## KF5OBS #6: FreeGiveaways

Up for grabs are the following 3 give-aways:

Ti TMS320VC5505 eZdsp USBtick
Altera Cyclone II FPGA Starter Development Kit
Analog Devices BF592 EZLite kit

To participate, you must subscribe to my YouTube channel, like this video, comment under this video and send an eMail to giveaway@kf5obs.com stating your name and that you would like to participate.

If you want to double your chances, just make this giveaway known through a social media site of your choice. Make sure you submit proof (e.g. screenshot or link) to me via eMail.

## How to entertain a toddler

Working from home can be a challenge if you have children. Naturally, a little toddler is always curious and wants to play with Daddy’s “toys.” So how to safely entertain a toddler in an electronics lab?

Naturally, the more effort one puts into prohibiting the handling of certain items, the more interesting they seem. The electric drill is a great example for that. But an old drill bit was the solution to the problem. It’s sharp enough to cut empty paper boxes but not even remotely sharp enough to do some serious damage.

Drilling holes in empty boxes

The oscilloscopes, just like pretty much any other test & measurement device, has always been interesting to her. All the colorful buttons and traces. Plus one can actually watch Elmo on it. One of her newest ideas is to use the LabNotebook function of my Teledyne LeCroy HDO4024 oscilloscope as a modern Etch A Sketch®.

Lena’s face speaks volumes; She’s a happy little engineering student

LabNoteook is a feature of many upper-level LeCroy oscilloscopes. It allows the user to store waveforms, a screenshot and the current channel / acquisition status. The user can also graph over the screenshot and place text labels. Paired with the oscilloscope’s touchscreen display and a stylus pen, this makes for a perfect pastime activity for a toddler.

Selecting colors and shapes is a piece of cake

After the screen gets too crowded, she usually hits the “delete all” button and starts over. Every once in a while, I actually manage to save some of her artwork. Here’s an example:

Lena’s LabNotebook report

[1] LabNotebook, Teledyne LeCroy http://teledynelecroy.com/

[2] LabNotebook Report, Lena http://jaunty-electronics.com

## Look Inside: Teledyne LeCroy HDO4024

True beauty lies within, or so they say. For an exclusive look under the hood, I decided to break the warranty seal on my Teledyne LeCroy HDO4024.

It took me a while to debate with myself whether or not I should void the warranty on this scope. Curiosity won and I couldn’t resist, I just had to open the box up. After removing the plastic case, a big metal shield is exposed. The metal case is very sturdy and not some floppy budget shield that you see in some of the test equipment around nowadays.

Solid metal cage under the plastic cover

After removing the metal case, the PC side of the scope is visible. I call it the PC side because that’s basically what it is. The scope contains just about anything that you would normally find inside your PC, aside from maybe an optical drive.

Top view of the opened HDO4024

To the right, you can see a standard off-the-shelf PC motherboard. On the left side under the black cover is a switch-mode power supply. The fan in the center brings plenty of air into the acquisition board section of the scope. One thing I noticed immediately is LeCroy’s effort to make everything mechanically stable. Even though in the picture I already disconnected the acquisition board from the top section of the scope, one can clearly see that all cables are mechanically secured and the connectors are well-chosen.

The red PCB in the back carries a few interfaces on the right side and the power mains connector plus a varistor for transient suppression on the left side. The BNCs are the 10 MHz reference input and the external trigger out connectors. Even though it is one solid PCB, there is a clear electrical separation between the peripheral side with a full copper pour on the back side and the power side.

10 MHz reference input (right) and trigger out (left) BNC connectors

The device class USB connector goes straight to the acquisition board.

