- Thread starter
- #16
whether the renard code can handle 250k is a question for p. short. i have followed the documented standard and only allow 115200 as a max. very easy to change.
ed
ed
EthConGateway - LSP/Vixen/DMX/Renard
- Thread starter j1sys
- Start date
I am developing the newer 16 channel code with phil and the ability is there to try but i have been anable to test it as i couldnt get vixen to go higher.
Hopefully now with this card the ability to run 250K (or other similar speeds) it will make high channel counts easy.
I look forward to this board being available so i can test it.
Hopefully now with this card the ability to run 250K (or other similar speeds) it will make high channel counts easy.
I look forward to this board being available so i can test it.
- Thread starter
- #18
Scorpia -
I will add an optional 250Kbps Renard speed for your testing.
----
All -
Update on Gamma board status. Not good but not bad ....
Gamma #1 is 95% built. I haven't added the RS485 DMX/RDM resistors yet and haven't tested the RS485 output yet but expect to today.
I have run into a couple of very minor problems so far.
On the ECG-M24H the holes that I designed for the new bottom entry connector are too tight. You can install the board but with 32 pins of parallel alignment through the tight board holes and the connector it is EXTREMELY hard to disconnect. I am worried that people will flex their boards too much and then start having problems with broken traces. I hand drilled the plated through holes out (since I didn't need them) and enlarged a little and it made it much better. My original design was to allow some OEMs (other designers) to purchase the ECG-M24H and have the option of using MANY different connector combinations. I will redesign for larger holes and still keep them plated through in case they want to use a different connector but it will be very loose for soldering and they will probably need an alignment jig for their connectors. Anyone using our bottom entry connectors will have no problem and we will make both the connectors and risers available for other peoples use.
On the ECG-DMXRen8 I had a little more trouble. I had redesigned a new buffer/decoder section to the board to isolate and cleanup the SPI signals so we could speed back up to 8-10Mbps. I made a design error. The error precludes using ICSP while a 74AC138 is in its socket. If there were no other problems we could release these boards and you could either always remove the two 74AC138s before reprogramming the slaves or you could apply an Engineering Change Order (ECO) that fixes the problem. One of the attached photos shows the required cut and tack soldering of one lead of a small resistor.
After fixing that problem I continued testing and on Gamma #1 had a problem with Slave #6. After hours of signal tracing, chip swapping and trying to find the problem we isolated it to a simple short between two traces. We tested the same traces on 4 other boards and all similar traces and found no shorts. We cut the traces and replaced them with wire and the problem went away. So it was a one time defect that happened to occur on the first board I grabbed. I don't like the odds of that. I will reduce the signal trace sizes from their current 17mils to a smaller size to allow a larger clearance than the current 6mils. I believe that this will substantially reduce the chance of this error occurring again.
So, I know that the season is rapidly approaching but with these mechanical and electrical problems I believe I need to bite the bullet and order a Delta round of boards. I believe I can finish testing in the next 2-3 days and have another the Delta round in my hands within 2 weeks. I don't think I/we can afford the delay to ship some of these boards out to Gamma Testers and then wait for their response BEFORE ordering the Deltas. That would make Deltas at least 4 weeks away.
Your comments and understanding would be appreciated. As I have shown in some previous posts, the three Beta boards have been rock solid for weeks for me and the two beta testers. I expect this gamma board to be rock solid for me but feel that fixing these problems now is in the best interest of everyone.
I've attached a few photos of the ECO and shots of the new bottom entry mounting and the entire board. It is looking very nice.
-Ed
I will add an optional 250Kbps Renard speed for your testing.
----
All -
Update on Gamma board status. Not good but not bad ....
Gamma #1 is 95% built. I haven't added the RS485 DMX/RDM resistors yet and haven't tested the RS485 output yet but expect to today.
I have run into a couple of very minor problems so far.
On the ECG-M24H the holes that I designed for the new bottom entry connector are too tight. You can install the board but with 32 pins of parallel alignment through the tight board holes and the connector it is EXTREMELY hard to disconnect. I am worried that people will flex their boards too much and then start having problems with broken traces. I hand drilled the plated through holes out (since I didn't need them) and enlarged a little and it made it much better. My original design was to allow some OEMs (other designers) to purchase the ECG-M24H and have the option of using MANY different connector combinations. I will redesign for larger holes and still keep them plated through in case they want to use a different connector but it will be very loose for soldering and they will probably need an alignment jig for their connectors. Anyone using our bottom entry connectors will have no problem and we will make both the connectors and risers available for other peoples use.
