advice choosing servo amplifiers. Brushed DC motors, +-10V signal input

Tom_R2E3 wrote:

Todd Zuercher wrote: I've just been comparing it to the X axis drive (which still works) and the working drive looks much worse. When I measure those resistors they measure 1.5kOhm.

Well, obviously, your definition of worse is backwards. Because the one with the lower reading is the one that works.
But, without understanding what this resistor is connected to, or what function it performs, you really don't know what
might be wrong.
{quote]
Jon, it's the power rating which surprises me. It's only a series 1 Bridgeport but 40A seems bigger than most of the drives I've looked at so far. Is it because the motors are old and inefficient? There are loads of Amps in the range of 15A constant and 25A peak.

No, these motors are efficient. If not, they would burn up! That is a LOT of concentrated power in a SMALL package, it better come out as shaft output. You can look up the torque constant of that motor to find out how much torque it produces at 45 A, and then compute the linear force it applies to the machine table. Then, if you know the weight of the table, you can compute the acceleration.

Jon

Yep all makes sense.

I've just been reading the manual for the NC400. The model which I have (NC414) is quoted as:
30A peak (5 seconds max)
14A RMS
14A continuous

I've got brushed SEM motors on my Shizuoka and used the cncdrive DG4S16045 Work very nicely indeed and although they aren't industrial gear, they seem good value at 120 Euros each.
www.shop.cncdrive.com/index.php?productID=366

As you note, these require step/dir (no analogue input) ie you'd need to change the 7i77 to 7i76.

That board seems to have been made around 1984 - 1985 or so, judging by the date code on that resistor (1984 week 14). I don't recognise the markings on those transistors. may be tricky finding replacements. You can probably expect further compt failures at any time judging by how hot they have been running. I assume they are linear amps rather than modern switchmode drivers.

You can junk the tachos if using a modern digital drive and use a higher resolution encoder (I have 1000PPR) - Chinese Heidenhain clones are available from ebay etc.

EDIT - seems I was wrong about the power devices. The controls are analogue but the power stage is a switch mode circuit. But I see that they are using an NPN / PNP complementary pair. You will struggle to find any PNP switching transistors these days, as they simply aren't used. Back in the 80s, they had a different view.

I take it you've seen this manual?
manualzilla.com/doc/6876616/operating-an...series-servo?page=12

So it's been a couple of weeks and I wanted to give some feedback, firstly to say thanks to everyone for their input. Never fails to impress me how good the support is on the Linuxcnc forums.

After seeing the cost of replacing the three amplifiers I decided to have a go with a second hand NC414 amplifier from ebay (US). The unit was very cheap so I thought it was worth a shot. Arrived after a week or so and didn't work, similar to my broken unit - nothing happens when power is applied.

I tested as much as I could on both units with a multimeter (some of the tests are described in the manual which Muzzer linked above) but couldn't find anything wrong so I decided to desolder the suspect resistors and test them off the board. Turns out one was blown! Nothing to lose at this point so I desoldered one of the resistors from the Ebay unit, soldered it in to my unit and it's working again (for now). I finished the program I was running which was about 30 minutes of milling and no problems so far.

A year or so ago that drive overheated and triggered a temp limit. When I investigated, the fan directly cooling that unit had failed. I replaced the fan with one of the case fans, just to get me going again, and after cooling down everything worked again. Perhaps this caused some lasting damage to the components...

In the future I'll probably invest in some modern drives, I just don't want to do it right now because there is also a chance that i'll have to sell that machine this year. Let's see how I get on this week, got a lot of milling to do in the next week or so.

Tom_R2E3 wrote: so I decided to desolder the suspect resistors and test them off the board. Turns out one was blown! Nothing to lose at this point so I desoldered one of the resistors from the Ebay unit, soldered it in to my unit and it's working again (for now). I finished the program I was running which was about 30 minutes of milling and no problems so far.

Fantastic detective work! Amazing that there were no other faulty parts on it.

Jon

There is the Viper drives that wpork with these servo motors and not expensive.
Plus step/dir input. will have to change the encoders though.