IEC320 power connector with integrated filter

No need to reinvent the wheel on the power filtering side. There are plenty of good IEC320 connectors with integrated filters out there. LeCroy chose a Schurter 5120.4074.0 IEC320 conenctor with inlet filter, available from DigiKey for about $8/piece. 2 GB of RAM Even though the RAM shows “Product of China” in big capital letters, the manufacturer, Micron Technology, is actually a US-based semiconductor company. Crucial V4 Hard-drive Some other manufacturers, like Agilent with their 9000 H-series oscilloscopes, do offer SSDs as a pricey option. I am very happy that the Crucial V4 solid-state drive (SSD) is standard for the LeCroy HDO. Not much needs to be said about Crucial. I mean, Crucial parts are pretty much everywhere you look. This mysterious board is home to a PLX NET2282 PCI to High-Speed USB 2.0 Controller The board pictured above looks very mysterious at first due to all the cryptic stickers on them. It turns out to be a carrier board for the PLX NET2282, a PCI to High-Speed USB 2.0 Controller. It seems that most of the communication between the motherboard and the acquisition board happens via PCI Express (PCIe). There is a jumper cable from the PCIe slot on the motherboard to a PCIe slot on the acquisition board. Judging by the width of the jumper, only one PCIe lane is used, but that should be plenty of throughput for the scope. PCIe counterpart on the acquisition board Even though it’s not captured on the pictures, the PCIe jumpers are mechanically secured with a little clamp. So even though there should be plenty of tension to hold the jumper in place just from the bus connector alone, LeCroy didn’t want to take any chances. LPC2378 is an ARM7TDMI-S based high-performance 32-bit RISC Microcontroller The PCIe bus connector on the acquisition board is wired straight up to a NXP LPC2378, an ARM7TDMI-S based high-performance 32-bit RISC Microcontroller. I suspect that microcontroller’s sole purpose is to serve as a PCIe bridge and dispatch data/commands between the motherboard and the acquisition board. EDIT (10/04/2013): The previous statement can’t be true as the NXP LPC2378 is way too slow to handle PCIe even remotely. Upon further inspection, it appears the PCIe bus is actually connected to the FPGA just above the big ASIC in the center of the scope. But before I get into details, lets zoom out and get a general overview of the acquisition board. Overview of the acquisition board On top of the fact that I really love the red silk screen, I think the board looks very neat and clean. The PCB layout clearly groups the components in functional blocks. The top-center and top-right areas of the board seems to be primarily reserved for power conditioning and power distribution. The top-left corner is host to the PCIe interface and the NXP LPC2378 we talked about earlier. The bottom-right corner appears to be mostly the domain of clock generation and distribution. I have to admit, I am a bit surprised that the clock section isn’t shielded at all. I am sure the scope is EMC compliant, but at the same time, this looks like an accident waiting to happen where external interference could make its way into the clock generation section of the scope. Teledyne LeCroy may want to revise this and see if my suggestion is relevant for practical applications. While I am writing this, I do remember that I can definitely “see” a 125 MHz (and 625 MHz) peak when using the spectrum analyzer option. Looking at the crystal oscillator’s frequency, it’s apparent where this spike is coming from. Sure, this is an oscilloscope and not a full blown RF spectrum analyzer, but at the same time, this is a high-end piece of equipment. And with 4096 resolution steps, even tiny interference makes it onto the screen somehow. Another overview of the acquisition board The bottom-center and bottom-right of the board is occupied by a large metal shield. Under the shield, there are 5 front-ends. Why 5 on a 4-channel scope? The fifth is for the external trigger input. It seems that the front-end for the trigger input is pretty much identical to the actual input channel front-ends. But we’ll get to the front-ends later. Under the 5 metal cans, which are located right above the big front-end shielding, are the ADCs. The first 4 are for channels 1-4 and the fifth one is for the external trigger input. The first 4 ADCs are connected to the FPGAs in the