On the ECG-DMXRen8 I had a little more trouble. I had redesigned a new buffer/decoder section to the board to isolate and cleanup the SPI signals so we could speed back up to 8-10Mbps. I made a design error. The error precludes using ICSP while a 74AC138 is in its socket. If there were no other problems we could release these boards and you could either always remove the two 74AC138s before reprogramming the slaves or you could apply an Engineering Change Order (ECO) that fixes the problem. One of the attached photos shows the required cut and tack soldering of one lead of a small resistor.
After fixing that problem I continued testing and on Gamma #1 had a problem with Slave #6. After hours of signal tracing, chip swapping and trying to find the problem we isolated it to a simple short between two traces. We tested the same traces on 4 other boards and all similar traces and found no shorts. We cut the traces and replaced them with wire and the problem went away. So it was a one time defect that happened to occur on the first board I grabbed. I don't like the odds of that. I will reduce the signal trace sizes from their current 17mils to a smaller size to allow a larger clearance than the current 6mils. I believe that this will substantially reduce the chance of this error occurring again.
So, I know that the season is rapidly approaching but with these mechanical and electrical problems I believe I need to bite the bullet and order a Delta round of boards. I believe I can finish testing in the next 2-3 days and have another the Delta round in my hands within 2 weeks. I don't think I/we can afford the delay to ship some of these boards out to Gamma Testers and then wait for their response BEFORE ordering the Deltas. That would make Deltas at least 4 weeks away.
Your comments and understanding would be appreciated. As I have shown in some previous posts, the three Beta boards have been rock solid for weeks for me and the two beta testers. I expect this gamma board to be rock solid for me but feel that fixing these problems now is in the best interest of everyone.
I've attached a few photos of the ECO and shots of the new bottom entry mounting and the entire board. It is looking very nice.
-Ed
Attachments
All
I have been a little remiss in providing a public updated on the beta testing but would like to support Ed's comments that the beta boards have been rock solid.
Ed and I have been working through some DMX software refinements for the Ren/DMX Slave board but the Master and Slave boards have been completely stable.
It is my belief that the time wasted to send the Gamma boards out and wait for feedback would just add a completely unnecessary delay to the final production release.
Having talked extensively on the phone with Ed discussing the couple of what are quite minor issues on the gamma revisions that, not sending out the Gammas and doing an order of Delta boards with a corresponding delay of say 2 weeks will result in a production release being available sooner than if the gammas were distributed for testing.
I have every confidence based on the current beta that a fully working product will be available in sufficient time for this season and I for one would rather see a short delay than a rushed release.
Cheers
Phil
I have been a little remiss in providing a public updated on the beta testing but would like to support Ed's comments that the beta boards have been rock solid.
Ed and I have been working through some DMX software refinements for the Ren/DMX Slave board but the Master and Slave boards have been completely stable.
It is my belief that the time wasted to send the Gamma boards out and wait for feedback would just add a completely unnecessary delay to the final production release.
Having talked extensively on the phone with Ed discussing the couple of what are quite minor issues on the gamma revisions that, not sending out the Gammas and doing an order of Delta boards with a corresponding delay of say 2 weeks will result in a production release being available sooner than if the gammas were distributed for testing.
I have every confidence based on the current beta that a fully working product will be available in sufficient time for this season and I for one would rather see a short delay than a rushed release.
Cheers
Phil
- Thread starter
- #20
Just a quick update and back to work ...
I've been testing the boards, finished all assembly of Gamma #1, finished assembly on 4 more Ren8 units to use for testing and finally hooked up a complete ECG connected to 8 Ren8 units.
I downloaded a little video program for my iPhone since I don't have a video camera and shot a 12 second video of the boards all hooked up and running.
