THis thread is dedicated to all the posts which use the original PCB design for the UHU, this is also meant for people who are using PCB from manjeet and others following the old format PCB.

I have just finished testing my new power supply for 3 standard UHU's running ametek servos (38v, 12amp max) The power supply uses 2 Laminated core transformers (350va?) in paralall, 2 large caps and 10a fuses between each UHU. It produces 50v no load, and when I draw 8 amps from it it drops to 47v My question is can i use this power supply with my ameteks (38v) or will that cause damage to the servos?

Thanks for your help on this one.

Regards.

I would not!

The power supply for your 38V motors should be around 44V (+15%). 50V is not so far above so I think they will run just fine. Surely they will heat a little more but this should be examined during testing.

As a reference I have my 56V-rated motors running with a 66V PS. Of course if you are not going to use the full speed potential of your motors you could definetely go with a lower voltage PS - say 35-40V.

Todor

Thanks, thats what i thought. I definately will need the full speed potential of my motors as they are only rated at 900rpm. On my Y axis of my Mill Drill i m converting i am using a 2.5mm pitch ball screw. I will need to direct drive this axis to come even close to fast enough rapids. (37mm/s)

Hi Mike,

do let us know the results!

RGDs
Irfan

Hello,


For almost 2 years I'm trying to start building a new CNC machine.
When searching the internet I came accross the UHU controller.
Last year I ordered the Controller and Crytals with Uli Huber.
And with the last PCB action on the german Message board Peters' CNC ecke
I ordered 6 PCB's (the old layout 20/10/2005).
Now I want to order the parts for putting the UHU board's together.
I found so many threads on the net CNCzone and CNC ecke about the controller, that it's almost impossible for a newbie to filter all that info.
Now is my question. Does someone have single and clear parts list for ordering all the elctronic parts. I know that one can find such a list on
http://people.freenet.de/Mode1/UHU_P...uecklisten.xls.
But for me there's still too much not clear. For instance
I want to use the opto3 option: HCPL2231(stated by Uli as the best option).
Then in DIY servo controller thread, tomkroscavage (05-23-2006, 08:04 PM)wrote when using this optocoupler: "If you are using the HCPL-2231 just do not insert R18 and R25 as this opto comes with internal schmitt-trigger and
push-pull-drivers and is already TTL compatible."

So is there someone who can give me a non-ambiguous list for building a UHU controller using IRFP260 (150V DC and 30A) and HCPL2231

Greetings Sam

when using this optocoupler: "If you are using the HCPL-2231 just do not insert R18 and R25 as this opto comes with internal schmitt-trigger and
push-pull-drivers and is already TTL compatible."


If he is right then it works like this:

Many opto couplers work just like a hand switch.
A switch does not give any voltage or current.

Imagine a hand operated switch between pin 7 and pin 5 of IC11 opto3.

If the switch is open, then there is no current from +5 volt through R18, through the switch to GND.
As a result the voltage on pin 7 will be +5 volt.

If the switch is closed, then there flows a current from +5 volt through R18, through the switch to GND.
As a result the voltage on pin 7 will be 0 volt.

The voltage on pin 7 comes from the +5 volt via R18.
Not from pin 7 itself.
R18 is in this case called a 'pull up' resistor.

This is how many opto couplers work.

With the HCPL2231 opto coupler it works different.
It does not work like a switch.

The pin 7 of the HCPL223 works as a small voltage supply.

If the HCPL2231 gets a ON signal, it will create its own +5 volt signal on pin 7.
Then there is no need for a R18 pull up resistor to create a +5 volt.
If the HCPL2231 gets a OFF signal, it will create a 0 volt signal on pin 7.

With a HCPL2231 you do not need the R18 pull up resistor.
And R25 is also not needed, it does the same for the other channel of the HCPL2231.

That is, if he is right.
I have looked at the HCPL2231 datasheet, with my limited knowledge i can not see if the pin 7 indeed works like a small voltage supply.
Hopefully someone with more knowledge can tell us :-)

On the other hand, leaving the R18 and R25 in is no problem with the HCPL2231.
It probably works fine.
Can someone "officially" confirm this ?

http://www.datasheetcatalog.com/datasheets_pdf/H/C/P/L/HCPL-2231.shtml

Vroemm.

