Retrofitting a 1986 Maho MH400E

Today I had a play around with thread milling.


Unfortunately, this again indicated that the Y-axis creeps. I notice that the cutter was only touching at the front (-Y). After centering it better, I reoccurred.
I ran a test and such enough, there are drop outs on the B channel of the Y encoder.


It was cold in the basement so I took a break. Since I now have a single channel 602EXE on Z, and it is wired with the differential signals, I will use that to test whether it is an EMI noise issue.
Mark

Mark Sorry to see the Y axis scale is still a lurking problem. I was wondering if your new 5volt switch mode powder supply might be causing EMI issues? Would it be worth rigging up a bunch of batteries ,say flash-light type for a test ? Don't remember the 5 volt range of the scales,but could use a length of small steel wire as a dropping resistor. Please don't burn out the scales bulb.
Could you recruit some old analog electronic hobbiest person/codger that has analog test gear , like a real analog scope an analog variable power supply. -- be sure to quiz them about making shop/house visits.
The cold Maho -- I have used a rigged up tent made from rigid electrical conduit frame with U clamps and plastic film type cover that used a electric heater . Takes a while to warm up the metal such that my tongue doesn't stick to the Maho..
Was useful for painting also.

Thread milling

---

be very careful as you might fall into that need of spindle indexing trap

Hope you find the scale issue very soon
Bob who is --- cold old analog

Hi Bob,

Hanno, one of the German guys who is also converting a MAHO also identified that cheap chinese 5VDC PSU as a potential cause. He pointed out from the datasheet that the isolation at 100MOhm/500VDC is not very good.

He also suggested a metal EMI shield over the 7i77, like I put over the EXE.

Here is a my plan for trouble shooting:
1/ Switch the Y and Z axis encoders exe's. This is easy as I just switch encoder and 7i77 connectors. Z has a stand alone 602 EXE, which only takes 5V power. The three channel board mounted EXE on X and Y also has 12V supply which I pulled from the ITX MoBo PSE. Since the 602EXE also has the differential signals output from the EXE--> 7i77, I am hoping to isolate with this step whether the problem is in the EXE or the Encoder.

2/ Make up a battery PSU per your suggestion. I bought an assortment of resisters at some stage, so it will be easy enough to bodge up a battery pack, tuned for 5V output. Thanks for that suggestion. If that works, would replacing the cheap 5VDC PSU with a name brand rail mount PSU be expected to also work, or do I need to do more to solve a problem like this?

3/ According to all the EXE datasheets I have been able to find, the differential signals should be available on the three axis board. They were just not brought out to the Phillips on the output cables. If the interference is is being picked up on those last cables, I could try replacing them and and bringing out the differential signals on twisted pairs.

Mark

Thanks to this post I got the matchbox virtual keyboard installed and working with Gmoccapy.
forum.linuxcnc.org/gmoccapy/34409-virtual-keyboard#109691

I once read a post about changing the size of the letters/keys on the Matchbox virtual keyboard, but can't find it again. If anyone could point me in the right direction I would appreciate it.

Mark Some of my thoughts .............. The Phillips 432 control used a switching power supply + 6 volt battery (memory back up ) for the computer an control boards. So switching power supplies can be used, but there can be big differences of say a switcher designed for some industrial use vers one designed for say consumer / home use. Do I think switchers are a good thing = very big Yes .
Do I know your switching power supplies are causing a / the problem = No . Consider that I am a old analog geezer and that is what I am more familiar with . EMI shielding ........ yes it can help , but my thought would be to reduce the amount of or remove the source of EMI generated. I have used a portable ( battery powered )AM radio to get some feel/value as to amount. VFD drives are just evel. The indromat servo drives you/we have use SCRs , and they generate some EMI noise also. Might try looking at scale error of just moving Y axis only an also see if the Y axis servo motor wires are close to the scale wires. Might check the shield wire of Y axis servo motor wires.. I wonder if there is Y axis scale error when only the Z and/or only X axis are moving ...you may have tried that, I don't remember.
My thought , If the 5 volt battery lash-up makes a difference then consider making up a analog supply using say a 7805 Ic

---

maybe even consider also using a 78xx series Ic for 12 volt DC. No I don't think ITX power supplies are good enough for Real DC,
The Maho has a 24 volt DC supply , might be able to use that as a DC source for a cascaded 12 volt an then feed a 5 volt regulator Ic Pro = saves needing a 240 volt to say 10 volt transformer + rectifier for a analog supply. Con = having a single phase 240 volt input would increase isolation and could be powered up with extension cord/wire for checking of circuits/troubleshooting. Suggest you bounce this off Mr. Hanno for his input. What size Maho is he converting ? any chance you could drag him onto this thread?
I have a Siemends "STOP modular 20A - 24 volt -120/240 rail (DIN ?) mount " switching supply from e bay. The specks looked good enough for me to try it ,but as of yet haven't done so . Siemends makes high to low quality stuff , but usually have product information on their web site. Might try finding one of their supplies on the used market.

Tracking down electrical noise issues can be a very deep rabbit hole and your circle program showing the repeatable Y axis scale input mess-up is a huge help. That is why I wonder if moving X an Z axis only, will give the same Y axis scale mess-up ? Your neat circle program

---

if the size of the circle is smaller would that change the number/amount/position of Y axis errors? Could point to break in a wire or shield wire. Do you get the same results if the control cabinet door is wide open ? EMI temp. shield ---- I have used aluminium cooking foil covered with plastic film and a clip lead wire for testing stuff .
.
To me,your trouble shooting steps look very useful . Now it's the time suck job of crawling around in that rabbit hole trying to find real repeatable cause an effects.

Hope you nail the real problem source. Keep us current on this misadventure
Bob

Thanks Bob, Your advice in this area is greatly appreciated.

Yesterday I got as far as switching Y and Z's exe's (remember the Z axis EXE is the stand alone unit with differential output). The problem followed (now on Z).

The behaviour is similar to that which I thought I had solved with more voltage (flat batteries posts a few pages back). For the first few few minutes it is stable, then it starts moving in steps (you can hear the change in frequency as it moves - this is with no motion commanded), and after about 20 minutes, it is moving constantly. This is in the video from 1:20.


I added up the loads on that crappy chinese PSU which is rated for 2.4A, and ist has the three encoder bulbs of 120mA each + three EXE electronics at 140mA each (although not sure on that as the recycled three axis EXE also has 12V supply, which I took off the ITX PSU), then the Logitech powered USB hub, the ELU Monitor Touch stuff, and the Blue tooth adapter for the keyboard are all on it.

Lets say 1A at 5V, I was wondering what batteries I can use. But I still have the 6V lead acid batterys removed from the MAHO, so I'll throw them on a charger and see if they have any life still in them, and add a dropping resister to get 5V.

Mark