I've include the links to two different formats of the video below. One is the raw Quicktime from the iPhone and the other is an SWF conversion of the same. For some reason I can play the .mov and .swf on my laptop but can't play them over the web. So I suggest right-clicking on the link and saving the file then playing it locally. That seems to work. The swf seems to cut off my last word 'text' and reloop indefinitely. Annoying, but I don't have time to play with fixing it right now. Each of the files are only about 2MB.
www.j1sys.com/Tutorials/ECG/ECGVideo.mov
www.j1sys.com/Tutorials/ECG/ECGVideo.swf
In the center of the video (sorry for the quality) is an ECG-DMXRen8 with an ECG-M24H mounted on top.
The blue cable is the Ethernet feed coming into the unit from my Ethernet switch. The little zip cord black cable connecting to the green pluggable screw terminal is coming from a 2A regulated 5VDC wall wart.
If you look close you will see jumper wires screwed into the 'injectors' at the front of the DMXRen8. This is supplying 5VDC on pin 3 of the RJ45 connectors to power the Ren8 units. I've built up a minimal circuit on each Ren8 and added a jumper to take 5VDC from pin 3. Instead of the optos I have plugged an LED into each socket.
Even though the Ren8 units are only displaying 8 slots/channels of dimmable values the data stream coming out each of the 8 cables is a full 128 slots Renard stream at 57.6Kbps at 25ms update.
I hope to finish testing tomorrow and get the Delta boards all ordered.
-Ed
I've been testing the boards, finished all assembly of Gamma #1, finished assembly on 4 more Ren8 units to use for testing and finally hooked up a complete ECG connected to 8 Ren8 units.
I downloaded a little video program for my iPhone since I don't have a video camera and shot a 12 second video of the boards all hooked up and running.
I've include the links to two different formats of the video below. One is the raw Quicktime from the iPhone and the other is an SWF conversion of the same. For some reason I can play the .mov and .swf on my laptop but can't play them over the web. So I suggest right-clicking on the link and saving the file then playing it locally. That seems to work. The swf seems to cut off my last word 'text' and reloop indefinitely. Annoying, but I don't have time to play with fixing it right now. Each of the files are only about 2MB.
www.j1sys.com/Tutorials/ECG/ECGVideo.mov
www.j1sys.com/Tutorials/ECG/ECGVideo.swf
In the center of the video (sorry for the quality) is an ECG-DMXRen8 with an ECG-M24H mounted on top.
The blue cable is the Ethernet feed coming into the unit from my Ethernet switch. The little zip cord black cable connecting to the green pluggable screw terminal is coming from a 2A regulated 5VDC wall wart.
If you look close you will see jumper wires screwed into the 'injectors' at the front of the DMXRen8. This is supplying 5VDC on pin 3 of the RJ45 connectors to power the Ren8 units. I've built up a minimal circuit on each Ren8 and added a jumper to take 5VDC from pin 3. Instead of the optos I have plugged an LED into each socket.
Even though the Ren8 units are only displaying 8 slots/channels of dimmable values the data stream coming out each of the 8 cables is a full 128 slots Renard stream at 57.6Kbps at 25ms update.
I hope to finish testing tomorrow and get the Delta boards all ordered.
-Ed
- Thread starter
- #21
Alec -
It is all dependent on speed and update rate:
At 57.6Kbps at 13ms update you could attach 1 Ren64 per slave = 8 Ren64s
At 57.6Kbps at 25ms update you could attach 2 Ren64 per slave = 16 Ren64s
At 57.6Kbps at 50ms update you could attach 4 Ren64 per slave = 32 Ren64s
At 57.6Kbps at 100ms update you could attach 8 Ren64 per slave = 64 Ren64s
At 115.2Kbps at 13ms update you could attach 2 Ren64 per slave = 16 Ren64s
At 115.2Kbps at 25ms update you could attach 4 Ren64 per slave = 32 Ren64s
At 115.2Kbps at 50ms update you could attach 8 Ren64 per slave = 64 Ren64s
At 115.2Kbps at 100ms update you could attach 16 Ren64 per slave = 128 Ren64s = 8192 dimmers = Max slot count allowed
-Ed
It is all dependent on speed and update rate:
At 57.6Kbps at 13ms update you could attach 1 Ren64 per slave = 8 Ren64s
At 57.6Kbps at 25ms update you could attach 2 Ren64 per slave = 16 Ren64s
At 57.6Kbps at 50ms update you could attach 4 Ren64 per slave = 32 Ren64s
At 57.6Kbps at 100ms update you could attach 8 Ren64 per slave = 64 Ren64s
At 115.2Kbps at 13ms update you could attach 2 Ren64 per slave = 16 Ren64s
At 115.2Kbps at 25ms update you could attach 4 Ren64 per slave = 32 Ren64s
At 115.2Kbps at 50ms update you could attach 8 Ren64 per slave = 64 Ren64s
At 115.2Kbps at 100ms update you could attach 16 Ren64 per slave = 128 Ren64s = 8192 dimmers = Max slot count allowed
-Ed
Hi all
A little video showing the ECG24H and RenDMX slave receiving 4096 channels of data and driving a 300ch/100RGB pixel string
EthconGateway and 6803 RGB Pixels
Cheers
Phil
A little video showing the ECG24H and RenDMX slave receiving 4096 channels of data and driving a 300ch/100RGB pixel string
EthconGateway and 6803 RGB Pixels
Cheers
Phil
neilric99
Full time elf
Phil, this is great to see the EthConGateway in action.
