What's new
What's new

Should it spin by hand? Hrt servo

I think your parameter reasoning is correct, home on both Z channel and table zero switch and inverted Z.
I do not have one of those controls. I presume it rotates backward to find the table zero switch then rotates forward until the next Z index.
Unless you can change the parameter the new encoder will have to have the same Z polarity.
1/2" shaft might be hard to find, most are smaller. I had to turn down the motor shaft to put the encoder on.
I've contacted Haas, but they don't service machines this old, nor carry parts or even info, and couldn't help me other than to provide the info below and recommend that I find an aftermarket solution.

"This is the description I have for the encoder on the motor. It is all together as 1 unit."

1657251468097.png

I'm finding a few companies making replacement Renco encoders. I will contact one of the encoder manufacturers tomorrow.

In looking at old posts from Magno-grail, I see he's encountered something like this quite a few times and there are some solutions out there, which makes me hopeful.
 
Last edited:
I did not scroll up to see your pictures. The LM339 is a quad comparator, so yes, TTL output, sometimes referred to as Open Collector or Single Ended.
Haas encoders at the time were 2000 line but you can use a different count by changing the parameter.
The HEDS 9040/9140 (what I used) are cheap ($44) compared to RENCO
There is a 1/2" shaft option but they are hard to find. 8mm is more typical.
There are replacements for the RENCO but run a few hundred dollars.
 
I did not scroll up to see your pictures. The LM339 is a quad comparator, so yes, TTL output, sometimes referred to as Open Collector or Single Ended.
Haas encoders at the time were 2000 line but you can use a different count by changing the parameter.
The HEDS 9040/9140 (what I used) are cheap ($44) compared to RENCO
There is a 1/2" shaft option but they are hard to find. 8mm is more typical.
There are replacements for the RENCO but run a few hundred dollars.
Thank you. This is helpful too. Does one normally buy the encoder and the encoder wheel separately?

I see what you mean about finding a 1/2" shaft. One site shows the inventory history and it appears the 1/2" shaft models dried up at the beginning of 2022.

Do you have a picture of your set that you could post? I'd like to see how this all looks in action.

Thanks again.
 
A HEDS encoder can be bought as separate pieces or as a unit.
Here is the encoder base plate and housing:
P1010037 (2).JPG
You mount the base down with the centering tool then snap on the housing then tighten the code wheel.
 
Here's an update as of this morning.

Haas Corporate was not able to do anything more other than give the number of Uptime, which services CNC machines http://www.uptimecorp.com/ I spoke with Warren, a very helpful guy at Uptime, who is going to look around for an old motor to see if he could salvage an encoder, but thought it was doubtful. He also connected me to Encoder Products - https://www.encoder.com/contact-us

Encoder Products evidently has a huge cross-reference database and might be able to help me out. I spoke with Steve and sent him some pictures via email. I'm waiting to hear back.

Finally, I spoke with Quantum Devices. When I mentioned 1/2" shaft, he said they were backordered about 52 weeks on any encoder with a 1/2".. He offered an alternative, which seems to be similar to the one suggested by @magno-grail - https://www.quantumdev.com/encoder-store/qml35/. Since it does not come in a 1/2" shaft size, he suggested either 8mm or 0.250". Same again as magno-grail. It's $37 for the encoder and you can get a 5-wire cable for $10. There is one option to which I don't know the answer, which is whether the index is 90 degree A & B High or Low. @magno_grail - do you know?

I'm going to wait until I hear back from Encoder Products before moving forward. Given all I've learned, I'm currently leaning in the directions suggested by magno_grail.
 
The HEDS encoder is Z index high. In the parameters there is a switch for Invert Z Index.
It is not hard to turn down the motor shaft for the encoder.
 
The HEDS encoder is Z index high. In the parameters there is a switch for Invert Z Index.
It is not hard to turn down the motor shaft for the encoder.
Thanks. I looked for an Invert Z Index, but did one did not jump out at me. Though, per my prior post, it will "home if inverted Z" so it must be there somewhere. I have a 1630 Victor, so turning down the shaft shouldn't be a problem.