center of the PCB through matched length trace pairs. The kind of ADC used in this oscilloscope is naturally the key element for the scope’s performance. Just as important, however, is the analog front-end. Big power connector for the main supply The acquisition board’s main power supply seems to be a single voltage rail. Two big wires feed the single voltage supply into the board’s own power conditioning section. Power conditioning and an I/O header In the middle of all the power conditioning elements is a 14-pin header with standard 2.54mm spacing. The connector is wired to a WT245 (4-BIT DUAL-SUPPLY BUS TRANSCEIVER) and a LCX244 (Low Voltage Buffer/Line Driver). I assume this is some sort of programming interface, possibly JTAG or something similar. A spot for some unpopulated ICs All over the board I found plenty of little stickers with 4-digit numbers on them. I wonder what they are. I doubt they’re to keep track of parts revisions as you’d expect something like that to be integrated in the silk screen. Maybe employee IDs from the factory identifying who assembled and tested the scope? I guess we’ll never know unless LeCroy lets us in on that secret. There are also plenty of spots with unpopulated parts. Could it be that they use the same PCB for the HDO6000 series as they use for the HDO4000 series? Either that or the parts were deemed unnecessary and the layout itself is from an older revision. At this point, I may want to mention that my Teledyne LeCroy HDO4024 is one of the first production HDOs leaving the factory. I am sure that since then, there is at least one hardware revision. Clock generation and expansion slot As an RF guy, I of course find the above picture very interesting. It’s the clock generation (left) and clock distribution (right) section of the scope. The little black coax cable is connected to the external 10 MHz reference input. A Crystek CVHD-950 125 MHz VCXO [1] seems to be the scope’s internal frequency reference. Crystek promises an ultra-low phase noise performance of better than -135 dBc/Hz with a distance of 1 kHz to the carrier. Two expansion slots can also be seen in the previous picture. They are most likely for the planned Mixed Signal option. I am not sure if that option is already in production but I know LeCroy has shown demos of it before. Channel 2 front-end Just as important as the ADC is the analog front-end. An ADC is a great “Garbage In/Garbage Out” part. That means if you feed a low-quality signal into it, you’ll have a low-quality output. Therefore, an ADC can only be as good as the analog front-end permits. That was reason enough for me to crack the shielding open to see what’s inside. The left side is where the signal comes in from the BNC and then exits to the ADC on the right. Channel 1 front-end I noticed that the front-end of channel 1 looked a bit different from all the other channels. There is a little coax wire from the clock distribution side going into the shielded box connected to a connector that’s soldered onto a trace. It doesn’t really look like the PCB was designed for that connector to be there. This actually does look like a post-production modification. Someone probably messed something up and the LeCroy engineers decided to bodge this little connector onto the PCB. Looks like someone bodged something onto the front-end section of channel 1 It looks a little ugly, but that’s just how it goes with bodges. As I circuit design engineer I do understand that often modifications need to be made. And electrically this modification is probably great. However, I wish they’d at least clean all the solder flux up a bit. On the bright side, that seems to be pretty much the only real complaint I have about my scope. ADCs and FPGAs for channels 1 & 2 The defining parts of this oscilloscope of course have to be the ADCs. So what ADCs is LeCroy using? They certainly didn’t make it easy at all to get a glimpse at the part number. The solid metal shielding made it rather difficult to figure out what part it was. My x-ray vision enabled me to figure out, that it is – to my surprise – an off-the-shelf part. The secret is a Texas Instruments ADC12D1600RFIUT [2], an extremely powerful 12-Bit, 3.2 Gs/s RF Sampling ADC. Texas Instruments mentions a steep price of$2399.47 per ADC at a 100 piece quantity. Now obviously, Teledyne LeCroy pays significantly less, but I still imagine the ADCs eat up quite a bit of the budget.