So what you have is the EthConGateway with a ETHDMXREN8 to the DMXtoSPI interace driving the string?
Cannot wait until these boards become available, looks like my next purchase.
Neil
So what you have is the EthConGateway with a ETHDMXREN8 to the DMXtoSPI interace driving the string?
Cannot wait until these boards become available, looks like my next purchase.
Neil
AussiePhil said:
budude
Way behind schedule - again...
Wow - that is beyond words... Not just the coolness factor but also puts the magnitude of sequencing this many channels into perspective. Very Cool stuff!
neilric99 said:
Neil
keep an eye on this board http://forums.auschristmaslighting.com/index.php/board,34.0.html as progress is made updates will be posted.
Cheers
Phil
budude said:
Brian,
Under no circumstance would i attempt to actually sequence that in Vixen, LSP will be the tool. Vixen was just a easy way to do a random sequence
Cheers
Phil
- Thread starter
- #27
Attached below are the long awaited PRELIMINARY releases of the source code for the Master and Slaves.
WARNING - This is preliminary code and it will probably not compile on your target machine. This is not done to make it difficult for you. This is a simple ZIP generated by MPLAB of the two projects. The projects rely on a lot of other source code that must be installed for this to operate and, currently, it will probably be dependent on the directory structure that we use to develop Microchip code.
We have a lot of revisions and cleanup to do this week for release next week. We will also work diligently on making the release of the source code an easy to use package that can be installed on your computer for development.
So browse the source code as presented and start to get a feel for the code and the communications that we perform between the TCP/IP and via SPI with the slave.
MUCH, MUCH more to come ASAP for developers use. As an end-user you don't have to worry about this. We will be releasing all updates as HEX files like most firmware and you can then burn it to your unit with a PICkit or other programmer.
Early fall we expect to replace the need for the PICkit with a TFTP boot updater that will allow you to update the firmware via the TCP/IP.
Thanx,
-Ed
WARNING - This is preliminary code and it will probably not compile on your target machine. This is not done to make it difficult for you. This is a simple ZIP generated by MPLAB of the two projects. The projects rely on a lot of other source code that must be installed for this to operate and, currently, it will probably be dependent on the directory structure that we use to develop Microchip code.
We have a lot of revisions and cleanup to do this week for release next week. We will also work diligently on making the release of the source code an easy to use package that can be installed on your computer for development.
So browse the source code as presented and start to get a feel for the code and the communications that we perform between the TCP/IP and via SPI with the slave.
MUCH, MUCH more to come ASAP for developers use. As an end-user you don't have to worry about this. We will be releasing all updates as HEX files like most firmware and you can then burn it to your unit with a PICkit or other programmer.
Early fall we expect to replace the need for the PICkit with a TFTP boot updater that will allow you to update the firmware via the TCP/IP.
Thanx,
-Ed
mrpackethead
Full time elf
nice work Ed. Will get busy and start reading...
- Thread starter
- #29
Updated Status - New Product Announcement - Ordering Window Opening Soon
We wanted to give you a quick update on progress of the EthConGateway line of products.
