Viper step dir drives for DC servos

Yes, our servo amps can be run from a Mesa controller. But, they need a "boot up" sequence to operate. After they are enabled, they need a short PWM pulse in each direction to clear the shutdown latches in the FET driver chips. I put a state machine in my PPMC driver to do this every time you came out of E-stop, but the Mesa driver doesn't have this feature.

A few people are using our servo amps with Mesa boards, so you could get the logic from them. I think it was done with a little custom HAL component.

Or, our PWM controller does use an FPGA, but it is controlled via the computer's parallel port as the communications channel.
It does not use it to send steps, it uses it to read/write a bank of registers on the board.

Yes, 250 cycle/rev encoders are pretty low resolution. I'm guessing this might be a Bridgeport V2XT or something like that. I think they have a 2:1 belt reduction. So, 250 cycle/rev X 4 (for quadrature) X 2 (belt) * 5 (screw TPI) = 10,000 counts/inch, which doesn't sound too horrible. But, more is always better.

Jon

jmelson wrote:

Yes, our servo amps can be run from a Mesa controller. But, they need a "boot up" sequence to operate. After they are enabled, they need a short PWM pulse in each direction to clear the shutdown latches in the FET driver chips. I put a state machine in my PPMC driver to do this every time you came out of E-stop, but the Mesa driver doesn't have this feature.

A few people are using our servo amps with Mesa boards, so you could get the logic from them. I think it was done with a little custom HAL component.

Or, our PWM controller does use an FPGA, but it is controlled via the computer's parallel port as the communications channel.
It does not use it to send steps, it uses it to read/write a bank of registers on the board.

Yes, 250 cycle/rev encoders are pretty low resolution. I'm guessing this might be a Bridgeport V2XT or something like that. I think they have a 2:1 belt reduction. So, 250 cycle/rev X 4 (for quadrature) X 2 (belt) * 5 (screw TPI) = 10,000 counts/inch, which doesn't sound too horrible. But, more is always better.

Jon

It is a bridgeport Series 1 R2E4 Boss 9 with SEM dc servos.

I know mesa stuff pretty good and have done quite a few retrofits, large and small. I was leary to take on a different system and have to learn everything over again. But at the price's you said they were, I guess I should go that route.

I should be able to use the existing power supply. I dont see why not. I will look into your universal controller and servo drives.
It should be able to use an older PC? No step/dir pulses so it shouldnt tax the computer that much. Is there sample configs using this setup for hal files I can look at?

Well, "older PC" is relative. You certainly want over 1 GHz CPU speed for later versions of LinuxCNC. The CPU load to run our PWM controller at 1 KHz servo thread rate is quite modest. Yes, we have sample configs included in LinuxCNC, look under by_interface/pico/univpwmv That one turns on velocity estimation which makes the PID run a bit smoother.

Jon

Following a serious accident with a mouse while browsing on ebay, I now have a fairly large (3t) 3 axis machine with SEM brushed drives - this while I still had no workshop. However, nobody died and I replaced the ancient servo "amps" with 3 of the DG4S-16035 drives powered by a 3kVA isolation transformer rewound to give 100Vac. Although they are driving DC brushed motors, the controllers are digital inside and seem reasonably well sorted. The SEM motors have Haidenhain clone encoders (1000PPR) and the whole thing works rather nicely. The CNCdrives drives are not entirely industrial quality but (rather like DMM-Tech products) they seem well priced for amateur users.

I sincerely hope you did not injure your hand with that accidental mouse moment. Sounds like you need it to fight Tommy off with an Iron Bar

Tommy, my now deceased father once told me that if you wanted to see a good cat fight, you grab 2 cats, tie their tails together and throw them over a clothesline. Sadly, neither he nor I have ever had the nerve to attempt this but we have had a lot of laughs just visualising what it might look like. Please give it a go on your side of the world and report back.... video evidence preferred.