Cheap Power Supply

Note to self:

Do spend more than $8.00 for a power supply which is critical to the operation of a system. In this case, these cheap USB power supplies are failing after six months of continuous operation.

Busy Summer 2016

image of two racks We have another extremely busy summer this year with 10 systems under construction in-house, and more in the planning stage for installation later this year.

Here are two out of the 10 racks under assembly by my associate Dan M.
Two additional racks are already complete and undergoing final testing prior to classroom deployment.
Our deadline for all 10 is September 1st.


Dan is also CTS Certified and does extremely tidy and serviceable rack layout according to industry standards.

With the exception of the system controller, all of these systems are now designed around Atlona core products.

stacks of boxes






More later. Cheers.  -Pete

RF remotes for camera presets

As posted previously, the IR triggers for camera presets were a fail due to false triggering. I wanted to provide manual pushbuttons at the three, main locations to allow the users to manually control presets. The issue which caused us to go down this road was this: no available pathway from stage-right to stage left for any type of low-voltage wiring. I wanted to try RF remotes.

I found a great DIY Build site called adafruit which stocks numerous RF transmitter and receiver modules. I selected a little 4-channel receiver board, and a compatible 2-button (2-channel) transmitter keyfob. This would allow me two wireless pushbuttons.

I needed to add a few external components (relays, driver transistors) to the receiver board, so I found some surplus perfboard and parts and soldered it all together. RF-board

The little receiver board is about half the size of a business card and is seen on-edge in the photo here:

For the transmitters, I disassembled two of the keyfobs. The batteries take up most of one side of the circuit board.

Fob2  Fob4

On the reverse side, under the mechanical dome buttons, are the contacts to which I soldered a pair of wires to extend to a hardware-store doorbell button.

Fob3  doorbell

I mounted the doorbell to one side of a single-gang wall plate with the wires extending through a small hole, then used — gaffer’s tape — to attach the small circuit board to the rear of the plate. I mounted the plates to shallow junction boxes at my two remote locations.

I found a surplus power supply and gutted the case. It was a perfect size to mount the receiver with the rest of our AV gear and the Vaddio camera controller.

Case1  Case2

I wired the receiver to the Vaddio controller and the remotes work perfectly! The whole mess cost less than $25.  -Pete

Cheap IR Triggers, Part-3

Good news: The triggers worked! Bad news: We tried the IR triggers for a couple of quarters and found it too difficult to narrow the window to directly in front of the whiteboard.

The IR triggers were subject to reflections off of the surface of the whiteboard; student movement in the front row would cause a trigger and unwanted camera movement. After many complaints, the IR system was scrapped and replaced with a manual, RF trigger system. And it has been working great. More later.


Cheap IR Triggers, Part-2

IR sensor plateAs our summer installations progress, I have had more time to consider where the IR triggers for our Vaddio camera systems will be installed. One is wall mounted, the rest are on the ceiling above the whiteboards. I wanted the ability to vary the coverage angle, but the heat shrink solution, mentioned previously, was too cheesy.

I decided on using the shell of a surplus Switchcraft D3M connector with the sensor mounted recessed into the rear of the shell–a perfect fit.

IMG_3424 (Medium)


I hot glued the circuit board onto the rear of the faceplate (and my finger. OWW!).

For now, I am just going to tape the little sensor board onto the rear of the connector shell until I have finalized the coverage angle to mitigate false triggers.

Parts cost is still under $10.

Middle Atlantic Products Fail

I have been spending way too much time trying to get Middle Atlantic Products (MAP) to replace some stupid, defective horizontal lacing bars. I am getting the run-around, so I am escalating here in order to publish the pictures in one location.

We have previously had very good results with Middle Atlantic’s various products, and we purchase a lot of them. The lacing bars mount to the rear rack rails of an equipment rack and allow an installer to dress the cables properly, and the distance between the mounting holes, horizontally, is a long-time industry Standard. Here’s the deal:

  • According to the EIA/ECA-310-E Standard, horizontal, hole-to-hole spacing should measure  18-5/16″, +/- 1/16″ tolerance. (18.307″, +/- 0.059″) So, 18-3/8″ is the maximum width to still be in tolerance.
  • Our rack spacing is correct: 18-5/16″
  • The MAP lacing bars received are too wide at 18.5″ and cannot easily be bent into submission (and I should not have to);
  • I called MAP Tech Support and he said they are “…within tolerance.”
  • I spoke to a MAP salesman at InfoComm. On display there was an improved version of the lacing bars with oval holes, allowing for width tolerance. The salesman said “…just call and get an RMA.” (Easier said than done)
  • The MAP website illustrates the improved version with oval holes!

MAP originally shipped us three packages (10 bars to a package); they were all too wide, measuring 18.50 inches between mounting holes. That is NOT “within tolerance.