The Rev 1.2 ECG-DMXRen8 boards have been received, several assembled, and fully tested. The Rev 1.2 ECG-M24H boards are arriving in a few minutes. We expect them to quickly pass testing. The ONLY changes to the ECG-M24H Rev 1.2 were to enlarge some hole sizes and a silk screen change to denote revision #.
With these two boards in hand and all of the parts needed to package the kits for sale we expect to open the ordering window for ordering the complete ECG-DMXRen8 packages in approximately 24 hours. Please watch your forums for instructions. We are going to be shipping these approximately 10 per week for the next 4 weeks. Then, based on demand expressed, we will order larger quantities for shipping early September. We will be offering a CIA (Christmas In August) sale on these items with a nice discount for those people ordering and paying in August. These units will be available in 4 slave and 8 slave kits and you can order an upgrade kit at a later date to add the additional 4 slaves. For those who don't already have an ICSP programmer or want to update to a new model we will also be offering these in bundles with PICkit2, PICkit3, or ICD3 PLUS during the CIA sale the discount will be larger with the PICkit2, PICkit 3, or ICD3. We are doing this because the current release of software REQUIRES an ICSP capable programmer. We expect to update the software frequently over the next 2 months as we find things that need to be enhanced based on customer feedback. We expect to update the software to TFTP and remote updating that will preclude the need for the ICSP programmers but most of you will continue to use the ICSP programmers for your lighting controller boards. Again, please wait 24 hours for more information on ordering.
We want to also announce today that we are almost ready to order a prototype/beta run of a new product - ECG-DR4. This product combines the ECG-M24H and a 4 slave version of the ECG-DMXRen8 onto a single 12cm x 12cm board. This board will use all SMD devices except a few connectors that we still prefer using through hole technology on. This board will be fully assembled and tested and should be available for approximately one-half the cost of an ECG-DMXRen8 package. So in essence the price will be the same. For a small additional fee it will be available mounted in a 12cm x 12cm red aluminum case that a lot people call a 'car amp' case. We feel that this is NOT a replacement for the ECG-DMXRen8 but a complement to our product line. We expect working prototype/betas to be in the hands of testers within two weeks and expect production quantities by the end of August. Since this is really only a repackaging of the existing design we expect it to fly through the beta program. Again, initially, it will require an ICSP programmer and we will be selling similar packages with similar discounts as a CIS (Christmas In September) sale on this product. The only thing you lose on the ECG-DR4 is the injector connectors across the front and the fun of breathing rosin core solder fumes for 8 hours.
More details and photos of the ECG-DR4 should be available in about 12 hours but we will definitely publish them before opening the order window on the ECG-DMXRen8, even if we have to slip that window by a few hours, so you can make an educated decision on whether to order an ECG-DMXRen8 or wait for the ECG-DR4.
Of course in a complex layout you can mix ECG-DMXRen8 and ECG-DR4 units in your topology and drive them with one Vixen/LSP plugin.
And looking ahead to late September we expect to have the TFTP option working on ALL products and we expect to introduce our Pro line of EthConGateway. It will start with ECGPro-DR4 which will be similar to the ECG-DR4 but will include POE or local power, 1.5KV isolation on power input, 1.5KV isolation on logic power, 1.5KV isolation on EACH RS485 power line, isolated RS485 drivers, transient protections on RS485 lines, and XLR-5 connectors. Of course the price will probably be quadrupled but it will still bring a cost effective 4 port gateway to the professional community. If all the stars and planets line up correctly we MAY even introduce this at LDI in Las Vegas on October 22nd.
Thanx for all your support and positive comments. I assure you we are working as hard as possible to get these to market as soon as possible. We have one FedEx arriving today, one tomorrow and one Thursday. We have two DHLs arriving today, one tomorrow and one Friday. We have two UPSes arriving tomorrow, one Wednesday and probably one more on Friday. So things are hopping as we start to prepare all of this for shipping to customers.
-Ed
We wanted to give you a quick update on progress of the EthConGateway line of products.
The Rev 1.2 ECG-DMXRen8 boards have been received, several assembled, and fully tested. The Rev 1.2 ECG-M24H boards are arriving in a few minutes. We expect them to quickly pass testing. The ONLY changes to the ECG-M24H Rev 1.2 were to enlarge some hole sizes and a silk screen change to denote revision #.