ADCs and FPGAs for channels 3 & 4

Each ADC is connected to an FPGA using matched length differential pairs. It is absolutely imperative that all signals are 100% time-coherent when they make their way from the ADC to the FPGA. Therefore, the wire length between each ADC and FPGA needs to be exactly the same. As the ADCs of channels 1-2 are closer to the FPGAs than the ADCs for channels 3-4, the designers had to make the differential pairs for channels 1 – 2 a bit longer. That’s what the meander pattern is for.

Channel 1 & 2 FPGA section without heat sinks on the SDRAM

If you take a close look, you’ll notice that the SDRAM for the FPGA group of channels 1-2 does not have heat sinks, while the same circuitry for channels 3-4 does. This is not a mistake, though. The cooling fan is situated right above the channel 1-2 FPGA group and can cool this area more efficiently than the channel 3-4 group.

Channel 1 & 2 FPGA section with heat-sinks on the SDRAM

Some sort of ASIC

There’s an ASIC marked “HTT711 38C12″ on the board. Judging by its position on the board, I suspect this is the timing and trigger brain of the HDO. Since this is a proprietary ASIC, I have no additional information on this part though.

Alright, that concludes the teardown for now. I am considering making a teardown video. Naturally, video can convey more details than this little post here. If you’d be interested in seeing a teardown video of the Teledyne LeCroy HDO4024 on my YouTube channel, let me know in the comments below!

Update (10/06/2013): Due to the large amount of positive feedback on this teardown, I also made a video of the HDO4024 teardown.

[1] CVHD-950 VCXO Datasheet, Crystek: http://www.crystek.com/

## IC-706 MKII Extended Tx Modification

Expanding the transmit range of an amateur radio transceiver is probably the most common modification. Because of its use associated with the Military Auxiliary Radio Service (MARS) or Civil Air Patrol (CAP), this modification is often referred to as MARS or CAP mod. This article shows how to perform this extended transmit modification on the Icom IC-706 MKII.

Asides from participating in MARS or CAP activities, there can be plenty of different reasons to extend the transmit frequency range of an amateur radio transceiver. I for instance, like using my amateur radio gear as signal generator for various experiments. A few days back, I got an IC-706 MKII in to perform the MARS / CAP modification on it. To my surprise, many online instructions on how to do this were either presenting the outdated main-board or poor graphics. So here’s the high-resolution instruction for the newer versions of the IC-706 MKII with main-board (B5209Q).

First off, one word of advise. If you have problems working with surface mount parts, this modification is not for you. The one diode that needs to be removed, is extremely small. How small you ask? Well, take a look:

Comparison of diode D2030 next to a cent coin

To get to the diode, set the IC-706 on the table (LCD facing forward), remove the lid, and disconnect the speaker. You should now see the following:

The main board of the IC-706 MKII

In the top right corner of the main PCB is the diode D2030 that needs to be removed for an extended transmit frequency range. Here’s the close-up view:

Diode D2030 (red circle)

Simply remove the diode and re-assemble the IC-706. Yes, that’s it. The removal of a single diode will convince your radio to transmit out of band.

Diode removed

## Review: ADSP-BF592 EZ Lite Kit

Blackfin processors from Analog Devices (ADI) precede their reputation. Anyone who has even remotely thought about Digital Signal Processors (DSPs) has come across ADI’s Blackfin processor series. ADI considers the BF592 as the low-cost entry point into the Blackfin portfolio of processors. To make evaluation a piece of cake, ADI offers the BF592 EZ-KIT Lite evaluation kit. Let’s find out how good it is.

The ADSP-BF592 EZ Lite is one of the many things that has been sitting on my shelf for months awaiting review. It has been about 8 month since this board shipped to all RoadTesters by Element14. RoadTest is a program sponsored by Element 14 that promotes real product reviews by real people. Reviewing such a board is rather time-consuming as this is a professional development tool, not your common Arduino or other AVR microcontroller toy. This is probably the reason why nobody else has submitted a RoadTest review yet either.

Because of this fact, I decided to keep this review a bit on the entry level by giving step-by-step instructions on how to set up the IDE and program the board. However, the most important rule for working with a product like this is RTFM[1].