That is, if he is right.
I have looked at the HCPL2231 datasheet, with my limited knowledge i can not see if the pin 7 indeed works like a small voltage supply.
Hopefully someone with more knowledge can tell us :-)

On the other hand, leaving the R18 and R25 in is no problem with the HCPL2231.
It probably works fine.
Can someone "officially" confirm this ?

http://www.datasheetcatalog.com/datasheets_pdf/H/C/P/L/HCPL-2231.shtml

Vroemm.

Yes, the HCPL2231 does not need a pullup on the output ( stated in bold on the first page of the datasheet under features ). It can directly drive 4 TTL loads.
It should work with the pullups in the circuit if the resistor value is not to low. The only difference should be slightly slower slopes of the signal.

Klaus Leiss

Yes, the HCPL2231 does not need a pullup on the output

Thanks for the info.
I have added it to the UHU wiki.
http://gsst.wikispaces.com/UHUget

Vroemm

Yes, the HCPL2231 does not need a pullup on the output ( stated in bold on the first page of the datasheet under features ). It can directly drive 4 TTL loads.
It should work with the pullups in the circuit if the resistor value is not to low. The only difference should be slightly slower slopes of the signal.

Klaus Leiss


True, but your UHU board will require both pull-up resistors if you are using a 74HC14 chip instead of the 74HCT14 or a plain 7414N (or 74LS14) chip.

Kreutz.

True, but your UHU board will require both pull-up resistors if you are using a 74HC14 chip instead of the 74HCT14 or a plain 7414N (or 74LS14) chip.


Oeps :-)
The parts list says the IC10 is a 74HC14.
Best to go from there.
I have removed the comment, about removing R18 and R25, from the uhu wiki.
http://gsst.wikispaces.com/UHUget

Thanks Kreutz !

Vroemm.

Hi there
I want to build this one but where can i find the PCB G-codes for the UHU servo card ?
Thanks i advance.
Per

are you planning to route out the pcb on a blank board

Yes, on my router.

there was a plt file some where you might search the old UHU thread for that in the initial few pages, also there were pdf files.

look into vroemms wiki, you might find all the files there.

Yes, i have found them but i need them in another format eg G-code or Eagle files so i can handle them in a proper program and run it in Mach3.

not that I know of!

Hi everybody !

I got the parts from Reichelt.de and made the PCBs, after verification the PCBs have only a 35µm copper layer on them not the 75µm claimed in the parts list.

Since it's my first "high" voltage/current project, i was wondering how much amps could a 35µm UHU PCB stand with no problem ?

I didn't remove the toner transfer yet because i don't know how to protect the copper from corrosion & heard that tin plating a PCB lowers the circuits frequency so, does anyone has a working solution ?

Is there a way to improve the PCB to make it more reliable ?

For those using the old PCB, can you please post some pics of your setups ?
(bottom views apreciated)

Thanks !

cnc2

Hi everybody !

I got the parts from Reichelt.de and made the PCBs, after verification the PCBs have only a 35µm copper layer on them not the 75µm claimed in the parts list.

Since it's my first "high" voltage/current project, i was wondering how much amps could a 35µm UHU PCB stand with no problem ?


Have a look at http://de.wikibooks.org/wiki/Bild:Current_vs_Width_vs_Temp_(1oz_ext).PNG



I didn't remove the toner transfer yet because i don't know how to protect the copper from corrosion & heard that tin plating a PCB lowers the circuits frequency so, does anyone has a working solution ?

This is not a problem at the frequencies used by UHU.


Is there a way to improve the PCB to make it more reliable ?

cnc2

If you find that an trace can not carry the required current at 35 µm thickness, you can solder a wire on the affected trace.

Klaus Leiss

Hi cnc2,

If you look through the old thread I have posted a picture of my first board. The link is here: /post 593/

http://www.cnczone.com/forums/showthread.php?t=14217&page=50

The other 4 boards are somehow different since I have used rectangular wire, something like 4x1mm if I remember correctly, but I do not have pictures of them. It is a problem when trying to solder this big thing on the board and if you get it too hot for too long time the board is going to bend which is not that nice.