With these two boards in hand and all of the parts needed to package the kits for sale we expect to open the ordering window for ordering the complete ECG-DMXRen8 packages in approximately 24 hours. Please watch your forums for instructions. We are going to be shipping these approximately 10 per week for the next 4 weeks. Then, based on demand expressed, we will order larger quantities for shipping early September. We will be offering a CIA (Christmas In August) sale on these items with a nice discount for those people ordering and paying in August. These units will be available in 4 slave and 8 slave kits and you can order an upgrade kit at a later date to add the additional 4 slaves. For those who don't already have an ICSP programmer or want to update to a new model we will also be offering these in bundles with PICkit2, PICkit3, or ICD3 PLUS during the CIA sale the discount will be larger with the PICkit2, PICkit 3, or ICD3. We are doing this because the current release of software REQUIRES an ICSP capable programmer. We expect to update the software frequently over the next 2 months as we find things that need to be enhanced based on customer feedback. We expect to update the software to TFTP and remote updating that will preclude the need for the ICSP programmers but most of you will continue to use the ICSP programmers for your lighting controller boards. Again, please wait 24 hours for more information on ordering.
We want to also announce today that we are almost ready to order a prototype/beta run of a new product - ECG-DR4. This product combines the ECG-M24H and a 4 slave version of the ECG-DMXRen8 onto a single 12cm x 12cm board. This board will use all SMD devices except a few connectors that we still prefer using through hole technology on. This board will be fully assembled and tested and should be available for approximately one-half the cost of an ECG-DMXRen8 package. So in essence the price will be the same. For a small additional fee it will be available mounted in a 12cm x 12cm red aluminum case that a lot people call a 'car amp' case. We feel that this is NOT a replacement for the ECG-DMXRen8 but a complement to our product line. We expect working prototype/betas to be in the hands of testers within two weeks and expect production quantities by the end of August. Since this is really only a repackaging of the existing design we expect it to fly through the beta program. Again, initially, it will require an ICSP programmer and we will be selling similar packages with similar discounts as a CIS (Christmas In September) sale on this product. The only thing you lose on the ECG-DR4 is the injector connectors across the front and the fun of breathing rosin core solder fumes for 8 hours.
More details and photos of the ECG-DR4 should be available in about 12 hours but we will definitely publish them before opening the order window on the ECG-DMXRen8, even if we have to slip that window by a few hours, so you can make an educated decision on whether to order an ECG-DMXRen8 or wait for the ECG-DR4.
Of course in a complex layout you can mix ECG-DMXRen8 and ECG-DR4 units in your topology and drive them with one Vixen/LSP plugin.
And looking ahead to late September we expect to have the TFTP option working on ALL products and we expect to introduce our Pro line of EthConGateway. It will start with ECGPro-DR4 which will be similar to the ECG-DR4 but will include POE or local power, 1.5KV isolation on power input, 1.5KV isolation on logic power, 1.5KV isolation on EACH RS485 power line, isolated RS485 drivers, transient protections on RS485 lines, and XLR-5 connectors. Of course the price will probably be quadrupled but it will still bring a cost effective 4 port gateway to the professional community. If all the stars and planets line up correctly we MAY even introduce this at LDI in Las Vegas on October 22nd.
Thanx for all your support and positive comments. I assure you we are working as hard as possible to get these to market as soon as possible. We have one FedEx arriving today, one tomorrow and one Thursday. We have two DHLs arriving today, one tomorrow and one Friday. We have two UPSes arriving tomorrow, one Wednesday and probably one more on Friday. So things are hopping as we start to prepare all of this for shipping to customers.
-Ed
- Thread starter
- #30
Well a lot of stuff arrived today and after opening and sorting it all I had a few unforeseen problems that have taken all day to correct.
So the work I expected to do today didn't get done and all release and opening of the ordering window will be delayed by 24 hours.
So the new target is to document/release information formally on the ECG-DR4 about 18:00 EDT USA Tuesday and open up the ordering process about 08:00 EDT USA Wednesday.
Thanx,
-Ed
So the work I expected to do today didn't get done and all release and opening of the ordering window will be delayed by 24 hours.
So the new target is to document/release information formally on the ECG-DR4 about 18:00 EDT USA Tuesday and open up the ordering process about 08:00 EDT USA Wednesday.
Thanx,
-Ed