Geekworm Board Integration

This post is a collection of correspondences between myself and the support rep at Geekworm named Cindy. My original question I posted on amazon.com under the product to ask if the board was “quiet” enough to run video streaming at the same time. On other board I tested proved to generate a video scan line down the image while streaming to our surveillance recording system. I wanted to make sure the Geekworm PWM/DC Servo Motor HAT didn’t exhibit the same problems. They confirmed it would not and they were right.

I ordered their board, installed it, and confirmed it worked well with both the PWM and DC Motors. However, the next steps caused boards to fail. First, after installing board separators (also known as standoffs and offsets), the PWM feature on the Geekworm HAT failed to operate as can be seen in the video below.

After this happened, I sent an email to Cindy explaining the problem as documented in this Report I sent them with the following message:

From: Frank Gould
Date: 2018-08-02 04:59
To: gk
Subject: Geekworm Raspberry Pi Full Function Motor HAT

On amazon.com, I received the following reply to my question also below:

Frank Gould: “My pwm was working with this board but now has stopped and the LED dims and blinks. Where do I go for help?”

Harry Huang: “Please send your order number and problem details to gk@geekworm.net, it is better to attach some photo or vedio [sic], thank you very much!”

Order number: 112-5159520-0237800
Ordered on July 24, 2018

Problem details: The Full Function Stepper Motor HAT worked great with my initial tests but after installing metal separators the PWM feature stopped working. I’ve written more details and posted video on YouTube to view off my pdf report, link below.

https://drive.google.com/open?id=1bqxKBHkpxsmQtBlA7F6ZQBfe5DdbFKxC

Would you please let me know if the metal lifts caused a short or maybe too much stress on the board and shorted something out? I can send more pictures if needed. I will pay for a replacement but I need a solution to the top board that needs to be secured and not damage the HAT with metal separators.

Regards,
Frank Gould

Then the following email messages went back and forth in discussions.

On Aug 2, 2018, at 12:05 AM, “cindy@geekworm.net” wrote:

Hello sir,

Thanks for your email and PDF report details. We are so sorry about the inconvenience.

  1. Do you mean there is no problem if you do not install the metal separators? Maybe it is the metal seperators problem, is there any way that you can replace them with plastic ones to have a try? By the way your “metal separators” means the screw packs you use to install the board on, am I right?
  2. We see you use another board on the Full Function Stepper Motor HAT, if this board use PWM chip with the same address, then it will be out of control.
  3. I’m a little confused, the Full Function Stepper Motor HAT is I2C interface, how does it control or related with the led light?

Waiting for your reply.
Best regards,
Cindy, cindy@geekworm.net
Shenzhen Tongban Science and Technology Co., Ltd.

On Aug 2, 2018, at 6:41 AM, Frank Gould wrote:

Hello Cindy,

Thanks for your quick response and questions. I am sorry for any confusion regarding this report. I have attempted to answer your questions below.

1. There was no problem until after I installed the metal separators. From then on, I got the problem I reported which is why I was wondering if the metal separators short circuited some PWM components and continued to fail but not the DC controller which did not fail. I will gladly try plastic separators with a new board but I don’t know where to buy them. I found the metal separators at a local electronics store, called skycraftsurplus.com.

2. No other boards in this configuration use I2C nor PWM. The MoPower UPS and RPi 3B are the only other boards in this configuration. You can see the MoPower UPS specs at the link below.

http://www.allspectrum.com/mopower/ 

3. All I am using is the code I sent in my original report (pdf). I’m assuming the LED light is an indicator on the Full Function Stepper Motor HAT that an error has occurred. My code did not do anything to trigger the flashing LED other than sending PWM to the board.

I hope this helps. If you can find and share a link to plastic separators, I will purchase another Motor HAT and the separators to see if the metal ones caused the failure. I have not searched online for plastic separators because I thought metal ones were okay. Please advise.

Regards,
Frank Gould

On Aug 3, 2018, at 12:18 AM, “cindy@geekworm.net” wrote:

Hello sir,

Thanks for your reply.

1. You can search the plastic screws on Aliexpress:
https://www.aliexpress.com/wholesale?catId=0&initiative_id=SB_20180802200857&SearchText=raspberry+pi+plastic+screw
But you need to find the right dimensions of what you need.

2. For the MoPower UPS, it is similar with our ups hat board, maybe you will be interest in our board:
http://www.raspberrypiwiki.com/index.php/Raspi_UPS_HAT_Board

3. I hope you can figure out the problem, all the product we send the Amazon warehouse is new and no problem. Maybe you can take out your MoPower UPS board to test our Motor HAT board if this possible, thank you!

Best regards,
Cindy, cindy@geekworm.net
Shenzhen Tongban Science and Technology Co., Ltd.

From: Frank Gould
Date: 2018-08-03 17:03
To: cindy@geekworm.net
Subject: Re: Geekworm Raspberry Pi Full Function Motor HAT

Hello Cindy,

You have not answered the original question. Did the metal offsets damage the board? Your UPS does not support NiMH batteries and is not a solution. Why are you sending me plastic batches of offsets when I only need a small quantity?

Please let me know what you recommend for board separators so I can order a replacement board because the Geekworm board failed.