I got a response back from Encoder Products with some additional useful info. Their replacement unit is $344 or 9-10x the price of the other options. It does accommodate a 1/2" shaft, but still that's a ton more dough. The notes from Encoder Products are below regarding my unit after inspecting some photos I sent them. EPC refers to their replacement and MCD refers to my unit.
  1. Mechanical: ½” bore, 2 pt flex on 1.181” BC
  2. Cover: Blind hollow, EPC = 1.19” deep; MCD = 0.80” deep
  3. Electrical: 4000 CPR, A,B,Z index, A leads B for CW rotation from mounting face. NPN Open Collector
  4. Pull-Up resistors: MCD has internal pull up resistors. External pull up resistors (2.2k Ohm recommended) required for EPC Open Collector output. Install inline unless receiving device has internal resistors.
  5. Connection: MCD = individual wires. EPC = shielded cable. According to MCD data sheet: Yellow = A, Blue = B, Orange = C index, Red = VDC, Black – Ground. EPC wire color and function are on the label.
I've requested the data sheet they have for my MCD unit. It's nice to have confirmed the color coding.

It looks like I'd need to add resistors for both this option and the HEDS option.
 
The HEDS datasheet says 2.7K resistors within 4 feet of the encoder.
Your encoder appears to have 4.7K resistors so you would need to add them. Digikey has a resistor array 4604X-101-272LF-ND
2.7k Ohm ±2% 200mW Power Per Element Bussed 3 Resistor Network/Array ±100ppm/°C 4-SIP
 
After looking all this over and coming up to speed, here's what I'm finding, and please correct me where I'm wrong:

With the HEDS-9040, I would need to buy the 9040 encoder (~$30), 6140 encoder wheel (~$40), and 8906 alignment tool(~$72), so it adds up to about $140, though I'm sure I could find them cheaper. I'm not sure if that actually includes the housing that holds the encoder and wheel. I've looked on Mouser, Digikey, Broadcom and Newark, and I don't believe I've seen the components together as one unit. Also, I believe the max resolution is 1000 lines. It turns out that my encoder has 4000 line resolution, not 2000 as was thought earlier. The wheel can't accommodate a 1/2" shaft diameter and I would need to turn down my shaft to 8mm.

The QuantumDev QML35 (https://www.quantumdev.com/encoder-store/qml35/) costs $37 for the whole package and I can get it at 4000 line resolution. The wheel also can't accommodate a 1/2" shaft diameter and would need to turn down mine to 1/4".

It seems like the QuantumDev QML35 is the cheaper, easier solution and I can get the same resolution as before. Please let me know whether my conclusion makes sense.
 
The Broadcom encoders with the higher resolution are AEDM. They are available complete with housing, about $55. The QuantumDev are cheaper and the shaft has to be turned down in either case.
You do not need the commutation output (UVW) since the motor is not BLDC.
 
The Broadcom encoders with the higher resolution are AEDM. They are available complete with housing, about $55. The QuantumDev are cheaper and the shaft has to be turned down in either case.
You do not need the commutation output (UVW) since the motor is not BLDC.
Thanks again Magno. I checked out the AEDM 5900 series and that looks like a potential solution. The shaft on my motor sticks out 0.815", so I would need to think through the potential for interferece.

I'm going to order the QuantumDev unit and give it a go.

I need to test the motor now that it's been cleaned and I have new brushes. What would you expect the max DC Voltage to be when running the motor? I have a 60V DC power supply and, if I need more, I have an autotransformer which I could run through a rectifier.
 
The servos on a VF-1 run on 160VDC but it is PWM so the effective voltage is much lower. You can run the motor on much lower voltage. I have run them on 12 volts when checking the encoder output.
 
I hooked up the motor to my DC Power Supply last night and, again, it started to emit smoke, once I got to about 20 Vdc. This time, I removed the rotor from the motor body (stator) and reinstalled the brush holder in order to see what was causing the smoke. I'm glad I did this because it took a bit of the mystery out of things. A smoking motor is like the boogieman to some people, including me.

In my case, the smoke was coming from the plastic cage that holds the commutator in place. In the pictures below, you can see how the plastic is charred and that pieces of the commutator can be pulled out. There is some pretty noticeable wear on the commutator as well.

PXL_20220714_015124838.jpg



PXL_20220714_015313459.jpg



PXL_20220714_015321298.jpg
 
I considered replacing the commutator and was able to find something that might work, but first I did a little more detective work by measuring the 180 degree resistance and side-by-side resistance of the commutator bars. First, I measured the resistance between all of the pairs of opposing commutator bars (I have 21 bars, so as close to opposing as possible.) The readings were all 0.5 ohms. Then I measured the resistance between neighboring bars and most were 1.0 ohms except for two that were 0.8 ohms. In both tests, all of the various readings should be roughly equal, so that all seems pretty good. Perhaps the two readings at 0.8 ohms may indicate an issue, but I don't know enough to know whether it is an issue.