It will be interesting to show my boards since they are made like noone has done here on the thread - the FET's are on the other side of the board. It's a problem I don't have my camera here now, maybe will do it with the phone, quality will be poor, but enough to see.

Todor

Have a look at http://de.wikibooks.org/wiki/Bild:Current_vs_Width_vs_Temp_(1oz_ext).PNG


This is not a problem at the frequencies used by UHU.


If you find that an trace can not carry the required current at 35 µm thickness, you can solder a wire on the affected trace.

Klaus Leiss
Thaks a lot Klaus !

Unfortunately i don't read German :(, but the graph says it all and in detail.:)

Thanks !

cnc2.

Hi cnc2,

If you look through the old thread I have posted a picture of my first board. The link is here: /post 593/

http://www.cnczone.com/forums/showthread.php?t=14217&page=50

The other 4 boards are somehow different since I have used rectangular wire, something like 4x1mm if I remember correctly, but I do not have pictures of them. It is a problem when trying to solder this big thing on the board and if you get it too hot for too long time the board is going to bend which is not that nice.

It will be interesting to show my boards since they are made like noone has done here on the thread - the FET's are on the other side of the board. It's a problem I don't have my camera here now, maybe will do it with the phone, quality will be poor, but enough to see.

Todor

Thanks Todor !

Waw that's a monster PCB for a monster motor, did you finish building your machine yet ?

Some pics & vids please !!

What did you put on the copper to protect it from corrosion, i mean in the low voltage side where it isn't tin plated ?

Thanks !
cnc2.

Hi cnc2,

Since I don't have one of those fancy spray stuff that protect the pcb I've done it the old-fashioned way - a cotton tampon with ethyl alcohol which dissolves the resin. This makes a very thin layer of resin all over the board which manages to do the work and helps soldering later. I know there is a better way to do this, but that's how I've done it.

The copper wire was a bit of an overkill, thats why I've used thinner wire on the latest boards.

The machine is still under construction, took more time than we thought in the first place. We are doing the control box in the moment which will house both the computer and the servo controllers. Looks like a small fridge in size. :D

Todor

Hi cnc2,
The other 4 boards are somehow different since I have used rectangular wire, something like 4x1mm if I remember correctly, but I do not have pictures of them. It is a problem when trying to solder this big thing on the board and if you get it too hot for too long time the board is going to bend which is not that nice.


If you want to add a large cross sectional area try braided wire instead of single strand wire. I have used desoldering braid for such a use, it has a decent cross sectional area and already contains flux. You should clean of the flux afterward.

Klaus Leiss

Hi cnc2,

Since I don't have one of those fancy spray stuff that protect the pcb I've done it the old-fashioned way - a cotton tampon with ethyl alcohol which dissolves the resin. This makes a very thin layer of resin all over the board which manages to do the work and helps soldering later. I know there is a better way to do this, but that's how I've done it.

The copper wire was a bit of an overkill, thats why I've used thinner wire on the latest boards.

The machine is still under construction, took more time than we thought in the first place. We are doing the control box in the moment which will house both the computer and the servo controllers. Looks like a small fridge in size. :D

Todor
Thanks Todor !

I think i don't understand your solution, what resin are you talking about ? do you mean the photosensitive resin ? in this case i want to tell you that i used toner transfer method on a copper board that has no photosensitive resin on it.

Thumbs up for your machine it's hard to wait to see it running !

Thanks !
cnc2.

If you want to add a large cross sectional area try braided wire instead of single strand wire. I have used desoldering braid for such a use, it has a decent cross sectional area and already contains flux. You should clean of the flux afterward.

Klaus Leiss

Thanks Klaus !

Your answer makes me happy because i've thought about using desoldering braid in this manner but wasn't sure it won't bring up frequency or resistance issues, now i'll go with it !

What are you using to clean the flux ?

Some pics or vids of your setup ?

Thanks !
cnc2.

Your answer makes me happy because i've thought about using desoldering braid in this manner but wasn't sure it won't bring up frequency or resistance issues, now i'll go with it !

You need higher frequencies to get any issues with this. Regarding resistance all that is of consequence is that the resulting resistance should be as low as possible however you achieve that. The only critical point is that you should avoid any points where the resistance is higher than elsewhere in the same circuit. There you can develop local hot spots that can be trouble later. If your braid is not long enough to cover the whole length of the trace you should overlap the shorter sections a bit.