Analog Devices ADSP-BF592 EZ Lite close-up

#### Overview and Tech Specs

Here are the tech specs as provided by ADI:

• Analog Devices ADSP-BF592 Blackfin processor [2]
• Core performance up to 400 MHz / 800 MMACs
• 64-lead (9 mm × 9 mm) LFCSP package
• Two 16-bit MACs, two 40-bit ALUs, four 8-bit video ALUs,
40-bit shifter
• RISC-like register and instruction model for ease of
programming and compiler-friendly support
• 68K bytes of core-accessible memory
• 64K byte L1 instruction ROM
• Flexible booting options from internal L1 ROM and SPI memory or from host devices including SPI, PPI, and UART
• Four 32-bit timers/counters, three with PWM support
• 2 dual-channel, full-duplex synchronous serial ports (SPORT), supporting eight stereo I2S channels
• 2 serial peripheral interface (SPI) compatible ports
• Parallel peripheral interface (PPI), supporting ITU-R 656
video data formats
• 2-wire interface (TWI) controller
• Event handler with 28 interrupt inputs
• 32 general-purpose I/Os (GPIOs), with programmable
hysteresis
• Debug/JTAG interface
• On-chip PLL capable of frequency multiplication
• Analog Devices AD5258 TWI digital potentiometer
• Analog Devices ADP1715 low dropout linear regulator
• On-board hardware debugger
• Battery interface with on-board battery charger
• 16 MB external SPI flash memory
• Audio codec
• Analog Devices SSM2603 stereo, 24-bit analog-to-digital and digital-to-analog converter
• 3.5 mm Line In/Out, headphone and microphone jacks
• DB9 female connector

Sounds like plenty of horsepower to play with. The stereo ADC/DAC including 3.5 mm jacks makes this board an easy entry point for people wanting to learn how to implement real-time audio effects using DSP technology.

#### Unboxing

As usual with ADI products, the board comes well-protected in a large protective foam filled box. The box is supposed to contain a CD with software, but in my case, it didn’t (See: Software Installation).

Analog Devices ADSP-BF592 EZ Lite, fresh out of the box

If you have absolutely no experience with programming Blackfin DSPs, I highly recommend that you download all the documentation available from ADI. ADI generally does a really good job at supplying high quality documentation. For practical reasons, however, they often split documentation relating to one product over several different individual documents. I highly recommend reading the ADSP-BF592 EZ-KIT Lite Evaluation System Manual and for beginners the document labelled Getting Started With Blackfin Processors. Both are available from ADI’s website [ 3].

#### Software Installation

At the time this product was shipped to me, it was supposed to come with a 90-day trial version of VisualDSP++. However, it didn’t. I contacted ADI and was informed that VisualDSP++ is, despite still being supported, no longer recommended for new designs. They shipped me the missing VisualDSP++ ASAP and also provided me with a full license of the recommended CrossCore Embedded Studio for review purposes. Special thanks to ADI’s Bob Olson and Denis Labrecque for handling this request within less than a few hours.

I’m not sure if newer versions of the BF592 EZ-Kit Lite will be shipped with CrossCore or not. In any case, you can get a trial version of CrossCore can be obtained from ADI’s website.

CrossCore Embedded Studio uses Eclipse, a multi-language Integrated Development Environment (IDE), as workspace. Every serious programmer will come in touch with Eclipse sooner or later. Therefore, I personally find it very comforting to be able to use the familiar Eclipse workspace and not having to learn yet another IDE.

#### Example Programs

In order to access the ADI-provided code examples, it is necessary to install the ADSP-BF592 EZ-KIT Lite® Board Suppport Package [4].

Once installed, the demo projects are available through the Help -> Browse Examples menu.

Accessing code examples through the Eclipse IDE

A list of samples will show up and all one needs to do to load the sample code is to double click on the example. That’s it.

Code examples from the ADSP-BF592 EZ-KIT Lite® Board Suppport Package

The first program I tried was the Audio Loopback example project. This program basically copies the ADC values straight into the DAC. There is absolutely no processing on the DSP side involved. Just a plain memcopy().

