Skip to main content

I’m referring to this post from Adrian’s thread “Weak DCS signals…”, a few years ago.

I would like to do (i.e. have done by an electronics expert tech) both the upgrade to a larger basic rapid TVS Zener than the one on the Rev L TIU board, and the “slow” TVS board designed by Adrian! and shown in the above post that was (so far) yielding a long life for signal strength at their club.  Understand his explanation of the need for placement within 1/10 of a very short wavelength of the protected device, hence inside the TIU within an inch or so.

Is there a design for a small board with either of the diode types, those used in Adrian’s first draft version, or the smaller board and SMD diodes suggested by GRJ in his question beginning the post?  I have a tech who is willing to do the complete basic TVS upgrade job, but wants me to supply the “slow” suppression boards for him to install.  So, I need a feasible design I can fabricate myself and supply.  Just need to make two of them.  Fairly small soldering with a temp-controlled station is within my skills and toolset, though experience with attaching SMDs is limited.  I’m sure Google can help there.  A suggested compact design would be very helpful.  GRJ’s idea of a very small (O.5 inch square) board and compact SMD diodes sounds very interesting.  Do not know what parts to look for to execute that idea.mceclip0

I’m puzzled by Adrian’s parallel circuit arrangement of the diodes; it looks like a roughly 7 volt AC drop (ten silicon diodes in series for each current direction) as opposed to a Zener-type opposed-diode suppression device, so I don’t think I understand the principle underneath it yet.

Thanks for suggestions, and help with understanding.

Attachments

Images (1)
  • mceclip0
Original Post

Replies sorted oldest to newest

Adrian's design with diodes is simple, the spikes were trying to stomp on are far in excess of 7 volts.  One benefit of the diodes is the low capacitance minimizes the shunting effect of a TVS.

My board looks to accomplish the same thing in a much smaller and hopefully easier package to install.  My protection board is connected directly to the 74ACT244 chip, this is the chip that usually takes the hit from voltage transients.

Referring to the diagram below, the large 1500W TVS diode is placed directly across the DC supply to the 74ACT244 DCS driver chip.   There are no capacitance concerns with the power supply, and it provides a handy reference for the driver output protection.  In the MTH TIU, each of the 74ACT244 have four of their output gates paralleled to provide greater current drive.  The Schottky diodes D2 and D3 provide rails for maximum and minimum voltage at pin-18 of the 74ACT244, that is part of one of the output gate groups.  Similarly, the Schottky diodes D4 and D5 provide rails for the other output gate groups at pin-7.  The Schottky diodes were selected for their fast switching capability.  They provide a path to shunt the transient through D1, the TVS diode.

IMO, this is probably going to be the easiest solution to install that should keep the 74ACT244 chips healthy.  I put this on two TIU's, and both are still alive and well with full strength signals.  Hardly a definitive test, but it's at least promising.

The small PCB is designed to have the large TVS glued directly to the top of the 74ACT244 chip and the four wires from the solder points on the DCS Protect board are connected to the appropriate

Here's the BOM for the Protection Board.

Here's a ZIP with the Gerber files to make the PCB's.  I recommend using a thinner board, I use the 31mil thick boards for stuff like this.

MTH TIU Signal Protection Gerber.zip

Attachments

GRJ did design a PCB with schematic and parts list in the thread you referenced here for DCS signal protection. It's a long thread so it could  easily be missed.  Here's a link to the info: GRJ's post with the PCB info.  I don't recall if anything was ever posted about use or testing, but it's been quite a while since I've looked at the thread.

Edit: So much for that, I see GRJ beat me to it.

Last edited by rtr12

Thank you both.  Am I correct that this includes the basic fast Zener at the center, rather than being added to it separately?  The technical aspects of creating the above are out at my personal fringe.  Is this a board that you offer for sale, GRJ?  If so two of them at any price would probably be welcome!  If not I’ll need a little more to go on to order what I need and muddle through it.  Creating the PCB of a suitable size to fit the components nicely is square one there.  G cubed has agreed to install a suitable board if I can provide it for him with my TIU’s.  Otherwise he only does the basic larger-Zener upgrade.

The board with the TVS on one side and the four Schottky diodes on the other side is the complete package.  Secure the little board to the top of the 74ACT244 with something like Alene's Tacky Glue with the large TVS down and then solder fine #30 wire into the four drops to the chip.

I don't offer these for sale, I don't want to be stuck building them by hand.  If I figured I could sell a large quantity, I'd have had them made  However I'd have to make a few hundred to make it cost effective to have them produced, that would be too much money tied up with an uncertain future.

You can get the bare boards made at places like JLCPCB for peanuts and then just buy the components at local electronic parts distributors.  A lifetime supply of the bare boards will probably cost you around $25 including shipping.

Thanks again to GRJ for so much good information. For anyone else following with same interest, here’s a good 18 minute video tutorial for SMD hand soldering. Separately found references recommending about 410-435 deg F fine tip pencil temp for SMDs. 60/40 tin/lead rosin core solder melts at 375F.  The 410+ temp is needed in part to activate the rosin flux. Video:

https://m.youtube.com/watch?v=f9fbqks3BS8

I recommend 63/37 solder, it is an eutectic alloy with a melting point of 361°F.  Also, it transitions directly from a liquid to a solid which minimizes cold solder joints.  There's no good reason to use 60/40 over 63/37 solder that I'm aware of, but there is good reason to use the 63/37 solder for it's eutectic properties as well as the lower melting point.

I use 500F for soldering SMD boards, almost all the SMD parts are rated at soldering temperatures of 260C (500F) for ten seconds.  I've been using 500F for years soldering SMD parts of all descriptions, it's worked out fine.

IMO, you may find the 410-435 deg F too low for quickly heating the connection, and it is likely to make the job more difficult.  Years ago I used to try to solder SMD components at 450F, but the job got a lot easier by moving up to 500F.  I've noticed no adverse effects from increasing the temperature.

The board with the TVS on one side and the four Schottky diodes on the other side is the complete package.  Secure the little board to the top of the 74ACT244 with something like Alene's Tacky Glue with the large TVS down and then solder fine #30 wire into the four drops to the chip.

I don't offer these for sale, I don't want to be stuck building them by hand.  If I figured I could sell a large quantity, I'd have had them made  However I'd have to make a few hundred to make it cost effective to have them produced, that would be too much money tied up with an uncertain future.

You can get the bare boards made at places like JLCPCB for peanuts and then just buy the components at local electronic parts distributors.  A lifetime supply of the bare boards will probably cost you around $25 including shipping.

GRJ: Since it's my first board order to JLCPCB, not sure about all their detail specs. This is what I have saved to my cart, with your Gerber files uploaded.  Does it look OK? Thanks

Attachments

Images (1)
  • mceclip0

Add Reply

Post
The DCS Forum is sponsored by
×
×
×
×
Link copied to your clipboard.
×
×