What are you using to clean the flux ?

Ethananol or commercial flux cleaner if it is necessary. You should look on your solder or desolder braid if a no clean flux is used. If yes no cleaning is needed. If this is not clearly labeled, but instead it is stated that the flux is rosin ( colophony ) you have a no clean flux. You can also use rosin solved in ethanol as a varnish that protects you copper traces from corrosion. If you want to remove excess rosin flux for aesthetically reasons you can also use ethanol

Some pics or vids of your setup ?

I'm still deciding what I want to build , I'm looking for something that can mill small aluminum parts and can engrave some printed circuit boards.

Klaus Leiss

Thanks for your answer Klaus !

For the desoldering braid i'm using a "Pro's kit 8PK-031B" i found on the net that it has a "no clean flux", thanks for your tips on this one !

I'm still deciding what I want to build , I'm looking for something that can mill small aluminum parts and can engrave some printed circuit boards.

Klaus Leiss

Like you, i'm stuck with the same question: what to build ?

I have to choose between a big router (2.40m X 1.40m) or a little mill that can do lots of things, i think i'll go for the little mill to make as much parts as i can for the big router (motor mounts...etc) while i gather the money for the long and expencive ballscrews.

I want this mill to have 4 axes, but i also want the 4th axis to be a lathe, like the one on the vid but with the chuck to be convertible to a lathe by belt and motor change...a mixture of the two mills of the vids...yes i'm a dreamer sometimes !

Thanks !
cnc2.

YouTube - Taig mill 4th axis cut
YouTube - Taig CNC mill/lathe

Hi cnc2

I was talking about the flux that is used to solder. I am not so sure about all those technical terms, so I may have made a mistake here.

I guess you have to decide what you will use the machine for and then make it to do the job. My work here is dedicated mainly to wood and other soft materials, that's why I look for longer axis travels. Metal working is not possible with it, anyway.

Todor

Hi cnc2

I was talking about the flux that is used to solder. I am not so sure about all those technical terms, so I may have made a mistake here.

I guess you have to decide what you will use the machine for and then make it to do the job. My work here is dedicated mainly to wood and other soft materials, that's why I look for longer axis travels. Metal working is not possible with it, anyway.

Todor
Hi Todor,

No problem bro ! no one is taking an exam here :)
Did you test run your controller ? becausei've read that flux should be cleaned (http://www.finishing.com/Library/flux.html)
from a circuit.

Me too, i want to build a large router (to make kitchens...etc) but need to make the metal parts for it, in here metal shops charge you for "what you look like" so, there's no fix price.

Thanks !
cnc2.

cnc2 - so just dress like what will fetch you the max discount :D

for cabinet making wood router may be you can look at www.mechmate.com

cnc2 - so just dress like what will fetch you the max discount :D

In this case you'll have more chance to be tagged "hippy" and have your questions ignored :confused: Today i saw a metal shop that i didn't know & tried to ask them for a price list or the price of an hour of work nothing then i asked for the price of making a timing pulley they said give us a schematic, on what basis do you charge for pulleys ...nothing not even an approx price :confused:


for cabinet making wood router may be you can look at www.mechmate.com

Thanks for the link Irfan !

Did you build yours following these plans ?
Where did you get the ballscrews from ? (and if possible:for what price ?)

Thanks !
cnc2.

I finally did it!
I managed to blow up 2 uhu's!
But let me first tell you what happened.
These are 4 uhu boards that I got from Manjeet.
After I solder them I picked one up and tested itas I read.
Everything went o.k. so confident for
my soldering, I connected another 2 to make a
complete test as a 3axis system without testing
them one by one.
But unfortunately the other two didn't moved.
The shaft was free to move it with my hand and
if I turn it too much the led was lit.
Then I tried to test them one by one and as soon as
I gave them power the two ir2184 were blown.
I don't know what went wrong and to tell the truth
I was pretty panicked. After I blew and the second
one I stopped saving the last one for testing it the next
day as it was pretty late and I was pretty confused.
The next day with my head clear I wired the last one
and thankfully it didn't blew. But it didn't worked either.
As I said when I give it power the shaft it doesn't hold
it's position and when I turn it with my hand it turns freely
and if I turn it too much the position error led lights up
properly. And the dos program does communicate with the
controller properly.
So my question is: how do I debug it?
I'm totally in the dark here and any help would be really
appreciated.