Next I measured for continuity between the commutator bars and the outside of the armature and I do have continuity on every connection I tested, with resistance ranging form 2,000 to 20,000 ohms. There should not be any continuity, so I believe that is a problem. I think that implies that one (or more) of the wires in the windings have breached their coating allowing current to escape to the larger body of the armature.

My rotary table initially gave me a "Hi Current" error. V = IR, so if I is too high, then perhaps R is too low. Perhaps, then, some of the flow of electricity is bypassing the maze of windings (and the resulting resistance R), resulting in a higher current (I) being read by the table.

So, as it stands, I can think of several options:

  1. Take my motor to a shop and get it rewound (there is one a few blocks away)
  2. Try to find a cheap replacement motor on eBay or via Google. However, I've been searching the web without luck. The model is ME3258-227F. I can find many ME3258 motors (Torque Systems, which seems to have been a division of Cleveland Motor Company, EG&G, ITT, and now MTI motion) but I have not been able to find a 227F. I haven't yet found a datasheet either (the motor is ~30 years old) that informs me what the 227F stands for. I'm guessing F is one of 8 winding variations.
  3. Walk into a motor shop with a big wad of dough and ask for a brand spanking new replacement motor (not my way).
Likely to be continued...
 
A quick update here. I called around for a motor rewind, and the quotes averaged about $1000. After doing more research on the motor, I figured out the most important specs are the 3528 series and the F windings. I scoured the web and was able to fine only one 3258 motor with F windings. It's not in the correct format (has a tach, wrong base, shaft too long) but I think I might be able to use the rotor with my existing stator, base, and brush holder. It came from a working machine. I'd still need to get an encoder.

Also, I found the model number for the encoder finally, but no spec sheet. It is branded Motion Control Devices:
  • Encoder Model # M21-DM-500-5-SE-50-A-W
  • Encoder Part number - 14414-358
After all the prior back and forth, it appears to be 500 PPR from the model #, rather than 4,000. My motor controller parameter is set at 4000 encoder steps per motor rotation, but perhaps that means something different that what I thought. I found another seller that can supply one with a 10 day lead time for $275, including the 1/2" shaft:
http://www.digitaldevicesinc.com/digital_devices_21SG_series_modular_encoder.html

I'll nail down the motor first.
 
You have to be careful on how they state the resolution. Some use lines, others use PPR (pulse per revolution). I may not be remembering right but I think lines refers to the lines on the wheel. So a 500 line encoder has 500 lines on the wheel. But a quadrature gives four changes in AB (00, 10, 11, 01) in pulses per revolution so a 500 line encoder makes a 2000 PPR encoder.
 
On a related and very interesting note, I just learned about EG&G, whose name appears on my servo motor as the owner of Torque Systems. As I may have mentioned before, Torque Systems has made its way around the industry and was owned by Cleveland Motor Company, ITT and is now owned by MTI Motion. More below the picture.

PXL_20220715_191205250.jpg
I plugged EG&G into Youtube, looking for info on a potential replacement servo motor that I found for sale, with a tachometer, and found a video about Operation Sandstone - EG&G


I watched the first few minutes and found out that EG&G were three colleagues and friends Harold Edgerton, Kenneth Germeshausen, and Herbert Grier that incorporated in 1947 as EG&G, Inc., at the request of the Atomic Energy Commission. They designed and operated systems that timed and triggered nuclear bomb tests. Their first venture was part of Operation Sandstone, which was the third series of US nuclear tests in 1948. The first two were Trinity in 1945 and Crossroads in 1946. Pretty interesting.

The film that I was watching indicated that it was created at Lookout Mountain Airforce Station in Hollywood CA. I actually live in Hollywood, though I don't work in "the industry", and thought, "There's no friggin' Air Force Base around here!" Well, it's not an Air Force Base, but an Air Force Station and it existed from 1941-1969. In 1941 the Air Force commissioned a station up in the Hollywood Hills, on 2.5 acres, as a World War II air defense center to coordinate Los Angeles area radar installations. From 1947–1969, its use was changed, though still an Air Force Station, and it produced motion pictures and still photographs for the United States Department of Defense and the Atomic Energy Commission.

