, and a 5v version exists specifically to run a Stamp also.
Either one would work for most Stamp boards, depending on what pin you connect on the Stamp. The 5v would be more efficient running a Stamp as you can bipass the regulator.
They greatly extend the usable charge on NiCD and NiNH batteries. You can even run a Stamp on 2 cells, if you want.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
I use NiMHs with my Propeller Protoboard Boe-Bot. I use the standard 4 cell AA battery holder that mounts under the chassis between the servos and the back of the chassis. I mounted a single AA battery holder with pigtail leads on the top of the metal chassis along the rear edge (using glue) and wired it in series with the leads from the original battery holder. I use a Maha Universal Charger (MH-C777Plus-II) to charge it. The battery pack unplugs from the Protoboard and plugs into the charger with an adapter I made for it. The battery holder and power connectors all came from the local Radio Shack.