its really difficult to tell - did you follow the procedures to assemble the old type board,
chk resistor values etc.,

I have 3 kits from Manjeet which worked easily first time.

whats exactly blown : chk the old thread there is lot of data about these old drives there. :)

So after 6 months of silence the old thread is given a life. :)

Koslaz, I guess there is something wrong with the output stage. Can you tell me what are the power FETs and the voltage you are using. On the "blown 2" board I see the upper right CAP on the logic rail has its cap bumped, or it just looks like that.

It is surely a mistake somewhere, a component put backwards, smth like that.
Try to power it with a lower voltage, I always use 12V with a power resistor in series to check if the boards are OK and functional. I never afforded to switch it first time with the proper working voltage.

Check for interconnections between the traces on the board. Some of them are quite thin and close. I've made 2-3 of these on the logic side, glad it didn't let to smoke, it just did not work.

And one other thing - you should always test them one by one firstly.

Todor

Hi.
I have finally assembled my old uhu board wich i had routed out from a blank pcb board.
I have failed to connect to the terminal, i have tried different computers and terminal program without any success. And yes, it is a strait cable.
When i tested some points on the board i saw that there is no -12V from TL082CP on pin 4 but i have +5V into pin 8 and i can adjust the voltage to pin 2 & 6 between 0 - 1.13V with R12 but there is no change to the led D21,it is just 3,8V.
Is this critical for the MAX232 so it doesnt work without the -12V ?
Is my TL082 broken ?
Sorry for my English
Per

Hi.
I have finally assembled my old uhu board wich i had routed out from a blank pcb board.
I have failed to connect to the terminal, i have tried different computers and terminal program without any success. And yes, it is a strait cable.
When i tested some points on the board i saw that there is no -12V from TL082CP on pin 4 but i have +5V into pin 8 and i can adjust the voltage to pin 2 & 6 between 0 - 1.13V with R12 but there is no change to the led D21,it is just 3,8V.
Is this critical for the MAX232 so it doesnt work without the -12V ?
Is my TL082 broken ?
Sorry for my English
Per

I have tested PCB with only +5V and GND and communication with PC works (of course, whole controller not, since there is no +12V). I must added -12V because I can't find any connector for it on original PCB.

Hi.
I have finally assembled my old uhu board wich i had routed out from a blank pcb board.
I have failed to connect to the terminal, i have tried different computers and terminal program without any success. And yes, it is a strait cable.
When i tested some points on the board i saw that there is no -12V from TL082CP on pin 4 but i have +5V into pin 8 and i can adjust the voltage to pin 2 & 6 between 0 - 1.13V with R12 but there is no change to the led D21,it is just 3,8V.
Is this critical for the MAX232 so it doesnt work without the -12V ?
Is my TL082 broken ?
Sorry for my English
Per

The -12V is generated on the Max232 internal charge pump circuit. It will never be -12V exactly but anything above -5Volts will work. Check that you have the specified Max232 chip (including suffix letters), they require different capacitor values depending on the suffix letter..

By the way, without a negative voltage present you wont get reliable RS232 communication either.

Thanks for your reply kreutz.
It is MAX232CPE chips with 10u caps around it.
I will change the caps (10u-0.1u) to se if that was the problem for me.
Best regards:
Per

Thanks for your reply kreutz.
It is MAX232CPE chips with 10u caps around it.
I will change the caps (10u-0.1u) to se if that was the problem for me.
Best regards:
Per

The capacitors should be 1 uF/ 16 - 25Volt for that chip. Observe polarity.

Best regards,

Kreutz.

Also Think Tantalum's would help

Ihave tantalums 10u in there.
I am no expert in electronics so be cind to me.
Is it good with electrolyts and shold i change only C10,C11,C12 & C5 or should i even change the 5 one below the four ones between MAX232 and the D-sub ?
Per

Ihave tantalums 10u in there.
I am no expert in electronics so be cind to me.
Is it good with electrolyts and shold i change only C10,C11,C12 & C5 or should i even change the 5 one below the four ones between MAX232 and the D-sub ?
Per

You can also use electrolytic capacitors, although I prefer tantalum for that job. 10 uF are not a good choice for that chip.