Cheaper IR Solution

We have been installing nice Vaddio HD-19 cameras in our classrooms to enlarge lecturer’s whiteboard scribbles onto the large projected image. Our first outings with these cameras included Vaddio’s HotShot preset controller as well as those heavy rubber trigger mats placed in front of the whiteboards and podium. The mats don’t work well with a chair sitting on top of them (!), so ceiling-mounted IR sensors should work great. Vaddio wants hundreds of dollars EACH for their IR triggers, so I decided to make our own, for under $10. each.

IR sensorMy intent was to find a sensor which could be mounted behind a single-gang wall plate. I found these little sensors ‘with relay’ on eBay for about $6. each, so we bought eight of them. The turn-off delay is adjustable.

In initial testing I found that the relay output is not ‘dry,’ but is +12Vdc. That’s not compatible with the input to the HotShot controller, so I traced out the circuit to see what could be done.


I located a trace which can be cut to remove the Vcc from the relay output, and lifted a little chunk of copper from the board with an x-acto knife.

I just added a short jumper wire to the relay contact and now I have my dry contacts: jumper wire



The next step is to add a small, 4-contact barrier strip of some kind.

We always have parts drawers full of leftover hardware, and I found a handful of small connectors from some Extron project.

I super-glued the connectors onto the end of the little relay. Super glued

These will make a perfect barrier strip. The next step is to trim and terminate the four wires:







This first board is my prototype. I drilled out a single-gang faceplate and hot-glued the IR sensor and relay board to the back of the faceplate. faceplate-rear


The blue outline is the opening into the Wiremold  2400-series junction box I expect to be using. The whole wretched mess has to fit within that rectangle.

And it does:

IMG_2972 (Large)



The PIR sensor has a wider angle than desired, so I need to experiment with various tubing, heat-shink, etc. to narrow the vision range.

In this case, black 1/2″ heat shrink works, but it’s unattractive and more difficult to fine-tune. I’ll post additional images as the project continues later this summer.

Heat shrink

I have tested the board and it works great with the controller.

Poor ‘Best Practices’

old room layout


Today I had a chance to dig around inside a large Crestron DM installation which was performed here just a couple of years ago. The department owners/users have had significant problems with the installation ever since the job was completed. Unfortunately, due to politics, my own department and I were never consulted and we were unaware of the project’s scope until after the installers were on site. The installation was performed by a large and well-known design-build firm. The owners/users are generally happy, but there are random failures and at the moment all audio has died, prompting the request for my consult. The owners were never provided as-built drawings. This is a full DM matrix with support for video conferencing and multiple displays.


I spy some significant installation errors and poor practices. What do you think? The system is in two parts: a lecturer’s console with rack, and a full-size distribution rack in a nearby closet.

Instructor's view of consoleThe console casework is nice, but the rack is barely large enough to support all of the hardware needed for the front-end of the system. The desk is height-adjustable, electrically; but is not used since stuff stops working whenever the desk is raised (see below).

It wasn’t until after the system was in operation that the users discovered there was not appropriate horizontal real estate for laptop users. DOH!

Console rack frontThe console rack is 15-RU and four of the spaces are taken up by horizontal power strips. There is room in the rear of the rack for vertical strips which could have possibly allowed tidier cable dressing between hardware and a traditional cable dump at the bottom.


This is a mess.

There are a number of jumper cables which are too long, so they are coiled up and hung from self-adhesive tabs(!) Several of the tabs have failed (they all do, in time), causing the weight of the cables to put stress on the connectors and cause random disconnects–particularly any wretched HDMI connectors.

Inside console rack

Here is a look up inside of the rack.

Standard plastic cable ties are used on all cables including category and DM.

It is not so obvious in these images, but power cables are often seen cable-tied to signal cables–very poor practice and definitely contrary to the contract specs. Nobody called them on it.


Under the desk proper, things could have been dressed better, and cables with generous service loops bundled more securely to allow for the desk height adjustment without stretching cables and connectors like a bow string. I can see how raising the desk will cause failure.  Under-desk wiring

The main distribution rack is located in a very small closet behind the whiteboards:

Distribution rack

The rack has to be on casters to rotate for access in this tiny space. In this image it is turned for access to the rear of the rack.

I seem to recall the wall-mounted air conditioner was added after-the-fact to mitigate overheating.

At first view the in-rack wiring looks quite tidy!


Let’s look more closely. Rack rear




It appears that power cables are on the left, and signal cables are on the right. Excellent.

Hmmm. I don’t see any Velcro straps used on the category cables yet. Are those cable ties really as tight as they appear?

IMG_0846 (Large)

Here again, power cables are tied alongside the DM, audio and LAN cables. No Velcro ties.

Connector farm




Shielded, twisted-pair cables are nicely terminated and labeled, but no cable ties were used to secure the cable to the connector.

Cable ties

This is beneath the Ethernet switch. Category cable cinched down with cable ties alongside the power cables.

Along the side of the rack, there are bundles of power and signal cables kinked together:

kinked cables


I will have a look around again later. The owners have asked the installers to return and repair the intermittent problems, as well as to provide the as-built drawings the original contract required. The system is “out of warranty” so time-and-materials charges apply. What do you think? Would you ever use these installers?