Thanks,
Frank Gould

On Aug 3, 2018, at 5:35 AM, “cindy@geekworm.net” wrote:

Hello sir,

In theory the metal offsets (metal separators) will not damage the board. But you can check the problem by exclusive method (by take off the metal offset one by one).
We are so sorry, we do not sell such plastic batches of offsets to install the board.

Best regards,
CIndy, cindy@geekworm.net
Shenzhen Tongban Science and Technology Co., Ltd.

From: Frank Gould
Date: 2018-08-03 17:41
To: cindy@geekworm.net
Subject: Re: Geekworm Raspberry Pi Full Function Motor HAT
Hello Cindy,

The reason for the offsets are because the PWM connectors point upwards forcing the PWM feature to take up space between boards. That is a design flaw and the only way around it is to have separators lift a top board higher to clear the PWM connectors.

The board I purchased does not work now without the separators installed so something failed by using them. Then you don’t offer any solution to fix the problem caused by the PWM connectors.

Guess I’ll have to find another solution because the Geekworm board is a failure.

Thanks for nothing,
Frank Gould

On Aug 3, 2018, at 6:33 AM, “cindy@geekworm.net” wrote:

Hello sir,

So sorry about the inconvenience, maybe there are some misunderstanding before.

1. Maybe I understand your meaning. I guess that you use the GPIO of motor hat, so no enough space between the boards, please try to use 2*20 Pin Female Header to increase the distance between boards (photo attached). Then you will have enough space.

2. I want to share my opinion with you, usually very few people use the GPIO pin of motor hat, increase the height is good solution if no enough space.

In addition, this motor hat is only a development board, it is not a final product. Please understand us.

Best regards,
CIndy, cindy@geekworm.net
Shenzhen Tongban Science and Technology Co., Ltd.

On Fri, Aug 3, 2018 at 6:46 AM, Frank Gould wrote:

Cindy,

I have done exactly what you suggested I add below in #1. That’s why I need the separators to get the MoPower UPS level on top. Because that board doesn’t have GPIO pins extended above it so I can put the motor HAT on top of it, I’ve written them about this, since their board is a prototype being field tested.

Regardless, the motor HAT is a design flaw requiring the PWM connectors to take up space uselessly between boards and the PWM pins are not easily accessible when stacked. The PWM pins should be right angle connectors and NOT pointed vertically upwards. These right angle connectors should also be moved to the end of the board so they are accessible when stacking another board above the motor HAT. They could sit on the same end as the VCC, SDA, SCL, and GND pins pointing outward.

At this point, the motor HAT fails to deliver a product that doesn’t require multiple work-around solutions to make it operate. I intend to return this board if you don’t replace this one. If this is a development board, make it fit into a configuration without senseless work-arounds.

Frank

I never heard back from them with any solutions, so I returned the boards to amazon.com who immediately refunded my account for the board. After removing the Geekworm HAT, the MoPower UPS board failed to boot, so I contacted the manufacturer. Moses’s response was essentially RTFM (Read The Fricking Manual) where it said to hold the power button down to force boot the board, which it did.

Automation Reconnaissance

After I spent a day converting 13 straight bin connectors to 90 degree bent connectors then testing them, I ordered the Geekworm HAT on Amazon.com again and received the following message:

Hello sir,

Thanks for your order with Geekworm again.

We see you order the Full function Motor HAT Robot Expansion Board again, but you have return and refund one you bought with order number 112-5159520-0237800 before.

And today we also get the Amazon negative customer experiences notification with the following reason:

  • Doesn’t driver motor.
  • Not the correct board I need.
  • This board has a design flaw that requires multiple work-around solutions to make it operate with other boards because the PWM pins face upwards consuming space, per the manufacturer whom I contacted.

Could you please tell us why you order it again as the board is not meet you need? We are a little confused.

Sincerely waiting for your reply.

Best regards,
Cindy, Geekworm

My reply 10 August 2018 at 6:35:05 AM:

Hello Cindy,

Yes, your board failed after my first workaround attempt and contacting you. I have since had to rewire my configuration due to two design flaws, as I mentioned when I contacted you last time. Because the Geekworm HAT failed after installing the separators and you offered no way to fix it, I returned it due to the connector that had to be mounted straight up.

I have since rebuilt the jumper connectors so that they are at a 90 degree angle and am going to try to install the Geekworm HAT without the height for the straight connectors. You can see in the attached photo my attempt to replace the straight connector (left side) with the 90 degree angle connectors I created (right side).

The Geekworm HAT is the only product I have found that is 1) quiet (no video interference other boards exhibit), and 2) both PWM and DC motor servo drivers on one board (with 4 PWM channel pins instead of 16 channels that most PWM boards offer). I am hopeful that the engineers at Geekworm fix the straight up PWM pins in the center of the board and put them accessible on the side, like the SDA, SCL, VCC, and GND pins are on one side of the board. I offer to test this for the next production batch.

I don’t understand why this is confusing. I documented this clearly in our previous communications.

image1.jpeg
Best regards,
Frank Gould

 

Cindy’s last reply at 7:46AM, Friday, 10 August 2018:

Hello sir,

Thanks for your kindly reply.

The office workers get off work now, we will send your description details and photo to the engineer next Monday when we come back to work.

Really thanks for your suggestions. We will try our best to meet each customer’s demand like yours.

Have a good day!

Best regards,
Cindy, Geekworm

I will update this as this saga plays out. Hopefully, they’ll fix their design and I’ll get a chance to test it.