Best regards,

Kreutz.

So Mr kreutz i should use 1uF Tantalium for the four caps C10,C11,C12 & C5 ?
I will replace them tomorrow and let you know if that was the problem.
Per

"Huston, we have contact"
Replacing the four caps from 10uF Tantalium to 1uF Electrolyt did it so i made contact with the UHU chip.
But there is only ~-6.5V from MAX232 to TL082 chip, could that be a problem and can it be better with 1uF Tantalium caps ?
I found the 4 Electrolyt caps here at home and i only wanted to test with them.
Per

"Huston, we have contact"
Replacing the four caps from 10uF Tantalium to 1uF Electrolyt did it so i made contact with the UHU chip.
But there is only ~-6.5V from MAX232 to TL082 chip, could that be a problem and can it be better with 1uF Tantalium caps ?
I found the 4 Electrolyt caps here at home and i only wanted to test with them.
Per

The voltage is probably going to rise a little, due to the tantalum capacitors having less leakage, but -6.5V is OK.

Congratulations!

Kreutz.

It is thanks for you kreutz that i managed to do this. :cheers:
This is my first PCB that i ever made and i am 47 years old
Best regards:
Per

Good - now move on fast - :)

I have tested old UHU with various motors and several times I blown mosfet drivers. First I tried with 12V zener diode, but without succes. Later, I have used P6KE15A, but controller doesn't work !! When I removed P6KE15A it works.
What is right solution to protect drivers.

Kodalo,
It sounds as though your installing the P6KE15A with the polarity reversed, turn around in the circuit ............
Paul

Kodalo,
It sounds as though your installing the P6KE15A with the polarity reversed, turn around in the circuit ............
Paul
I have installed zener diode with same polarity direction and it works.

What MOSFETs are you using? In most cases the FETs have internal diodes but they are quite slow and the switching spikes are more obvious. With the external diodes the signal has a better shape.

Todor

What MOSFETs are you using? In most cases the FETs have internal diodes but they are quite slow and the switching spikes are more obvious. With the external diodes the signal has a better shape.

Todor
I am using IRF450. Since I have plans to work with higher voltage (~90 V) I think is better to start using HP_UHU. Is there any usefull PCB? I have found UHU rev1.0b.rar and UHU_MODIFIED_Dated1.4.08_R1.0b.pdf inside, but there is no acceptable print layout - I must resize, convert to black.... Any other solutions are also welcome - good PCB, getting bare board...

anyone have any problem with hpcl 2231, cause my servo motor turn only at one direction , the signal from the computer is good but why my servo turn only at one direction ?

anyone have any problem with hpcl 2231, cause my servo motor turn only at one direction , the signal from the computer is good but why my servo turn only at one direction ?
Check MOSFETs. I have similar problem and one MOSFET was wrong.

Hello everybody,

I've been away from this project for too long, i was going to assemble my first board when i read "0,6mm silver wire is used for jumpers" so my questions are:

1) is it possible to replace the silver wire with 0.6mm copper wire from telephone line ?

2) why is the silver wire used & what limitations/troubles can result from the use of copper wire instead ?

3) if it's critical, where can i source silver wire & what kind of shops usually sell it ?

EDIT:
Ok guys, i agree that this was a stupid one, but i'm really lacking knowledge...so... help is needed please !

Thanks !
cnc2.

CNC2,
Never a dumb question, I ask when something bothers me. that's one of the nice things about CNCZONE & all of the good people out there.
use what ever wire is handy ! I generally save some leads I snipped off resistors and use those for jumpers. they work just fine
Paul

Thanks for the reply Paul !

Yea man the zone is great for that, i agree.

Thanks for the tip, i'll go looking for the leads remaining from previous projects.

Did you run your boards yet ?

Thanks !
cnc2.

I just soldered the jumpers & saw that the soldering iron was burning the flux from the solder on thin tracks, making it hard to get a clean stuff...