The Air Force Station is/was located at 8935 Wonderland Avenue up in the Hollywood Hills. It has been in private ownership since the late '70s and has been owned by a few actors and developers. It is now owned by Jared Leto who is an actor who has been in movies since the mid 1990s.

And all I wanted to do was figure out what the inside of the tachometer looks like on a servo motor, and can I easily remove it.
 
Last edited:
I bought a used replacement motor on eBay for $214 delivered to my doorstep. It's also an EG&G PM DC servo with model # MT3528-102FF. The original was ME3528-227F. Both are the same sized motors (3528). The new one has a Tach (T) and the old one has an encoder (E). 102 and 227 are just internal numbers for client specific built motors. The first F is for the motor windings and the second for the tach windings. This was the only F winding motor I could find after several days of searching and it seemed like a good idea to get the same windings.

I ran the replacement on my DC power supply and it is noticeably quieter and smoother than the original, although I did not think the original was loud or unsmooth. I ran it up to 40Vdc and there were no issues. Next, I took the motor apart to contemplate swapping it over to the rotary table. I can essentially take the rotor and stator/permanent magnet from the replacement motor and install the belt pulley, base, and brush holder from the original motor. The only hitches I see are 1) that the replacement motor has a tachometer which needs to be permanently removed in order to 2) fit the encoder.

Regarding the issue of removing the tach, the motor shaft is 1/2" on both motors. On the replacement motor, the tach has a separate commutator and small set of windings that are just above the commutator for the motor windings The tach sits on a OD=5/8" hollow tube which is pressed over the 1/2" shaft. I think/hope that I can remove the 5/8" hollow shaft and tach.

There is also some kind of blue epoxy on the shafts, and also the outside of the rotor windings. You can see it in the pictures below. My inclination is to use a gear puller on the tach windings and try to pull the entire tach assembly off. Has anyone removed a tach like this before and, if so, can you offer any advice?

Replacement motor, with original motor posing in the background.
ShciOPKcUgyR95VxLvjPoq7NPa-6gN_iMo7oi7GpvJh5vz_8KooeXMlY53wKi1MffFPQwm_EPLZfuVZHF6LrcmGklc8SpH5vdOJDGLJFT0BpjUt1Z4mqt9agjI9Ca9UmH1hzTJ0tR9bgSlGWbRESmyXz1Q6Zi_xAcLi0RdfmVtCINslt27um0N_op2RFsKS5yYhx4GftuJVRhkcATZXq6CkXlno53Z0UZ25dD-SkQpt4flJM1HsbklSzpta3NPRobEmz4axsBoa9YgG9XOM5vmhNy2SBg_BkiUt1VjUPF3_G4DwiEiLBWyexby5zxsXX-pfkfncb-CRMw4Aun3-IxiZFAVddn8wcx3oW0hseztH27vVOlnaokfGlokywS6UW8atjEi2SM7sz6z6bitSJ-B26FO8S9j00IbGCqMQpPLiGgz2WMuvOubQk8K3k5_PmHhh7CerL2plcdkUTVv7s1vQ0v0pxQn3vT97i8tF_045W1LEtKKqVzy3lJ4N7rfqn8Bthb8cxLY9OEpe-eW4OnUNaZBzJa0rvzPf33g9NEIUFx8BXi8VdjPQd3BDIys13tyLVVUUCnHC_s4ncEH1mdgxEsHCltnR4doF-VjqVcIfYeQ0HSOC0FDVf6w1M8PkvspqkXJHu0GSOu3LLPDhWHWjY1yfjEa37SM6Ps6K9Qx7Okpdw9xIxRaTH5EVuqdfsLYKzbooM0jIYxKpeDVZu5Tpou9hArR1m_nfwuVTqE0tPEVjsDc873Y9rOaIexD8=w690-h1226-no


This is the commutator and windings for the tach, which are under the top cover, above. The wear on the tach commutator is significantly less than what I found on the original motor's motor commutator. Unfortunately, I cannot remove the brush holder for the motor windings until the tach is removed, so it is not possible to see what's going on with the shaft between the motor commutator and the tach windings.
1658964848757.png

And here's the end view of the shaft. The 1/2" shaft has a center hole that is 3/4" deep.
1658964998735.png
 
I'm back on track. I got the tach off without much effort at all. I reconfigured the new rotor/stator with the end caps from the original motor and everything went together well. I'll need to mill a slot in the shaft at one end for the pulley's 1/8" keystock, and turn down the other end for the encoder.