Example code implementing an audio loop-back between the line-in and headphone out connectors.

If you just want to test the program real quick, simply connect the board via USB to the PC, build the entire project and use the On-Board debugger to run the program on the target environment. The next section will describe how to burn a program into the permanent brain of the board.

#### Programming the Board

Programming the board requires quite a bit of pioneering. Even though there are a lot of references to programming in the datasheets and auxiliary documentation, there’s not a single functioning example to be found anywhere.

After a while of reading the documentation, I designed myself a little batch file that would take a compiled .dxe file as argument and program it to the ADSP-BF592 EZ Lite Kit via USB.

Here’s the code:

elfloader %1 -b flash -f hex -width 16 -dmawidth 8 -init "C:\Analog Devices\CrossCore Embedded Studio 1.0.2\Blackfin\ldr\ezkitBF592_initcode_ROM-V02.dxe" -o Outfile.ldr -si-revision 0.2 -proc ADSP-BF592-A

 

cldp -proc ADSP-BF592-A -emu KIT -driver "C:\Analog Devices\ADSP-BF592_EZKIT-Rel1.0.0\BF592_EZ-Kit_Lite\Blackfin\Examples\Device_Programmer\serial\bf592_m25p16_dpia.dxe" -cmd prog -erase affected -format hex -file Outfile.ldr del Outfile.ldr

Make sure you change the file path for the init-code and the device programmer driver if they differ on your system. Also change “-si-revision 0.2” if your silicone revision is different from 0.2.

Save the code as a .bat file and place it in a convenient spot on your hard drive. If all the path settings are correct, simply drag the .dxe file you wish to program to the BF592 onto the .bat file and watch the magic happen.

Programming the BF592 using a batch file approach

This approach works but it’s a little unfortunate that ADI didn’t integrate the programmer option in the IDE. Since it’s based on Eclipse, I imagine it wouldn’t be too difficult to write a plug-in by yourself if needed.

I highly recommend this board to anyone trying to get into DSP programming. It may take you a while to get into it but the time investment will definitely pay off.

[1] Wikipedia, RTFM: http://en.wikipedia.org/

[3] ADI, BF592 EZ-Kit Lite: http://analog.com

## FT-101E and IC-718 Repair Impressions

Over the past weeks I had KF5FBR’s Yaesu FT101E ad N5WSA’s Icom IC-718 on my bench for various services. The FT-101 needed a general health check and brand new capacitors. No output powers was the problem the IC-718. The latter was caused by burned out final transistors. While working on them, I shot some pictures that I would like to share here.

Output filter section of the Icom IC-718

One of the two burned out final transistors of the IC-718

Driver section of the IC-718′s final amplifier

And now on to the Yaesu FT-101E:

Plenty of old caps, cobwebs, and dust on this FT-101E module

Yaesu FT-101E SSB filter and modulator board

## News and Status update

I haven’t been writing any articles in almost 3 month. I apologize for this big gap and appreciate your patience. Recent job development and my move to a new location offered little time for the blog or electronics in general.

I expect to have the new lab set up within a month or so. My target is to produce some sort of content at least once a week starting in August October. The plan is to produce a written article every 1st and 3rd week of the month and uploading a new video every 2nd and 4th week of the month. That means two videos and two classic text blog articles per month.

Frequency domain screenshot of a LeCroy HDO4024

There is a big bunch of manufacturer’s samples sitting in my lab. All of those are awaiting a review on the blog. It is probably enough material to fill the blog the next two years with content. However, i simply do not have the time to cover all and more stuff comes in every day. Therefore, I will be cherry-picking the most interesting bits and pieces.

For a noncomprehensive list of articles to come, see the preview page.

Remember, unless I receive input from my readers on what YOU want to read, I will be picking at my own discretion and according to my own personal preferences.