What kind of soldering iron are you guys using ?
(i'm using a cheap 40W soldering iron)

Is there a way to get it done by pluging the soldering iron into a light dimmer ?

Thanks !
cnc2.

Hi to everyone !

Please i need some help.
I was looking at the parts list & saw that i had recieved the wrong part# for the P6KE12CA diodes... instead of P6KE12CA i recieved P6KE15CA:confused:.

So my question is obvious :
1) Is it possible to use P6KE15CA instead of P6KE12CA on the old UHU board dated 23-10-2005 & how will it affect the circuit ?

EDIT:
I've attached a snapshot of the datasheet & it's a gibberish to me due to my POOR knowledge of electronics.

Thanks !
cnc2.

Hey,
I hope H.O. or someone who knows what their talking about will jump in here. From what little I know they should interchange with no problem. you may have to do a little creative bending with the leads. the only stat that I'm unsure about is the Max clamping voltage 21.2 vis 16.7. all the other specifications mean a more substantial component.(I think)
Hope this helps
Paul

Thanks for the help Paul !

I'll try to PM H.O and see.

Paul tell me what do you mean by "creative bending", do you mean the 15CA are bigger than the 12CA ?

Thanks !
cnc2.

Hello,
I'm quite confused actually....The schematic that you linked to in your other thread shows P6KE120CA.

IF 12CA is what is supposed to be there then you can probably use 15CA if that's what you got but if it indeed should be 120CA as the schematic says then you most likely can't use 15CA.

Thanks for the help Henrik !

I re-checked the schematic & i can read P6KE12CA, i agree that the C looks like a Zero (120A) zoom it to 600% then it'll apear as a C. It's more readable on the parts layout on the PCB.

Thanks to you, i'm happy to be able to use the 15CA i've recieved:wee:

Thanks !
cnc2.

Ah, that clears it up a bit :-) Then you can probably use the 15CA safely - although I can't say for sure. Good luck!

Thanks for the help Henrik,

i've finally found that the assembly instructions use the 15CA, so now i'm sure it's safe & working.

Thanks !
cnc2.

Hi guys,

I need some help on the color of the LEDs on the UHU board.
I don't remember why i ordered 6red +3green +3yellow
I know there's a green LED for +5V, a yellow one on I-lim & a red one for error but i have no idea about the color of the LED for H-On & what it means when it lights ?

So if you have the answer or can point me on a user's manual, it would really help me.

Thanks !
cnc2.

The color doesn't matter - red, green, yellow, take your pick - anything will work.

I think that the H-On LED you mentioned is the Ready signal from the servo controler chip. It'll turn on as soon as the chip has "booted up".

/Henrik.

Thanks for the reply Henrik !

The color doesn't matter - red, green, yellow, take your pick - anything will work.
I ain't that dumb man ! :) especially after you explained me how to feed LEDs :D

Nevermind, i agree i heared that some colors used a bit more current than others but the reason i asked the question about colors is that i don't want to search for the meaning of the red LEDs (if two onboard) every time one of them lights up.

After your explanation it'll be two green LEDs for H-On & +5V(to mean everything is ok) a red one for error (to mean encoder signal problem) & a yellow one for I-lim (means doesn't bother me too much)

Thanks for the HELP !
cnc2.

I purchase and built 3 UHU boards from Embedded Acquisition Systems. Two of the boards work fine. The third though is giving me a headache. I keep getting the error LED as soon as I power up the board. It immediately comes back on after a reset. The tuning display shows a lot of errors from the encoder even though nothing is connected. I only have the 12V power and the serial cable connected at this time. Any suggestions as to cause would be very welcomed.

I have repair old machine and added UHU controller. Both axis are driven by tooth belt. During summer days it works fine, but now I have problem with longer axe. There is lot of overruns (about 120 on 10 repeat G0 X2000,GO X0). Error is about 2 mm. Shorter axis has length about 1.8m and encoder is attached at motor shaft. X belt is about 3 meter and encoder is attached to belt using timing pulley. I suppose problem is belt and temperature, but I don't know how to check and solve it.
Does anybody has experience with such long timing belt?

kodalo,

Can you tell me at what step frequency and motor voltage you are driving the UHU board? From your post I understood that your encoder is not mounted directly on the motor, but through the timing belt, is that right?