This was so simple and I thought it was going to be more of an issue.

Pulling rig
hP68yK5UpYrCQgWbJA0i78Q7uNOo1Ryava-OZcNaURgexwX7fmWvJQyS-VzoVKZspO5Oyzsbnq1-fQrZbSqctC-_EjQ0riKTWrNtcaGXOIz3DggGl-LzZN9IHWxYn26mCKgaK7dxlI8=w2400


Success
BnZCqYt1WnhLYr_O_kysaXMag-YNa9SqPRc4ODuJ2tfRl6mR3ZgJ2vsyV3KcF_k55-z2YcmCe76naU2yBwgX6M48UCip2Js4RyBYBS8m8RRsroRrPp8g6YxFHUhixYiPTxQEv4R9LWE=w2400


Close-up of the knurling
oBXPCWH1W8A-X-eTnN6yuA9CqZ9UG9PbSBuRCr9PZSBy4N91PhVYshX9ClyQElTdSA6sokaPUJsADMSojMt0pKvC9Wo6m-qKuUeqBARTkQMihLgaBUDbrv2YJNw99DTibQt3qQ0ymKo=w2400


The replacement motor is above the original. The difference in shaft lengths will not matter in my case and I will turn them down to the correct length and diameter as needed. Everything buttons up quite well.
eCZrxS3u-52TEbGL1MEMV3plXDyAvJlIjk4Ra-w6jJQHkNglFXCUU5JqjLPFP5mXH9RuqYabJkTlm5_WL7mOK2tW98y9fcx6-QN9SRfooTTDvGUV66ihYzOgB3EWn4_F-BY7TfKxq6Q=w2400
 
I'm back. I finished the motor a few weeks ago and was waiting on the encoder. Due to a shipping issue, the encoder didn't arrive until the day after I left for a 2 week trip. And I just got back a few days ago. I turned down the motor shaft to fit the 1/4" encoder collar. I then wired everything up and tried it out, but no joy. Well, I got partial joy. With the encoder disconnected, I can now jog the motor without getting the Hi Curr error. So the new motor is doing its job.

With the encoder attached, when I hit Zero Return, the motor moves a little bit, then stops and I get either a Ser Err (Servo Following error too large) or EncodEr (Z channel fault - bad encoder or cable).

Using an older post response by @magno_grail (https://www.practicalmachinist.com/forum/threads/haas-14-pin-brushed-rotary-connector-pinout.393236/) I traced the wires from the encoder all the way back to the Servo Control PCB. There was continuity the entire way and I confirmed that the wires are correctly connected to the encoder, according to the wiring diagram in the aforementioned post. The only issue I found is that Pin L for the thermostat is not connected at the PCB. It's cut off about 2" from the end and has no connector on it.

The encoder has power, which is indicated by a red LED, and I have 4.98VDC at the encoder. As mentioned, I have A, B, and Z correctly connect. The new encoder has a driver, I'm told by the seller, so push-up resistors shouldn't be required. I've tried changing the resolution in the Parameters. I did determine that the original factory encoder had a resolution of 500, per the model #, but the head unit is set at 2000, which seems to be the factory setting. There must be some multiplying factor for gearing or something else. The only other thing that comes to mind is that this encoder is considered 90 degrees A & B High. A & B Low was another option I could have purchased. I still haven't found the way to invert the Z Index.

Any thoughts?

SHBPilKTwbjAc4pf8vRcYuF6qghWiGX97RViGTEGQ70Uv692IYoXU90prbhLzI1OHl8-qToYr7exS2-R8YyHSO3SXZNNSdprbkNVvgsCCdomkR1PIa_8TBkRnHlEcUbS1NkKQ0j0Oiw=w2400


7dnXGtp6_Q07cCt_aSTutZoObCDVtKRgieEfpWVSHtcqrLc4AtmOwfieXauz_TCbvzSLpDQKMHqDssTJL05LrF_fsHwJ32hjavud-qPvEuudeB4LjmCfsfpT7Xm4V2PgDvkO8IMR2jI=w2400


Xh0L1Z3eiaEzZMgPmMwEcqNNOdyxFw8x72hkegdcmTfXpqekSJ8RxcuMuJo0r7vSE7LiWnwFR9vjHhlGzTqp6X7UmmB18OiMqg8VHY9NBkbTox5DTQrKvj7K2KmZfoFNlcRfGF425wY=w2400
 








 
Back
Top