If the encoder is not enough temperature stabilized this could lead to changes in the pulse widths maybe between summer/now time. It is strange, cause now is colder, I thought high temps tend to be the problem for electronics.

As for jjobezo, check that something is not shorted around the emergency-in pin. If it is grounded surely you will get the result you are saying.

Greetings,

Todor

UPDATE: The encoder input is pulled-up with resistors, but somehow you get a high noise level that counts false pulses. The error led will also light when the max deviation from the position is exceeded.

If you're getting O-counts then you most likely have noise on the encoder signals or you're trying to run faster than what the UHU-chip can keep up with so it misses pulses. Why this works in the summer but not now is the question and Todor might be onto something even if it's a longshot.

The length of the belt shouldn't have anything to do with it.....Or, perhaps...if the belt expands with temperature and it is now colder it means there's more tension on the mechanics which means the motor needs to work harder which pulls more current and might produce more noise that gets injected in the encoder cables/circuitry. Another loooong shot....

Try to think about if you've changed anything on the machine or if you added anything to the shop that might induce noise in the cables. Fluorecent lights, variable frequency drive etc.... Are you using shielded cables for the encoders, shield connected to GND at the drive?

What type of encoder, any datasheet for it?

And finally, regarding the error-LED, Todor is correct - the drive will fault and light the Error-LED when max error is exceeded. However, that relies on the fact that the encoder signal integrity is OK which it seems isn't in this case here so the drive doesn't know the TRUE position.

/Henrik.

EDIT: Ahh, I see, the comment on the error-LED might be targeted at jjobezo.....

If you're getting O-counts then you most likely have noise on the encoder signals or you're trying to run faster than what the UHU-chip can keep up with so it misses pulses. Why this works in the summer but not now is the question and Todor might be onto something even if it's a longshot.

I have tried to decrease speed, but result is same.



The length of the belt shouldn't have anything to do with it.....Or, perhaps...if the belt expands with temperature and it is now colder it means there's more tension on the mechanics which means the motor needs to work harder which pulls more current and might produce more noise that gets injected in the encoder cables/circuitry. Another loooong shot....


I changed belt tension, but nothing....


Try to think about if you've changed anything on the machine or if you added anything to the shop that might induce noise in the cables. Fluorecent lights, variable frequency drive etc.... Are you using shielded cables for the encoders, shield connected to GND at the drive?


Environment is absolutely same, nothing added, nothing removed. All cables are shielded at very beginning.

What type of encoder, any datasheet for it?


I dont have any datasheet, but oscilloscope show normal shapes.

Can you tell me at what step frequency and motor voltage you are driving the UHU board? From your post I understood that your encoder is not mounted directly on the motor, but through the timing belt, is that right?


Motor is 110V nominal, but I am using 88V, since higher voltage leads to high oscillations.



As for jjobezo, check that something is not shorted around the emergency-in pin. If it is grounded surely you will get the result you are saying.


What exactly you think? I suppose in that case, Error LED will light always??




The error led will also light when the max deviation from the position is exceeded.

Error LED is OK, and don't lights.

Yes, the error LED will light always, but that is not your case. I was thinking about something else before.

From my experience the UHU chip is rather sensitive to encoder signal pulse widths and counts more errors when it doesn't like them. My encoders have pots on every line - A,B,C so the pulse width can be precisely adjusted with a scope and I had the opportunity to "play" a little with this feature. :)

Henrik is right that you search for a change in the machine setup. Even if you haven't done anything a short-circuit or a loose connection is still a change though. Have you checked the brushes?

Another thing - do you have a differential connection to the encoder or it is done the simpler way?

Todor

Problem is solved (I hope :) )

Some people on some forums says the Intel can be problematic. But I think, since one motor works fine it is not true for me. I have tried DriverTest.exe and got lot of "piano keys". I change CPU, take AMD ant test again. DriverTest gave me flat line (well, not exactly flat, but there is no piano-like). Then I have tried motor but there is O-error still. But when I start to change UHU parameters system is more cooperative (unlike Intel) and I quickly found good parameters. Now it works fine.

CONCLUSION: Always check with DriverTest, piano-keys are very bad sign. And finally, AMD is the best !!!